Predicting Credit Risk of Customers.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predicting the Credit Risk of Customers (Good or Bad)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The objective of this case study is to fit and compare 4 different binary classification models (classifiers) to predict whether a customer has good credit history or bad using the German Credit dataset. The dataset is sourced from the UCI Machine Learning Repository (https://archive.ics.uci.edu/ml/datasets/statlog+(german+credit+data))\n",
    "\n",
    "The dataset contains 20 descriptive feature with 1 binary class feature. Out of the 20 descriptive features 7 features are numerical while 13 are categorical. All of the categorical features are represented with some coding format which is explained below. The class feature has two classes 'Good' which is represented as '1' and 'Bad' which is represented as '2' respectively. The whole dataset contains 1000 observation instances.\n",
    "\n",
    "This report is organized as follows:\n",
    "- [Section 2 (Overview of the Data)](#2) Provides details of the features in the dataset along with their meaning. \n",
    "- [Section 3 (Data Preparation)](#3) Summarizes the data preparation steps taken to fit the data in classification models . \n",
    "- [Section 4 (Data Exploration)](#4) Explains relationship between different columns using various graphs and visualization techniques.\n",
    "- [Section 5 (Feature Selelction and Ranking)](#5) This section helps to rank and select the important feature for the further modelling.\n",
    "- [Section 6 (Data Modelling)](#6) This section first fits data into 4 different binary classification models and build classifiers.\n",
    "- [Section 7 (Hyperparameter Tuning)](#7)This section describes the hyperparameter tuning process for each classification algorithm.\n",
    "- [Section 8 (Performance Evaluation)](#8) This section calculates the performance of each of the classification model after HP tuning. \n",
    "- [Section 9 (Limitations and proposed Solutions)](#9) This section defines the limiations of the models and proposes future scope and solutions.\n",
    "- [Section 10 (Summary and Conclusion)](#10) provides a result of our experiment and a brief summary of our work in this report.\n",
    "- [Section 11(References)](#11) provides the reference of sources used for this project.\n",
    "\n",
    "This is Jupyter Notebook file contains both narratives as well as the related Python 3.6 codes.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Comaptibility of Altair to Jupyter Notebook.\n",
    "## The code below require Altair version 3.0 or above which indeed require Jupyter lab version 1.0 or above. \n",
    "## To process and update the same the following code needs to be uncommented and Vega package needs to be installed.\n",
    "## We need to execute this code each time loading this page. We have to Restart Kernel and clear output after executing the below commands.\n",
    "\n",
    "##%%capture\n",
    "##!pip install --upgrade altair\n",
    "##!pip install vega vega_datasets\n",
    "##!pip install --upgrade pip --user"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Overview of the Data. <a class=\"anchor\" id=\"2\"></a>\n",
    "\n",
    "### Information about the data and its features.\n",
    "\n",
    "The dataset classifies people described by a set of attributes as good or bad credit risks. The Dataset contains 20 Descriptive features whoes values are encoded in the below mentioned format. The class feature variable is 21st feature which have two different classes 1 and 2. '1' represents 'good' credit risk while '2' represents 'bad' credit risks. Below are the Feature Values and Attribute Information :\n",
    "\n",
    "Attribute 1: Status of existing checking account (categorical)\n",
    "- A11 : ... < 0 DM\n",
    "- A12 : 0 <= ... < 200 DM\n",
    "- A13 : ... >= 200 DM \n",
    "- A14 : no checking account\n",
    "\n",
    "Attribute 2: Duration in month (numerical)\n",
    "\n",
    "Attribute 3: Credit history (categorical)\n",
    "- A30 : no credits taken/ all credits paid back duly\n",
    "- A31 : all credits at this bank paid back duly\n",
    "- A32 : existing credits paid back duly till now\n",
    "- A33 : delay in paying off in the past\n",
    "- A34 : critical account/ other credits existing (not at this bank)\n",
    "\n",
    "Attribute 4: Purpose (categorical)\n",
    "- A40 : car (new)\n",
    "- A41 : car (used)\n",
    "- A42 : furniture/equipment\n",
    "- A43 : radio/television\n",
    "- A44 : domestic appliances\n",
    "- A45 : repairs\n",
    "- A46 : education\n",
    "- A47 : (vacation - does not exist?)\n",
    "- A48 : retraining\n",
    "- A49 : business\n",
    "- A410 : others\n",
    "\n",
    "Attribute 5: Credit amount (numerical)\n",
    "\n",
    "Attibute 6: Savings account/bonds (categorical)\n",
    "- A61 : ... < 100 DM\n",
    "- A62 : 100 <= ... < 500 DM\n",
    "- A63 : 500 <= ... < 1000 DM\n",
    "- A64 : .. >= 1000 DM\n",
    "- A65 : unknown/ no savings account\n",
    "\n",
    "Attribute 7: Present employment since (categorical)\n",
    "- A71 : unemployed\n",
    "- A72 : ... < 1 year\n",
    "- A73 : 1 <= ... < 4 years\n",
    "- A74 : 4 <= ... < 7 years\n",
    "- A75 : .. >= 7 years\n",
    "\n",
    "Attribute 8: Installment rate in percentage of disposable income (numerical)\n",
    "\n",
    "Attribute 9: Personal status and sex(categorical)\n",
    "- A91 : male : divorced/separated\n",
    "- A92 : female : divorced/separated/married\n",
    "- A93 : male : single\n",
    "- A94 : male : married/widowed\n",
    "- A95 : female : single\n",
    "\n",
    "Attribute 10: Other debtors / guarantors (categorical)\n",
    "- A101 : none\n",
    "- A102 : co-applicant\n",
    "- A103 : guarantor\n",
    "\n",
    "Attribute 11: Present residence since (numerical)\n",
    "\n",
    "Attribute 12: Property (categorical)\n",
    "- A121 : real estate\n",
    "- A122 : if not A121 : building society savings agreement/ life insurance\n",
    "- A123 : if not A121/A122 : car or other, not in attribute 6\n",
    "- A124 : unknown / no property\n",
    "\n",
    "Attribute 13: Age in years (numerical)\n",
    "\n",
    "Attribute 14: Other installment plans (categorical)\n",
    "- A141 : bank\n",
    "- A142 : stores\n",
    "- A143 : none\n",
    "\n",
    "Attribute 15: Housing (categorical)\n",
    "- A151 : rent\n",
    "- A152 : own\n",
    "- A153 : for free\n",
    "\n",
    "Attribute 16: Number of existing credits at this bank (numerical)\n",
    "\n",
    "Attribute 17: Job (categorical)\n",
    "- A171 : unemployed/ unskilled - non-resident\n",
    "- A172 : unskilled - resident\n",
    "- A173 : skilled employee / official\n",
    "- A174 : management/ self-employed/highly qualified employee/ officer\n",
    "\n",
    "Attribute 18: Number of people being liable to provide maintenance for(numerical)\n",
    "\n",
    "Attribute 19: Telephone (categorical)\n",
    "- A191 : none\n",
    "- A192 : yes, registered under the customers name\n",
    "\n",
    "Attribute 20: foreign worker (categorical)\n",
    "- A201 : yes\n",
    "- A202 : no\n",
    "\n",
    "Attribute 21: Credit Risk (Class Feature)\n",
    "- 1 : Good\n",
    "- 2 : Bad\n",
    "\n",
    "---\n",
    "\n",
    "### Methodology\n",
    "\n",
    "We are considering the below 4 binary classifiers to predict the target feature:\n",
    "- K-Nearest Neighbour (KNN)\n",
    "- Decision Tree (DT)\n",
    "- Random Forest Ensemble (RF)\n",
    "- Naive Bayes (NB)\n",
    "\n",
    "Before we start the modeling process, we try to find the relations between some descriptive features using some graphs and visualization techniques. This report includes 3 One-Variable graphs, 3 Two-Variable graphs and 3 Three-Variable graphs to explore the relation between them.\n",
    "\n",
    "The Modeling starts with the cleansing of the data and then transforming it into the classifier readable format. This transformation includes encoding the categorical descriptive features and then scaling them using one of the SKlearn scalars. First we split the dataset into training and test sets with 70:30 ratio. That means:\n",
    "  * The randomly selected 700 observation instances of the data are used during HP tuning phase called as **Training Data**\n",
    "  * Remaining selected 300 observation instances are used for performance evaluation and comparison called as **Test Data**\n",
    "\n",
    "In the next step we try to find, the most important features in the training data. We select those features using **RFI (Random Forest Importance)** method. We consider top 10, 20 and full set of features after encoding (59) after encoding the categorical features. In the next step to tune the hyperparameter in each classifier we use 5-fold stratified cross-validation using ROC_AUC as a performance metric. As the class feature contain more customers with 'Good' credit risks (imbalanced class issue), ROC_AUC is crucial to test the performance of each of the classifier as it is robust to this issue. We also tried to find out sensitivity of each model by plotting the ROC graph of the performance of each model.\n",
    "\n",
    "Classifiers with the best performance with hyperparameter values as identified via grid search on the training data are called **tuned** classifiers. When we find 4 tuned classifiers, we fit them on the test dataset  and we perform paired t-test to check the statistical significance of each of their performance. And in addition to this, we try to compare the confusion matrices of classifiers on the test data.  \n",
    "\n",
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Preparation <a class=\"anchor\" id=\"3\"></a>\n",
    "\n",
    "### Load Dataset\n",
    "\n",
    "We load the dataset from the local repository using `read_csv` function of a `Pandas` package. The Original dataset dont have feature labels associated with it. So we provide appropriate feature labels as per the feature information stated above."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns \n",
    "import matplotlib.pyplot as plt\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "import matplotlib as mpl\n",
    "%matplotlib inline \n",
    "\n",
    "pd.set_option('display.max_columns', None)\n",
    "\n",
    "\n",
    "attrnames = ['Existing_CA_Status', 'Duration_Months', 'Credit_History', 'Purpose', 'CR_Amount', 'Saving_Acc', 'Present_Employment',\n",
    "             'Inst_Rate', 'Gender_Status', 'Other_Debtors', 'Present_Residence_Time', 'Property', 'Age_Years', 'Inst_Plans', 'Housing',\n",
    "             'Exsisting_CR_Count', 'Job', 'Liability_Count', 'Telephone_Status', 'Foreign_Worker', 'Credit_Risk_Status']\n",
    "\n",
    "dataset = pd.read_csv(\"S3779009_Data.csv\", header = None, names = attrnames)\n",
    "\n",
    "#print(f\"Shape of the dataset is {dataset.shape} \\n\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Existing_CA_Status        object\n",
      "Duration_Months            int64\n",
      "Credit_History            object\n",
      "Purpose                   object\n",
      "CR_Amount                  int64\n",
      "Saving_Acc                object\n",
      "Present_Employment        object\n",
      "Inst_Rate                  int64\n",
      "Gender_Status             object\n",
      "Other_Debtors             object\n",
      "Present_Residence_Time     int64\n",
      "Property                  object\n",
      "Age_Years                  int64\n",
      "Inst_Plans                object\n",
      "Housing                   object\n",
      "Exsisting_CR_Count         int64\n",
      "Job                       object\n",
      "Liability_Count            int64\n",
      "Telephone_Status          object\n",
      "Foreign_Worker            object\n",
      "Credit_Risk_Status         int64\n",
      "dtype: object\n"
     ]
    }
   ],
   "source": [
    "#print(f\"Data types are below where 'object' indicates Categorical Feature: \")\n",
    "print(dataset.dtypes)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The full data has 1000 observation instances. It has 20 descriptive feature including (13 Categorical features and 7 Numerical features) and \"Credit-Risk-Status\" as a target feature.\n",
    "\n",
    "### Checking for Missing Values\n",
    "\n",
    "Lets make sure that we do not have missing values in each of the feature of the dataset. We can do that by using `isnull()` function. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Count of missing values in each feature of a dataset: \n",
      "Existing_CA_Status        0\n",
      "Duration_Months           0\n",
      "Credit_History            0\n",
      "Purpose                   0\n",
      "CR_Amount                 0\n",
      "Saving_Acc                0\n",
      "Present_Employment        0\n",
      "Inst_Rate                 0\n",
      "Gender_Status             0\n",
      "Other_Debtors             0\n",
      "Present_Residence_Time    0\n",
      "Property                  0\n",
      "Age_Years                 0\n",
      "Inst_Plans                0\n",
      "Housing                   0\n",
      "Exsisting_CR_Count        0\n",
      "Job                       0\n",
      "Liability_Count           0\n",
      "Telephone_Status          0\n",
      "Foreign_Worker            0\n",
      "Credit_Risk_Status        0\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(\"\\nCount of missing values in each feature of a dataset: \")\n",
    "print(dataset.isnull().sum())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will take look at some sample observations of the data by using `sample()` function."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<b>Table 1: Sample of a dataset</b>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Existing_CA_Status</th>\n",
       "      <th>Duration_Months</th>\n",
       "      <th>Credit_History</th>\n",
       "      <th>Purpose</th>\n",
       "      <th>CR_Amount</th>\n",
       "      <th>Saving_Acc</th>\n",
       "      <th>Present_Employment</th>\n",
       "      <th>Inst_Rate</th>\n",
       "      <th>Gender_Status</th>\n",
       "      <th>Other_Debtors</th>\n",
       "      <th>Present_Residence_Time</th>\n",
       "      <th>Property</th>\n",
       "      <th>Age_Years</th>\n",
       "      <th>Inst_Plans</th>\n",
       "      <th>Housing</th>\n",
       "      <th>Exsisting_CR_Count</th>\n",
       "      <th>Job</th>\n",
       "      <th>Liability_Count</th>\n",
       "      <th>Telephone_Status</th>\n",
       "      <th>Foreign_Worker</th>\n",
       "      <th>Credit_Risk_Status</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>842</th>\n",
       "      <td>A14</td>\n",
       "      <td>18</td>\n",
       "      <td>A32</td>\n",
       "      <td>A45</td>\n",
       "      <td>1943</td>\n",
       "      <td>A61</td>\n",
       "      <td>A72</td>\n",
       "      <td>4</td>\n",
       "      <td>A92</td>\n",
       "      <td>A101</td>\n",
       "      <td>4</td>\n",
       "      <td>A121</td>\n",
       "      <td>23</td>\n",
       "      <td>A143</td>\n",
       "      <td>A152</td>\n",
       "      <td>1</td>\n",
       "      <td>A173</td>\n",
       "      <td>1</td>\n",
       "      <td>A191</td>\n",
       "      <td>A201</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>A14</td>\n",
       "      <td>36</td>\n",
       "      <td>A32</td>\n",
       "      <td>A46</td>\n",
       "      <td>1819</td>\n",
       "      <td>A61</td>\n",
       "      <td>A73</td>\n",
       "      <td>4</td>\n",
       "      <td>A93</td>\n",
       "      <td>A101</td>\n",
       "      <td>4</td>\n",
       "      <td>A124</td>\n",
       "      <td>37</td>\n",
       "      <td>A142</td>\n",
       "      <td>A153</td>\n",
       "      <td>1</td>\n",
       "      <td>A173</td>\n",
       "      <td>1</td>\n",
       "      <td>A192</td>\n",
       "      <td>A201</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>308</th>\n",
       "      <td>A12</td>\n",
       "      <td>8</td>\n",
       "      <td>A32</td>\n",
       "      <td>A42</td>\n",
       "      <td>1237</td>\n",
       "      <td>A61</td>\n",
       "      <td>A73</td>\n",
       "      <td>3</td>\n",
       "      <td>A92</td>\n",
       "      <td>A101</td>\n",
       "      <td>4</td>\n",
       "      <td>A121</td>\n",
       "      <td>24</td>\n",
       "      <td>A143</td>\n",
       "      <td>A152</td>\n",
       "      <td>1</td>\n",
       "      <td>A173</td>\n",
       "      <td>1</td>\n",
       "      <td>A191</td>\n",
       "      <td>A201</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>881</th>\n",
       "      <td>A14</td>\n",
       "      <td>24</td>\n",
       "      <td>A32</td>\n",
       "      <td>A41</td>\n",
       "      <td>9277</td>\n",
       "      <td>A65</td>\n",
       "      <td>A73</td>\n",
       "      <td>2</td>\n",
       "      <td>A91</td>\n",
       "      <td>A101</td>\n",
       "      <td>4</td>\n",
       "      <td>A124</td>\n",
       "      <td>48</td>\n",
       "      <td>A143</td>\n",
       "      <td>A153</td>\n",
       "      <td>1</td>\n",
       "      <td>A173</td>\n",
       "      <td>1</td>\n",
       "      <td>A192</td>\n",
       "      <td>A201</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>350</th>\n",
       "      <td>A14</td>\n",
       "      <td>9</td>\n",
       "      <td>A32</td>\n",
       "      <td>A44</td>\n",
       "      <td>1236</td>\n",
       "      <td>A61</td>\n",
       "      <td>A72</td>\n",
       "      <td>1</td>\n",
       "      <td>A92</td>\n",
       "      <td>A101</td>\n",
       "      <td>4</td>\n",
       "      <td>A121</td>\n",
       "      <td>23</td>\n",
       "      <td>A143</td>\n",
       "      <td>A151</td>\n",
       "      <td>1</td>\n",
       "      <td>A173</td>\n",
       "      <td>1</td>\n",
       "      <td>A192</td>\n",
       "      <td>A201</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Existing_CA_Status  Duration_Months Credit_History Purpose  CR_Amount  \\\n",
       "842                A14               18            A32     A45       1943   \n",
       "68                 A14               36            A32     A46       1819   \n",
       "308                A12                8            A32     A42       1237   \n",
       "881                A14               24            A32     A41       9277   \n",
       "350                A14                9            A32     A44       1236   \n",
       "\n",
       "    Saving_Acc Present_Employment  Inst_Rate Gender_Status Other_Debtors  \\\n",
       "842        A61                A72          4           A92          A101   \n",
       "68         A61                A73          4           A93          A101   \n",
       "308        A61                A73          3           A92          A101   \n",
       "881        A65                A73          2           A91          A101   \n",
       "350        A61                A72          1           A92          A101   \n",
       "\n",
       "     Present_Residence_Time Property  Age_Years Inst_Plans Housing  \\\n",
       "842                       4     A121         23       A143    A152   \n",
       "68                        4     A124         37       A142    A153   \n",
       "308                       4     A121         24       A143    A152   \n",
       "881                       4     A124         48       A143    A153   \n",
       "350                       4     A121         23       A143    A151   \n",
       "\n",
       "     Exsisting_CR_Count   Job  Liability_Count Telephone_Status  \\\n",
       "842                   1  A173                1             A191   \n",
       "68                    1  A173                1             A192   \n",
       "308                   1  A173                1             A191   \n",
       "881                   1  A173                1             A192   \n",
       "350                   1  A173                1             A192   \n",
       "\n",
       "    Foreign_Worker  Credit_Risk_Status  \n",
       "842           A201                   2  \n",
       "68            A201                   2  \n",
       "308           A201                   2  \n",
       "881           A201                   1  \n",
       "350           A201                   1  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import display, HTML\n",
    "display(HTML('<b>Table 1: Sample of a dataset</b>'))\n",
    "dataset.sample(n=5, random_state=999)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Checking for ID-Like Features, Constant Features and Other Redundant Features\n",
    "\n",
    "Sometimes Data may contain some ID columns such as CustomerID or RecordID which is just for the recording purpose. This feature is not useful for the classification purpose. Hence we have to drop such features from the dataset in the preparation step only. Our Dataset doesnot have any ID-Like feature present.\n",
    "\n",
    "Next step in the Data Preparation is to check Constant as well as other Redundant features which we have to drop before going forward. We check the Unique values in each feature and drop the feature if that feature has only 1 Unique value. That means its a constant feature."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1000, 21)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset = dataset.loc[:, dataset.nunique() != 1]\n",
    "dataset.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Shape of the dataset is still the same after removing constant features that implies we dont have constant features in our dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Summary Statistics\n",
    "\n",
    "The summary statistics of both Categorical features and Numerical features are shown below. We can find some interesting facts about the data by exploring it. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<b>Table 1: Summary Statistics of Categorical Features</b>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Existing_CA_Status</th>\n",
       "      <th>Credit_History</th>\n",
       "      <th>Purpose</th>\n",
       "      <th>Saving_Acc</th>\n",
       "      <th>Present_Employment</th>\n",
       "      <th>Gender_Status</th>\n",
       "      <th>Other_Debtors</th>\n",
       "      <th>Property</th>\n",
       "      <th>Inst_Plans</th>\n",
       "      <th>Housing</th>\n",
       "      <th>Job</th>\n",
       "      <th>Telephone_Status</th>\n",
       "      <th>Foreign_Worker</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "      <td>1000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>A14</td>\n",
       "      <td>A32</td>\n",
       "      <td>A43</td>\n",
       "      <td>A61</td>\n",
       "      <td>A73</td>\n",
       "      <td>A93</td>\n",
       "      <td>A101</td>\n",
       "      <td>A123</td>\n",
       "      <td>A143</td>\n",
       "      <td>A152</td>\n",
       "      <td>A173</td>\n",
       "      <td>A191</td>\n",
       "      <td>A201</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>394</td>\n",
       "      <td>530</td>\n",
       "      <td>280</td>\n",
       "      <td>603</td>\n",
       "      <td>339</td>\n",
       "      <td>548</td>\n",
       "      <td>907</td>\n",
       "      <td>332</td>\n",
       "      <td>814</td>\n",
       "      <td>713</td>\n",
       "      <td>630</td>\n",
       "      <td>596</td>\n",
       "      <td>963</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Existing_CA_Status Credit_History Purpose Saving_Acc  \\\n",
       "count                1000           1000    1000       1000   \n",
       "unique                  4              5      10          5   \n",
       "top                   A14            A32     A43        A61   \n",
       "freq                  394            530     280        603   \n",
       "\n",
       "       Present_Employment Gender_Status Other_Debtors Property Inst_Plans  \\\n",
       "count                1000          1000          1000     1000       1000   \n",
       "unique                  5             4             3        4          3   \n",
       "top                   A73           A93          A101     A123       A143   \n",
       "freq                  339           548           907      332        814   \n",
       "\n",
       "       Housing   Job Telephone_Status Foreign_Worker  \n",
       "count     1000  1000             1000           1000  \n",
       "unique       3     4                2              2  \n",
       "top       A152  A173             A191           A201  \n",
       "freq       713   630              596            963  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "display(HTML('<b>Table 1: Summary Statistics of Categorical Features</b>'))\n",
    "dataset.describe(include = np.object)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<b>Table 1: Summary Statistics of Numerical Features</b>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Duration_Months</th>\n",
       "      <th>CR_Amount</th>\n",
       "      <th>Inst_Rate</th>\n",
       "      <th>Present_Residence_Time</th>\n",
       "      <th>Age_Years</th>\n",
       "      <th>Exsisting_CR_Count</th>\n",
       "      <th>Liability_Count</th>\n",
       "      <th>Credit_Risk_Status</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "      <td>1000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>20.903000</td>\n",
       "      <td>3271.258000</td>\n",
       "      <td>2.973000</td>\n",
       "      <td>2.845000</td>\n",
       "      <td>35.546000</td>\n",
       "      <td>1.407000</td>\n",
       "      <td>1.155000</td>\n",
       "      <td>1.300000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>12.058814</td>\n",
       "      <td>2822.736876</td>\n",
       "      <td>1.118715</td>\n",
       "      <td>1.103718</td>\n",
       "      <td>11.375469</td>\n",
       "      <td>0.577654</td>\n",
       "      <td>0.362086</td>\n",
       "      <td>0.458487</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>4.000000</td>\n",
       "      <td>250.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>12.000000</td>\n",
       "      <td>1365.500000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>27.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>18.000000</td>\n",
       "      <td>2319.500000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>33.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>24.000000</td>\n",
       "      <td>3972.250000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>42.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>72.000000</td>\n",
       "      <td>18424.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>75.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Duration_Months     CR_Amount    Inst_Rate  Present_Residence_Time  \\\n",
       "count      1000.000000   1000.000000  1000.000000             1000.000000   \n",
       "mean         20.903000   3271.258000     2.973000                2.845000   \n",
       "std          12.058814   2822.736876     1.118715                1.103718   \n",
       "min           4.000000    250.000000     1.000000                1.000000   \n",
       "25%          12.000000   1365.500000     2.000000                2.000000   \n",
       "50%          18.000000   2319.500000     3.000000                3.000000   \n",
       "75%          24.000000   3972.250000     4.000000                4.000000   \n",
       "max          72.000000  18424.000000     4.000000                4.000000   \n",
       "\n",
       "         Age_Years  Exsisting_CR_Count  Liability_Count  Credit_Risk_Status  \n",
       "count  1000.000000         1000.000000      1000.000000         1000.000000  \n",
       "mean     35.546000            1.407000         1.155000            1.300000  \n",
       "std      11.375469            0.577654         0.362086            0.458487  \n",
       "min      19.000000            1.000000         1.000000            1.000000  \n",
       "25%      27.000000            1.000000         1.000000            1.000000  \n",
       "50%      33.000000            1.000000         1.000000            1.000000  \n",
       "75%      42.000000            2.000000         1.000000            2.000000  \n",
       "max      75.000000            4.000000         2.000000            2.000000  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "display(HTML('<b>Table 1: Summary Statistics of Numerical Features</b>'))\n",
    "dataset.describe(include = np.number)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see from the Table 2 that, target feature 'Credit-Risk-Status' has two classes '1' represents 'Good' credit risk factor while '2' represents 'Bad' credit risk factor.  We can check the target feature balance by checking number of class values of the target feature.\n",
    "We will use `value_counts()` for this purpose."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Percentage of class values in the target feature are: \n",
      "1    70.0\n",
      "2    30.0\n",
      "Name: Credit_Risk_Status, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "print(\"\\nPercentage of class values in the target feature are: \")\n",
    "print(dataset['Credit_Risk_Status'].value_counts(normalize = True)*100)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see from above result that, more number of observations have 'Credit-Risk-Status' as 'Good' than number of 'Bad' 'Credit-Risk-Status' observations. This is an example of typical class imbalance issue. That gives us idea that, we have to use some different performance metric while model evaluation than 'Precision' or 'Recall'. Hence, we are using 'ROC_AUC' metric which is robust to this problem."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Exploration <a class=\"anchor\" id=\"4\"></a>\n",
    "\n",
    "\n",
    "### Univariate Data Visualisation\n",
    "\n",
    "\n",
    "#### 1. Box Plot of Age\n",
    "- `Figure 1` shows the Box Plot of Age. \n",
    "- It can be observed that the Age numerical attribute is Right Skewed.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(dataset['Age_Years']).set_title(\"Figure 1: Box Plot of Age\", fontsize = 17)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 2. Pie Chart of Foriegn Worker\n",
    "\n",
    "- `Figure 2` represents proportion of foreign workers in the data.\n",
    "- The data is depicted using a pie chart.\n",
    "- It is observed that approximately `96%` values are A201 that is `foreign workers`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "a = dataset['Foreign_Worker'].value_counts()\n",
    "x1 = a.index\n",
    "fig1 = plt.figure()\n",
    "ax = fig1.add_axes([0,0,0.5,0.5])\n",
    "ax.set_title(\"Figure 2: Proportion of Foreign Workers\",fontsize = 17,x=0.5,y=1.6)\n",
    "ax.pie(a,autopct ='%.2f%%',labels = x1,shadow = True,explode=(0.3,0),radius = 2.0)\n",
    "ax.legend(loc = \"right\", bbox_to_anchor=(2,1.3,0,0))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3. Pie Chart of Present Employment Duration\n",
    "- The following graph shows the proportion of people in the data set and thier respective Employement Duration.\n",
    "- It can be observed that the Maximum proportion of people approximately `34%` are in between `1 to 4 years`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "a = dataset['Present_Employment'].value_counts()\n",
    "x1 = a.index\n",
    "fig1 = plt.figure()\n",
    "ax = fig1.add_axes([0,0,0.5,0.5])\n",
    "ax.set_title(\"Figure 3: Proportion of Present Employees\",fontsize = 17,x=0.5,y=1.6)\n",
    "ax.pie(a,autopct ='%.2f%%',labels = x1,explode=(0.08,0.08,0.08,0.08,0.08),shadow = True,radius = 2.0)\n",
    "ax.legend(loc = \"right\", bbox_to_anchor=(2,1.3,0,0))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### BiVariate Data Exploration\n",
    "\n",
    "#### 4. Box-Plot of Housing and Present Residence Time\n",
    "\n",
    "- The following Figure 4 shows the Box plot of Credit Amount and Housing.\n",
    "- It can be observed that the people with `free housing` have the `maximum range of Credit Amount`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "A152    713\n",
       "A151    179\n",
       "A153    108\n",
       "Name: Housing, dtype: int64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset['Housing'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(dataset['Housing'],dataset['CR_Amount']).set_title(\"Figure 4: Box Plot of Housing and Credit Amount\",fontsize = 15);\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 5. Credit Amount and Credit History\n",
    "\n",
    "- The following graph shows the range of box plots of Credit Amount and Credit History.\n",
    "- It can be observed that `A33` category that is the population having an history of delay in paying off in the past have the maximum range of Credit amount, however there are no comparisn instances that can be drawn form this graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.boxplot(dataset['Credit_History'],dataset['CR_Amount']).set_title(\"Figure 5: Box Plot of Credit History and Credit Amount\",fontsize = 15);\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 6. Gender and Age\n",
    "\n",
    "- The following graph shows the Age in years on Y axis and Gender-Status on X axis. \n",
    "- It is observed that the population considered for this data set is in between `20 to 60` years of age."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.violinplot(dataset['Gender_Status'],dataset['Age_Years']).set_title(\"Figure 6 Violin Plot of Gender and Age\",fontsize = 15);\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 7. Duration Of Months and Credit Amount\n",
    "\n",
    "- The Following graphs shows a scatter plot of Duraiton of Months and Credit Amount. \n",
    "- It can be observed that the higher credit amounts have longer duration of months."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(dataset['Duration_Months'], dataset['CR_Amount'], alpha = 0.3)\n",
    "plt.xlabel(\"Duration of Months\")\n",
    "plt.ylabel(\"Credit Amount\")\n",
    "plt.title(\"Figure 7: Duration of Months and Credit Amount\",fontsize = 15 )\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Multivariate Graphs\n",
    "\n",
    "#### 8. Age and  Credit Amount colored by Job Types\n",
    "\n",
    "- Although there is no clear overall pattern, however, it can be observed that unemployed/unskilled non residents have the least credit amounts usually less than 7500 and the skilled employees and the management/self employeed have the highest credit amounts. \n",
    "\n",
    "\n",
    "- A171 stands for  unemployed/ unskilled - non-resident\n",
    "- A172 stands for  unskilled - resident\n",
    "- A173 stands for  skilled employee / official\n",
    "- A174 stands for  management/ self-employed/highly qualified employee/ officer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1da37e636c8>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.patches as mpatches\n",
    "\n",
    "plt.title(\"Figure 8: Scatter Plot of Credit Amount and Age colored by the Job types\", fontsize = 15)\n",
    "for n in range(0,1000):\n",
    "    if dataset['Job'][n] == 'A171':\n",
    "        plt.scatter(dataset['Age_Years'][n], dataset['CR_Amount'][n], color = 'red')\n",
    "        plt.xlabel('Age')\n",
    "        plt.ylabel('CR_Amount')\n",
    "    elif dataset['Job'][n] == 'A172':\n",
    "        plt.scatter(dataset['Age_Years'][n], dataset['CR_Amount'][n], color = 'blue')\n",
    "        plt.xlabel('Age')\n",
    "        plt.ylabel('CR_Amount')\n",
    "    elif dataset['Job'][n] == 'A173':\n",
    "        plt.scatter(dataset['Age_Years'][n], dataset['CR_Amount'][n], color = 'green')\n",
    "        plt.xlabel('Age')\n",
    "        plt.ylabel('CR_Amount')\n",
    "    elif dataset['Job'][n] == 'A174':\n",
    "        plt.scatter(dataset['Age_Years'][n], dataset['CR_Amount'][n], color = 'yellow')\n",
    "        plt.xlabel('Age')\n",
    "        plt.ylabel('CR_Amount')\n",
    "Red_patch = mpatches.Patch(color = 'Red', label = 'A171')\n",
    "Blue_patch = mpatches.Patch(color = 'Blue', label = 'A172')\n",
    "Green_patch = mpatches.Patch(color = 'Green', label = 'A173')\n",
    "Yellow_patch = mpatches.Patch(color = 'Yellow', label = 'A174')\n",
    "\n",
    "plt.legend(loc = \"upper right\", handles = [Red_patch,Blue_patch,Green_patch,Yellow_patch])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 9. Customers Credit Amount and their Age by Instalment Rates\n",
    "\n",
    "\n",
    "- In the next plot we are going to explore relation between three features: `Customers Credit Amount, Instalment Rate and Age in years`. We are using `3D Scatter Plot` to plot those features. We can find out the Maximum values of the three features from the graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'Age in years')"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAvcAAAHBCAYAAADttJSWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOydeXgUVdb/v9X7lk46+wYB2REQkCUKIoqIjMCICjqCjgsu4zKKwyDvOA7oq/Mqo+M6guvPERURZQQGEHHDUUFAGRGEQEBICNm37izd6eX8/uhU2Xs6SXe6k5zP8/ST1K2qW/fe2r731LnnCkQEhmEYhmEYhmG6P7JYF4BhGIZhGIZhmMjA4p5hGIZhGIZheggs7hmGYRiGYRimh8DinmEYhmEYhmF6CCzuGYZhGIZhGKaHwOKeYRiGYRiGYXoIilgXgAkIxydlGIZhGKa7IMS6AMwvsOWeYRiGYRiGYXoILO4ZhmEYhmEYpofA4p5hGIZhGIZheggs7hmGYRiGYRimh8DinmEYhmEYhmF6CCzuGYZhGIZhGKaHwOKeYRiGYRiGYXoILO4ZhmEYhmEYpofA4p5hGIZhGIZheggs7hmGYRiGYRimh8DinmEYhmEYhmF6CCzuGYZhGIZhGKaHwOKeYRiGYRiGYXoILO4ZhmEYhmEYpofA4p5hGIZhGIZheggs7hmGYRiGYRimh9CmuBcEYYUgCBTg90nrekXr8h3RL25kEARhpCAInwiC0CwIQqUgCP8QBEHfgXwUHu1xTYD1N7Wuc7Q374ULFyI/P19a3r17Nx555BG/7f785z8jMzOzvdl3itzcXCxbtqxd+wQrf0d59dVXIQhCwN/jjz8eseOIxKKdP/roIzz33HN+6b7XRrywevVq9OvXDwqFApdccknIbb///nvMnz8fmZmZUCqVyMnJwc0334zDhw9HtYxJSUl49NFHpeXJkyfj2muvlZaDtXkojh07BkEQ0L9//4iVM1a8++67ePPNNzu0b2FhIQRBwEcffRTRMnXm2fHJJ59AEAQcOXIkomXqLO25ziZPniw92xQKBfr374/f/e53qK6ubvdxjxw5ghUrVsBsNrd731hQV1cHQRDw1ltvhbX9K6+8AkEQcNNNN0W5ZB2npKQEM2fORGJiIgRBwFdffRV0W6vVir/97W8YPXo09Ho99Ho9JkyYgFdffRUtLS0RK5PVasWKFStw4MCBiOUZjwiCMLBVk13WyXyeEwShWhCE1ADrXhcEoUwQhKTOHKPTEFHIH4AVAOoA5Pv8hnlskw8gva284uEHwASgEsBOALMA3Ny6/H4H8lIAIAAWAB8GWL+9dZ2jnXnTggULaOLEiSTy9NNPk1wuJ1+Ki4vpu+++80uPJt9//z0VFRW1a59g5e8or7zyCgGgnTt30q5du7x+JSUlETuOyIMPPkgZGRkRzzcU9957Lw0YMMAvvbCwkA4ePNilZWmL06dPk0wmo/vvv5++/vpr+umnn4Juu27dOlIqlTR9+nRau3Yt7dy5k95++22aOXMmpaamRrWciYmJ9L//+7/S8qFDh+jYsWPScrA2D8XDDz9Mrc8B2rVrV8TKGgt+/etf07Rp0zq077FjxwgAbdu2LaJl6syzY8eOHQSADh8+HNEydZb2XGeTJk2iSy65hHbt2kX/+c9/6OmnnyaDwUAzZsxo93H/9a9/EQAqLi5u976xoLa2lgDQmjVrwtr+oosuIgBkNBrJarVGuXQd4/bbb6fc3FzaunUr7dq1i8xmc8DtGhoa6LzzzqOkpCRasWIF7dixg3bs2EErVqyg1NRUeuGFFyJWpva2cxBiru/a+gFQt+rVxE7mYwRwBsBrPunnAXABWBDruirC7AM4iGh3iA5C0HWRRhAEOQA5EXW023o3ACWA2URkbs2zDsAHgiCMJqL/diDPzQCuFAQhkYjqW/NMBXAxgPcA+Fn1I0Vubi5yc3Ojlb0Xzc3N0Gq1GDNmTJccLxwmTJgAjUYT62KEjdiGnWHAgAERKk3kOHbsGFwuF2655RYMHz486HanT5/GzTffjOuvvx6vvfaa17rrrrsO//73v4PuG4m28yVUWcPl3Xffxfnnn4/9+/dj7dq1cflVhenepKSkSNfV5MmT0dDQgIceeggVFRVIT0+Pcenig9LSUuzcuRPTpk3Dp59+iq1bt2Lu3LmxLpYfR44cwXnnnYeZM2eG3O5Pf/oTDhw4gD179ng9py655BLcfffdOHr0aLSLGrcIgqAloub27kdENgCd1qtEZBYEYTGAtYIgvEpEu1q16YsAPiOitzt7jLZosw3C6KGsAFAVYr1ovb7DI00G4DG4LeJmAK8CWNC6XW7rNpe0Lg/1ye8rAO96LL8F98m4CsBPAOwAzmtdlwdgHYBaAE0AtgEY1EZ9PgDwsU+avrUsD7az9ybW/XcAigHc6LHudwAKAdwHD8t9mPX2styLVmrPn2hd87Qom81m0mg09NJLL5Ev55xzDt14441E5Lay3njjjdSvXz/SaDQ0aNAg+stf/kItLS3S9qIVbu3atbRgwQJKTEyULEU5OTn0wAMPSNt+9dVXNGvWLMrMzCS9Xk+jR4+mtWvXSutDlZ+I6IcffqDLLruMDAYDJSQk0Pz586msrMyvDp6IeTY3Nwfd5p133iFBEOjzzz+X0o4fP056vZ6WL1/uVc93332XrrvuOjIYDJSenu5l3fVtZ5HCwkKaM2cOJSQkkMFgoDlz5tDx48el9Xa7nQDQM888Q/fccw+lpqbSkCFDiIho06ZNNG3aNEpNTaWEhATKz8+nHTt2eB3Pt81uueUWIvL/qkNE9N1339FFF11EWq2WTCYTLVy4kCoqKqT1Yj3ff/99WrRoERmNRsrJyaGHH36YXC5XyLYmInr22WdpwIABpFKpaODAgfTss8+GLGswC9BDDz1EGo2Gampq2jxmTk4OLV26lFasWEHZ2dmk0WikdV988QVdcMEFpNVqKTk5mW677TayWCxe+3/22Wc0cuRIUqvVdO6559Lu3bv9LPeTJk2ia665Jmg9xDYPxv79+wkAvfbaazRv3jzKzMwkh8PhtY14vjZu3EhDhw4lrVZLs2bNotraWiooKKALL7yQdDodjRs3zu+LTENDA911112Unp5OarWaxo8f73WdiO3keT8S+d8fogX7yy+/pKuuuop0Oh3179+fVq9e7VVO3/r73gehCGS5F8v25JNPUnZ2NplMJrruuuuovr5e2sZms9HixYspNzeXVCoVZWVl0dy5c8lut4d8dhw6dIjmz59Pubm5pNVq6eyzz6bnnnvO63r2tdyL9+Rzzz1H9913H5lMJkpNTaW///3vRET02muvUb9+/SgpKYkWLVrkZ/09efIkzZ8/n5KSkkir1dJll11GR48e9WuDUPdZe68zz2tUZNOmTQSA9u3bJ6W11R5iW3j+PL8etFW3QLTnXRLOs2fdunU0cOBA0mg0dOGFF9K3334btkX56aefJgB09OhRyszMpHnz5vlt09zcTLfddhsZjUZKSUmhpUuX0t/+9je/L0NVVVW0aNEiSktLI7VaTZMmTaI9e/a0WYZQ7wTx2gvW/p5YLBbSarW0dOnSNo8Z6N0kHmvVqlVS2oYNG2jMmDGk0+koKSmJJk6cSF9++WXAcsHj6055eTktXLiQTCYTabVauuiii/y8BXJycgjA4wAeBFAGt7fHSgAC3B4SP7VqwQ3wsZoDSAHwCoAKANZWLTSe/HXWvQCeg1tXHmldN6V1e3Prbz+AKym4ZhvYmtdlHmmnW8v+BwAlAGoAvA3AGCwfj323A/gvADmA37eWf7DPNgPg1p11ABoB/BtAf4/1cgB/B3CodX0RgDcApPrkUwXgEQCPAigFUBuybGEUfkVrpgqfn+DT8J7i/o8AHK37zgCwCm7x21FxXwmgAMB1AKYDyAaQ2npSvgMwD8BsALsAnASgDlGfLQD+7ZOmgvtTylseaY+iDXcaz7oDeArAdo91X7bm0WlxX1FRQffddx/J5XLJ9UR0e/C9sa+88kq6+OKLvW68goICr5fu/v37acmSJfThhx/SF198QatXr6asrCy68847pX3EB3JWVhbdfffdtGPHDvrss8+kG9lTTLz11lv0xBNP0JYtW+jTTz+lFStWkEKhoPfee6/N8h85coQSEhJo+vTptHHjRlq/fj0NGTKE8vPzKRTiS7+hoYHsdrvXz5Orr76a+vXrRxaLhZxOJ11wwQU0ZswY6eUj1jM7O5vuuOMO+uijj2jZsmUkCIKX8PFt5+bmZsrLy6OhQ4fSunXraP369TRs2DDKzc2l2tpaIvrlAZuZmUm/+c1vaNu2bbR161Yicovl5557jrZv304ff/wx3XvvvSSTyWj37t1E5Ha3uuaaayg3N1dqM/El4Svuy8rKyGg00vnnn08ffvgh/fOf/6SsrCwaPXq0Xz3z8vJoyZIl9PHHH9OSJUsIAH3wwQch2/rFF18kQRBoyZIltH37dlq6dCkJgkB/+9vfpLI+99xzUidp165dVFlZGTCvKVOm0JQpU0IeTyQnJ4cyMzNp+vTptGnTJtqwYQMREe3cuZOUSiVde+21tHXrVnrjjTcoMzPTSwAVFRWRVquladOm0ebNm2nVqlXUv39/0mg0QcV9qDYPxtKlS0mlUlFtbS1t2LCBANAnn3zitc2CBQsoPT2dzj33XNqwYQO9+eablJiYSPPnz6exY8fSyy+/TFu2bKERI0bQiBEjvPadP38+JSQk0AsvvEBbtmyhOXPmkEKhoG+++carncIV94MGDaLHHnuMPv74Y7rhhhsIgPSiLiwspClTptD48eOl+p8+fTqsc0UUXNz36dOH5syZQ1u2bKHVq1eTTqeje+65R9rmoYceouzsbHrjjTdo586d9O6779INN9xANpst5LNj+/bttHz5ctq0aRN9/vnn9NRTT1FCQgKtXLlSyjuYuM/JyaE77riDtm/fTnfddRcBoD/+8Y908cUX07///W96/vnnSalUStc4EVFlZSXl5OTQ2LFj6b333qNNmzZRfn4+5eXlSZ2AcO6z9l5ngcT9Cy+8QHK53KsD31Z71NfX0xNPPEEAaNOmTbRr1y7av39/2HULRHveJW09e7799luSyWQ0f/582rZtGz3++OPUr1+/sMX9hAkTaPz48URE9Pvf/560Wq2fy8udd95JWq2Wnn76adq6dStdddVVlJub6yXum5ubadSoUTRgwAB68803adu2bTRr1iwyGo1UXl4e9PjhvBN27dpFo0aNktysxPb35bPPPiMA0js3FOGI+yNHjpBSqaSlS5fSZ599Rlu2bKGHH36YNm7cSES/3CcrVqyQrkmbzUZERBMnTqSsrCx64403aOPGjTRp0iRKSEjwumZbxf1pAOsBXAbgL6065+8A9gGYC+D6VoH7Av2iezQAfoDbEHp9676bAdSj1dXbQ2eVAnindZuZAJJat3u9VRdeCmAJgFsouGYLJu6LAGwE8CsAt8Mtsp8Llo9Pfs0A/tpat4d91mfC3dkRDdRzAOwFcAyAkn7Rn6/D7eFxYevf7wDs88mrqrUNtgC4HMCvQ5YtjMKvQIBeHYBLfBr+jtZlJdw9sGd98vkYHRf3LgAjfLb7P7hFf5JHWgrcPu63h6jPs60nU+GRNqm1LFs90h4GYG2jbTzF/Xi4vyqkAcgRy4wIiHui4H6nvjf2e++9RzKZzMvy/cgjj1BKSoqf8BWx2+30z3/+k7RarbSN+EC++uqr/bYPJCZEXC4X2e12uvnmm2n69Oltlv/aa6+lYcOGeVl6Dh8+TDKZjD766KOAxyAK/DVA/P3nP/+RtqusrKSMjAy67bbb6KmnniK1Wk0//vijtF6s58yZM73yv/HGG6lPnz5e1jbPdn7++edJoVDQzz//LKWdPHmSFAqF9DIVH7Djxo0LWg8iIqfTSXa7nS6++GK69dZbpfRgfrm+18Yf/vAHMplMXi+yr7/+mgBIHSyxnjfddJNXXmeffTYtWLAgaNnsdjtlZGTQokWLvNJvvfVWSkpKkl4A4fo2DxgwgBYuXBhyG5GcnBzKycmRjiGSn59Pl1xyiVfa9u3bSRAE6fiLFy+m1NRUry87b7zxBsHHGu0rnNrjC+1yuSgvL49mz55NRERWq5WMRqOfFXbBggV+18rixYsJAL399ttS2saNGyXLIxHRgQMHCAC99dZb0jZOp5OGDBlCv/rVr7zaKVxx//DDD0vb2Gw2Sk5OpgcffFBKi7TPfU5ODg0aNMjra8Zdd91FOTk50vKMGTNCWijD8bkXnzsPP/wwDRo0SEoPJu49rx+Hw0FpaWmUnJzs9fVn7ty5dP7550vLy5Yto9TUVEmoEbktvAaDQTIEhHuftdfnfv78+WS326m5uZm++uor6tOnD919993tbo9gPvfh1C0cQr1L2mqTuXPn0siRI72s+StWrAhL3B8/fpwA0FNPPUVERN98843ffuXl5aRSqaSvNETudhoyZIjX9bV69WpSq9Ve4tVms1FeXh4tW7YsaBnCeScQBe6s+bJmzRoCQIWFhSG3IwpP3K9du5bS09OD5hHM537z5s0EgL766ispzWKxkMlk8urAtYr7AgAy+kXTfN+qifp6pP0dQInH8u1wW7vP8khTwW2k/T/y1ll7yVsz5bem6yiETvPZJ5i4Pwq3u7eY9gKA02Hmubw1z2MAND7rnobbNz/BIy0Tbk+T3wbJTw5gSGueYz3SqwD87KldQ/3CDYVZD7d49fx9G2TbPLgF7iafdN/l9lBERAd90i6B+5NIQ2vUGkVrOb8HMC5EXi8DyALwjCAIGYIgjADwDwDO1h8AgIiWE1HYztxEtBfAKQBXw93z+ilAmaPOrFmzoNPp8P7770tp69atw1VXXQWFwj3EwuVy4amnnsKwYcOg1WqhVCrx29/+Fs3NzTh9+rRXfpdffnmbx6ypqcHdd9+NvLw8KJVKKJVKvP7662H5BH7yySe48sorIQgCHA4HHA4HBg4ciNzcXOzbt6/N/b/++mvs3bvX6zd69GhpfWpqKl566SW8/PLL+J//+R88/PDDGDFihF8+vr6ZV155JYqLi1FaWhrwuHv27MH48ePRr18/KS0vLw/5+fl+0Q8CtWFxcTGuv/565OTkQKFQQKlU4rPPPuuQH+WePXtw2WWXISEhQUo7//zzkZub61eWSy+91Gt5+PDhfufck6KiIpSXl2PevHle6ddccw3q6upw6NChdpdXEISwt73kkkugUqmk5YaGBnz77beYP3++dL04HA5MmTIFMpkM3333HQB3m8yYMcNrPMaVV17Z7rKG4ptvvsGpU6ekaDtqtRpz587FBx984BfJYsCAAV7XysCBAwEAF198sV9aSUkJAGDv3r0QBAFXX321tI1MJsO8efNCRtgIhef5V6lUGDBgQMjzHwkuvvhiyOVyaXn48OEoLS2F0+l+3I4ePRqvvfYannzySfz4449h59vc3IyHHnoIAwYMgFqthlKpxPLly1FYWAiXyxVy32nTpkn/y+Vy9OvXD+PHj4fBYJDSBw4cKJ0LwP2smjFjBgwGg3TdJSYmYuzYsX7PqvbeZ23x3nvvQalUQqvVYvLkycjNzcXTTz/ttU1n2qM9dfOkPe+Sttpkz549mDNnjtfzIdx7du3atRAEAfPnzwcAnHfeeejXrx/Wrl0rbXPgwAG0tLRgzpw5UpogCJg1a5ZfW4wfPx59+/aV2kImk2HKlCkh26I974Rwac+zMhSjRo1CVVUVbrrpJuzYsQNNTU1h7bdnzx5kZWVh0qRJUprBYMDll18eqE6fE5HnhVYI4DgRFfmkZbb6pwNuHbcXQJGHjnPB7fngq+O2+Cwfg9vCvlYQhDmCICSGVanAfEZETo/lnwBkeZQzFH9r/fs8EVl91l0CYCuAZo/6VQH4ER71EwThCkEQvhUEwQy314sY3muwT37biSis6IvhinsHEe3z+VmCbCvGC6z0Sfddbg/lAdJS4fbjt/v8pgDoEywjIjoEt6X9Brg/l/wAt9X8xyDHaQ/rAFzb+nu3k3l1CK1Wi9mzZ2PdunUAgEOHDuHQoUNe4f6eeuopPPDAA5g3bx42bdqEPXv2SGHZrFbvazMjI6PNY15//fX44IMPsHTpUuzYsQN79+7Fb3/7W7+8AlFdXY3HHntM6hSIv6KiIhQXF7e5/9ixYzFu3Divn+cLGnALxNTUVBARFi1aFDAf30Fp4nIwcV9aWhqwbTIyMlBTU+OX5onT6cSsWbOwZ88ePProo/j888+xd+9eTJ8+Paw260xZkpK8o3OpVKqQxxTr75u/uOybf1vk5OSgqKio7Q19jiNSXV0NIsJtt93mdb1otVo4nU7pmikrK/M7pwkJCREdfL127Vqo1WpMmTIFdXV1qKurw+WXX466ujq/cJCB2t03XUwTz0dpaSkSExOhVqu99s3IyIDZbJbEcXto7/mPBIGO6XK5YLfbAQDLly/HHXfcgeeffx6jRo1Cnz598MILL7SZ75IlS/DMM8/gjjvuwNatW7F3714sW7YMRNRmmMBAZWqrbaqqqvD222/7Pau+/PJLv2dVpNv50ksvxd69e/Gf//wHS5Yswa5du7B8+XKvbTrTHu2pmyfteZe01Sbl5eVBn8NtsXbtWkyYMAEGg0G6F+fMmYOPP/5YChlaVlYGAEhLS/Pa13e5qqoKX331lV9brFmzJmRbtOc53BY5OTkA0K5nZSiGDx+ODz/8EMeOHcPMmTORmpqKhQsXoqqqKuR+7axTnc9yS5A0GdweHoBbx02Gv467Hv46zkufEVE13G7fGgDvA6gSBGGzIAj9QlYqMG2VMxQtPn89SQVwC/zrNwGt9RME4UK4ffKPwu16fh6Ai1r3931hha1Rw42W0x7KWv+m+aT7Lot3tconPRnuzySeUIDj1MA9eOKvAdaFDOJLRK8IgrAGwCC4G6sWQDXcFvzO8C6A/4F7EMl1QbZpT707xDXXXIO5c+eipKQE69atQ2ZmJi688EJp/fr163Httdd6xY4OFt+2LctBY2Mjtm3bhpdfftlLOIcrPEwmE37zm9/gxhtv9Fvn+9DtKEuXLoVMJkNaWhoWL14cMI53RUVFwOWsrKyAeWZlZeH48eN+6eXl5UhOTvZK823DgoICHDhwADt27PCKB9/c3O7B/1JZfMsvlsXT4tLRvAH/9ikvdz9jfOvaFlOnTsXKlStRV1fn97IPhG/bmUwmAMCjjz6KGTNm+G0vvhQzMzP9ymyxWCImZJ1OJ9avXw+bzYY+ffxtCWvXrvWyEHaErKws1NfXw2azeQn88vJyGI1GyRqu0Wj8xFt7xUQs0Wq1ePTRR/Hoo4/i6NGjePHFF3HPPfdg6NChIedLWL9+Pe6991788Y9/lNI2btwYtXImJydjzJgx+NOf/uS3zmg0Ru24gPu6HzfObeibPHkyKioq8NRTT+Guu+5CdnY2gM61R0fr1p53SVtkZGQEfQ6H4uDBg9IXRPH54Mn777+P22+/XZqnpLKy0qtOlZXedsfk5GTk5+fj+eef98srlHGgPe+EtpgwYQJ0Oh22b9+OqVOnhtw23Pt/9uzZmD17Nurr67F582YsXrwY9913X8g5BEK9W9pbpyDUwO2Pfk+Adb4Paz8dSERfA5ghCIIObr/7v8Ptyj05EoWLADUAvoB7TKYvYofiKgAniOh6cYUgCGcHyS+QFg5INGaoPQW3lf7XPum+bzpRyA4TE1p7XAPDPM6nAM4G8GOArwpt+jYQkZWIfiSiCrit+C64e38dhoh+BPAagBeJqDDIZh2qt0qlgtPphMPR9heZmTNnwmg0Yv369Vi3bh3mzZsHmeyXU93c3OxnDXz77Y5FbrJarSAir/zq6+v9QhoGK/+0adNw8OBBnHvuuX4W+Ly8vA6VyZNPP/0Uq1atwurVq/HKK69gzZo1AV94//rXv7yWN2zYgNzc3KDifuLEidizZ4+XZaWoqAjffvstJk8O/VwRRbxnm504cQK7d3tH6ArX2jdx4kRs27YNjY2NUtru3btx+vTpNsvSFnl5ecjIyMD69eu90t977z2YTCacfXawZ1Bgbr31VsjlcixdujTg+i1bfL+8emM0GjF+/HgcPXrU73oZN26cdL7Gjx+P7du3e7Xfhg0b2ixfuG3+6aefoqKiAk8++SQ+//xzr9+CBQuwadMmr/PRESZMmAAiwgcffCCluVwufPDBB17nNTc312/yrx07dnTomF1hyQ/F4MGD8fe//x0KhQI//fSTVKZAzw7f55jT6ZS+WEaDadOm4dChQxg5cqTfdTd4sO/X89B0tp0feeQRuFwuPPPMM1JaOO3h+3VIpKN1i+S7ZPz48di0aZPoXwwgvHv2nXfegUKhwLZt2/zuxbPPPltyzRk1ahRUKpXX85+I/N5V06ZNw9GjR9GvXz+/tgjk0inSmXeCL3q9HosWLcILL7yAgoICv/U1NTXS+yI3Nxe1tbWSwQUIff8nJiZi4cKFmDNnjtc9BvhfFxMnTsSZM2fwzTffSGmNjY3YunVrp98trXwKt+vJyQA6Lmy3ZiJqIqKNcEeZ6Xx848jxKYCRAH4IUD9RI2rhb/Vf0NkDR9xyT0R2QRCeBPBXQRCq4Y5gMxe/iFlX63YnBUHYD+AxQRBsrWX5E9w9nXB4Em7r+GeCILwA96CFTLhHG39BRO8F2kkQBBOApXC74jjh/vxxP4CbiajOY7uHATzQHr/71nrd2sb6DtV76NChAICnn34aU6dORWJiYtCHrkqlwhVXXIGVK1eitLQUr7/+utf66dOnY9WqVRg3bhzOOussvPnmmzh58mS4VfQiJSUFY8aMwYoVK6DXuyf5/b//+z+YTCYvv75g5X/kkUcwYcIEzJ49GzfddBNSUlJQUlKCjz/+GIsWLcIFF1wQ8vh79uzx8skG3Bag/v37w2w24+abb8aCBQskn/pbbrkFt99+OyZPnoyUlBRpnx9++AF33nknrrjiCnzxxRf45z//iX/84x9Bv1zccsstWLlyJWbOnIkVK1ZAEAQsX74cGRkZuPXWkJcAzj77bGRnZ2Px4sV45JFHUF9fj7/85S9+8xUMHToUZ9gtXqMAACAASURBVM6cwZo1azBs2DCkpaUF7PAsWbIEL730EmbMmIGlS5fCbDZj2bJlGD16NK644oqQZWkLuVyO5cuX46677oLJZMK0adPw+eef45VXXsHKlSv92r4tcnNz8frrr2PBggUoKirCTTfdhOzsbJSUlOCdd97B7t2727TWrVy5UvLfveqqq2AwGHDq1Cls2bIFTzzxBAYMGIDFixdj9erVmDVrFhYvXozi4mKsXLmyTbecUG3ucrmkmULXrl2L5ORk3HPPPX5toNPp8Pbbb2PTpk34zW9+06728WTEiBGYP38+fve736Gurg79+/fHyy+/jGPHjnnNETB37lzcf//9ePzxxzF27Fi8//77AQVBOAwdOhRbt27Fxo0bkZOTg5ycHGRlZeHVV1/FrbfeiuLi4ojPqzFnzhxMnDgRY8aMgUajwXvvuR/d4r0f7Nkxffp0PPfcc+jfvz+SkpLwwgsvhGX86ChLlizBO++8g4svvhh33303srOzUVZWhp07d2Lq1KmSv3c4hHtvByMvLw8LFy7ESy+9hD//+c8wGo1htYfYlqtWrcL8+fOh1+sxYsSIDtctku+SBx54AOeff770JffAgQN444032tzv3XffxWWXXYbLLvOfdPSGG27AsmXLUFJSgpycHNxyyy3485//DLlcjiFDhuC1115DY2Ojl/HrpptuwksvvYSpU6fiD3/4A8466yxUVVVh9+7d6NOnD37/+98HLEdn3gmB+Otf/4q9e/fivPPOw+LFizFp0iS4XC7s3r0bzz//PP7yl78gPz8fM2fOhEajwY033ojFixfj+PHjeOmll7zyevHFF7Fv3z7MmDEDWVlZKCgowIYNG3DLLbcAcD+z+vTpg3Xr1mHYsGFQq9U455xzcPnll2PixImYN2+e9F5fuXIl7HY7/vCHP7S7TgH4f3APqv1CEISnAJyA25UlH0AxEQWdxlkQhF8DWAh3lJtiALkAbgXwWSQKFiEeBzAfwCeCIKyCO9pNFty6cxsRfQhgB4BFgiA83vr/VLgjQHaOMEYCr0DH4tz/Fe6BAxYAa+CePIrgPWp4CNwzxTYCOAx3OMuAce6DHDsXwD/hdq2xwT3Ceg08Zs8NsE9CawPWwB3CaA+AOQG2a1cozBDb3OebTxj19ouI4nK56P7776esrCwSBCFgnHtPtm3bRgC8Ir6ImM1muuGGGygpKYlMJhPdeuut9OGHH3pFlQg126RvdI6CggKaOnUq6XQ66tu3Lz355JN+5QpWfiJ3fOa5c+eSyWQijUZDAwcOpNtvvz1kGL5Q0XJuv/12IiK6+eabKTs72yumen19PeXl5dH8+fO96rl27VqaP38+6fV6SktL84oqEqydCwsLafbs2WQwGMhgMNDs2bO9ohsEijUssnv3bjr33HOl2NBvvvmm3zlvamqiG264gVJTUwkIHed+3759NHXqVNJoNJSUlEQLFizwCtsW7HwGyisQzzzzDJ111lmkVCppwIAB9Mwzz3itb+9MoPv27aOrr76a0tLSSKFQUHZ2Ni1cuNArNFyoqEzffPMNXXrppZSQkEA6nY6GDRtG999/v1f89E8++YRGjBhBKpWKxowZQ998803IOPdEgdvc5XJRS0sLmc1mqqmpocrKSkpMTKTbbruNHA5HwHkCBg8eTLNmzSKiwG0caJ6GQOeooaGB7rzzTkpLSyOVSkXjxo3zi3Nvs9no3nvvpfT0dDKZTLR48WJ68cUXA0bL8T0/vvUvLy+nX//615SUlOQVWejZZ58ltVrt1b6+hIpzH6rujz/+OI0dO5aMRiMZDAaaOHEibd68Wdo+2LOjtLSU5syZQwaDgTIyMuiBBx6gVatWhax3sHsyUASTQPd8cXEx3XDDDZSenk4qlYry8vJo4cKFUnjOcO+zYPd2IIJFVzl69CjJ5XIpEks47UFE9MQTT1CfPn1ILpd7Rexpq26B6My7JNB9sXbtWjrrrLNIrVbTBRdcQLt37w4ZLWfXrl0EgNatWxdwfXFxMclkMnryySeJyN3ut912GyUkJJDJZKJ7772XHnzwQUpJSfHar7a2lu6++27KyckhpVJJOTk5dNVVV7U5A3Vb7wSi8KLliDQ3N9MTTzxBo0aNIq1WS1qtlsaPH0/PPPOMV4jSzZs307Bhw0ij0dCUKVPo4MGDXtf5V199RTNnzqTMzExSqVTUr18/WrZsmVcksq1bt9KIESNIrVZ7RVQqLy+nBQsWUFJSEmk0Gpo6dWrQOPfkrXP8tBuARa3vaY1HWhKA5+H2amhp/fs+gHwKobPgNhh/ALewt7XutwoeERR9fwgR576tcobIM6QOhDvIjBjS3QZ3xJs34DEnE9wRd84AaIB7AO6I1jxv9NimCsCKtsoj/sRY9VFHEIQ3AFxARPE3vWb80TUnhUFhYSEGDRqEbdu2BbT8ML0XIpLcQcT/xbEkvs9NmUwGhUIBhUIBmUwGuVwesUgX8cCCBQug0+nwyiuvxLooDBNRpk6dCrlcjk8//TTWRenu9JwHXg8gGgNqIQjCOQCuhHughAvugPs3wO3+wjAME9eI0VxEVxyZTAan0ym55XgiWkpaWlq8BraJIl8ul0OhUHRrwb97925s3rw51sVgmE7xySef4LvvvsPYsWNhs9nw7rvvYufOnX5jrhimuxMVcQ+3u8mFcE/Hq4d7kO0SInom5F4MwzAxxOVyweFweAn5tgR5KMFfVVWFyspKKYa9KPhFsd9dBH+gKCAM090wGAzYsGEDHnvsMdhsNgwZMgRr1qzp9Ngkhok3uswth2kXfFIYpgshImnCGiCwYPe05IdLfX09ysrKMGTIEMmVx3dCIV/BL5PJvAb4MQzDdAPi30rRi4iW5Z5hGCbu8fWrD2Wp76yFXdzfc7ZWUfDb7Xa0tLRI2wiCIIl90Y8/nK8IDMMwDMPinmGYXgcRSS44+/fvx+DBg6VQrpFEEAS/wbe+64HAgt/hcPgJ/kAuPSz4GYZhGE9Y3DMM06vwHSwbTXHclrgPto/4V3TPEfNwOp2w2+0s+BmGYZigsLhnGKZXEMyvviMCvKsJJPgBb7cicTmYDz8LfoZhmN4Bi3uGYXo0ogC22+0A/AfLtkfci3754RLtjkOwSD2i4D9+/Dhyc3OhVqshl8uhVCpZ8DMMw/RwWNwzDNMjEf3q7XZ7yMGyMpksqgK8q78KeNbTZrNJ1n6XywWr1eq1rWjhFwftsuBnGIbp/rC4ZximxxFoEqpgRNO6Hi9COZiF3+VywWazwWq1eg3u9XXriZd6MAzDMG3D4p5hmB5DOPHqfREEwS/2fKSIZ3/+UILf6XT6zbbraeHvLpNvMQzD9EZY3DMM0+1py68+FPEswLuaULPttrS0BBT8oqWfBT/DMEx8wOKeYZhuS7h+9aGItltOLDsOkRDbbQl+T1jwMwzDxB4W9wzDdEva41cfinAFeEdEaqzFPRCdAb3BBD8AL8F/+vRp9OnTR3Lp8YzUwzAMw0QHFvcMw3QrOuJXH4p4EOA9gUCz7VZUVKBPnz6w2+1+s+2KYl/04+fJtxiGYSIDi3uGYboFkRb1Ij3ZLSceCDTxFgA4HA4/wc+z7TIMw3QeFvcMw8Q1npMyddSvPhTRjHPP4t7fnSnQbLtiG4mDolnwMwzDdBwW9wzDxC2iX/3333+PMWPGRMVXO9oCvLeL+3AIJPgB746duCyG4vT14WfBzzAM44bFPcMwcQcRwW63w+l0AoDfzKqRJNpx7pmOE2zgrqfgF5HL5VAqlSz4GYbp9bC4ZxgmbgjmVy9a16Mh1mLpcy/Wh92CwifU5Fu+ncCWlhZotVpoNBoW/AzD9BpY3DMME3Pa8quXyWRwuVzd0i0nEE7IUNPkQLnFhgS1AmkGFfRKduHpKMEEf1FREVJTU5GYmOgVzcfXrYcFP8MwPQkW9wzDxJRw4tV314g2gfJ2QoZvf67Dl8cqYHO4IADISzHgijHZSFRFpRi9ElHwizH2gV8s/E6nM+Bsu2JYTp58i2GY7gyLe4ZhYoKnX72n+00gRMt9NOjqSawqGx34/Gg5HE53OgE4Wd2AXSdqcNmwVICiU8/eiG/btzXbbiDBz7PtMgzT3WBxzzBMl9KRePXRtq5Hq+Pgi0wmQ1GNRRL2npyqbkCjPRl6fipHjHDGabQl+D1hwc8wTHeAXyMMw3QJnYlXHw+W+0jkTUTQqwI/dtUKORQyAW5bPhMJOjoIO5jgBxBQ8AcKzckwDBMrWNwzDBN1wvGrD0W0w1V21UBWIkIfkxapBjWqGmxSukwAxvVLhlYhwOWKXFl6u1U5khGWPAfkeuYPAHa73W+2XVHsi378PPkWwzBdBYt7hmGihsvlgsPhCMuvPhTRnEU22jPU+pKkkeHqc/vgy2OVKK1rhk4tx7i8FAxN10elA9ObI/BEK3yqSCjB73A40NLSgsOHD2P48OE82y7DMF0Gi3uGYSJOR/zqQ9Fdo+UEwuVyIdMgx5Wjs9DY4oJSLkCv7Dq//95ELDo2ntZ7QRBgtVqlL1VOpxN2u91rGxb8DMNEGhb3DMNEjM741Yci2j73beXd2NiII0eOoKmpCRqNBgkJCTAajUhISIBarW73MYkIchCMKgCgiLriMN7EUij73gO+Lmm+s+0SUVAffhb8DMOEC4t7hmE6jRg/3OFwdNivPhSxstw7HA4cP34cNTU1GDx4MHQ6Hex2OywWC+rr63H69Gm0tLRIgl8U/SoVB6yPB6LtltPZ4wcbuOsp+EXkcjmUSiULfoZh2oTFPcMwncLlcqG6uhpyuRw6nS4qkUK6OloOEeHMmTM4efIk+vbti8GDB0tx+TUaDTQaDdLS0qRtrVarJPiLi4ul7RISEqTBliz4u55Yi/uOzKocTPC7XC5YrVavdNHCLw7aZcHPMAzA4p5hmA7i6VdfUVEBvV4PvV4flWN1ZbScuro6FBQUwGg0YsKECVAqlQCC+28LggCtVgutVov09HRpW1HwO51OHDp0CHa7HVqtVrLwJyQkRF3w93ahF2txH6njhxL8NpsNVqvVa3Cvr1tPb78OGKa3weKeYZh2IboN2O12AL8MCozm4MVoR7QhIthsNhw9ehQ2mw1nn302DAZDp/IUBf+pU6cwZswYEBGam5thsVhQW1uLoqIiL8Ev+vCLnYlIwdFyur+4D0Qowe90OqV4/C6XC2azGenp6ZKFnyffYpieDYt7hmHCQhQOdrvdb6BgNN1mxPyjJVKJCPX19fjuu+8wcOBApKWlRUX4CIIAnU4HnU6HjIwM6dii4K+ursbJkyfhcDig1WolsR8Nwd9biLW474hbTmcIJPidTidOnTqFpKQkrwm4eLZdhum5sLhnGKZN2pqEKtriPlpuORUVFSgoKIBCocDEiRO94pV3BaEEv9ls9hL8Op3Oy8KvUPDjuy1i/dUi1p0LsQyiePdMI6KAs+2y4GeY7g+/HRiGCUq48eplMpnkphMNIh0tp6GhAQUFBVAqlRg6dCjKy8vbFPZdJXI8BX9mZiYA93loamqCxWJBZWUlTpw4AafTCZ1O52XhZ8HvT6wt97EWx4G+HgRz6QHgJfg9Owa+oTkZholf+E3AMIwfgfzqQ4mUrnDLiUT+DocDhYWFqKurw5AhQ2AymWCxWOJ+AilBEKQBy76C32w2ewl+vV4vif14r1e0ibXlXIxbH0vCdQ0KNduuGPHJM16/KPZFP36efIth4gcW9wzDSITyqw9FvLvlEBFKSkpw6tQp5OXlYciQIV5CJdbuGx3BU/BnZWUBcAs5Twt/TU0NLBaLlzuPwWDoNRb+eBD3sRa8nfl6EErwOxwOP8HPs+0yTHzQO57wDMO0SVt+9aHoCsu90+ns0L51dXU4cuQIkpKSvEJbinRXcR8ImUwGg8EAg8GArKwsuFwuZGVlQS6Xw2w2o7y8HIWFhSAiycJvNBphMBi6fLxBVxBrcR3r44tliPSEcuJfMV/x/hG/9rHgZ5jYwuKeYXo54frVhyKa0WzE/H1n7GwLq9WKo0ePoqWlBSNGjAga2rIniftACIIgCX4Rl8uFxsZGWCwWlJWVwWKxSILf08LfEwV/V9LV0XJiVYZAgh/wnm33yJEjGDRokJc7D8+2yzDRgcU9w/RSIiHqRaI5yVR783e5XDh58iTKysrCCm3Z08V9IGQymeSXLyIKfrPZjNLSUjQ0NICIYDAYJAu/Xq/vdoK/t1vuY9nB8HymNDU1SV/NRMHviVwuh1KpZMHPMBGAxT3D9DI8rWnt8asPRVf43LclwIkIlZWVKCwsRGZmJvLz88MeSNjbxH0gggn+hoYGWCwWlJSUoLGxURL8nhb+WFun45V4jZYTCzw7OsEm37JarV7pooVfHLTLgp9hwoPFPcP0IjrjVx+KWEfL8QxtOXbsWGg0mrDzZnEfHJlMBqPRCKPRiJycHABuq6to4S8pKUFDQ4Pk+iN2Dljwu+lO0XJiSajZdm02G6xWq9fgXl+3Hhb8DOMNi3uG6QUQEex2uzQoNdKD2mIVLcdut+P48eOoq6vD0KFDkZSU1KG8e6q4j4bokcvlkuAXcTqdXhZ+UfBbrVacOXMGCQkJ0Ov1cS8yI01vd8vxpL3tEErwO53OgLPtihZ+nnyL6e2wuGeYHozoV3/8+HFkZ2dDpVJF5aXXFZZ7TwEeKrRlZ/PuaXRF3eRyORITE5GYmCilOZ1O7N27F06nE8XFxWhsbIQgCJJ1vzcI/ngQ1vFQhkgRTPCLs+0GEvw82y7TG2FxzzA9EF+/+rq6OmRkZETt5daVbjm1tbUoKChAUlISJk6c2OmY7V1huff1N472AOR4QHSZ6NOnj5TmdDphsVhgsVgkwS+G7xQH7ep0uh4jRuPFch/rMogCPBq0Jfh9cTgcSEhIYMHP9GhY3DNMDyOQX32sfeI7iyAIcDgcOHDgAOx2e8jQlh3JuyuEtiAIsLpkqGqwo6qhBUlaJdITlNDJoyd84g25XI6kpCQv9ymHwyG59Jw6dQpNTU1eg3tFC393FGHxIO6JKOYRjrp67EEwwd/U1IQTJ05g+PDhUlogH/6e0rlkei8s7hmmh+DpVy++3DwHoXV0EqhwiKa4d7lcOHPmDCorKzFy5Eikp6dHNP9oW+7F/BsdAr44WoV1e0+htskOpVyGi4am4/qJfZGoitrhY0o4wlahUAQU/KKF/9SpU2hsbIRcLpfEvmjhj7Vwbot4sJrHg1uO0+mMeRnETrwo4oFfXNbsdrvfbLue8fjFQbuxPpcMEy4s7hmmmxNOvPruaLn3DG1pMpmQmpoacWEPRC8OukwuR02zC5mDRqHOLkOL04G3dv+MBpu7k2V3urDjpzLkpegxZ2QaKIqdr+6GQqGAyWSCyWSS0jwF/88//4ympiYoFAovC3+8Cf54sJrHg7h3uVwxbwfA3cnwLIen8UNEFPwOh8NP8PNsu0x3gcU9w3RT2hOvPh7i0LeHhoYGHDlyBCqVCmPHjoXD4UBhYWHE8o82dsix/5QZ//q+BEdKajA424Qpg9Nx2YgsvP/daWk7IuC/xbW4eHAK9FF4Gvck4RFI8NvtdknwV1VV+Ql+l8sVU9eYeHDLiRdxH+syAP7iPhCBZtsVn21OpxN2u50FPxP3sLhnmG5Ie+PVR9stJ1IvM7vdjsLCQtTX13uFtmxqauo2fukymQzFNTas+vw4qhttsNrs+LmqEeWWYlw+MgtnpRlworIBAKBVymHUKGFzEFqcBINKBpWsayLc9ASUSiWSk5ORnJwspYmC32w2w2azYe/evVAoFNKkWwkJCdBqtV0iwOJB1MZDGcIR1V1Vjo60RSDBD3gbWMRlMRSnrw8/C36mK2FxzzDdCJfLBYfDEdCvPhTRttx3FiLC6dOnUVRUhH79+mHo0KFe9Yr38osIMhksduCnUgsUChmykrSoqHVCp5Kjf6oeBWVmDM8y4kRlA3QqBUx6Ffom67H5QAnO1DZjUEYCLh+ZBbuTYHMQFHIBiRoZZNTxuve2joKn4K+srMT48ePR0tIiWfgrKirQ3NwMpVLp5cOv0WgiLsDYch8/ZQAi38kINnDXU/CLbNu2DSNHjsSIESMidnyGCQaLe4bpBoTjVx+KeBbHYmhLk8kUNLRld5hoShAEFFZZ8X1RHUrqrThUUo9ckw7XTxoImQwgEjAiOxEKhYAknQqWZjs0KjkMahkOnm6EiwAnEXb9XIuDJXWobbRBq5TjnL4mXDgwFRp5fJ6/UMTLOVOpVEhJSUFKSoqUJgp+s9mM8vJyWK1WKJVKLwt/ZwV/vIj7WJchHgbUiuWI9heEYIJ/586dyMjIiOqxGUaExT3DxDHt8asPRbTdcjqC1WpFQUEBHA4HRo4cCb1eH3Tb7jDRlLkF2H6oFE0tLpzTJxk61RlcMSYbXx6rhF4lx/DsRKzeeRyCQBiUkYDzBqSgvsmGL46Uw0WAWiFDtkmPV3YeR59kHVRyoNnuxO7jVUjWqTAxzxi3HbTuSDDBbzabYbFYUFZWBqvVCpVK5WXhV6vVYd+D8WCxjpcyxINbjsvlgkrV9aGpBEFAY2MjEhISuvzYTO+ExT3DxCHiNOsOhyNsv/pQyGSyuBH3TqcTJ0+eRHl5OQYNGoS0tLQ294mXSZ9CiboGmwOVFhsAwGxtwf2XDsGJykb8UFyH+6YPxstf/gyZAAggEICS2iZcPDQdOpUCDTYH0o0anKhsQFOL/3k6dKYeo3ISoI698bNdxNpi3F5UKhVSU1ORmpoqpdlsNsmlp7S0FDabDSqVysvCH0zwx4PlvqtjzAciHjoYQGx9/xsaGmA0GmNybKb3weKeYeKM9g6WDQeZTAa73R6B0nUcIkJFRQUKCwuRnZ2N/Pz8sOsWz25FInKZALlMgNNFOHS6FtNHZMNJhCvG5OJUdZMk7AGgrskOvUqBn0rNyDXpcKTMDCJAgACjVgmlXADg/aWiIxIx1sIylkTqS49arYZarfYT/KKF/8yZM7DZbFCr1X4W/ngQ9/EgrONpQG2symGxWFjcM10Gi3uGiRM661cfCrlc3iXiOJiYEUNbqtVqjBs3Dmq1ul35dgef+0SNHGelJuBYhRkE4HRNI2wtTuiUAhpbXJKwBwCVQgalDNCrlEgzalBQbkZtYwumj0hBcXUjHC5v6/3I3ERoFAJcrtY8ZHKYrU44iaCSC0hQBe/8xKrdYn2+oims1Wo10tLSpK9ORORl4S8pKUFLS4s0qZzD4ZAs/F1NPIj7eCgDwJZ7pvfA4p5hYozoVy9a1qMRJ7kr3HJEAe5ZdjG0pdlsxpAhQ7xmIW1v3vFAKMGokQOXj8rEzqNynKhsQKPNjkuGp+PrgjIMzzLi68IqAIBSLoNJp4SLXDg3LxGD0/U4JzcRMgFIUCugHJuDLwrKUWmxwaBWYHTfZIzISpDEewvJsPd4Lb47VQOz1Y60BDWmDE7HsHQd0ImoOj2NruxcCIIAjUYDjUbjJfgPHToEvV6P+vp6nD59Gi0tLdBoNF4W/mj7gMeDsI6HMgCxFfdWqxUajSYmx2Z6HyzuGSZGiH71ZWVlUCgUSExMjN5sqV3g1iIeQxz8Giq0ZU+EiJCoAmaNyIClJQ2CIEApl0EGgs0JXD4yG3tOViPFoIZeJcOonCQMSNVBJbiQohHbxokRmTr0NeXB5gTkMiBJLZNEu0wmQ1GlFd+eqILF5v7CU1Zvxab/nkbKpLOQrvMWUHq9Hk6FDrU2QKMQoFd23diFWFvugdh2CsVOenJyMgwGAwB3m1itVlgsFtTX16O4uBh2u10S/KIffyQFfzwIa6fTCaVSGdMyiOWIpXtQrM8D03tgcc8wMcDTr95isUClUnXYqh0OXeGWI4r72tpaHDlyBMnJyUFDW/ZkZHDBoBTw/WkLjpabMSjNAIVCjktHZOJXo7LQ0GyHSa9Eik4BpeA+J6IIFTt8BiVgELVQq7B3koCKBieOVzXinL7JICIcOF0Li9UBq92FgjILMgeapPPsgAzHGxT4/nAxrC4BqQa3hX9QmgboIoEfS3EdD/7uvoNZBUGAVquFVqtFenq6tI0o+Gtra1FUVAS73Q6tVitZ+Dsj+OOhHeKhgwHETtzHQ0eX6V30rrcuw8SYQH71CoUi6i4zXeGWQ0Q4ePAgiAijRo0KGdqyO0JEqK6uhkajgV6vDymYKhod2PNzNYbnJOEfnx9FTZMLgkzA8Cwj5p2biyHpWsjgAgQZ6qwuVDZYoVLIkKpXwaD0FwMOkmHXyTp8ebQSh8sssDTbMSo3CZeenYGvjlWgxeGC0/WLiBMEAScqm7Hp+2Io1Wqo1Sqcrm3Cv74vxk2Tz0Katmd/RQHiQ9SGU4Zggr+5uVkS/KdOnYLD4ZAEv2jhD9caHut26E1x7oMRDXdLhgkGi3uG6QJC+dV3RSSbaLrliKEtzWYzhg4ditzc3KgcJ9qEEmLi1wi9Xg+73Y7m5mYoFAoYjUZJaIkTHsnlchSU1aFPsh5fHq1EWb0VKqXb6lpc24zth8qQl6JDqk6OwxVN+OjHUpit7vOfY9Li16Nzka7zjutf2WjHzgJ3PHyTTgVzsx0/nK5DjkmLfil6nKhswJDMBKkD5yQB3xfVwkkET/nXbHfiSKkZGR4W/mgRa2tlvIj7johaQRCg0+mg0+mkiY9EwW82m1FdXY2TJ0/C4XBAp9N5Wfh9BX+szwMQX3HuY9HJsNlsceGWxPQeWNwzTBQR3SzsdnvQSajkcjmsVmtUyxENtxzf0JYpKSlRdS2KJoEGAwPul/LRo0dhs9kwatQoKJVKaTvPGU5LS0ulAXMZGRlwORXQq+UoqmmCGABTACATgBanCxUWKzRKHbYf/EXYA0BJbTM+O1yOq8ZkQyG4RZlMJsPJKjOcrZFyTDolmu0aVDfYcLTcgl+NzMSgjARkGJQARBcewO50tYbP9BZ3NkfXzVja291yIjk7rKfgz8zMBOCuY1NTZ1isVwAAIABJREFUEywWS0DBbzQa40Lcx4vlHojNNWmxWKRxFwzTFbC4Z5goEW68+q6YPTbSbjkNDQ04fPgwNBqNFNrywIEDcSEkOoLvDLgulwvFxcU4ffo0Bg4ciPT0dEnQi/jOcCqGQ2xoaMBZyWr8eKYBWhnB5XLC4QAUCjlS9AbIZUCCRonqxhbUN/t/sTld24SGFheSPKIm6tS/PKrlAiE3SYP0BDX6p+oxvl8yElUCBBBkMrd1VEaEYVlG/LfwjHc9BWBoZkKXDaqNJfEg7qNdBkEQoNfrodfr/QS/2WxGZWUlrFYr9uzZA71e72Xh78qxMPFiuY/V88lisfDstEyXwuKeYSJMe+PVd5W4j4Sg8wxtOXToUCQmJkb8GLHAM46+6IKTkpKC/Pz8sEWJZzjEVEEGhUYHmVKNk9WNUCrkSEvQQEE2JCsAwWqGHSrYW1qgUCq9rg+lXAaZx+XicrnQL0WHZL0KNY3uzoUAglYBTBqQApNaQIsTKKlvwaEzZqgUMozIScTInEQcH5COn8osAACtUo7zB6Yh06gCdUHIzFh39GJ9fCCylvtw8RT8WVlZqK+vx7hx49DY2AiLxYLKykqcOHECTqdTEvxGoxEGgyFqgj9eBtTGioaGBhb3TJfC4p5hIkRHJ6Hqikg2ne1AEBGKi4tRXFyM/v37Bwxt2d3FvdVqxYkTJ2C32zs/IJhc6JuoRLIuDdmJKuwvqgeB0D9Nj3F9TdAKdtghw7i8JHxyoAgAQalQQqlS4byzkpGg8p6h1qSR4epxffCfgkqUma3QqeUY3y8Fg9N1cLiA3Sfr8PmRMohzXO07WY05o3MxfYgJk4dkoMUlQ6JWgSSN3D2Qt4uIteU81sfvqM99pBEEAQaDAQaDAVlZWQDcglt06SkvL8fx48fhcrmg0+mkcSQJCQkRsbjHOgQlENsvOWy5Z7oaFvcM00nEwbIOhyOoX30ounKCqY5QU1ODgoKCNkNbdoW4j8YL2uVywWq14r///S8GDRokueBEIl+dHEh2VGPe6EwoVBroVTKYrYSCWhdOVJrRJyMFf5ybhZ1HSlFtsWJohgHDU+T4/rt9ICJJZBmNRmTp9bhyTDaa7C4o5b/ErK+xEXYdr5SEPeD2q//qWCV+NTgBekUTcpOTW89/9+x8dYTe4JYTzvGDIZPJAgp+0cJfXl6OwsJCEJGfhb+9Qj0eLPexPBdsuWe6Ghb3DNMJwvWrD0VXuOV05KXW3NyMgoICuFyusCzZ0Rb3ol98JF/QYseFiDB27NiohO+0Wq1wtVih1apR2+zCmm+LsONQWesgWwET+ifj9gsHQCUjGDVywOVEbtq5ANxuUI2NjSgqKoLFYvGK0IPWCD0NVieaWvyvn5pGG0hmAlFLTFxUYu0WE2thLRLLMrTXLUgmk0kWe888RMFfWlqKhoYGSfCLnc+2BH88iPtYfj1gyz3T1bC4Z5gOQESw2+2SKO9MDOOuEPftwel04ueff0ZFRQUGDRqEtLS0sPaLtrgXBCFiIsFms6GgoEBywSkoKIia+BC/msjkchwuN+PjQ2UQIMDteEP49udqjO1rwqwRqXA6nRBkMtRZCSX1NhwtM0OlkOOcPkMwJEEBh80qRegpKyuDzWZDSt9BaGlugkuQQaVUQSZ318OoVUJGThBiJ7JjbbWOB3EfSyLhFuQp+LOzswH8IvjFSFGi4DcYDJKFX6/XS2K6t4t7ttwzXQ2Le4ZpBx31qw9FvIh7IpJ8b3NycpCfn9+uF3JXWO47m7/L5UJRURFKSkqkjot4DqNtaXaSgEMlZrgIcLjcYSrlMgFymYBDpfW4dHgqFIKA6iYX9hXV4c1vfobZ6r7O0hPUuOPCgRjX14CU1ig9gPuctbiA84dkY+fhM6htqIXL5YJKpcC0wQPgaKyDU6uVyiAIAhrtgNnmvt6Majn0ASbN6gn0xDq1l2iJ6mAW/oaGBlgsFpSUlKChoQEAYDAYYLfbYTabYTAYYibyY225F12fGKYrYHHPMGHQWb/6UMTDQFSLxYIjR45Aq9VKoS3bS1dY7jsj2EQXnNTUVL8oOOHm3ZEyiPvIZQJSDCrUNdvhanWQl8sEJGiUSDGooRAEOElAUa17YqsEjRIpejWqGltQYbFh28FSDEofgETVL/k2OQQU1TVjQGYiTAk6HK9sgFYhYFSOEWlqJwoLfobL5cLJkyeRnJwMeWIWth+uQLmlBYIgIMekw+WjspFlkPdIMdzbLfddaTGXyWSSy1hOTg4At6BubGxETU2NJPjFwb2eFv6uKGMsxX1jYyPHuWe6FBb3DNMGkfCrD4VvjPWupKWlBYWFhbBYLH6hLdtLvFrurVYrCgoK4HA4cM4550Cn0wXMO1rnQBT3FqsDI3IS0cekw6nqRgCA00XQqWS4cHAaQE7YHAStSo6LhmVALhMgEwTkmrRotDqw++cqNLY4kahyC5QGO/Dhf8+gsMICAUBmohaDM42Y0M8EvYJAREhOTkZSUhJMJhMabE68820RCs5Uw253ACDU1irRYrNhwcS+MKjlEb22Y+0WE+vjxwOxdoeRy+UwGo1QqVQYNmwYALfIFi38xcXFaGxslAS/6MMfDcEfa8u90WiMybGZ3gmLe4YJAhGhvr4earVastR3d7EgCh7f0JbDhg3rdN2iHfWnvVZzXxec9PT0iOXdXgRBQL3VgaOl9fjzrGH4+FA5jlc0IMekw8yRGcgyqkHkhEYphwABu45X46dSM0rrm6GUyfDb8/thQv8UGNTuWWhlMhmOVTSgsMIdw54AlNY3o7S+GSq5gAsGmLzOhUwmg10QUNnslGYRFseNnKoyo7bRhoKDhyEIgleEHp1O122veRb3sRf3Ip73llwuR2JiopchQRT8ZrPZS/B7TrrVWcEfa3HPPvdMV8LinmF88PSr379/P/Lz83uESBAt3/X19SgoKEBKSkrI0JYdyd9u959xNVK0x3JfU1ODI0eOID09PayJqMTBuh1FEAQ4IcBic0EuAEb1L2UVOw5KmYDvi2pwurYR55+VjKlDUmGzO3GwuBb9TFpo1YDTBew7VYPS+mYUVTcBAOxwYu2eIjx4+XBY7U4Yle5O5onKhoBlOVndiPP6m+B7xQpwfwlwtQqt/8/euwbHdd5nnr9z6T59v6Ebd4AAeAMo3imKlCzbsiUncuLLKHEsezOJN5NsUtmd2szuVu3qy9Sm/CGRt2pmNsmkNq6Uk2gymTi2x2U5iZPYViw7tmRJFEVKJAESBEHc0egG0PfrOefdD81u4kqicWtQPL8qV1kN4pz3XBrnef/neZ+/JEnY7XZUh4bX42Hf2bPoul5bsDs6Okoul1uW0OO9k9Cz0e+DVblvLI1oorWSjVyHtQS/ruu1Cv/4+Di5XK4W31m9F10u14YFv7Wg1uJhwhL3FhZ3WMtXX20wtReqX9vB5cuXAda1p2yFveC5r1pwDMPg5MmTGz7GrVTuJUlioWDywxtxxuezqIrE0Y4Aj+0L4lAq58M0TXwOhQPNXq7PpphJTtd+/1xvGL9DBmGSLRnMJPIUyiYht5182UCSKt78TKHMVCJP6x3vbrPPAVOrx9Pk0VAkgSmWH5dXk+mLeLkRTS37930RL15NBkxUVSUYDBIMBms/L5VKyxJ6CoUCmqbVqvtV28VKGu3ht8T93miitdm/n6qqEggEam+agNrkM51OMzY2RjabRVGUVRX+ta57o8X9ViyPFhb1Yol7CwvW99VXk2x2qi37blCNtsxkMhw5cqQWZ7fd7PTagXtNHkzTZGxsjOnp6ftacNZiK+I+p8PfXp5hNH63kv6DoSiGKfjIoaaa0HAo8PGjrXg0lVuxDIosMdDu53xvEElUjkuVJZx2BVkCt6bgvOOvVyTwOdTaGA3D4JF2H++ML7CQLdX263WonO4KrHme7LLg2aMVL391rL1hDx870oxdFqx3+PY76TxLE3qKxSLpdJpkMsnExATlchmn07lM8DcaS9zvDVvOdo5hrcnn/QR/1V7WaHG/F74TFg8PD65isbDYBkzTRNf1Sr74Gr56VVXRdX1T6TH1sJ0Z7lVWRltWF1fuFFu1tmxk+2sJ8Pn5ea5fv75hC049294Ii3md2/OrLTJXphKc6Q7Uti2EwG+HTxxtJlUMI92x74gl58xrlzjbE2JoJk26UK5Za460+5Ek6A65aue4ySnzucf28dbtBaKpAmG3ncd6m2h2r518I4Qg5JD5hZNtpEvmnf3J2GRR17FLkoTD4cDhcNR6IAghyOfzpFIp5ufnGR0dRdd1yuUyExMTm+5suhX2wpuDRrMXxP1Oi+q1BH+5XK55+Kv2Ml3Xa779qqVntyZ/VlqOxW5jiXuLh5KN5tXvVgb9dtt/lkZbnj17FrvdTiqV2tFj2e20nEKhwNDQEKZp1mXBWYvNintJkiiWjVVVb5tSiZnUTdC8QTT7kusqTHz26v+9ezxCkonnDfrCbv7XZw4xnykyly6SL+n0hd0EXDbCLgXEXatPxCnxc0cilAyBXZWQxL2FuhACVRIEa3NVc92Kfb3nweVy4XK5aG1tBSodjgcHB1EUZVWjo2p1f6djEB92z/9eEPeNGIPNZlsl+G/cuIHNZqNYLBKPx2vrSZZW+J1O545cM9M0G/bWwOLhxBL3Fg8V9ebV76a43w77z72iLfdqVOVGqQrwqgVnZmamrg66G9l2vQghCLrs+Bw2UoXKYmKnTeH8/giTi3m+dmGSYiHHmZ4mznokNLnigTaRsKsKuq5XzpkkMxjNMZcucjOW4fJEEpsi0xt287EjLfQ0OXEpYu3zK0zsMmA2shftaiRJQlVV2tvba1aw9WIQdyKhp9HiutH7h70h7g3DaPgYoHI/VmNhq5TL5dp6klgsRj6fry0gr4r+rQr+vfAGx+LhwxL3Fg8FVVFYFVMbzavfLXG/1RhJ0zSZnJy8Z7Tlgy7uZVkmkUhw/fp1WlpaOHfu3LZVw7Ziywk4ZH7maCv/+N4MmaLOkY4Ab48vEk0Wcdkk0ukMM8k8ZRTO9YYYj+cZjedI5MscbffT0+QEBNemk+hC4tuX7i62jaYK5Eo6/9OTPTgcD5Z/fK3zuV4qynYm9CzdfyPF9V4Q1ntlDHuhar2WPchmsxEKhQiFQrXPqgvI0+k0c3Nz5PN5bDZbbfK52fux0RM9i4cLS9xbvO/ZShOqqud+p9nKJGJ+fp4bN27cN9pypycqOynuC4UC0WgUVVU5deoUTqdzW7dfT4faVQiTIy0u2v29zKVKODWVC7cXsasSugDlzvV4ZzxBV5OH//zPI7wzkUAIgVtTef5sF5843o7TrvKj4fiyTedKOsl8mdF4lnC3r+GdjOtlI4JmOxN6ltJocd/o/cPeEfeNHgNs3Pu/cgE5rH0/2my2ZRX+9QT/Xjl+i4cLS9xbvG/ZqK/+Xuy2Lace8vk8169fRwixoWjLB7Fyb5omt2/fZnZ2Fr/fT3Nz87YLe9h60o8kBEFNoqnFyehiJbIyVSijSBJBp0rIbUOzKbw3magJe4BsUeeH12Oc7g7ic9owzOVjUCQJiUr+fVYHhITbtvGFyzvdnGun2ExCj9frxWaz1bbR6OPeK+K+0VXzRqbUbNc41hL81fsxnU7XBL/dbl/m33e5XGQymU0vpr1+/TrPP/987b9v3brFF7/4RX71V3+V559/ntu3b9PT08PXvva1ZZNjCwtL3Fu876j66qsNlbbSWXYv2nIMw+DWrVvEYjEOHTpEOBze8D4eJHEfj8e5ceMGLS0tnD9/ntu3b++YYNsuEZwtCxZyJUwhyJcq1zORKWC3+ekIurgRTa/aT65kcHMuzeEWL8c6/NyIpms/C3k0HHYFp6byl6/fRpIkTnUHOd7uwy7v7Sr+dl6rjSb0GIaB2+3G5/NRLpcbWjHdCxXbBznnfrvZ7kmGpmlomrbs729V8KdSKb7xjW/wJ3/yJzQ1NZHP5/nWt77FmTNn6Ozs3PDz6PDhw1y6dKk2/o6ODp577jlefPFFnn76aV544QVefPFFXnzxRb70pS9t27FZPPhY4t7ifUPVV18ulze0WHYjKIpCsVjcphHeez/3E8Yroy3Pnz9f10PzQbHlLH0jsdSCs5OTk+2K8ZzPlrk6tcgnT7TztbcmyBTvvDUCjnf6GbvTdXYpHk3FZVc50OxGsylkijpv3Z7HabdxIOLhkQ4f16YTzCQLAMwk8pim4HyPf8/bdHaycr1WQo9pmuRyOVKpVC0RJRaL7WpCT5W9UrlvtLDeKwtqd+MNwlLB/9u//dv81m/9Fj/84Q/5D//hP3Dx4kX+9E//lImJCbq6uvjjP/5jenp6NrztV155hf3797Nv3z5efvllXn31VQC+8IUv8NRTT1ni3mIZlri3eF+wFV/9vdgrtpx0Os3g4CBut7sWbVkve71yv9SCs9YbiZ20mNSz7fVEmyRJZIsGI3MZDrf6+J8/coDJxRwygq6gi4jHzmO9QV4ZjBLPVCaMAZed3oiLI+1evDYYaHbSFdzHLz3aiWlWusz+8+AsI7G7OfoCeGdikUfafbitv+DLkGUZj8eDx+NBkiR0Xae9vX3XEnqWYon7u2PYC7acRrzFqN6PBw8e5Itf/GJtHFNTUxt+41rlq1/9Kp///OcBiEajtLW1AdDW1sbc3Nz2Dtzigcd6NFg80GyHr/5eNHpBbalUYnh4mGw2S39//5a6HG41kWcj29+suK9acFpbW9d9I7HTlfutThyEEIS9dpw2heuzKUZiacIejWw2Tz5f4FxvkOOdfn7vF47xz4NzpAplDrR4ONcTojuoYZqVa+NWwO2UkCSZa9H8MmFfJV8y0M36xlsWMsmCQb5s4HWo+DW51hl3J6hngfJOTNqq4no3E3qWsleE9V4Yw17o8N2oNRjpdBqv11v7b0mS6OzsrGsbpVKJb3/72/z+7//+dg/P4n1K479xFhabYDt99feiUZ570zSZmJhgcnKSvr4+jhw5si0Wo1KptNWhrstmFqXm83mGhoYA7puCs5MdcLfrrUDQofDU4WZeGYxSMkxmkwVcsuBDhyIomLhlwYk2F0fb+sgUDTRFwq7AYsEkUzSxqxIhl4oNEyEEzT4Np00hX15+D3YEXXgq4fYbGldWl/jetSjXZhKUDYHPYeOp/mZOtHuRN7iNzbDePWsIiYW8QTRdxKupRLx23Or2CrB7Vc53KqFno/vfLfaCuDcMY8c7fO9lVor7zfAP//APnD59mpaWFgBaWlqYmZmhra2NmZkZmpubt2OoFu8jLHFv8UBR9dVPTk4SDodRFGVHH6C7acupTlSq0ZbhcPie0Zb1sltNpjaCaZqMjo4SjUY3vChYluXaOdputkvcy5ic6fazP+IhWdBRFRlVBk3cnVRV7icDj1qppr96M8G33plifCGH36ny88c7eLo/QlCTCDllPjrQwvevzVLUK9euxefgw4ciqJK4b3dZSZLQNI13xhNcnlysfZ4qlPnHKzN0BJw0u3ZX/BnIvH47wb8Mz1HSTSSgJ+zh0yfb8dfvNluXem0Y25HQs3L/lrjfG2OAxuXMZzKZLYv7v/7rv65ZcgA+9alP8dJLL/HCCy/w0ksv8elPf3qrw7R4n2GJe4sHhqW++omJCUKh0I6/7t1NcZ9Op3nnnXcANhRtWS87Le43SiwWY3h4+J4WnLXYac/9dp0bBRNVkXh7bJGb0RSpTIbOoJt/dbaPLr+tth9FURiJF/ivP73NXKqEQBDPlPjL128T8Wic6/WjSSanO330NLmZTuRx2hVafRp+bePXsoTC9eji6s91k9F4ltYdWpi78lrJskzJEESzOj+6MUfZqOxTAKPxDG+MLvAzAxEwt+f7ttV7ZTMJPdXM8+oC+UaLWmsMFRoZi7rVyn0ul+N73/seX/7yl2ufvfDCC3z2s5/lK1/5Ct3d3Xz961/fjqFavI+wxL3FnmctX33VC7+ZhaX1sBvi3jAMotEoCwsLHD9+fFmW8nayWxOV9ahacCRJ4vTp0zgcjrp+f73JiSRJpIoGJd3Eocp4NKXuh7ksyxtaW7GR6p+JzA+GolybTlT+W8BYPMPfvTvFL5/bh9cmkSnDQqrEj2/OM5ss0d/mZX+zBwW4PZ/lwtg8Jzp9aPbK24CwU6LZ7UEIUXt7tVEkQJbXsaes8/l2UT1fOjLD0Rxj8SyposnYQoGIx45dkRBUrtXteJZcKYRrG59K212tvV9Cz+zsLDdv3kQIgc1mQwhBOp3etYSelVQDBhrJXsi5b2RiTyaTqdtjvxSXy8X8/Pyyz5qamnjllVe2OjSL9zGWuLfYs9xrsexuCdWdXFArhGB2dpZbt24RCoVoaWnZMWEPjavcL7XgHD58eNPHuFblXjdhcDbFayPzJHNlQm47Tx6McKjZRb26tbrtdNEkkS9jmAK/00bQWZ8wSZVMRmKVrPqQ286hsAskQaYMiVwZ2W3jm+9MUyhXKpqfONlGsWxy4fZiJf6zO8CpriArPTebvXaq0DnZGWBiPsvSLXo0ld6we8fvCVmWuTqV4duXJgl7HLT4Xcwk8+RKOr1hN8qd66TZ5DuTje2psu6WLWZpQk97eztQEZNTU1PMz8/vakLPSqyc+7tjaNQEYztsORYW9WKJe4s9R3WxrK7r6+bV71aKzU4J4lQqxdDQUC3aMp/PMzk5ue37WcpGsvS3m6oFp62tre5c/pWsdS3GF/N8650pykZFEGaKOt+8OMkXnuihow4DtyzLyLLMbLrM6yNxTFGRmLphcmZfiIOR+rvinuwO4rCpvHlrnqKuc7K7Cc2uMrGQ4+p0moVskX/3sUN8571Z/vL1MVS1Im6HZtP4nDZOdfmoLpgVkkyyaGKa4LTV16W2WCzySFuAdLGFS+OLZIs6zT4HT/U30+RSEBvYjizLZMsCUwg8dnlDv1OdLGVKgjdH5zEFxDIFjnYGaA84mUnmKZRNPHYZSRI8uq8Jhyph1pkCdK/9N6pqrSgKTqeTQCBAb28vsDsJPSvZC8J6L+TcN/LtwXYsqLWwqBdL3FvsKTaaV79b4n67xcF60ZbFYnHH30TsdBTmUnK5HENDQ8iyvCkLzlqsrNwL4OL4Yk3YVymUDd6dStIZaN6wPUeSJFTNye35HCVD8M5ExVJzojPAVCJP2GMn6KykDRWLRZxO56p7Q1EUTCSCTolzfWFmU0Veem2EYlnHNE1uxnJoNpmuoItMUWc2VaBQNphOFAi67aQKZZw2lZ4mFyNzGQplE5cCOUPiRzfiXJlOUCgbtPmdPDPQQm9I27DAt8uCDx0IcbLTjyEEDlXe8AShaMq8O5Hi3ckEumlysNnLYz0hPLaNRV2WTUH2TjMvIeDadILPP9bNT27GMUxBu1/j0Z4Qh1tc2zr5bPSC1pX7342EnpXsBXG/F3LuGynurcq9RSOwxL3FnkAIQblcronP+0VbKoqyK+J+u7hftOVu2Ix2w5YjhGBkZGTLFpy1WDl+U0C+tPbx5Ip6XQtwJUlCsrt592acb70zVTOGXJ5I8Inj7Rxu8TARn2ZsbAy73V4T+D6fj3A4TMnm4epMgnfGFmkPOHjyUDPfuTqMeWfcHk2lu8nNlckk/a0+HLaK4IpnSkwnchzr8KGpCkjgtCkoCEqGiSSrvDWyyBuj8dpYJxdzfOvSJF94vJfAfRIGl50D08Bnh4oLX2ysQi4rvHFrgR8MRWsfzSYLzGdKfOpEK5p8/224bBItPifJfLn2+7nSPMc7fZztacKnbXyiUQ+NFvcbEdbbndCzmTHsNA975T6TyWypP4mFxWawxL3FnqCaVrLRvHpVVRu6OLQelkZbnj9/fs2HzG6I+5225cRiMbLZLLIsb9mCsxYrxboqSwy0+RiOplFkiZ4mFxGfA1mSONLuq2/RqSRRNiV+fDO+yvH9w+tRfnYgDMUijz32WO0+LRQKpNNp8tj5q5/c4juXp5BkGUWRiXgdqJJEX8SDYQgUWcKuyGSKOhJwsjPA2EKOWLqAwyZzZTrFQKsXTZWxaTJBlw2XTSZdMhmcSS4fKxLJvM5MskCwxbnhCYwsy7Xv1kbvtVTB4NL46qSd69EUiUKElg3EaNpl+PDhCLF0gcVcJRI0W9TxO22EnTKKtMGJRp00MiGluv96JxdbTehZyV4Q91bl3hL3FruPJe4t9gz1NEFSVXXHMs/XYjMP6lwux/Xr14H7R1vuhh9+p2w5VQuOoih4PB66u7t3RFCsrNwLIehv9TC1L4jTrjK2kONvL8/g1lRSBR3n4QgR98b+xEmSBMJcvoZVVGwTBVRsNpV9Bw5gGEZN3DudTjweD+/O5PiXkQUcTgemaeK0KURTBZL5MslckWiqSKaoE/E56A46yZUNfvZYK20BJ5OJPL/yeA/fuxalWDbwO1Q0VeKpwy34HTLpool5Z1ASEmUTciWDom6SK5voyCjc+5rKskxWl7gxl+bmXJYmj50THX7Crvt/3wwhyJdXvyEzTEGhbCBJ6ycTVb8zQgg6vCq/+kQPo7Es2ZJOX8RDs9uGIu3sPb+XbDmbpZ6EHo/HU7PzuN3u2u83kr0wwbDEvcXDhiXuLR5IFEUhn8/vyr6qonKjDwdd1xkdHSUej3Po0KENWVN2ww+/3bYcwzAYHR1lbm6uZsF58803dzSLfuW2PXaZjx9t5XuDc1yeSODSFDRV4spkgnShxPNnuvBo9xcWkiQhykWOd/p5dzLJfDpPqVwi5HVxel8In2O9P5USU4l8zfcvyzJ+t8aPb8Z58mCEHw/HGYnnKOkGc8k8v3Kumyujs1yRJH7hVAeiz4+kqBzv8DOxkEU3TXrDHsJ3Frp6NIWDLT5i6RhFQ3B7Pke6UCbospEvG3x/KMZHDoaxy+tfV0nz8PKlaYbn0rXPLk8s8j88to9Wz71jQ102mRafk7H57LLPfQ4bAadtw9daCEHADme6Kt7jyn248xaxRttydmr/6yUbtzEDAAAgAElEQVT0ZDIZ0ul0LaEnm81y48aNXU3oWUmjrwM0VtyXSqWHukOvRWOwxL3FnqEej/Ru2nKqlpn7PRyWRlt2dnZy7ty5DVes6nlrsVm2swnU3Nwcw8PDtLe3L7Pg7KSvf71tZ0sGw9E0IddyH/J4PEc8V8Kj3X8xryRJSMU0Tx1sIjq/SJPTjc0WwK0pfOxIC0GnjbUiGoUw6Qw4UWUJ/Y61RJEk3ptKcaIrwL95spd/vOKgpAuOdvjoDbt4dShNNlckns4TG79JuVzG7XbTEQ7j8Xhw2E0kURFDkjB5Yn8TfWEPibzOrViawZkkj/Y0MRpLMxrL0N/qpSdgX/Pa2u12ZjPlZcIeIJUv88ateT55vAXpHtGTDgU+2t/CNy9O1DzzmirzkYEWfJoM4t7XeqWo2820pkaLSiHErgpKRVHw+/34/f7aZ2+++SaRSGRXE3pW0mhhD40T9422hlk8vFji3uKBZLfScmBjfvhqtKXH4+Hs2bM73lxrM6x8yFb/u54HUC6XY3BwEFVVOXPmzKoUnO3s9LqS9SYnphAU9dX7FECxvLGxCCGYmpoi2FTk3z5zhJH5PEJAf6uXVp9GVdivJVa7gg4+OtDCd6/OAqCbgrBHI+Jx8L1rs0wt5lFlie+8N8MXHu8hXzaxa3Y8Hi+dp04hhKhZLCYnJ0mn0wghiEQiKL5m4nnBrfk8700liXg1Pn2qi1S+yFujGQBGY1n6Qo4171FVVZmKrf2Gay5ToKgLHPfQPEII9gVs/I9P9DKxmKdkGHSH3IRdCtJ9hH2jhc1eEPeNFraSJN03oSefz+NwOLYtoWcvYhhGQ4+p0feBxcOHJe4tHkh2My3nXhOJUqnEjRs3yOVyDAwMPBCRZ5IksZDTGY1nSeTK9IRdtPk1nOr6bxkMw+DWrVvE43EOHz5MKBRa8981onLv1RTaAw6Go5nlnztUmjyrH+hLH7RCCGKxGIODg9jtdg4fPIAsy/RGPLWf3w+fHX7lXDcnuwNcHEvQ7NU40RXgu1dnuRnNMBrPkC6U+cUz3SxkipR0k0MdAbyaDFSsG263G7fbTVtbG1A534mCwdvjSS5PpvjnoSjZsqCsm1wYXeAXT3fQE3JzeyGL17G+PcYwKtGZkFz1s7DHgaZKq5plrUQIQUCDUHvFw22a5n0r9nuBRovrveA1X4udTuhZSaMneWBV7i0ePixxb7FnqOdB3AhbzlJM02R8fJypqSn2799PS0vLA1GdUVWV2XSJr781wVy6CIB0HR7tDfFMfwSnbbUYqVpwOjo67ms12klxv17l3q5I/Nyxdi74FplN5BhfyGNXJJ450krIufxP3EJe59p0iomFPC0+O90eCT09z8DAANFotHZs9TyUhRD47fDhvgBPHQgxnSzyL8MxTncHCThtPL6/iSa3nUSuzHuTCQ61+Hi6vxlVEuvqalVVWSjouJwO3hwbx26zg2ySMCrRmW+MznO208NEOU+bG+LxOD6fD1VdfrylUon2gJ++sIdb8buTH69D5VxvCBmx4X6wjehuvBUaLe4bvf+Nst0JPXuRRsVxFgoFnM76G+BZWGwVS9xbPJA00pYTj8e5ceMGzc3N60Zb7lW8Pj+vj8zXhD1UDCdv317gaLuP3qa7D6KqBcdms61pwVmLRlTup5Il3rq9QDRdpMXn4PMHmwm5VQIOFXmJtlrI63z1zQlmEnkymQzZXJaDXS382gcPoZn5LY/bMAwk06Skm1ybTjIcTdMZdOF3qLhtgq5OP88ebcWvSWjKvScQkiRRNirNn6qWI5si43fZkSWJXNmktzXE2f0RHHqG+BIBtjQxRQiBKGZ47lQHg7NpRmIZmtwaJ7v8NHvUDXWa3SwPirjdKR7k499KQs9SEb2Ti4rrwTCMVRPf3SCVSuHxeHZ9vxYWlri3eCDZjVz4lfta2nX11KlT216RqfrVd7LCJFSN6ehqD7YpYHwxR1/Yha7rG7LgrMVuV+4nkyX+20/HyNzpgDq1kGN6Icfnz3UvE/aSJDE4nWYsliSxmMDhcNDc3EK6JHF5MskT3a5teYUuhCDisdHf5mdkLs1cusD1qTzlcoF/85GjtLgr52fprhRFZb5gUiibKBK0eG2YepmIVyOZL+PW1DsdXgWaKtPk0ehv8bI/4sKtCoS466k2TZNsNksqlSKZTFLERrkkUVyc50hI40RrGIdmr4zhAavE10ujxXWjbTnbLaw3mtAjSVKtur9S7DeKRlXure60Fo3CEvcWe4Z6HkS7kS5TRZIkJicnKRaLG4623AzVrPudfAhJwsCtrb2wLOiyE4/HGRoa2pAFZy12s3JvAm+NLtSEfZXZVIGbcxke6wnU7hHDMBicmCOZTBEKhZb5h2cSecQ+z7bdT5osePZoG9PJIHOpIn6nSlCDzpBz1bkxJJV3p7P81U/HeHcyQcSr8a9OdfCzR1pocgpSXgcfP9rKNy9OIUng1Wz0Nrn48KEwHhurmj/JsozX68Xl9fPudJo3R2bQ7CUCbif7mxS85jzRmRk0TatVW3dqAWWj/caNFveN3v9uVM3XSujRdb22YDcWi5HJZLh48eKuJvSspFGNtNLptCXuLRqCJe4tLNZBCMHMzAwTExOEQqFNid16qL4h2MnXx6VchnO9Bxifz1E27grNiEfFZeSYmory6KOPbjqXeTcr9yVdsJAtrflv59IFoHINp6enWVhYoCcSZixlILFcVLT6HSjS9onRrC7xj1dmuBFNUzYEDhke7wvQEvCwmBfkSkZlsa9LZTZV5ve/c43xhRxCQDJf5j9+7wY+h42f6W+ir0mj2dfBub4w12eTtPlddDe58GnrCyNJkrg9X2A6UaCvLcSNaBqKJhlTobOlm57uborFYq26X11A6XK5lgmw7RBDD7O43gv7b0S1WlXVWkJPPp/n5s2bHD58eFlCT6FQQNO0XUvoadSCWkvcWzQKS9xb7Bn2gjezSjKZZGhoCK/XS19f3648KHejkVWpVKI/pPH82S7evL1AOl+myQHHWp00+534uo5vafs7Le6Xoqky7YHVDZYAukJustksV69exe1209/fT0dZ5nqsQDRVqP27Jo+dE11+TFPfFnEvyzIXxxe5Nl1Jp5GATLHM94bieNxu/vwnt5hKFGj1OfiND+2nUDbwOGwcafdTLJvE0pXOtt8fnONcbxCfrZKQMzi1SE/Yw3Asy8uXpoh4Nc71Bhlo9WGaAlUBnyYjTBMTiVi6xFy6xDcvjGEIUck2d9r4d88c4vy+SrfMSCSybAFl1U8djUZrfuql4svtdu+p7+j9aLS43gu2nEZbYqpj2O2EnpU0StxbthyLRmGJe4sHmu1+gFejLfP5PEeOHMHr9TI7O0s2u1pAbje7sY5AlmWEoXOo2UXIrhONzRP2e4iEmzCQSRdNnDaZe6Ri3nf7u5WqIiE4sy9APFMglddZzJYoGSb7I24CSpF33x1mYGCgZhkI2eD5x7p4byrJ9GKeVr+T450+mj128nljW8R9riwqlfKl45QVppJFfnprAaddxTAFU4k8f3t5msd6QyxkS4zGsngdKt1NblRFpqgbGKZAFxL/PDSHIkv86Gac71+LAqCpChMLeQ61pgi7bUwv5njiQIQTHV4kAQ67wj9dncEwBdUXFal8mR8MzXG83Yt9xVdmvTjOTCZDKpVibGyMbDa7rAGSz+e75yLrvWDLafT+H+bJxb3GsNsJPVbl3uJhwxL3Fg8s22ljuVe05W4t3t2N/SiKQiaT4cqVK5Vc98OHsWsa44sFXh+ZZzFbIuzVON/XRFegfmvOrop7SUJVZI60+4llSjR7NVwqSIUUUinHuXPnVomrsEvlo4fCGEKg3LH5VN/KbFQM3kuwybKETVkuZgwBiXyZgMvGB4IRPnAgTLFsMjiTptmr0e53MBrLki7o3IymOdLu44n9YQIOmUTeYGw+y+MHInzt7anK/pHIFHVShTKv3Yzzmx/q473JBN95b5qwp4e+JieGWbH/LB835Ms6uZKBfQOXdi0/dbUBUjKZZGZmhmKxiMPhWCb4l34fG13pb7Qt5mEX9/WI6u1K6LnX9ncbS9xbNApL3FvsGer941uNw9yquL9ftOXK2M2SAZmSjl2R8WrKtlUId1oY67pOJpNhaGiIRx55pJawMpUs8ddvjNfE4EyywO14ll99oocWT32vxXeyQ+1KJhIFvvn2FPFMEcM00At5njnazpMHIzjXWTQMd7zIrK7sbsd1dChwqjvI7flMLRFHQuJQi4+T3UG++uY4V2eS+Bx2PnwoQsht57lTHYzEMswmi8gynO0J8cxAM5IwkSSwKRKyJN1JzAEkKOiVn6UKZRRZRgIMU3BpIklvyEGbv/K/qYUsphCoskSzz0Gzx4GqSCwWBYYp0FQJv7bx+24te8Va1VaPx1P73mynyJRlmUzJxBDgtcv3bKb1sIvrRu9/O8awmYQen8+Hy+Vq+MQSIJvN0tXV1ehhWDyEWOLe4oFlq11qNxptWa2oS5LE+GKBHw3HmUsWcGkKZ3tCHGv3YduGN747VbkXQhCNRhkZGUFVVfr7+wkEAnd+KnFxfHFVlTdd0HlvMknrQKQu0buVCcp6TarWomTAq9djxNIF0pkMuVyOgN/PWxNZjvc046zzpUO945ZlmbwuKJQFdlXCbatMakzTpL/Fzc8+0saliUUCTjstfgcHWnz8p+/d4AdDMQAEOcbiWbpDTjoDTv6vZweQJYh4Hbg1BSFMCrpEwKFwqNVPvmzQGXQyuViJMfU6VISAjoCTTLGMAAQSJUOQKYPfofKLpzv5/rUZyobAqdlw2mTO9oa4Op3m1etRCmWDiFfjo/0t9De7NtV1dr1qazabZXp6mnQ6zdtvv11L8alWW51OZ93iq2jKXBpPcnkyQVk36Yt4eHx/EyGnQrpYWSjtsd+dXDZa3Dd6/3tB3O9EBOX9EnpGR0fJ5XLLLGSV+Nndvx6W596iUVji3uKBZbNdaqs57vPz8xvKca+K7ni2zDcuTLKYqyS0LOZgNjGNXZU51u7dcuV3J8R9NptlcHAQTdN49NFHuXXr1vI4SSFI5Mpr/u58tlSX4IaK6K13wpUqGkwuFphLFWgPOGkPOPDY7y0IMkWd8ViKubk5HE4Hzc3NyJJEoWwQSxc39cZhw8cpyQzPF/jxcIz5TBG/087j+8P0N7uQMbFJJk/0BjjQ4uXKVIp4pog3WaCkm0S8GgvZEq1+B10hF5cnktgVicsTCzza08RX3xpDN0w0BfY1efi5Y2188GCYd8aTfOJEB3/z5jimgBafg2SuyMePdnI7lqZsSswk8zx5MMJXfnwLt6by9EArTU6Ji2MJvC4HZ3pCCODldyZqxxpLF/nby1M0P9lHyLE9wqcq5MPhMKqqsn//fnRdJ5VKkUpVrlk+n68rjlOWZd4aTfDK4Gzts3ShTKvfxaV8iRuzKQAG2n2c6Q7gUkTDxXWj978XxP1uRVAuTeipUrWQpVIpCoUCb7311q4m9EBF3Pt8vh3dh4XFWlji3mLPUO+DsN7KfTXacnR0lK6uLs6fP7+hfSqKgqZpjMazNWFfxRCCC2OLHGx2oylbe5Bvp7ivTmAWFhZ45OgxSrJGsgyuQGTZw1aRJfoiboZXLAIFONRSf/Z7vf0HEgWDb16c4nb87oLlgTYfnzjehldbW5iUSiUWFxYxS3lCTSFs6l0hLwFue/1ioh5xH8sZ/PcLE+TLlWuVLuh8650Jfvl8Dz0BO0IIFvIG37gwzmyygNNuI1fSuRXPcrTDjypLlAwT3RCVBcDNXvaF3Xz9rUlSeR2/Q8EUgtvzWb55aZoTHX7m0nn6W/387qePEksXUCQJr0Pl3clFJhcL5Eo6HzoUoVg2mM8Umc8U+c67U/ziiWa6fSGaAgEMJC5Npji1rwlNVcgWy1ybTpAu6NyezxHu8u6YpUpVVUKh0LKJdLFYJJlMkkgkGB8fR9f1Whyn3+/H4/HU7tVE0eTyxOKybfa3+fnhjTlmkgVctsp3byaZJ1sw+Jkj4YdeXO+F7rCNPAdVC1kwGGR+fp4zZ87sakIPWJ57i8ZhiXuLB5Z6KvdLoy3Pnj1bV8VGUSq++mxx7YlEoWRQNsSWxf12RGEuteB0dXVx7PRj/GRkgcsT0+RKBk6pzLPHOznml5CoVDcfafcxHE0zkyygGyZlQ9Df5uNAZHPifqMCUZIk3ptMLhP2AIMzKY60+zjZ6Vu2fyEEpVKJt956i4GBAT56oo9/ujLL0hF2h92EvfUvBN6ouFcUhStTyZqwr1I2BG+PLbIv2IosSVybSbCQLdET9qDKMh1BJ+adHPuyYZIu6LT6HJQNk3+4MsuxDj9Ti3l8TpVKey4oGYKfDMc50x3kymSC9yYTqLLERw63cO5AENMUNHtaONcXYTqRZyye4eLYfG1M0VSBhVyZAHkkKcDQTIaXXr/N+HwORZY43uHnmSMtvHYzVvf52gj3O5+aptHc3Exzc3Pt31e7687MzJDJZAAqKSnNXaRzRRCAVJnEtficfPWtCTqCLiRkxJ074d2pRc71hRou7vfC/htduW9USs3KMciyvOsJPWCJe4vGYYl7iz1FPRXUlQtd16JYLDI8PLws2rJeqgkzvQf3o0gxjBXj6wm78dgVYOu2nHJ5bYvMRqgultU0jbNnz6JpGq/dWuTHw3fF23Q6xzcvjNMa9NDssSFJEoos8dGBFuKZErIkEXTZCHtsuG31C4N6xH3ZEIzG144YvTmX4WTn3dfZ6XSawcFBTNPksccew2azcdonsCky74wvUtJN9oXdPLG/qbLQsk7qEWEr1ydUKZQNTFF5G6Ib8IGDLVybTpErleiNePjdTx3hT14dIZkr0RVys5At8aFDEf7sX27R/jOHuTyZ4GR3EL+jMv6yIXDZVbwOGx/ub+HGbIrpRJ63xuY52uHDZ4eABjIy3xieI1VYfe+YpgAZEgWT716bpXppDFPwzkSCkEdjoM3HvibXjlTt6zmvkiStuXgynU5TMssENLgyFkOSJZyahmmalHQTh3pX2ENlsm2KxkdhNrpyvpWqeXXcWz2HpmnuaFO+rY5hpxN6LFuORaOwxL3FA8u9bDn3irasl6qPvMVn58lDYV67OU/ZMJGAfRE3j/UE2aqwh8rxFAqFdX4qIRDIa0x+dF1nZGSEhYUFBgYGaotlMyWDK1PJFVuRSBfKjMaztHiDzGXKvD4yzzcuTpLO67jsCh84GOb5R7twb+LtdD3iXpEl/M61dxJy25cd2+LiIgMDA1y9erX2oHaoEo/1BDjS5iVdNMmVdN6bTNERdNLmd+BZGea+CSRJwhR3z7sQgoE2HxfHFlZd8YFWHza5sk+HpvJ/v3yV2WQBAXz1zTH+95/t599/4gg35zLIsszkYo6vvT1ONFPkZizNkQ4/M4k8vlYPmaLBSCxDZ9DFD2/McXFskc+c6cQUgrJhoqoyecNEliT8LoW+iIdLK2wrYY9GwKEg6zIzyTyZQpmwx05JN1nIFdENwa14hudOddDkVDa1oHYzVKug1UWO9/u3gUAASZL49FkPhqIxl8xTKpcolsqc7PSRSKbIGDo2mx273UZ32IfjzpPtYa6cb0bcC2QWCgYzyTwOVabFp9WVprSSnVhQWy+6rtc1hu1M6MlkMssW/lpY7BaWuLfYU9RbuS8Wi6s+j8ViDA8P09LSsma05WbRFIkPHmhioM1HNFXE51Rp9tjxatuz/bU897pZiXy8OLZIUTc52uFnf8SF2yavsuCsXEMgxJ3K7RIkqTJJMExB2YCxhTx/9cYYqUJlklTQDf7hvVna/A4+fbwVuU5tVJctB8GZniCDM6ll1XC/08Yj7T7m5ua4ceMGXV1dHDp0CEmSatuvXlMhBOmiwZ/+eJTXb8YRVDrXPneqk1843YF3CwI/milzaSJBPF2kJ+zmkXYffk2mM6Dx+IEwb43OUzYEiixxtCNIT8TDUDRHk9fBT0cWlll3yobg7y5P81hPkGS+xHy2zB//YASJSjb+vwzH+dfn9nFxvCLQdcPkyQNhTnUH+a8/vU0sXeTv35vhUyfaiXg13hxdZDiawqbInOwK8tH+Fgplg7lUgaDHjt9p43hnALWUQAhRy95XJEFn0EGLT0MAbX4nTS51R4T9qqhRSWY2Xebq9CK6YfJIh582rx1Vuv++hRB0+Gz8yvl9zKSKFMsGvWE37U0e/v7SFIlcgXKpjE0yONvpYnx4kGw2y8zMDIFAYMOV1u2k0bacesW9kGTenc7w3aszte9js9fBc6c7afNsLvJ3txbU7vQY7pXQk06nlyX0OBwO3nzzTR5//HFyuRwul2tT+0wkEvzGb/wGV65cQZIk/uzP/ozDhw/z/PPPc/v2bXp6evja1762bBGxhUUVS9xbPLCoqrqsc2w2m+X69ev3jbbcCnZFot1np8Ovbftr/5XiXpIkrs6k+PalKcpGZV9DMynO9oZ4ar+foavv4XQ6111D4LHLHG7zMpXIL9umXZHoi3jIlnRux7M1YV9FN02uTKX4yKEIAWd9D8V6IyXbfRqfe6ybn96qNNBq9jk4uy/AwuQIxWKRRx99FE2766FfOfkTVITxazfjtc+KusnXL0xwqjvA8XZPXWOvMp2qZP8n8xWry/XZNIMzKX7pTCcOyeQjB8Oc6AywkC3hd9kplg3+8rVbLObKPHe6m8uTCQ42e0jkymSKOm5NodmnkSkaXJtJ83hfGFkCU4AsVbrH/uXrt/mPnz2JokhkCjqJfAm3pvJrH+jlP79yg9FYlr6Il5/cjHFj9u4bmfH5LB97pI1fONPJ7fk8Q7NpvJqC264iyjKmUaY16CDidRBLF5AQVNccn+ry47VLqyaB20VV3MqyzOXpLH93eYqSUbk/3ro9zzNH2jjf40fawOTCNE18dvBHKl1xhRD0R5w0f7CXiYU8EtAVchF0yJhGE6lUJUGnWmmVZblWafX7/Tgcjh0X3w+SuE8WTL5/bXbZRHsuXeAHQ3N85nQbtk0cSqMXFcPO+f7XS+iZmZlhaGiIv/mbv2FycpJPfvKTnD17lkcffZSzZ8/W7D/343d+53d49tln+cY3vkGpVCKXy/F7v/d7PP3007zwwgu8+OKLvPjii3zpS1/a9mOzePCxxL3FA0vVc7/UlrKRaMvtYCf8vIqiLBPGqYLO6yPzNWEPYAqTH165zX5fDwcPHlySV782Z7qDLGbLXJ1KUjJMQh4HTx2OEPHYyJUqfuWl2FWZgNOOw6aQ001sJQl3HR72uvPiJegJOegIdFAoG2QS80yOXKGvr6/WKOle208WdN6dTK76d4YQXJ5McrLTd8/xSJJEqqCTLBh4WntJFw08mspbtxdqwr7K+HyuknjT4kQSJs0umRa3k/m8yX95bZzCnUp9STfxOVQujC3S5LajqTIIwVyqgAR8/GgbLrvCb36oj5cvTSFLEqos8UuPdlIyTL74ratcj2Zwayplw+ALT/Tyv32sn1cGo0gS3IplMEVFaaly5U3M22OLBFwaX79wu9Y868JonM+d7cRUTGLzOT7S38prI/NEU3k8doVjnX4e7Q7sStOxVEnw4+FYTdhDZWLz2s0YA61e6mmGvOy7J0xCmkS4ozKJM00TREVQyrJMZ2dnTWCvFce5tLuu1+vd8WjE3aSeirUkScxni2TWCA2YTuTIlkRd16jKXllQu1tjsNvt7Nu3jxdffBEhBB/84Af58pe/zIULF3j99df5wz/8Q37913+dz33uc/fcTiqV4kc/+hF/8Rd/Uduu3W7n5Zdf5tVXXwXgC1/4Ak899ZQl7i3WxBL3FnuKeipdsiyTSqV444036O7urlk3dmpcO12FWpmWU9QFiTvRmwJBPpcnlU7hcXuQnR4CAfd9t+nVZH7+WCtPHGiipJsoZhkzn0KRwOdQ2N/soc3vYCZZwK7KhD0aM8kCQZeN712pNED6+LE2OvwbEz2bbWKVTiwwNDRES0sLZ86cWfc8r4zatKvr+/aDLtt9J2ETiQJ//+4M04t5ZqNRTvf38KmTHcRSq+1eUOnee6zVtWy7k4v5mrAHGI4m+Pnj7VyaSJAq6ORKOi6bzC892s1wNI1HU2j2eHn2aCtPHYowlSiwr8lFs0/jK/8yyvBcFsMUyBIYJvy3N8b4f58/xfneICXdJJYpUdRNFFmiUDZx2RWQZLIlnaWH2xVy8+ZYksvjCxQMCd0UnO4O8ItnupAlwb6QEztbSGeSK82jgIo1zVx/W2XDZD67+pxmijq5skFgA9a2SndaQaZUaSjn02QcSkXsb+SeWxnHKYSgWCySSqVYXFxkbGwMXddrSSk+n29ZHOeDRj1/r4QQOGwKkgQrvzKaTWGzQWDv58r9RpAkia6uLrq6unjuuec2/Hu3bt0iEonwa7/2a1y+fJkzZ87wB3/wB0SjUdra2gBoa2tjbm5up4Zu8YBjiXuLB5JkMsm1a9fQdZ3HH398W7OJ16JqmdnJB9VKW46mygRcdpLZFIuJBKqqEIlEcNps+B0bP167Qq2pUzyeZj5TybQXQtDb5OTffvQg37w4xWK2RL5s8HNH2+gKOvnpyDyGEHznvRk+/1jXfRtLQf3ivlQqcf36dUqlEidPnryvP7U6yaritik8faSZ10Zi5Mt3P+8IODnZFbinuM+WTP7pyixTd7q+CmBqMc97EwmavBrjC7lVv9Pud67apmNFe+LhaIbuJg9f+sxx3hpdoFg2ONYZ4My+IH97aZImj5d///JVRmIZ2gNOHj8Q5ifDMX7rqf0MzqQAgarIyLKELFXeBEwlcjzTH2FsIY9pmney6bOYpkCzKTx5IIxHU/E7bSTzZWQJOkJuvvzqMO1BF7fns+RLBsPRNFenU3zwQBMTCzmePtSEuIcoX/fc6RI/HY1z/U7zqMOtPs73hnCry6NLq5NtRapMwhayy/tEOG3KqrdHayHLModRIeoAACAASURBVOPJMv94ZZrpxTySBD3hSpOvsHP93gr3muwvjUZcL44znU4jSdKyxkdrLZzci9S7oDfkUulp8jAazyz7/ERnJcFpM5P2h1nc67q+6aQgXde5ePEif/RHf8S5c+f4nd/5HV588cVtHqHF+xlL3Fs8UBSLRW7cuEGhUKC/v59bt27tuLCHu8J7J/e1Uty7bXCyzcm10QkCfj92ux0JONMTpMm9OfvASvHtUCXOdPnobXKRLhrciKaJJvO8cWu+Fvk5tZgjXdDxbMCysFFxL4RgcnKS8fHxutKMVlbuhRAcafXyfz47wHfemyGeLtLX7OZTJzrovM/bhkzJYGINAf/O+CKfPbeP0VhmWffe7iYXvWEXpqkviwps9Ws0ex3MpStJRwL46Uicz5/vhd4Qc8kCMwtpXiuX+fkT7VwaT6IqEgNtPuyqzKWxRUZiGX75/D46Ak4GZ1LYFQm7ImNz2pAliY6Ak0LZZGg2zSdPdvL//OP1WqxlwGnjyYNhfjQcozfi4dL4Ii67SraoU9RNTBPyS3zUY/NZnj3aynsTi5zdF8Bb5y1tIPP9weiydJ5YOkamqPOJoy0orL7+AYfMY71NfPfqDEvt/af3hfA7FFjjd5aSLpl8571pZu6sHxGiYk/63rUov3hqc37wtbhXHGcqlaotnLTZbMu66y5dF7JXqFdYu1T41Il2fnwzzu14Bpsqc7wjyOlu/5bSchr95qNRY0in03g8G1/zs5TOzk46Ozs5d+4cAJ/5zGd48cUXaWlpYWZmhra2NmZmZmqTUguLlVji3mJPsZ7AM02TsbExpqenOXDgAM3NzRiGsW0dXe/HdnaPvdc+qvGAs7Oz3Lp1iwOHDvN/fOoslycqjZOOdfjpaXKxiSasy/axFCEEQaeCbgpevxmnbKzxIN+geNqIuE+n01y7dg2fz8e5c+fqqm6trNwDaAo80RdkoNVLvmzgd6o41fsPWALkO1GXSynqJn5N4V+f38e7k0nmMkV6m9wMtHnxaTKl5cVnAprMc6c7efX6HNOJHA6bwseOtPHqUJThaAphCuYX5omEI+TL4s7C1hKSBEGXnX1Nbj58OIIkwSdPtvPOxCJ34ukREpztDdIZcHIrnuMHQ3N84kQ7/8tH9jOxmMOhKrT6HQzPphhfzLM/3AJAyTBxayodfidixWJVv9NGWTdp9mnoAtLlyuLrjSxqBUgVTQZnVq9zGJxO8qGDEYJr6FzTNDnT5SfktvPOeALDNDnWGeBgxL3mZGAl6aLB7JKF4VXG5jNkSmLNfW4X1TjOpetbSqVSzb8/NTVFsVisddet+vcbTd1pOaLiq//E0WbSpcqCb6998zGYmxnDTrDTRZn1SKVSm74PWltb6erq4vr16xw+fJhXXnmFI0eOcOTIEV566SVeeOEFXnrpJT796U9v86gt3i9Y4t5iz7NetOVuCO4q9XTD3SyyLFMqlbhw4QIul2tZCk5P6G7yz1YW896rC65XU+gJuxiOLn8t3xVy49M29qdiZWV9Kbquc/PmTZLJJAMDA5tq7rJeVKppmvgdcq0B1EbwOVR6I+5Vx7sv4sZhU5jLFHHaFXqb3HQFnTjXsY+YpkmbR+Ezp9rJlkwUGYoGjMbSdwYNCDAE/OhGjN/88H4WskVURabJY+fx/U28MbrA//fqCF94Yh9f/NRR/vn6HMlcmWOdfp46HOH1kTiyLBFwVawtf/LDm/SFvdhVme9em+Fgs5e+sJt9TW6mEx4cqsL+iJuzPQEuTyaxKzKlO70ZPnAgjN+pYggH33x7ikJZpzvk4gMHwjTdw+JSxTAFJX214CvpJoZ5p4Usq6MgVcnkUNjBgXAbAlAl6hCOUu08Lv+0MfYYu91OOBwmHA4DyzudxmIxRkZGyGazDA4O1t34aLvYtLAWZu1tzlYXW++FnPtGVe4zmcyWJnl/9Ed/xC//8i9TKpXo6+vjz//8zzFNk89+9rN85Stfobu7m69//evbOGKL9xOWuLfYs2SzWYaGhlBVldOnT+NwOJb9fDd9r/dqmLUdVIVvPp/n2LFjqxqfbFc6z70q6w5V4mcfaUUwy+1YFkHFivLsI60bTsxZa/tCCObm5rh58ybd3d0cPnx4Sw3FtivdxaFKPHu0DVmeZSyWxWlXOdzq5RPH27g6neIf3ptBv+MhkSX4SH8LT+5fO4lJCIEqCfx3Ksj5srgjdO9SMgSTiTyaKvPkwTAXxxN84lg7f//uDFOJHG0BJ7/78lWePtLMr57fh8su47VLxHMGF8cXCHs0PjbQiiEE53qbeH1kAaddoTvkxKlKnOsN0t/sZF+wE1UGGcHTh4JEPHbaonkWcyVOdPlpDziYS5d4bSROtQlxPFNkLl3kc2e7cN/nqeC0SbT4Hcwmlzdca/U7cd7HHyOEQEIgAfWkb/odCl1B16p1EPubK29T7mfr2WnW6nT65ptv0tHRQSqVYmJigkwmg6Ioy+w8OxnHuReq5o1u5AWNFfebteUAnDx5kgsXLqz6/JVXXtnKsCweEixxb7GnkCRpWbRlf3//nmjSsVNvCYQQzMzMMDo6Snd3Ny6Xa0c7Gq5ly1lKs8fGL53pJFM0EELg1RScto0/nFfaZnK5HIODg9hstlWZ9ZthO8U9QMSt8pnTHaQKOvPzAfo6WkgWdF4djlE0TJQ7wssU8NNb8xxudrIwNYrH47lnVrpHU9jf4mUkmkYg1TrdHmz2MDGf5fOPdfHbHz5QycB3qHQFnUws5Hh3MsGl8QSaIvMr57owjYpNS5HgQLOXRL7EXKrIJ090cKTdTyxdwKepnNkXZKDFja7rFf+5qBS5zVySMy0aTx7eR9moWI5A4vL4JPY7MZpVJhdzRNNF9ofu3cPBa5d5ZqCVv708VYsL9TttPD3QsszGsZ1xsS4Vfv54O9+9GmV8IYMsSRxo9vLMQDOqJFYlvOwFqh1Ml76hKpfLNf9+NBpdFcfp8/m2zUKyF8Q9NDbrHxrrud/M20kLi+3AEvcWe4p0Os3bb7+949GW9bIT4j6dTjM4OIjb7a5ZcCYmJrZ1Hyu5ly2nilOVcG4y5aF6vUzTZHR0lGg0Sn9//7b1Hqing/FG0RSJiNtGfCrNtRkXuil4Y2QeVZFp8zvxagqmEEzPzTO34CPk85HL5Zibm6NQKKBpGn6/H5/PRyAQIGcqLORKnOgKcrDFx+BUgng8RnfQyal9IWYTWTpCbv7L62MYpsHxriD//eIksiThtqs8d7qLbKGELsCU/n/23jw6jvM89/zV0vvejQbQ2BeSALgv4qLVkmzZsi3Jku3Ejq3E13dmEsV2nMlMMvE5yTjJzSRRMnEmZ3yOnXtvYl3fyY0TL1Jsx7FMx9a+ktRCkSAIEvve3eh9q+6uqvmj2U2AAEmsRFPp5xz+we7G931V9VXV873f8z6viM0It2+v58x0gqfPzCIKAl67kcPtXj60p4k2txGztLyEQtd1MpkMXq8XiwgYBBJ5yBeLi4h9GWmlCFx7AaZpGtvqzHz29k6mL0Xvm1xm3Ms4qmzU/avrOg02iV881ES6oCEgYDeKGER9U2pOrBdXG5PBYLhhdpyaplXN83MrsVXSoGok99PT07jdbqxWK9FolGQySSAQ2JKchBo2FzVyX0NVwWazceTIkRU/bG6E/zxsLLkvFotcuHCBRCJBb2/vpkbqr8RGR76XQ7FY5NVXX6WxsZFjx45t6LW5UtMfz6kUVB2LQcRuklZE9ARBIJVXSSsaogAui4xREohpZn7y1iS3b/PjtBiYiefIFlS21VlJxSL4PU4a/V5cBm2RW04ul6skV4Zy8M9vTDIbz2G2mPE47XxgVxN3djqp9zqJpAu4LAb++a1JXrgQ5Hc/2Mf/89NBJqJZrEaJPc0uXhoK8wcP7mQ8kuXN8Sh7W110+h1846UxdB1kWcBukplP5fm3MzM8eqwV4yo4nN0o0uy2LrGlNMkijctYfS4HXdNwGcFdb6mcB32T55Wu6xhFMJrKB6tVZcQeluYbXA3L2XFqmkYmk1m3HWc1SGKqAasp5rWRWK8sZyNRno9/9md/xm/8xm+wY8cOfvd3f5eXX36Zxx57jMcee2zNtp01VCdqV7OGqoIkSauKIpS18JtdWbJcDXc9WCjBaW9vp7e394ZH1jaT3CuKUvGsv+WWWyqe9RsZbS8v5vIqvDUZ48RIlGSuQJ3dxJ07/Gz3WxEFgXReRRIFLIbFiwFBEBiNZPm3/jmmolmMskhPo5P39Pp5azKGqopMRzPc01PPP54YJ5nOEjGLNPvruLu3Ea9VRi3kF7VnsViwWCwUdJHvnJomK5ix2STy+TzTcyH+ORnn04ebmRsLUlfnx2h3cG46yaEOL+dmEiRzRQQgk1cpqKXiVCPhNBZJIKMUiGdVYtk0ZlnkI/ubK/Icp1lGQEdbLtP0GpAEnTt3+AmlchXdvFEWuae3AZ/l+raUC3Gt63qt7wRBWNb56N2ClZL75SCK4obYcVaLLGerUSwWt0yWU064rhYcP36cr371q/z0pz8lFotx/PhxHnjgAT7xiU/g9/u3eng1bCBq5L6Gmxo3wsUGSouI/JUeiKtAWYJjt9uvuTOx2TsR13KzWSt0XWdiYoKJiQm2bdtGMpnEarWSVDSmYzmimTwBt4UGhxHzCiwqrwVBEJAkiYuhNP96+rJn+ngkw/GzszgOtXJ2OsZwMI1RFjnU4aG3wY7xUonNSKbID9+aJpgsVUst5lXeHI8iiQINLjtT6TRDoRQ7G238T7e28s5MGq/dxL299XTVWZEErlrTNZ3XmIylEQQwGA0YjAbsgh0QyItGGj0ewuEQpjojFrlUgGs8kkEQwO8wUdR0XBYDJllkIpJlb7Odw111/ODtaQ61+9je6EAHHn/6HLoOFoPIgTYPe1rcYJKI51RUDQwSuExXX8Tpuk69VeRTR9qYiStk8kWa3Ra8VhlxgxNTryS4giCQKsBsQiGeLdDktuCzShiEKg3BrxEbLYlZix3nVjvVVItcaisj952dnTe83+VQnotut5vnn3+er33tazz22GOVxaPNdv1q5zXcXKiR+xpuamxERH0lWKssp1AocPHiRRKJxIrsH8sJrzdLxC2RSHDu3DlcLlfFs35kZIRIpshTb00xFi65m4gCHOrw8t5eP9ZVJOheCVEU0QWJU+PRJW4rXX47/3RigmhGqQSyx+bTfHhfE0faS9Vqy64wV2IolOKubg+vj8WJxeNEojE6An56Gh3c09tAo8NQkp5cMxoNRkkki4qAQF6DWLaArusYTFYUkwFfq7eUfLozwEsX59neYOf42TlEARpdZtDhYjDFJ4+0EXBZGJhL0uKxEnCaCLjr+f2nzpDLq5iNEoIA0Uye14YjdPttPH1mmmSugM9u4q4d9exqvPoLW9d1HAZw+s0LIuibH0WPKTpPvjHFeCQNgCwK3LrNz53dXgzCuyeKfyMkMdez40ylUrzxxhuL5Dw30o5zPbsXG4mtep6mUqmq09x/7nOf46tf/Sr5fJ5777234sV/vcrgNdx8qJH7GqoKq30ZbLZF5cJ+VkPudV1nenqa0dHRVUlwygmv1a5/XJg3sHPnzkV+zg6nk1PjsQqxh5LbzMmRCH0BJ9vqLMs1uSKIooiq6RSu8Fm3GiUMksg7UzFaPJfbL/UbpbfBfikxdvl2VVXDY5FotAK6E6vFjKbreG0mXGZpibRnObjMIjubXLwyFCavwXAojUES6GtycWpsniffnGFvi4tbu7y8b2cDmq7jtBi5p9fP2HyGDp+NqViGB/YGCDhN/KcfvI2GxHA4xYP7mrm7t54H9zXR5rUiiwIWo0Q4maN/JoFBFirONaGkwg/fnqLO3nXdeXu9Bct6sKRdUeL1kXCF2AMUNZ2XLgTpabDT4jRUTbR3vdgKYnulHWc8HufgwYOk0+ktsePc6p2DhdiKRUYymayKYmZlFItFHA4H3/jGN3A4HOi6TqFQ4IknntjqodWwCahuBlHDv0usRqN9o2Q5q+lnpRKc5XAjC3OtBbquMzc3x9DQ0FUXLYLBzOhceunfUoqQb/db10ziBEFA1Iv0NDoYDV/uwyRLJHIFnBbjEvl5Jl8kr5Y+9NuNeKxGopnLEislr+B1QD4e5D/e00ciD8lcEUkU8NqMmKQVEgNN445uH5Io8PJQBJtJ4rZtdTS7Lfzjq8PkCioz8RzH+0MkchoBp5lmr5UDbX1MRbNcCCbprGujy2/jKz8ZIJWHRpeRgNvKXELBIIoYJIG/eW6IWCZPo9PM+3Y2sL/VTfGKqsL5okb/TIKDdaZ1ycnWi4VzI1tQGZ1PLfmNpsNoOEOry73yeSGIpAtaaZEjV592v1p23xYS+TIKhUJFzrOZdpxbJYepFqy3iNVGI5FI8Fd/9Vd89KMfBUr3ps/nw+fzbfHIatgM1Mh9DTc1bqQs53r9rFaCc7V+qpXcZzIZ+vv7MZlMi6rnXglR17Be5TuXZX2kQRRFcrkce5ubmIll6Z9OUNR0ippGT6OT/uk4RXUxQax3mrCbSiTDY5H58N4Ax8/OMhvPkk4m2Nbg4MOHOgmOXSBX1PnZQJDRYBpV1/HZjbx/VyN9DSvTpFplnff3+Wnz2TFIAtMxhX9+a5J7djbhsRrRgbcnovx8YJa7exqIZgt8/dkh2uts9DY4uTCb5L5djciSSLPHAgh0+GyMXdLm/+j0DDPxHLIoMBPP8v03J/nf3t9DIpUil8thNBgRpRKp1BdUiy0Tza0kwbIoYDUu/8pxWFb2KhIEgXBW45WhIJPRDBajxOEOHz31VqQtLmS1ENUiSVkOBoNhEakrOz4lk0kikciydpwOh2PVi5VqitxvBapNlqNpGna7nb/4i7/gnnvuwW63Y7PZcLlcN9SxrYYbgxq5r+GmRjXIchZKcDo6OtblgnO9IlNbgbJnfTAYXFFRMSWb5nBHE8OhNIUFEWWP1cg2v31d0ouyPtxhEnlgb4Bbu+tIKUXcVgMui8xdO+p57nywUh3WbTFw1w5/JfouCLDdb8Gxz084kcZua6PRbcNiEEnaHDw3GGZo7nJ0eT6V519Pz1B/ewc+6wofl7pOKlfkH09MsLPJwaE2L88NBpmK5Qi4zOxqcrG72cWugItUvsifPrKHqViGE6NRzs8mOdLlQ9V06h1GQCCpqPQFHLw6HKHNV1pkZAsqZoNEs9vCRDSL3SCTz+dJp9JomobZbKTD04SipJHsXk5Pp0nni7T7bPhtMrJwmXxu1ny78jqbJIHDHV5GwqlF1Xvr7CbavdYVjSOZhyffmGQ6lq18NhHJ8MiBVvYEVtbGjUA1k/srsdDxaaEdZ1nOMz09TSqVWrUdZzVE7rdS5lVtkftisYimaXzzm9/kySefRBRFYrEYvb29PPnkk1Wz21TDxqBG7muoOlSjLOdq5L6cUOp0OlctwVkOKykytV6sxpFnfn6e8+fPEwgEOHr06Ir+Jp/P0+sx8vFDLbw6Mk8qV6TBaeb2bT789vXpqkVRpFAoacvNskCzywhc3iW4tdNDt9/GcCiF1SjTWWejznb5MZfNZjl37hyyLNPb27to90GTzYyE4kv6jGcLBJNKhdxfjbgJgkBBE0gqOi8PzXMhmOTRY+38wQ/OcjGYQhIFCqrGwGyCv3n0EAgCF4Ipzk4naHFbePRYO2en42TyRTr9diKJLKIIuYKALAg4zDJKoUhnnY28qqEUNDL5Iqqmc09fE5F0kdlEDqfZwK1dHnwmmEsa+NaLA8ynFAxGAzaLhffvbaOn0cH5uQReq5F2rxWHcXOI0MLzVCp+ZeFjB1t5bWSetFIk4LJwx/Y6PMsUwFqurel4bhGxB1A1nROj82zzWzBdmp5brd1fDVESRbFyT271uMsQRRGHw4HD4aC5uRkokcOyHefw8DDZbPaadpzVQBa3cgzVFrlvbGzk+PHjlf9rmoaiKJX3zVZfqxo2FjVyX8NNDVmWUZSl7icbjSvJfaFQ4MKFCySTyTVLcFbSz2ag7HV/rYe5oigMDAygqioHDhzAYll5EqwgCKCp7AzY6aizki9qpYRXcf2k63oLP5Ms0Oo20eYxA5f703WdsbExpqen6enpWVZnKghwtVMiideOwgqCQEzR+Wn/LL1NLt4Yj9IXcDIRyWA1SnisBrIFFVEQ+OThNkJJhWfOh3jyjclKisBzgyF+674d/PTMLJ+9s5NgQiGayWM1SkiigEESeWs8yvm5JOFUHnQwGkQ+tCfAbELh44fbKBQ1HGYZCcgUNJJ6irYGL6I5Ty6fp6AL/PeXh/nk4RaefmMYXZTpbHDxyaOd+Cyb/3IX0djVaKWrzkJe1bEaRGRBX1HEXRAEkrnCst9l8kUKqr6I3G9l5Hwl/YuiSCSrcTGUIJEr0u23EXAaMVapLagsy3g8nkU7d+XquolEgsnJSQqFAhaLBafTWVm0bCVUVd2y3QNVVauu8uvFixc5ffo0siwjSRJWq5WOjo6qKbZVw8ahRu5ruKlxo2Q5ZUKs6zpTU1OMjY3R0dFBX1/fhvtZ3yhyvxwWetZv3769sk2/lvZ1XcciC1jkjXu5Xm3sSUVD1XWcZgmRxYuIRCJBf38/Xq+Xo0ePXvVlL6s5+hqdvDYaW/R5vcOE325a9m/KKOgCPzk7y7mZOF1+B26rgYlIhklflpRSpKPOhkkWcVuNHGxzE0wqPD8YwiCJ5NVSpdVzs0levhjmY7e08MRLo4RTOQpFHZtJ4hcOtRJwmfnwviair44RyxRodJp5b1890XSeZwaCfGBXPX31Vi6Eszx7PshUNMdMPMuRTi/HtrmYTym4rGZSShGXzcqO1kaG5qIMzUT4+TsiB7wahUIel8tVsU1cz9y+GsHVNA2TyCUivvIqs5qm0ewpOQUVr7A9avFYsRlF0LVr9n2jcL3+BUFgOlnk2yfGK8ndL16AY1113LOj7qaxBTWZTPj9/koBJF3XK9V1g8FgJdJvt9u3xI5zq8h9tezAwOW5eObMGb761a9y/Phx4vE4LS0tnDlzht/6rd/iK1/5ypYuhGrYeNTIfQ1Vh9W8lG+ULEcQBFRV5fXXX98wCc5yuBHk/mp9xONxzp07h8fjqXjWrwWbWQX3ysh9pqDx5kSc0xNxCqpGp9/G7d0+vNZSovX4+Di5XI49e/Zct1BLJp3mWGcjmiBybjpBvqjR5Lbwvp0NeK1y5SVZHsPCeZpUNGQR7u5txG6W+fihFv7fn12gzm4kpRSYiaZp9dlpdJowGyTm03kKmo4siSBAUdVRVY14tkAkneeHb0/T5bfhsRo5MxVnLpHjjx/ejUEUeGhfEzajTFopcnY6xqmxKAfbvcQyecIZE0+9MUlKKVLUIJxU+O7JcX77/l3YLTrfOTVJLKPwwT0B9rR58TotGEQRj81Ae4ebTCJKIpFgbGyMdDpdkV2UCf9qK0FvNMGus8rcsd3PixdCFYLf6DJze3cdInplF+RK69KCJpDKqwiCgMt0eRGwWbjezlhRF3jxQmiRa5Ouw+sj8+xpdhGwr49kbRW5FAQBm82GzWZDlmVSqRRtbW2kUikSiQTj4+Ok02lkWcbhcOByuXA4HO9aO86t3rmAy+T+mWeeoampiSeeeILjx4/zp3/6p3zta1+rEPpqGGsNG4caua/hpsaNcMspS3ByuRz79+/f1CSpG6G5v5J8l48vlUqxa9eudR/fZpL7xW0LvDYS5ZmBYOX7cEohks7zQJ+XcCRKSnASFZ2QhFa5lIh7Nei6jkHP86FdDdzW7UPTwWYUsRquX9VXEgWSeY1/eXmUWCbPp4+2879/oIdsvsjjH93LXEIpFdFK5tB1nW6/nZJnp46ul4p8OSxG/A4T52eTyJKAzSRjNkjUO8zEMnnGIxlcFgPfem2c6fhi3Xm3306d3cxQKE1KKd0PBknA7zQTywi8eDGM12ZkOpZFlkTOTiXo8NnIFTSePj9Lm8+G22qkr96zyDkjn88Tj8eJx+NMTExQKBQqLioulwu73X5DyZMsaNze5aWnsSR5cphlWtyWZXMGyguxUEbjZwNBJiJpZFFgZ5OLO7b5sEqbR4CvF7nPFDTmErkln6uazkw8R7PTsa57qBr07uVIsCRJSxxZFtpxzszMkMvlsFgsixJ2N8qOcytqhlRT5L6MbDaLxWJhfn6eyclJACKRCOl0yVK4Gsdcw9pRI/c13NTYTFnOQglOZ2cn0Wh0090PJEmqJIxuFhbKZmZnZxkeHqa9vX3DJEY3KnIfzRY5PbFYQqOqKqfOT3DXNi8vTRcZnQ9Vvuvy23jkQDMu8/JR0XJSoySycmccSsc7HE7xb/1zlUJSf//aGA/ta+Y9PX6ePR9iLpFDlkRu7fbR4bMSz6l8cHeAH5+ZRRI1CqrOriYnt2+r43+8No6ug1LQeDMYxWYqJQabZImAy8Jt2+r4l9PTKAUNHZ29LS52NJR2BWYWJJvquo7LJGI22Iik8zS5zLR4LJgMEruanPzg7WnSisr4fBpBEPjW6+N8/GAzh9uclQWm0WhcIrsou6hMTU2RSqUQRXHZokibRRZkQaPRJtLkcFaKcF3ZVZlcp4vwg7enmIhcLqj2ylAYgPf31m1aBP965N4oCdjNMuHU0nwht3X9xbw0TdvySOy1FhhXs+NMJBJEIhFGR0dRVbVi1eh0Ote0kCwWi1uyyEmn01WjYy/Pg127dqHrOn19fTz99NM8+uijRCIRPvnJTy76XQ3vDtTIfQ1Vh2qQ5cTjcQYGBnA6nRWJyujo6KZreSVJIpdbGtHb6D5SqRTnz5/HbDZf07N+tdB0sLj9qIK8KQRv4cJB1XSyhdK11ymRzlQqxe6OJkYiCqPzi6Pbw6E052aS3NrlWXZcax2vouq8MxmnxWNF0zOklAIei5FGt5n/8vwwmbzK+Zk4BQ1+dm6Ox97TTaPTzH+8vYOjXV7OzSRoclnY1+JmPJKmxWOho85GZ7dr1wAAIABJREFUIlugqOkkc0U0TSebV/m/fzLAQ/ub+T8f2MXFYBKvzUjAbaHBbsQm63TX23luUKBwyetfKxYwyEZ2t3opFFTavRaUgsbeFheCIBBwW0hk87w5HiWYUDgxGqWnwY7tKm8GQRCw2+3Y7XaampqAEoG6siiSxWJB13VsNhvFYnFR9FSSJPKqhiyW5DFrnSPXWkDquo7BYCCaKTC5gNiXcW4mztFOL66NmfbLju1apNIiCxzrqmMymllUl6Gzzk6jw7Tu+0bX9aqJ3K8EC+04GxoagMV2nOWFpCAIFd99l8uFxWKpSjvOcp5BNaD8XPvwhz9c+ez3fu/3eOaZZzh27Bh9fX1AzS3n3YYaua/hpsZGy3Ly+TwXLlwgnU6zc+fORZH6slZ9M7d5N1tzr6pqJTq2e/fu63rWrwZzqQIvX5zn/EQQjzPJ7Tsa6am3YVhphdcVYCEBt5kkGt0Wzk9FicaiGAwG6uvrafO7GI8srZALpQq5xzqXP+a1kntdB03XsMgC3X4bBknkzu1+NCCTV0krKrmiRkEDURR4fjBEX5OT/pkERknAaZKJZ/LEswW66uz4HSYsBomf9s/ithho81m5p7cekywyFcvwlz8Z4J6een7hlhYCDhMGUbu0EwN+m8z7dwZ4bjBISilikET2tbvpaXLy/TcnQYdDHV7CqTzPng+hodPoMPGennr6p+Jk80WUoo5ZLrnYrOR8yLKM1+vF6/VeOh+lKOzIyAjJZJK33noLXdfxeDw4/M0MhnOMRbI4LQZuafcScMoIG7wILJPbQlG/smAxUMpx2EwVwvWCAJqmsd1v4ZeOtHNiJHLJ/tTBwVYXNgPrHls1yHI0TVtX0OB6dpxDQ0MVO85ydP/KvJCtShKtNo97QRBIp9O88MILvPrqq/zyL/8yn/3sZxkfHyeTyWC1Wrd6iDVsMGrkvoabGqJ4fT30SnClBGfnzp1LXs43O7kPh8MMDg4iyzLbtm3D4/FUjrEsSVkrYjmV752cZDaRIxbPMJ8tMpcs8rFDLexucmxYBH9h5N4k6hxttTE8PoPb5cZoNGI2SOxqdjIUSjMcTC35e6/dhCAsT57WSu7NssCuJhej4TQmSeDOHX4uhNIkc8VKDoCGgFES0PUS4c/mVWYTOYyiwE/6g1iNEreF0vziLS38znfe5hNH2vjVu7oRBFCKGtIl/finj3YwMJvgpaEwt2+vwyyL1Fkuz1MJjUOtTrbV28kWVQR0CkUNQZS5t6+RqWgGTdf54elpQikFpagyFk4zEc3yG/du5/RUjHOzCcbnM9zS4aHLZ1l15ddyFNbhcODxeAgEAqVFpaLynZMT9I+HK3KJk4NWfuXO7XR4jBt6X+m6TrFYxGsz4LEaFyWuArR5bZfyLzZHlrMSci3qGt1eEx2eJgqajlkW0DfI675ayP1Gj+FadpwL80KsVitOp5N8Pr9hu5KrQSqVqprIffk6PPHEE5w+fZrvfve7HD58mO7ubj7/+c/z+c9/nvvvv3/LHaZq2FjUyH0NVYcb/YApS3BcLtc1XWK22qZyrcjlcgwMDKDrOgcOHGBqagpN08gWdUbm07wzGccki+xvc9PiMiOv8n0sCAITkQyzlxIEBQHQoajpvD4aodtvwyxvzDUtE/Byca2Ojg7+1wcOMhXLkS9qtHgt1FkNOMwG3pmMkcxd3tVxmGUOtLqvSp7WSu41TWN3wEk4mSeWVRgKpXhhMMynjrbx8sUwRllkJpaj7pLcYlu9nYlIhhaPheFQGqtRwmGWyeZVCppOq9fK8bNz9DTYONzhY2AuyWsjEXRdJ5ktcP/uAA/ubULVdFRdRxQldARE4ZJURdfwmAWySfjeyTHG5uI4nQ4a3FYe3NfEU29M4bYYaXZrXJhLIYoCaaXIVCzLrV1eXrkYIpLOMxRK8siBFvYEbGuakwvJgizLzMeKzKRUPN4SMdNUjXw+z/PnZ5HajIRmpzcsWbd8HV1mkft3B3j6zAzRTB4BaPVaube3/tLOxJqaX1H/K3mO6bqOSMmfX9c2bjDVQO5vVNT8WnacoVAIRVGYm5tblKy7XpvX6yGZTFZN5L58L3znO9/hqaeewmQyLUpW3orFTw2bjxq5r+HfLcoSnEwms0SCsxy20qZyLdB1nfHxcSYnJ9mxY0fl5SeKIqLBxPODYV6+GK7IFt6ZjPHAvmYOtjpXTXLjC0i0IAgVU8KMopJXNcwb5HWvqiqRSIR8Ps/+/fsr28leq30ROW90GPilI228NhIhlFSosxs52uWj0XF1B4715AiYJY37+uqIKxr/eGKSFo8FTdf56MFmXh+NkFaKaLrOLe0e9rW4+f9eHeVIp4+/f3Wc3KUiX42uEvl/YF8T52eS7Gl14TQbGAqmqLObODcdRylqfPfUBH/w0G4ssohJlugPZskXdeK5Am6LgXaPGbNB4MdnZpiYT6EDqq4zHU0zEkojCWA1CDS7LdTZTSRzRURRwG01MB3JEEmXotyqpvPayDxddVYs67x8giCUCm8tgCiJmC1mCoJMfVMLHa3NpNNp4vH4dZN1K22IIhogslSDLwilSHhPvYXG2zqIZAoYJBGPRb4kfdk6t5yNRHn3cuHxVAO536oxLLTjVBQFq9WKz+er2HGWbV5lWV5SXXejrlkymaya6rTlY2pqaqK/v5/BwcHK2FKp1IZKM2uoHtTIfQ1Vh7U8YFfzMtV1ncnJScbHx+nq6lpWgrMcboSn/kaR+7Jnvdfr5dixY4siaJIkkVFFTo5GFumRC6rOK0Nhuv1WnKaVszld12nzWJEEAVXXgcskudljwW5cP7HXdZ2ZmRmGhoYwGo0cPHhwyTW7kqy1uE0E9gXIFTXMsoi0DM9YSOjXmwAsoeEwCjhNIu6AHU0rjemjB5p5cG8TRllEFgVCKYVfvaub//bSKG1eK+mCit9uYlu9A5tRZjSU4pGDzUTSCi9emCepqNTZjexr9RDL5jGIIrPxHMc6Pbw4GEKURP7pxDjhVB6jJHKk08cjB5spqEuj7ePzSdrr7MzEMxhEMBgFHCYTmYKKwyTz8kRk0e9TuVLl1/WSe13XCbjNXNrUWQS/3YzFICLoWiVZd6HGupysOzs7W7FMrG9opGBycWY6SSJXZHuD41Ii8OWKxOX5oWsaTiO4TMZL13dz9fZwY9xqVETC6SLTsTQOs0yjy4zj0qKlGsj9VnvMl8ewWjvOMtl3OBxrtuOspsh9+Rp87nOf42c/+xmjo6O88MIL/PVf/3VFngM1t5x3G2rkvoabHmUpy0q2gMuk1+12r7pQ042ohitJ0rpkOWXP+nQ6ze7du5fVfcqyTCRXRCku7SeazqMUdbh2QdYlqHcYOdbt49Wh+YqmvdFl5rZuH+I63xmZTIb+/n7MZjP79u1jaGhoxS8iSSx51V+JdF5jOq4wE89SZzfR4jFjNBrXLYkySQKHO30Ekwpf/fkQk7EsJllkV5OTe3vq2dnkotllxmSU+OL7tjMaTlPvNOMyy0xGMxRUjXv7GvnW6xNYjBLhlMLgXBKT7ELXdbJ5jYKoYzKIJHJFTEaJJ9+YqkTF86rGwFyC4/0SO+rtjMxGF41vIpLlvTsDRFI5LgYT6AiYDSJ399YTzeTIFRYvLOudZqyGtWnTFxFsXafBYWRPi4fTk5fH5LIYuHWbD4nlE1+XS9bN5/NMxIt88+f9RJNZQOfVAQP7O+t55EALDpO07GL/Rvp467q+qZKUoi7y8kiUFy8EK85IzR4LHz3QgtcsVAW53yqnmoW4ljToWnac8/PzjIyMoKrqouq6K5WKVVtCLcCtt96K3++nrq6Oc+fO8cADD/Doo49u9bBq2CTUyH0NNz3KjjnXepHk83kGBwfJZrNXJb3XQzXLcsqR7ZGRETo6Oq7rWW8zSphkcQnBd1uNGNfgbmOWBe7eUUdfwMngdASnRWZHkxf3VTzlVwJN0xgdHWV2dpa+vj48Hg+5XG7dBDylaPzr2VnOTsYrhLK9zsZ921zoavaafwvXjvBrmkZPvY03JmLU2Y0c6vBgN8mYZJG/f32ce3vqeXh/M994aZQTI1GMsoBRkri7x8/+NjdppchENEs8m+fcTJaHD7QwEUkzFc2QUlRkScBplmn3Wjk7FcNuNjJ+hdVjLq8SjOe4a7sfEBaR5lafDbdZ5JEDTcwk6ggmFbw2Az6bialYhrfHo5Sl3y6LgTu3+zGIGxPpNos6H9zVQF/AwYW5FB6bkZ0BBz7LypPiBUFANlk4NTmNbLbhN9vQdZ1CoUD/RJQDLU70SCmpUlVVwuHwmirrrhebLcsJZ4q8MBisVOkFmIpmeWUowod2+6ve5/5GYTPtOMv/lrPjTKVSlTaqBX/yJ3/C3r17+YVf+AXq6+u3ejg1bDJq5L6GqsNqX0plcm8yLQ03XynBaWxsXPNL70Yl1K62j1Qqxblz57BarRw5cuS6W8mSJGEmz6EOLy9fDFc+l0WBW7t9uC3ymqKcJlmgzWPCUpTJ5/PrIvaxWIxz587h9/s5duxYhSSsN+FYEARGIxnOTMYXfT4WTtNvlzjcbF5z25U+0Gl2mvn4oRZeGopwPpqku97G73+oD6/NWEpWDabIqypJRaPeYeZvnhviC/duo8Fh4o3xGMFknryqMzaf5rO3d10qOFXE7zBxW7ePWCaPKJTOidkgkitcPidmo4TNLNPgNNMXcDE4G8NsEOmqc3DfrgaMIlwIZTh+dhZd10kpRWRR4IN7mvn1e7ZzZjKOzSSzrd6O17x2qdJyf2eWNHY2WNnVaKvIR1bbfq6gVfICoHRNjUYjRqORnG7k2IEDhMNhxsbGiMfjjI+PUywW110QaTXYTGIriiKT0dQiYl/GRDRNJu+76Xzuq3UMK7XjNBqNOJ1OxsbG2LFjx7pkOR0dHTgcDiRJQpZlTp48SSQS4ROf+ASjo6N0dHTw7W9/e9VaeY/Hw9e//nV+9KMf8YEPfIDbbrsNv99fS6h9l6JG7mu46XE10h2LxRgYGMDj8axagrOafjYSq7H2VFWV4eFhwuEwfX19uN3uFfeRSsS5a3s3LR4L70zFMUoiB9rctHos65YvFItFFGVp5c2V/u3g4CDpdJq9e/dis9kWfb8RhbEuLmORCTASSnGwef1+zwI69S4L/9e/9BNK53FbDNzd4+cHb0/T7bcxm1BI51U66+xomk5R0+iosxFOKexrcdHisfD6SIQWj4XJaJaH9gX44K5GZEkkX1RxmmXGI1na66xous6hNi8vD80DYJAF6u1mdgac1FklHtxdR25nPUZzKcF3Np4lkzcxHcuRzBZKYhsdFE3j+cE5PnNbB/fuKElg1kK8l5yLZRbSa12cldsyGwTq7CaCycXF3gSgwWmqFLAym80VPXG5su7VknVdLteGJlRuZuRe13VspuWfZWaDhCxVjyxnq8ewGQuMa9lxPvfcc/zRH/0RwWCQU6dOMTQ0xJEjRzh48OCqvOSfeeYZ6urqKv9//PHHee9738uXvvQlHn/8cR5//HH+/M//fFXj/uIXv8gXv/hFfv7zn/P7v//7/M7v/A6f+9zn+PSnP11ZuNTw7kGN3NdQlVgNibuykNVGSHCu1k8+n6+ML5otViKIXpsRzxoj3mtBKBRieHiYjo5OWjq6Ma/Cv7K8O2A1COxpcrA74Likk994L/rVYG5ujosXL9Le3n5VWdFa2l6o+wbw2JaPVDmtBsQN8D3XgP7pOAZJxCiJ3NvXwDPnQ5wcjfLRg83IosDpyTiddVZa3FZ0nVJegi5gMkg8vL8ZpaAyHErz0L4m/vsr4zw3GMJskNAuHcOvv6eb0XCah/YFaHJbqHeauRhK4bMZ2d/q4linF1nQUZUMyBZ+fHqakXDJOUdDoN1ro6/ZzTMDIUwGEatBIpoukCtoaGvLIdw0CIJApgihVJ6UUqTRZeHYtjrGI2lSyuX7vq/JRYPTWHGOWTh/FlbWvTJZNx6PMzs7i6IomM3mRQWR1koMN5vct7otNLrMzMYvL3BEAQ53+LDIApEqINbvVnK/HMp2nH/4h38IwGOPPcbDDz9MJpPhH/7hH/jt3/5t7rjjDv7yL/9yTe1///vf59lnnwXgM5/5DHffffeqyf3MzAwDAwPkcjk+/vGP89RTT3Hq1CmeeOIJfvM3f5PHHntsTWOroTpRI/c13PQok3td15mYmGBiYoLu7m4aGho29AVbTqgVBIHh+Sw/enuaYLIUoW5wmvnw3gCdvvVHvq+FXC5XkeC09OzjxGSC4NAkDU4zh9o9BJzG6/a/MGm3/NuNHPJqCXj5mCRJ4vDhw9fcJl7Nok9RdeYSeS4Ek5gNEtvr7dTbDfQ1Ojg5EiGeLVR+azFI3NLhJa9krtHiylBQIZUr0Oq10Oqz0O238c9vTgIwMJPgI/ub8VgNCAhkCyoj4TSqpvOR/Wa+9N3TPHygmY8dasEsS7wxXvLqP9LpZWw+zXwqj6rpvDEe5X++o5PB2RQj4SQ9DTb2NDtxWoy4LAbemYxjN8u0urycuBBiOJxaMD6d7705xa/d1UUsmycRKeC0GOhrdGJYzlJoDRBFEZvNtiGFqWKKzg/eurw4MckiD+xt5pdv7eT0ZIxopkBvo71UT0Fc6pZzNSyXrJvNZiv+6ENDQ+i6XvFHd7lcWK3WFT1TNluWYzcJfOxgGy8NhZmMlGolHO700dNgRdO0qiDW1SAN2qrzUN557Onp4TOf+Qyw8uCJIAi8//3vRxAEfu3Xfo1f/dVfZW5ujkAgAEAgECAYDK5qPKlUiq985StMT0+jKAqf/vSneeGFF4AS6b/nnntq5P5dhhq5r+GmhyRJJBIJhoeHN0yCc7V+VFUlkSvykzOzFWIPMJfIcfzsLJ862oZ9GXeW9ULTNMbHx5mamqKnpwfN7OJ/vDpeqbw5Gk4zOJvk08faqLdfO/S6Fl3/arBScl/24Z+ammLHjh2LtqGvhivJvSAIJHIlL32bUcJiKMmaihq8Ohzl2fNB1Eva5JdMMh891MJ2v5VPHGnj1aEwwaSCy2rk1i4vdi1FLrt+NySTBO0+GyPhFFajjCSApukYJQFRFHh+MMivv6eb6XiOs9NxtjfYuWu7n9FQiqRS5MdnZuios1HUNM5OxzkxGsHvMNHb6KTBWfJqb/FacVoMvD0ZQ5ZEhkMp3FYjJ8einBqL0uI2Azq3d/sYmr8c3RUQUDUdp9nAcChNwGUmkS2QyBborreX8iT0y3NDEASyRUgXNAQEHCYRo6hflaiIokg0p3F+Ns5YMMe2Bicmp4BNXjm5EUWRhKJR1MBukjgxGl60OFGKGt9/a5L/5a5uPtBXmjNl/X4Za4mcC4KA1WrFarXS2NgIlCK/qVSKeDzOyMgImUymoq8uE/7lclw2K3KvaCIX5tKcnoxhM8oc7vRx9w4/JqlUu6B8DjRN29RK2jcLtmqBkUqllvjcr3Q+vPTSSzQ1NREMBrnvvvvo7e1d93iKxSIdHR388R//MRaLBbg8R10uF3/7t3+77j5qqC7U7v4aqhIrjdDm83mCwSCFQoEDBw5sasnvMrlP5zVmYktdVaZjWdJ5dcPJfTm5tK6ujmPHjiHLMsfPBSvEvoxoJs/pyTj39fmvee42owruattPJBL09/fj9Xo5evTokq1zQRDI5FUyhZJHvX0Ze8O8pnN2OsFrwxGSuQJ1DhN37fDT7bMwnynw0sVwhdgDpJQiz50P0eRqocVl5OH9TWQLKiaDiFEUmJycX/euiyiKCILA4Q4PQ8EkE9EMAtDmtRLLFjDJIs+cD4EOv3Z3N72NDmbjWSwGiR0BJy1eKxajhNUgEc9ptPtsFDSNeLZAKKXgshi4GEzS6rHw7RMTNHus9AQcJDIFmtwW/m3gPPFsgQanCYMI8UyeXFGDssO8IDCfzuEwy+xodGA1Suxr9bCt3k69fXFCuiAIzKRUfnxmhslIGlEQ6PI7uH93A17z8rkh4YzKP54YJ5RUSCaTvDEWZU9Hhof3NWFdwdumqIu8NZ7g5KXCX7dtr+fsTBIBYZFZZlHTGQ6labS7lp1rG0Wul/NHL+urY7HYVZN1N4Pca4LIi0MRXrxwOWp7ejLKfTsD3Nq5+DxUg1vOv2esp4hVU1MTAPX19TzyyCO8/vrrNDQ0MDMzQyAQYGZmZtVuN263my984QvAZalSeX5YrVbuuOOONY21hupFjdzXcFNioQTH4/FgNps3ldjDZXIvCgKyJFT8pSvfiyLiBrxQBaEUgVNVlcHBQTKZzKLkUl2HYGL5hNVgQrmuxGa9XvrXw7XIvaqqXLx4kVgsxq5du5Z1lNB0uBhM88KFMJG0gtNi4NZuHzsbHSxMLRgOZfnBW9MVAp/MFQkmFP7D7R1EM/klnu0A4WSOTF7DahCRRXAsKNYliuKa6xgIgkCyAMPhFJPRLAGXhUcOtTIbzxFMKXzh3u386zszTEazNDrNOC0G8kWdk2MRehqc/PD0NG9PxjBKIm0+G5+/uxujLBJM5PjE4VaeHQghoDMSStHittBdb+e/PD+M2SBxT6+fcFJB1eBDewJ879Qk5Vk4GUmxp9nLqyMRxEsE3yiJ5Isa9Q4TfruJ10bm+acT47R6bXzsQBPtbiPoGukC/Oj0NJPRzKXrojM4l0ASBT66P4AsLJ5ogihycmyeUHlH69LXF+aSjEWy7Gq0LjsvBEEgr4FShEg2z4mR+cqu2Hwyx3xaoajqXFlX7WpJpbC5mveyvrpc8Xk5u0RFUSgUCuTz+Q2rfhrPqZwanV/0mabD6yPz7Aw4cC5Qs1WDx/y/Z5TzN1aLdDqNpmk4HA7S6TTHjx/ny1/+Mg899BDf/OY3+dKXvsQ3v/lNPvKRj6y6bVVVEQShYtpQW/y9u1Ej9zVUJa714ClHsn0+H0ePHmV+fp5EIrHpYypr+11mia56O+dnkou+395gx7kO+8cyRFFkcnKSiYkJOjs7l1TQFQTo8Nk4P5tc8rcdddZKEalrtb8VspxQKMTg4CCtra3s2LHjqtd4Npnne6cmyV4i58lcke+/OYXtWDvdPgsOh4OiLnByLLIoMg+QVooMzibpaXQuqJh7GXazAcNVfPwlSbpu5F7XdcLhMLquL5JkJPPwvTemGJ2/JB8RBGymkk/8SxfCNDoMfOJwKwVVRxIFxuYzvDwU5v5djfzw9AxDoRQ2o4ym65hlkX95Z5oH9zbzzVfG+NTRNn7rvh1ouk6uoBLNFPjeqUmCSQWHSWZwNsXRLh9Pn5nh9RGJo10+JiNpdF1jLpFjf4fEg/ubefFCmIKqcVu3g0aXhWxe5as/v8BkNIssCShFnb9/dYxfubWdNpeBhKIyFV2agzASTpLMN+C5wnk2X9SZjl3x+0unejySYXdgsfMRlBYEY9E8L10ME0oppBWNQx1u3NYcg3NJpuNZDrV5+fnAHGaboXJ9vDYjbV7LVReRN7Jg1XJ2iWfPnsXlcpFOp5mZmUFRlEXVT9eSrJsraJV7YiHi2TxX1qOrkbetx1rkQHNzczzyyCNASUrzqU99ivvvv5/Dhw/zi7/4i/zd3/0dbW1tfOc731l127XF3r8v1Mh9DTcNFEVhcHAQRVHYs2dPJVIvy/KmW1TC5ci9SRb44K5GzLLEcKhE5rbVO3hPjx/TGgpALUQqlapofK/mWa/rOjubnJyZiTMVuSwPavFa6As4r0tsNluWU955KENRFAYGBtA0jUOHDl0zoiUIAmen40tIjKrphBIKsihyMWvBFE6TyhUrapOFSOdV3BaZnoCD/unLiz5RgFs6PLgWLMA0HaLZIvPpPOgWzNfQjmSzWfr7+5FlGVmWGRsbQ1VVGhoamCrauTAbwyDLIJSI7sBsjEaXhaNdPmZjGZ54aZTpaJYvfbCXU6MRpuJZ7trhxySL7GhwoOk6XqsRWRKIZYpMRjMEXGa+c3KCue1+Pn5LC3/7wjBT0ZL8S9N0ZEnEbpTo8Fl5aF8Tmq5T7zQzGUmXjlkyUFR1xudT7GiwYzFIdNfbKao6r49GiWUK1NlNNLjMpHJ5kma5ZMO5v3Ell3oRDJKAz2ZibD695Lt6p3nZeRlMFfn2yXHSShFVF7gQTHF6Mspjd29jOpZlJpalyWXh4QMtDAUTpHIFmlwW7uzx4zFfex5vJbkVBAG32115Rl0tWXch2b9esq7FIOEwyyRzi3eXfHYThit421Yn1N7IxVW1jWE9/XZ1dfH2228v+dzn8/Gzn/1sPcNibGyM/v5+bDZbZb41NjauWT5UQ3WjRu5rqHqUEy8nJyeXdcG50gpzs7DQ595rlfnIvgDxSy9at6WUOLlWqKrK0NAQkUgEh8NBd3f3NYtReSwSn7illYvBNNOxDM0eK91+24oKR90IWU7ZjnBqaorx8XG2b99ekTFcD2llMbH32ozc1l2HIJbI33Q8R/ziPDaTTEpRsRsXSGsE2NFgxyjBB3c3EnBZuDCXRJZFDrV52NFwOXqs6XBmOslP++eIZwtks1laPGZ+6TbnoqTkhYm/PT09uN3uimuSpmkUCgVefGuOVDJJoVgoeazb3bgsBqZjWbp8No521TEayWIzywyFUiiqhstsIFdQOT+bZCqWRRIFIuk8HT4bXruBDp+NW7t9oOvc3VOP9dJxSiJIYqny7JFOD/fvDvDK0DynJ2P0Bpz0BZwc6/IRSuZo99kYD8W4EL6cVPv8YIhPHW2joKq0+axomk48k0cpqjS7IZkrUFB1nCaJZo+1Isspo7POgcMowhW2oQI6hzs9DM4lSCnFikY+4LKwzb9UkiNJEu9MR0lfsrOURSrn7I3xKC1eKwMzCU5PRPncvTs42u6iqOlYDSIGUb/mHN7qyPWV5Hq1ybpl/f7CZ4DbLHL7Nj8/7Z+t7FgZJZE7t9fjMolLNPdbSe63un/Y+iJa1bBzUr4PTp48yde//nUmJibIZDIoisL4+Dif//zn+fKXv7zl56qGjUeN3NdiBgQiAAAgAElEQVRQlSg/GKPRKAMDAxUJznIOEGWLys3GlUWsZBF8K8kSvA6CwSAXLlygpaWFo0eP8s4776xoJ8Jtljjc7oL2UrLfatxIyu1vRFGo5drP5/OcOnWK+kAzXXtuIavqzGeKeCxyydP9KtB1nb6AkzfHouhAwGVme6OTZwdDvHgxjNtqYH+TnR1uC5IkEEwqxDJ5BMAgiRzt9tHkKu0MuMwS79nu5UiHB0kUMMmLjzWSKfKTs7OLIqHDcwmeGwzx8P4mDGJpJ+Xs2bO43e5K4u/CuSaKIlarlfZ6NxcvudIUVZiI5RgKpWhym3nqrSmimQL3727klnaZH5+Z4d7eBtp9Vt6eiHG4w8vQiXGMkkhc1XhnKs6DewMYRIHt9Q68dgPd9XbS+SKfubWDp8/OMhnN0u238eC+Zr57aoIXBsMUNR2HxcCP35nhU0fauGObjydPjjEajHOwM4DLZkTXYS6eYzau0OSyEExcJv2OS5Kldp8Nkyxg0nU+vLeJp8/MMLEgofa+nfUYRH2J9EvXdQIOA48e6+DUWJSJoEhng4sj3X6cxuXnWXyBO5Gu69TZTWTzKqlcEYPHgtkgcecOPw6jgFRZTGjXzSvZanK/kv6vlawbjUYZGxujWCxit9sr0f1DrS6a3BbOTieQRYFdTU4a7IYlC52tJtdb3T9sHblXFKVqqr6Wcy++973vIYoix48fX/Z3NWL/7kON3NdQlVAUhf7+fhRFWbZS6ULcKFnORpOFbDbLwMAAoigukqusJrK+FmIuCALexmaG5rPMxrO0eG1YDRJGScBlllmic1lhm1B6oY6NjZFKpdh/5DZeHk1w+u0RlKKG02zgju11HGp3Y7jGe7/NY+FYt49ToxF2NDp56s1JUoqKUtQIJRV+8FYMh82M3Sizr8VFp99OVinisxvx2QwYr9hCsRgWF7EqI5hUFkschNKRj89nSOSKxGbGCIVC7N6zl5xgJJhWMck6TtPi9lVVZVeTk7fGI0TSBfKaRjCp4LYa2Rlw8V9fGKbeaeb5wRC/clsHgiDQ5rViMoi8Mhzmw3ubePRYOydHI2QLKm1eKw/tb+bvXxnlJ/1BvnBPN69cnGdgNsGeZhd3bqvDZTXS7DbTP5MklFS4f3cjXX47BklgLpbh1FiEB/eWfLFv72nmxeEo52cTGGWJQ+1ebt1WRzyT5z099Tw/GMJhNtDqsdDhtXKwzY1+af4F7BKfvKWFdEHDJEtIAhglrlr0TNc0GmwiH95dz/SsjsNqxm5c/NtyUp8gCPQ2Ojg9Ea18ZxSh3WflfX0NtHktvKfHj8csrbq42M1A7pfDtZJ1JycnSafPYzKZ2F9fj9VqxWQoIrKUmG01ua6GSPBWjSGZTC5rFLAVKM/B8o43QKFQqLjlVMPuQg2bgxq5r6EqkUqlqK+vp76+fkXFaG5E5H6joGkaY2NjzMzMLOvvvlEJr6oGCUVFFEqyoTLBSuc1XhpLM5cd50i3j389PcPFYIpmj5U9zU6OdnpxmlZGDHQE5pJ5zs8mSWZyNNsF6t1eGotFLswrnBiJVH6byBU4fnaWZo+FVrfpqm1aDAL39vo52O5hdD6DySAhCiXnmDJFPDUW5cG9TQyFUtza5cXmW70zhfGKqr7CJQG/rhUJzs5iBPYdOsKrI1FOjEyRUoq4LAaOdXk50OJAXrAI8llEfuloO2+MxxkJZzjY7mVXk5Mn35gkpRRpFqG30YFJEvjivdv4xosjPHKohYIKj/94gINtHu7tbcBiLC3sHCaZY10+7t8TYDiU5rWReRxmA98+OclQKM1sIsfv3t/LZDTDf7itg4vBNBeDSTxWI4c7fRSKKqKgc3RbPV975gJD4XJuRpFnzwcJuM10+608sLeJ9/U1EkzmaHCa6PJZcBgu+4Pruo5Z0inqIm9PxBgMJjFKIofaPWzzW5GvQrp1TSWdiGGUPItcrBRNYDSSo38mgcsis7/Vw4E2L29PRNB0QNDZ0+Rib7MDu6G8KFi9hGyryf1GkevlknULhQKJRKJiJLBcsu5Wk/ut7h9q5B4WX4dvfetbjI6Osn//fkym0vP3rrvuqsyrGt5dqJH7GqoSdXV1FAqF6/+QpXKZakZZZuT3+5f1d4eNOZ5gqsCLF8OMhTMYZZG9LW4OtbuwGkTGolleuxjk/Yd6eGMsxjPnS77Z8WyBWEYhrRT50O6GJRHwKyEIAoNzab57cpzp0DyqquLzeHnvzgCd7jpeG1vqYFTUdM7PJWnzLJ9gWYZJEmiwG5hLiNgMEgVRx242kMyV5kShqCGJAke7vDjMckXjvxr47UYCLjMz8ZI0Rdd1cjmFLo+BjkAdVquFszMpnh0IVmh8PFvgeP8sDQ4T7Z7L1YA1TcNikPDZTcyn84xHSrkQDx9oxmk2sL3BwXxa4ZXhedp9Ng53ejFJAp11Nl4bifDM+RCvDM1jNkp89EAzqqbzn58f4fZtXsYjWRpdFpK5IqqmU9B0YpkCAzMJjnX7+M/PDfHSxXnsZhmHSebN8Qj/x/29mCQBg1iqF2AxSCW9ulHGazPyb/1BxF0NfOPFs/hsRvY0u/H21BFNF7C4jITTKtOxNA6zTL3TwksXQ7w+8v+z957RcZ33ue9vl+m9YGZQBp0o7L2ItihZllxjObKtyLETx/eu5Nr3JOc6zfngdbPiD8dZ6ctZzjknx8s3sRMrTuLETbItW0cSJYvqFDuI3oEBZjC973Y/DDEESJAEJUKkZTzfOJyZ/e55N/Z+3v/7PM//cgzjWDzHB3Y2sz/qWreptaoL/HQgzmuTlxd8Z2fSPHKwnX1tPhazZYJOCyG3GZlaHwOL9MbM328FuRdEkUxZRzMMbLK4qonURh7fZDIRCAQIBAL1Yy2bdRcXFxkdHaVYLNZ26Pz+dZl1bzU0TfuFJvcbHcu8XizLWBsbG3nf+95HtVrlxIkTVKtVpqamaG1tpbm5+bYvhjdx67FJ7jfxc4+fh5tStVplaGiIcrl8Q5nRmyX3uYrG90/NMZW8bIT86YUYmmFwb0+A0cVawk/IbeFfXpliOU0yW1YJ63BhLsvR7iANjuvfHvIVjSfOTDExO4/b5cZmtyEgcGIsQee+RkShuubnpJuYr0aPFYdFplBRaQ/YiecqJAyVfe1+drd4MBB4fjSJKEB70EGDw3RdTf9KeG0yD+5p5qmLi4zFUsh6hft3RHlgdwc2s4hmGJyaTl8lUtJ0gzOzGToC4cveBUnmucE4//3pEXQEcmWVXFlhKlnkv9yzhS985zQht52yqnFyMo1JEmk+2k6L38anjrTx6kQKURQ40uWnJ+TiwlyGtqCdbFlD0WqRlh6bCatJwmGW6Gpw0OKzUahoLBUUHBYZvR4LKnB+LsuOiANFUfBZoMHtxDBqfyvjiQLaisXQUqHKM0OL7Ih6kASBsaUCzw8v1vs4BFx2dkfd2ExSPcVIN+CViSV6w05ucJnUkSiqnFohwQFIFxWevBDjV/Y1EfW4qOowuFDg9ekUug5bm9zsaqk1aFJ1A6dZwtBv/Lex0UkpVV3g5fE0r08lKVZVQm4b7+oN0e63YOj6W0qW1jLrnjp1ilAoRKlUWpdZ91bjTsjZv10LjDupcr8cPXzs2DGOHj2KzWbDbDZfdW3+PDxDN3Fz2CT3m9jETWA5IWW9D43lxJjJyUk6OzuJRCLrMtq9GXKfyFeZTl6dT35mOs2eqAef3YzJVBNUxLIVFK1WbXRaZSqqjq4bV+XHX4lischMIsdsIkuoIbTq96goOplimd1RL+Px/CpybDVJ9EZc6yZfAbuJ922P8OSFBdJllbaAnYOtTh452MJ0ssTjp+frhNNulvjgrib6I06SRYViVcNulvDZTMhrTJdhGITsIkcbBQ5EQjgdbWiFzKoOw9eaqitfThY1nrq4gEEtNcZllTFJArIocmomTdhjx22VUfIaTovM8GKOmVSRF0YSSILIXd0BnBaZpXyFn1yI8e7+MC6LzGy6yHu3NfI/jg8TclnY3uymrGjEc2VkSeTZoTiyKLC92U2pquGxm7DKEsmCQl6BBpcFj1WmemnAVc2gqur0N7qvSsERqAlgjg/G6/MjIDAWz6PpOh1BJxfmM/X358tqbQFwjafIlTr7WqOtq+c9WahQVAxMksDLE2meHIjVFiLUkpJyZZ3JpTwVVaU94ORwpx+v5cZG8I0iLKIocm42x1MDsfq1PbVU4D9PzvAbRzvwWW6/LMUwDLxe76qEquuZdT0eDw6H45aN+U6o3N+uBUY+n79jyP2TTz7JJz7xCf75n/+Z06dP43K5EEURs9lMoVDgD//wDzdlOW9TbJL7TdyRuFMrCcvm3fU8uHK5HAMDA7hcrmsm/ayFN6u5Lyn6mpbYiqqh6ga9EScNPi+zqRItPjsD87Wuo1aTRDxfocVnQ7pG+XulX6Br6y6ikQYWViSuQC2mUUCnL+Lk3v4wJyeS5CoqQaeFu3saCLvWnyQhCjWtuttmZi5TwiJLeMwGumbw9OAiJVWrZ90Xqxr/+2JNYvTY6bk6uT/Q4edoVwCrvPqcYrEYo6OjdHZ20h4Jkc1mmc5mgGjt2MDuqJfB+SwrOakkCuy8VE0WRZF81aCqGWTLGgYCogCqZiCKAg0uC3PpEmZJYD5TI+S6YWA1yyxkK2wJu/j6iUlOz6bpjbhJ5Mp8ZF8Uj91EVdVxWUzkKwr/5zu6WMpXSBWqlFSNB7ZF+PH5GIc7A/zH6zM4LTJ7ol6KVY2zsTRWWeDZ4TidQTsfP9LJD84ukCxUscgCe9u8HO4M8o0T4/VzCrutBJ1mZlLlVaTZoNZUazyRpze8WmrQ4LJeMitfm2Sv/Dv2O8yIAlzJ7z02MxZZIFPWeXUiWU/CiQYclFWDfzgxTG/EhUWCWKZMLFPmY/ubsV+Ht21k5byoGJyeTl111tmywsRSgUCL67bLHNZaXFzPrDs9PU2hUEAUxXpl3+12v6Euq8vffbsr96qq3jZyf6dkx7/zne/E4XCwe/duGhoa0HWdUqmEoiikUilsNtvtHuImNgib5H4Tbxu8FQ/U5SjE621pq6rK6OgoqVSK/v7+VVF36z3Gev0Ga6HBZa5LWVaiyWfHbZExywIf3tvCyeksH93XzL+/YiBLAsl8LeHlga0RUsUqIadpFdHLZrPMLSSwuv1072zBYzNxoMPPD8/MrSJsW5vcmNQSVlngWLefnc1uVB0ssrDK2LsSqg65iook1hJ7lt+jGfDieJKnBxYwyzXdeDmf4dP37eBiLE++rOKyyjgtMoIgMJEoUqxqFKu1xVGxqvHsYJxmr52+sB2AcrnMwMAAkiRx4MCBemzdWtdOR8DOu/rDvDyWJFtW8NnNHO700+iqzf9kqsoTF+bZ3x6gu8HJxVgOu1lC0w2ShSq5skJfxM1/vDZNWTHoCbuYShbx2U10h5w0e21YZJHZVAmrSeLjB6K0Bex89/VZ5tJlzs2lOTuT5tPv6ODDu5sxgOeG45ycSqFoOlVF59iWBl6fTqNoBmdmMkTcZnZFvXzzxCjNXhu/+Y42Ht4VIq+CzWLGZjHx7FAc6RL5a/HZeHB3M/FsmajfzmsTiVW/gVkWCLoc6CvmzWmROdbTgE0WVsiBrg3DMGhwmuhv9HB+7nL13yKLHO0OYpMFCopBrnL5um/02PjJhQUUrSbJWW4QN7mUJ56r0r7C87DW8TbqXmAYBuo1znl5Fwxub4FiPTsH1zPrZrNZ5ufnKZfL2O32enXf5XKtizDf7p2L5TH8ohtqd+zYAcDRo0dv80g28VZjk9xv4m2B5cScjdSRwo0lM8uZ9dFolJ6enjf0gJckiXK5fOM3XgN+u4n7+kM8eWGhTnLDbiv39jbUO1km0lmsJgm7WeK/vrubsXiBiqoT9duQBDg/m6U3VPMFqKrK2NgY2Dy8uiQzObyAphtEA7W0lYcPtPLaVApF1emNuNjR7GHg1CvQ3YEg1Bp+LWMtMraQV3huKM5MqoRFFtnd6mNP1INVFkiVVF4aW0IzoKRoCMCR3iZm0iUWMhUkSUCsaownCsiiiMsqYzFJhNxWppJFrLKIAZyeSdMXtjM9Pc3U1BQ9PT1XNdVSRRNmfyOxnILLKuEwiVhlgXd01ZJvVN2oxYVaajn+2arID87MEs9VEEjywZ2NTKeKnJ7JYJJEAk4zd28JEfXbCDqtTKeKiCIUKxphlwUBeHZokf5GN+/b0UihonFuNkMsW+Zod5CdLV6mkkE6Ghx0Bhz80wvjHOsL88S5GJmSgiQJOC0yR7oCvHNLkJlUiZ6wiy0hJydG4mxp9GGWRM7P57mvy0U+s8hCNosoihxoirL/l/ooKAaqZiBg0OS1IgkCYbd11W6MJBg8sDVc8z+YJWxmmY6gEwSD87EiQZeFgO3GcZUW0eD9O2qRnYOxLB6bmcNdAfw2GVEEmwkanFbmM7VkH1kSyZVVTJKIacVOkkFtIQjX3gFaTvvZCNjNIltCbubSpVWvW2SRjoBjQxvE3QzeyL3nWmbdTCbDwsICIyMjALhcrjrht9lsVx3rTonCvB1587lcjqamprf8uGtheZFrGEb9ubVyXu7UHfJNvHlskvtN3JG42ZvOMum+XeS+VCoxMDCALMvs37+/HjV2K4+x7s8LsCfqodVvZz5TxmoSCbut+GyXU2UibivPjScYT5ZYyldp8loxDChU3DQ4LbQFasR++YHe3dfPMxMlhhYL9eOMxwt89/VZfvVQK1sa7GgG2EziqgfJjZApa3zn5OwqohQ7Ow/AXZ0+yoq+Kos+5LaSrhqcHk5wpMvPmZkM52ZrqTx2s8TRLUG+d2qWbU0eXhhbIuCw4LXLCIbO8PAwmqZx5MgRJElatdCYzVT5wetznJ+IEQj4afU7eN/OCA1OM+mSCgb47DJmUUDTNARBYKmgEM9VAJhJFTk5ucRn7+kiX1apagZOi0y5qvIfr83w2Xu6GIhlEQWBXzvSxtZGN1VVx26WmFwq8s2XptANg9HFPHtafUQ8VkqKRn+jm2JVZSCW4139EUySwN42Hz+9sEBR0ZhLl/mbnw7xR+/rI+g08+LYEmG3hf4mDxdjOQIOEx1hLz6vC4+7Vk3UNI1cLkc2PY+kKHCJiKWKEoFAgIf2NHNiLMlMsoTFJHKwI8DWiAMJnXu3+CmqAt8/PcdgLINuXKq+b2ngSLsPWbhMbNeqntslg/1RF/uiHtJljVi2zEi8wFK+SkfQzof2tPAfr02RyFXIlxU6gg4ssnQpuak2XxZZJOyyXFdzv5GGWkPXOdDuZTFfYWAuAwK4LBL39oUvNbW7M8j9rcBKs25jY61vQv36yWYZHR2lVCphNptXyXnuhMr97VpgFAqFOyYtZ/nvTxCEdctCN/H2wOZsb+KOxc10T32rsu6vJN66rjMxMUEsFqO3t7de8XozuBU595IAIaeJkPPyYmflb+mzwNEtDSiDCcqKzkyqxNZGNztbvFycz7IlZOf1119HEAT2799PTpUYW0xcdZyZZJF0UaHZY8bEzZOqhVzlqgqoAbw+lWJ7kwurLK6SGLlsMhOJIseH4/y/H9xGwGlB0Q1kUeBAu582v52/+ukgug4dQQcD81msopUur0SksZmCbuK50TRVVacn4qLRbUHRdH50dp6pZBHJbCFb0Tg9m6GqG+yNenl6cBHDMOhocHJfXwiftfbAtJkkeiJu8mWF+UyJoYUcQws5ukJu3r01wp987yw7Wrx4HWb+6idDfGh3Iw/vixLLVvhfx0dJFhUOtPs43BnghdElZtIlrLLIe7ZF+PfXpnlwVzPffHGK0USeYlWlL+Jmf5uPd/eHcZhlTk6lCDgtHGz3o+kGugFRf81D8dMLC5RVHbdVZmopT8i1hRZ3bXEnSRJerxev11u/LsrlMplMhvn5eSRJYl9zIyGnTL6iMpnIYZEFehrsyEJNJjWwwlhbUXWODy7S1eCk2bW27GoldF0nXdU5N5fl1YkUzwzGaz4Ek8QD2yI8vL+NkYXagm1Pq58fnJ5lKV/bSZBFgbu3hC7tBl2fRG9UVVI1RCaTRZp9NjqCDkCgLWAj5JDgDqnabySuvH6gZtbNZDJ1s26pVMJqrcXd3mqz7npxO6Mw7xTN/Uq8+OKLWK1Wdu/eTaVSK0q8mSLUJu5sbJL7Tbwt8FaR+5XHSSaTXLx4kXA4zOHDh2/Zw+tmOtS+ESRLKi9O5nE6nXxoTxOFS2ZXsyRQUXUe2hUmNnaRlpaWeoOtrKqs0lyvxLUk1+vRPS/Lhq5ERdFRNAO/XeZgh7+eNZ+vqPgdZvx2CycnksiSyN5WL2ZZYi5d4uJclhafHVXTsZlEHJLBB3Y3s6szwHC8yHdPjtfTdU6MJrivL0xnyMFMqohqCExnqpTUKgGnhSfOx9gScpIrK4iCwPnZDBVV42N7mykotcjIpYJCyGWhr9HDaxNLZEoKHpuMzybxe+/pq6fDzCSLvKs3xOvTab70w4FLshIYmM9xfi7LQ/ua+d6pOd7V28BUssjuFi+KatDb6MRuEXl9Ks1irswTF2K0Bx0IgsED2yL0N7o4P5tGEmBXi4e2gJ1/e3WaB7aFKZRVlgoVNFXjueE4H93ThCxcPVmCIGCz2bDZbEQiEZJlg394fpylbBGlWqVarfK0oPPpu3tpdJu5OJu8am41vbbr0OL23ZDcS5LEWDxLoaLx1CUDNEBZ0Xh1Iklf2Mk9W/z1XaZfP9zKdKpEoaLRFrATdMhINyD2G6W5F0WRczN5fnB6Bt0AkyRgkkSCLgu/si+K/Rf0iWqxWOpNBwHGx8fr98Nls64kSfXKvsfj2XBiuUnua9B1nUcffZQnnniCixcv8sorr3Dx4kWeeOIJPv/5z9/u4W1ig/ALeivaxNsNy0bXt+I4lUqFs2fPUq1W2b17N3a7/ZYfY6OacuUqGt89NcfJwXmcLidOm5WdLV6CTgsnJ5MsJrO0BRy8s38rwRVVf5dFpjVgZ3ghv+r7Ih4bHtvVD9DlXZcbEaxGjxWrSaKsrD7fFr8Nt1VGFOBIp5+Ix8rp6TSGIbCjxc1UqsTIYp6dUR//+MIkfruZZp+NRK6CzSxyrDeI0yywv93H7qiPsmrw7GC8TuyhRkifH43TGqzNXzxfJZkvY7XWxjSVLFJSNEqKjsNcO8eZpSJLJZXvvDrNQrZMvqrz/EiCqL9mSo2lixzrCaLpOiMLOV6bSmGTZT55uA2TLPLMYLxmHBYE7BYZu7kWWzmWqPkDYtkK/U0uAg4LTw4ssJgt09ng5L/c2833T8+RLdUSWcpVjRafyJmZNE0eG4+dmeOlsSU+si/KlpCLiaUiYbeVD+9pplSpoCOh6mCSr78bJooiU8kcW8Jutrf4qKga8+kSsXSJgUSFjgY3aCqJRK2hldlkwmw2YzKbsJtXP06uN//5isbwYv6q14tVjdFEnkPtHkyXFiJuM2yP2OsxtOuRvWwUuc8rBq9OJusLWkUzUDSNqaUiC7kKnf7ry4V+UWAYBg6HY1X37ZVm3bm5OarVKjabrS7nWa9Zd724XeT+TorCBFhaWuIrX/kKf/u3f8tv//ZvAxCJRPjGN77B5z//+due7LSJjcEmud/EHYubleVsdJdawzDIZrOk02n6+voIh8P1m6KiQ6qoUKhqOMwSfvva2errwUaS+4Vclcl4AQQBgVp6S4vfzumpFOVKhXTFIDVfYrE0y8cPRvFYaw9HqyzwwLYIqjbP5FIBw4BGr4337YisSrdZhiiK69LdBuwm7u1r4OmBRcpqjbQ1emy8c0tD/fezygL9YQd9YQcgcnxglvv7Qgwt5mn123hkf5TpZJGKqmORBXY2uShXdX5yPsG+dh92k0S6rJLIV646fqFSM+k2+eycnbvcUVcQYEeLh/FEgWxZxWGR4NI5n5vJMJ8pIQjgtIh0h5xouoEoCHx4dwu6ofPMUIIXxpYYWcwDAk6rxDt7GkgWq1hlibbApeQeVb9U/RVoCzjIVzQCDgtfe26MTFklkatwcirNvjYfv7ynmR+fj7GtyUN7wE6hrJKrqDw/miBVVLinN8T3T8/xg9PzWE21cV2cz/LhPU28NBYnW1bZE/UiCgZem4xdvlpGJUoSGiInxpJMJAo0uCy8qy+Mx26momjYbBbu2d5KvFzrhVCtKlSrVayqjlcqc/r0GC6XC4/Hc83dJ8MwCDrNeGxX+2McFgmXxYQsCrBibDfbgXijCIuqGZSqaxcRarsxmzIHWDvn/laZdW9mDL/oUZhQSzlzu90cPHiwHm2azWZvOsVtEz9f2CT3m3hbYKNlOblcjgsXLiCKIq2trfVOkAAlxeC5kSVem0hSUq6frb4ebJQsRxAEciUFg5p2+ciWMIoGX3tulHi+yuHOIMf6G3llPMlcusRUssjOZnedVIWdJh7e30K2omIYNROhewWxV/Ral1sB8PmD6zoHWYQDbT46gk7mMyXsZpmI21JfVKwaPwA6rV4zz4xksMkyiqbx4O4miorGbKpEg8uC3SxxfDDOlrCTzgYH8XwFu0XGZTVRuYLgmyQBi0nkAzsbWciU+dlgLbFnS9jF7hYPXz8xgdN6mYQ2++zMpi83fzIMA1kAWRIYXSxg6AaCAE8OxFF0HZ/dxM4WH26rRNBpoSfsIl9RmU2VSBaq6IZBLFPiw3uaOdjmY6lQ5ZWJFDazjNUskyxUkY1ax9kP7Wqiu8F5KTO/iEkSaXBZuDCfxWM3UahqjMYLBF1mJEEgXVKYXirgd5jpbHDyLy9PcWo6xd5WL4lchYf2tqBoOvlK7Zr1WiWWsgr//uo0zw7FAZhKFrkwl+V3H+ihscGJWYTuBhu/tKuFVyaWyJdVQu4A9/SGaHHLFDw2stksi4uLJBIJstksPp+vTtasViu6rtPitVJS3KuSeWxmiUa3lf1tXkSM66Tn3xgbRe6dZpFmr51kYXX3ZbMs0uix3RFV+zthDOuJoVyvWddisYXviOwAACAASURBVNTlPDfTWXeT3NfgcDjYuXMn/+2//TdSqRRDQ0P867/+K9u3bwfemgjpTbz12CT3m3hbYKNkOaqqMjIyQiaTob+/n2KxSLF4mdwJgsBYIs/zw/E6GVnOVm/x2ekN3bxkZ6Mq94ZhEPJYMUkCvU1eppcKPD+6RCynUFF1fnIhxlK+wuFOP6+MJ1nKV6/6DodZxLEiXm6ZSCSKKk9fXGQiUUAUBMI2nWAT+NeRRCeLEHGZaHRfO7d8JUxqkY/uayZbrXVcfX4kzuB8hpJqEM8rBB1mPnmkDcMwODeb5tXxJJ+6q4397T5+cj62yiOwvdmLxypjlUUeORjlrk4vZrOZUlXFItcIuUkWwajFie5q8WAYBsOxFVV+BAqKjoHB0EKWqN9BqljFZzfxgX1RXptI0d/oZjJR4EO7GollSpyaSgM1Qvuu3hDz6RJL+Qr72nyki1VEQcBultgT9ZIsVtF0A5Mk8snDrZyaSuN3mplMllgqKOxs9rK92cNsqkRF0Wj22pBFkYuxHBaTRL6iYr8kKzo1nebe3hCabvDMUILBWIZ8WcEii+xp9dPe4CJZqBD125lNldANg7KqMTCf5Z6eLnRdQwb2tjjpCdlRdLCbRMxirbK+TNYikQiGYRAOhwHIZDIsLi5SKpWw2WwEg0G6A2F+7/4enrq4SKpYpcVn5+4tAaJeC4bx5ha3G0VYJMHgHVuCLGTLLOZqixKTJHBsSwi/TeJOSMq5E8jaG03LWcusWy6XV3XW1TQNh8NRl/Ncy6y7mZZTQyQS4bd+67f44z/+Y5qamvjMZz5DW1sbf//3fw9w21ONNrEx2CT3m7hjcTMPKFmWKZVKN37jOmEYBouLi4yMjNDa2kpvby+CIFCpVFYtInTD4PRs5qoqowGcncnQH3HedBX+VqTlXAsNdhN7mp0omsZ/nophCBJuKxSqtYr76Zk07+oPYZFFogH7ush2UdF5/Ow8oyv0+KMzCSw2Ox/e04Jpnc+O9RwrWVIZzklkE3H6Gr1UVY3XR2OkFInZdJlMqcoFA5xWGY/NxN5WH8cHF1jIVtjX6sVlNXFyOoWmGvQ1udjR5MEi1eRfzR4LwzOLXIgVKFRUehpd/P4DPWiGcSkK04TXJrMr6uHMTIp08dLiR6jJajqCDp65GMNpkTnQ5qXBbePvnh5hV9TH2dkM33xxkv/1a/v4yN4WdrV4WcxV6GxwIgnw2Nk5DncEyFdUdAPOzWUwSwKtAQcemwm/3USz18Y/vzhJT8RNxG1D0XSCTgv5sspPL8Toa3STLinYzRImk4DdLKJoAkGXhfFEbW6CTiuSKOC2mfn746N0NjgwibXEm5fGkyCIeKwmVN1gR4uHYlXDKovIooimGRQMcJlrkqvL5lGda03dsonS5/PV57guxZgaxSLLfLTfgyF6MUsCThugazVd1JvARlWvDcMg4qx5KGbSJQoVlahvtcn3raqcC6JEVdNrPQBWLIbebjGUVqsVq9VaN+sud9bNZDJMT0+Tz+eRZfkqs+7t+h3uhO68K1Eul2lqauJb3/oWExMT+P1+3G73HbHDs4mNwya538TbArdSllMsFhkYGMDr9dK/+wAaApmKjtcqXVVVFxBWNdhZCZO0fs/ASoiiuCE33kKhwMDAADsam1HNPp4aSqEhEHFbayZNRUPTDcpVjT2tXhrd62s9nymrTKzIv4fa73JxLkOmL0LQcWtuM/GCyr+9Ms3gdAwMGJxeJBQMIlvtZLKrzZkTS0UOtPlYzFXwOyxoulHT4ze76A3Xuq3aTKuz7i0SdFhLHD3Sg6IbOEwidvPq9xiGQdhp4hOHWnl9KsNCrkSj107AYeHEyGItNSae50N7Wnj8TIxcWaXFa+XCfA4d+McXJvHZTVQ1HafFxH+8Ns34UoGqavBrh9t4bSLFkU4/A/NZYtkysUyZYlXjM8e6OD+f4dGXp3FaZf7u43uQBJFHX5qmrGpsb/bQ3eAk6rORyFcJuixYZAlJ1NnR5OF/Hh8lU1LY3uylxW/nB6dmMVheQBvLJ8fEUoGOBieD5+aRRAGzJJItG9zjtjKZLHJiZJFtzR72t3mxide/Rte6hgVBwOFwYHc4CRlgEqFarZLL5cikMkxPZN5wV9S1jrUR0HUdlwm2hmxrmnw3unIuiiLxosZrk3HmMyWCTiv723xEnLVrVdf1n9vK/XqwsrPuMpbNuplMpm7WLZfLTE9Pb4hZ91q4E3ZNlrE8ltHRUf7sz/4Mm81GZ2cn5XIZm83Gtm3b2Lt3b10StYm3FzbJ/SbeFrgVshxd1xkfH2dhYYFtO3YykTV44pVZkoUqbpuJQ51++oK2K6rqBntafVyYy9ZjD6Gmad8V9d4R1RFd1xkbGyMej9Pf34/X6yWWSLG90clsTgFgS8hJtqRgMUn0N7oJu8zYTet7SKkater2CgiCQFXTb+H5C7w2mWI2ladULCHJMjZHCIdVJl/RsJgkSoqGLIloqk7AaaaoaAgCeGwmgk4zgiCSKass5SuYZAm/vSYpWTnmdDpN/wq9/1rjNwwDvwXe0x+4lEAj8sJYqt7QqqLqDC3kqKga+9p8JPJVmr02VN3g9HSa/+fdW3hyYIEfn4thkSU0w8DvqJlIY9kyharGrx9pYyFXIVdSaPLZ8VhlfnwuhkkSub8vzMsTKc5MpzkzmyFdrPLccILffGcHv3G0nalkEVUzONjhx2U18czFBTx2M16biWM9QRRVx2KSCDgsmMTLC1ADA5dFpjvk4rmhRQpVjbKhs6fVx742H08NzLGQrbCQLVNRdN7dF7hhrvuVRMcQRKbSFV6dTFGsqHSHXexoqlX216zuLywwPDyMIAh1su92u7FardclUW8FybqWyfdWHFsQBIpqLd0Kav6WZQN0sqTzLy9NsnRJ9z+RKDC0kOWTh9posIsb2p13vVjLULuRWMus+/LLL2Myma4y6y5fQ2/GrHsj3AkEf/k6nJiY4Pz58+zYsQNFUTh9+jRjY2M4HA4cDgd/8Rd/wa5du273cDdxi7FJ7jdxx+JmZTlvRsqytLTE4OAgkUiEw4cPM5tV+P6pCRStRl4S+Qo/PhfDta/5qgpQq9/K+3Y08tLYEuliFZ/dzJGuIC3e9VW+NxKpVIqBgQEikQiHDh2qP3ANpcyxngA/HkiSLFSxyiLegJ0HtkZo91tZ3owwEMiU1VozGpvMWo9rl1Ui5LKwmFthVhUEmn22WsrMLUBZ1RieWyKRWMJisWC1WVkqVNnW4iXgMAMG6WIVh1nGYYYjHQFOjMZ5Z3eQJq+NXEXjhbEF4rkKLX4bc6kimgG/tLPxUlfRm0tngkvVSUBTdbY2uhlayDMaz1FWDU5PZ9gSdvG9U3NMLBX46L4oHQE7I/E82ZLCQ3tb8NnNdATtdDY4aXTbyJSqRP12Tk2neWl8iY6gA92Anw0neGhvC7FsiU8eauP+rWH+v+fHsJgkwm4rubJKoazy6EtT/N4DvcynyvzKwSjz6RKyLLKvzc/uVh87W7wIQFXVubcvRCxdQlnxNyMKsK/NS3eDnT/76C6mU0V8dguyCC+NxVnIXp7fszNpDrT58NxEOIwoigzGS3z7tWmql5KRRuN5JhMFHtzdiPXSTsBaRktVVesxirFYrF7dv1aM4u2soL7ZyrkgCCwUNB4/O89MsoABtPjsvH9HE81uE+fnU3Viv4xsSeHkdJr39QfvCFnO7ZamCIKAKIo0NjZeZdbNZDJv2qx7LbzVi5r1YGRkhE9/+tP1GEyAr371qzidTvL5PN/4xjf4q7/6q9s4wk1sBDbJ/SbeFnijspxKpcLg4CCKotQz6wVB4MxMok7sl6FdqrweDK4m7WZR4GCbh56wE0UzMEvCmmkvbyUURWFwcJByubxmFr9hGLgp8WtH2phNl6iqOi0+Ow0OU53YZys6L44luTifxTAMusIu3tEdwHvFuXmtEu/Z3sjjZ+bqKSItfgf390dWVcbfKPL5PLNz83itMqFQiHw+X6uaAmem03zycCsji3lOTqVxWWtNrwzD4P861k3UZ+XkVJr/PDnHK5NJKqqO32Hmt+7uJJEr8dMLixzrCWKSxEtdT28Cgki+qiOJAh6ryEN7m5jPVBhaLOC2mnDbTHzn9VlsJon/fG2Gjx9qZSlfRZYEZMHg/zjawcmpFN8/Vcuvv6c3xAd3NtETdrKYq3JqOkWmqHCkO8hCrsy7t0aIpUuMJfK8MJrEbZMJOCz0hJ0k8lVEATxWExaziKbpXJjP8TdPDtIVdPHIoVb+7ZVp5jNlmrxWjvUE+eDuZl4aSRDPl3BbTRzsCNDVYEcyNFrdEu1eDwsFjf/x9PBVnhJV15FlAQ0BSQCBq6vYV/67pBqcGE3Uif0yBmNZFnMB2rzXNlTLsozf78fv99e/e7m6H4vFrqruK4pyc3N5C/FmK+dlFX58LsbU0mWp23SyyI/OzfOpw20sXkoXuhJL+QqacWdo7u+EMVyJ65l1k8kkExMT6Lq+LrPutZDP5+8oMy3UngUnT55kamoKq9WKw+Hg5Zdf5ujRo6iqWv+b2sTbC5vkfhNvC9wsuTcMg+npaaanp+nu7iYUCq2qtl1J7Ouv6wbaNaQIV5Le2wHDMIjFYoyNjdHR0UFjY+OaVURRFCkUCjQ1yQQd7vpnl6Hq8PRgnNcmkvXXlsaWyJcVPry76aqIzy0NNj51VxuJfBVJFBCrRXyWN2cKXikn2rZtG/eELMy9NEUhT10mbpIEQi4zveEw+9t8TCaLZEoKQacFHVjIKaSLCrqu10llqlDlsdNzfGBnI0+ci+G1mzg7k+ZYbwivv9Z0RwdKVR2zLK5pCM4pIi8MJRiL5zBJIrtavOxq8dDTYLvU2CpOSdX5lf1RilWVilbT+Hc3OJlK5emLeHj05Sn+8cQkmq5jkkT2t/v50bl5FrNlFvMVwi4rv3qwjWxJ4amhRXJlhR+di9EVdtHitzGRqGn1O4J2UoWakTZfUTjaFUSWRM7MZLCYas3HTk+n+e6pWQIOM06LxE/Pz3OgI8DDB6LohoGOQLGi8tpUlmavDb9DxiEYWCQBr91Mqni5UuyxmTjWG+bVyQxTS3m8NjP72n00uUyrjJ2wevetqhlkS1eTbgNIFau0+9bfAOp61f1MJkMqlSKTyeB0OvF4PHg8HpxO58+F7rqgGEwlC1e9PpMskK9qtAednJlJX/X/rX4HknDnEOvbKU1Z7+7JWmbdfD5PNpu9yqy7TPiv11k3m83eMQ2slq+BX/7lX2ZycpIvfvGL9Pf3c/r0aQDuuecevv3tb9PS0nI7h7mJDcImud/EHYubeTjcTHxkNputG2YPHTqELF/dWXNbk4dTU6lVsYkAO6NeyksT6x7XG8WyUe9mHtLLRmCLxcKBAwcwm6+dQ7ncZArW1pWnSyoDc5mrXh9eyJEpqVhdV29f+2wyPlvttxwenqEqe696z3qRTqcZGBggFArV5URu4FcPtfLsgEyqWKGrMcD+Ni9+u4lEQeFfX5km4DQjyxJfe36cxWyFnS0eBATu3x7h4mKOdEHBapIYTxRxWU2YZRHdMMiVVX5yPsav7A6wkFd4eTzJXKqE65LXot1vQ7p0ORZVg8fOzjOymKuPdy5doqIZHOvyEnKaMQCnWeZvfjrE7lYvB9v9bGlwIkkCu6IeFrIVnh9dIuK20Bly8o7uALIo8o8nJuhvdBPPljk9nWYmVeLjB6J8eFcz3zs1R2/ExWQiz319If7h+QkKVZWyorOjxcMD28I4LTJffW6M372/hzOzabobnBzpCvDPL9a6+Gq6gaobgMCrE0vsifqQRPiXl6d5bSpNvqzS1eDgg7ua2BpxE3ZJ3L81wuNn5yhUaovnw10NvDCWJJEt1dPoL8YyfPxQG1G36ZoE3SaLNLisV2XES6JAyGW9ZqrUeuVSK6v71WqVcDiMyWQim80yPz9PLperXUcrUlWWm/rcStwKcr3WrU8QBBRNZ0vIQUfQWU9AAmjy2tjZ7EbX9TuG3N9OvNG0npXXxzJWmnVnZ2epVqurDN8rF413WndaXdfp6uriS1/6EsePH2dmZoaHH36YPXv20NLSwh/8wR/c7iFuYoOwSe438bbASrJ6LaiqyvDwMNlslq1bt173Jhz1Wbm7N8RLo0uUFK2WA97moyfk5NRk/pqfu1VYb4dXqN3AJycnmZ+fp6+vb13brDdqlKUbBpVLlW6Dmkm0otYkKGX1xtGe65mPtbDcVyCbzbJjx46rtrijXgv3b3Gj6tAYCiLUyWWOTKnKjhYvX3l6hNl0CVU3KFY1zs5kiXisHGjzc2J0iaqmY5EEJEHg14+0Y5FFDncFWMyUKekyT7w6zeKyvjxdYjye5+OHWukM2ABIFhVGVxD7ZZydTrI36kEUBXa3+GhwW3hXX4iJRJ5YtsJSoUK6pDKbLLIj6kXXDD5+qJVMSWUmVaakqPzGXe2XOukaJItVhmK11JyB+SwmSSBXVjg+lOBoV4A/fE8v5+YyRNxWDnYEmEsV+Z8vjDK0kCNfVom4LJyZzfBLuxqpqDplRcNiEmvdX6k1gK2oOq9PpTg9k8FrN/GRvS0sFaqcm83S4LZiNVvoC9uIeDqZTZdwWGoG5mViLyCAUNvleHFsicZdjUiX5uRKQm4Saxnxc+kiufLlXbZ9bX6CDhNXZsSrhkiqpJEt1wztfptUj5u8EZalMcumwbWq+/Pz81Qqlbp2fzmZ580S4zdbuXdaRDqDLoYWsqtebw84cZpFrBJ8ZG8z40tFZlJFIh4bXUEHbjP1tJxNcn/rojjXMusWi8W6/yObzfLcc89x/vx52tvb6+95I9eApmns37+f5uZmHnvsMcbHx3nkkUdIJpPs3buXf/qnf7pu0eZKLF8HLpeLo0eP1v0GX/jCF/id3/kd9u/ff0cl/Gzi1mGT3G/ibYEbJWcsLCwwOjpKW1sbfX19N7yZWWWBu7sDbG9yk6+oOCwyPruM+Rqxl7cayzsRV+4qXIlMJsOFCxcIBoMcPnx43Q/1G2Xp280SjR4r06kSyYLCTLpIVdWJ+mzMZ8o4zNJ1Iy4vRwReG4IgUFF1ZFFAFGqm5osXLxKNRut9BdaCUimjKAoCgfpri9la5OVsukS6pFyqTkOurNLktfLqZIqP7mvh1ckUuYrKJw614rWbOD4cZyFdpqiovLO7Abul1s21O+QiV1YYjxfIV1ReHk/S6mtGFqFc1dGNq6ur3WEPz4+lePSlCd6xJUSuYuf8XAa/3czOqJcmrw1btsxcukTQaebjh6I8dTHOM0Nxdrd4uBjLcdwe5w/f00uhopIra+TLCiG3lWavFUU3ODuTJl1UeOLCAien0/RH3Oxs8ZLKV/nx+XmKiobVJDG0kOORg22M/WiAkYUcXQ0OXi1U6Qg6sZpEMHR8djNOq8zIYh5ZErivP8yjL00xnaw1aTs9k+ahvc3c3xvAZzHwR+yIosjTwzWplqILFCoqVVXHYZVJFhSqqoFtxWWxcg4NwyDqMfOpIx0MxHLkygq7oz7cNhlNN7CaL+++lXWR40MJTk4mqWo1edT+dj93dwexiDcm+NciLGtp94vFYj1CMZ/P16u3K5N5bgZvlixZRIP3bg8jCjCeyGNQI/YPbAtjlWrf75BhZ6ODXU3OOqFfXkttkvuNbWC1HOe6ctHY29vLiRMneOyxxzh58iS7d++mpaWFw4cPc+jQIe6///51XRNf/vKX6e/vJ5utLez+6I/+iN/93d/lkUce4TOf+Qxf+9rX+OxnP7vusY6Pj/PMM88wOztLPp/nq1/9Kn19fTz88MN0dXXVz2cTbz9skvtN3LG4FTedYrHIhQsX1iVVuRKyCCGniZDzzSUovBHcSGa0vAuRy+XWrHDfCDeqrLssEvdtDfPvr85weiaDoukEHGY+uKuJM9Mp5tJFHtzVWJeqQG2+smWNsqphcvowidcef7qscXYmw2i8gN0s0t9gwW/S2LbnAHaLGVG8thRjLZlGW9DOdLKIWRZZ2VFJ1WqxmAGHBZsscrDdz942HzujXv76J4O8OpGkyWvDKoscH4pjNUuMLOY4O5OhO+TiaHeAM9NpsiWFqqYjiyI+uwmXRSZfvVx9dlpkPHYz3zgxQbPPQcBh4Us/HCBgN/Obxzr5xgsTxHMV/HYLFpPAh/c00dfo5ms/G0cWBXIVlZDbQkXVeX5kiSavlbl0Eb/dQlnRUATY3erjI/ta+M/XZxhZzGM3SRzs8FOqqDw/moBLPRccFplsWaXFb+evH97F2dkM9/aFaQ84SBWqYOh47Sbeu70Rp1nEZpbY2ezlxMhSndgDCMCzQ3H2Rr2E7EKdRDZ7bZQ1g/FEoS7VEQToj7iRZRG49rwbuk7AJnB3l5eyJnBuLssT59PohkF/o4ddLR503WCppLBUqKBfmsuqqnNiJEGr30Ffg/WGMp31EuyVRK2pqal2zagqmUyGbDbL3NwclUoFh8OxKnf/euT5zZJrwzDwW0Ue2tNIvlo7T6f5chfglce51vF/0QnbW92d1u128973vpdKpUJbWxtf/OIXmZmZ4aWXXuJnP/sZ999//w2/Y2Zmhscff5wvfOEL/PVf/zWGYfDUU0/x6KOPAvCpT32KP/mTP1k3uf/c5z7H8PAw3d3dWK1WHnzwQR5//HG++c1v1ncYNvH2xSa538TbEivNmH19ffUM7Z8XXI/cLy4uMjw8vO5diGt9//XIvWEYdAVsPLw/Sm/EhaaD32FiPJ5nJlUiX1HJltW6xl4zatKY54YSpApVZKOW8nLQD+YreE6+ovO9U3OMLuYplooUC0VO+Xy8b2cLs7NxqprOoY4AXUE78hoc6cqFiWEYdDc4OTuTJug016IvF3I1wm0zM58u8eCuJqJ+O5mSwtRSgZKicWE+i0kSKSkaEbeViqrzo7Pz7G2tNb9azFWI58p8YGcjml7TjFdUA1kSuLsvzE8vLKCqGgYGnQ1OkgUFkyyyO+rjmaFFKqrOPX0hHjs9z/m5DIIg0Oy1cW9fiLMzWVLFKs0+Oz1hF6puYJJEhhdzzGdKbGty0x5w8N5tESYSOebTJRpcVvIVhd++dwtmWaSsaNhMMsWqiqLqeG0mAg4zTV47fruJLz1+gYjLwrv6Q/SEXfRHnKSK1fpCzWsVEYCDHX5m0mWeHY7Xf1Ofw1wzRgsQy5QIOxz1XPdmr42oz8652cuejJDLSnfIQTxfpdklr5kBLwgCOgKKZmCRRJ4fTfDs0GLt/xAYiRe5GMvR7LPx1MAiW8IujnQ38MJoHFWrfde52Qxbw/Yb+mveTPVcluU1ZRjXqu4vd0S9Fcde+R0mAXz1r712F+C1Pns7K/dvRI53q/FWk/tl5HI53G43giAQjUaJRqN89KMfXddnP/e5z/Hnf/7n5HI1yd/S0hJer7e+e9vS0sLs7Oy6x5LJZJicnGTfvn089NBD7N69+01HfW7i5web5H4Tdyxu9gG5LAVJpVIMDg7S2Ni4Ktv9VmKjdYprke9yuczAwACiKLJ///7rpjbcCDeS5SxD1XUm4gUEAc7NVOsGY1kUV51/LFvluydn6zr9fKHE905OEfK66AmtjuGcz5YZms+Qz+cRRBG3z8/gYgHOzrG/vdYQbDJR5OEDUfrCjqtI4lodfN0WkV/e08JCtsz/fW83xwcXmUuXKSsaH9zZxPZmD69MJnn05Smifju7ZZF8WcVlNRF0WrCYJGZSJYIOEy6rieW+rcOLeWRR5GC7l2RJ5cfnY4wu5OgMOnjv9gi6AY1uCzazTKJQZWfGi0kSmErW4kXbg3a+f3oOm1lCEgTe3R/mJxdizKTKfOquNl4aWyLgstDV4CSeq7Al5GJLyEl32EnUb8dqkkgWyrQHnYCBIIh84TtnqagadrNMqarx+w/08jv3baGq6mTKKnPpEm6biS9+aDsjC1lcVokfnp3hgf4IrZ7aw90wDAxdxwB2t7jxOyycDLsxyxJOi4zTKmORRcySgMt65WNCZ1+blxafjdF4Hr/DTFeDk/OzacCgxe2rz8/yNWIgMp2u8vJEknxF4UhXA6+MJxFqIZqoBkwli4zF8/zX+3pI5CtcmM/ygZ2NdIdcXJyvyRSs64xWvZV/n2tV99fqiLqs3b+d+e5w+2U5tzvjHm4vufd4PDf9uccee4xQKMS+fft45plngGt3eF4vvvKVr3D8+HG+853v1BtVjYyMEI/HNyv3vwDYJPebeNtAFEVOnz6NYRjs2bMHm822IcdZrx7+zWAl+V4Z29nT00NDQ8N1P6voNdNnoaLitMj47aarKuBryXLyVZ1sWUMAEGpddtsDdoIuM8MLq03E25rdeK0yNcIpcHY2g90s0xOxYZJEUjkzw/MZTk6l6G6ws9KqMLeUI1uqUNBNaJqAraTissosFaqYpdpAFU3nxfEl2v02LFfEbl5Lz++2iHhCDnqAnU0uUiUVRdNxW02ki1V6Qi4+e6yLs7Npmr02zHJtkeC2msiWFNKlKvvbveiGzpawi4qqY5FF3DYTkijw+mSai/NZCvkCzy/GGZx18KH9Xbw6mWZ0MUeuqqFqBvvafPRFXEwkCljlGqnPlNRL+fUVTk2lsZgkNN3gob3NPDmwSKao0OS1kipUiPpDfOV/DzO0mKev0cXv399LtqQwuVTCJIl88UPbOTGa4IfnYoiXutqWFY3nhuOcmckwvlQkV1a4qyvI3T0NyIJOWdVRDchWDFwWEVYk15tFiLgt/NKuRv7u6RHi+QrZskLYbaU14sJlt6AhIl9qMmUCRhfzpIoVQi4zFUXj6YsxNN3gSFewPkdurxckE6IkMZYs8W8vT1FSavPW1+jl3HyONn9td6aiGuTKCoZRM3OH3TYS+SqvjCf5xOE2Ls5nMUu1XZE7oTK8lsmyUCiQzWZZXFwkl8tRKBTquv0rq/sbiTuB3N9uzf/taiaVz+eJRqM3/bnnn3+e73//+/zwtxE1RQAAIABJREFUhz+sZ+9/7nOfI51Oo6oqsiwzMzNTX1yuBw6Hg/e///28//3vZ2BggG9961vs3buXL3/5y9x111185jOfue3ztImNwya538QdjfXE4BmGwdTUFLlcjt7e3jd0c70ZLHfD3Uhyv7yAyOVyXLhwAa/Xy+HDh29YjSoqBs8NJ3htIklZ1bGZJA52+jnaFViVTX/lzkCioPLyRBKf3cKzw3HOzdYqvh/Y2cj9WxtpC2Y5MZzALItsa/ZwqMPPSoLotsn0Nbl5cTRJqlih2Wvl2NZmyqqBbtS6nxYKBSYmJvB7W5nJKPXUnWJVw2M3cV9fiLJyeUzFskpVM9Yk91deEzo1XbZFlhAwcJhFHJf8FZmyzgtjSZ4fSeBzmOgMOukIOvjcfVs4PhSnqGhUVY0jnUH6Iy7OzeVwWWW0kkK2rFBWNL7+wiSNHittbonTKQWX00VPk5cfnpljIVvCZjZRUiFTUvjJ+Rgf29/CbKpISdFoD9a6zvaGXYwl8mgGmCSRx87McayngS0hFwu5Mh1BB1tCLmZSRaaSJXpCLtwWEy+OLpEtK/z43AKxbO19H9vfwsOXDMJ39zQwvJhnMJYjlq0wmyqhaDo/OD1Hm99G1G/DQOLfXp1BRKcr5OKuTj8eS+13TJU0BhcLDC3k+OThNs7OZilWVBq9Nt7RHeS/PzXM7lY/O5vdZEpV2gN27u4J8vUT48RXdCUOuay0+W3oBsznVJ6bVlBmFmgOOOlucHH/tiZmU0WGFrJUVY2g00yhquG1Sei6hmFA2G1F0w1MErQHHRQqKhZZpNln466uBprcJgzjjRtqNwqCIOB0OnE6nVitVpLJJG1tbasiFBVFWZXM43Q6N4Rc3W5yfbuq5leOYSPvz9fCG43C/NM//VP+9E//FIBnnnmGv/zLv+Sb3/wmH/vYx/j2t7/NI488wte//nUefPDBm/re5ftkf38/X/ziF/n/2XvvILvu+8rzc+O79+Xc/Tp3A2jkRBKBmRQVqWTLsk1Lsle21zOumdqd2a3aMLXeP+avLdfW7Hhqa2rHM1PWOq1nZI3GtnKgJEoMIAmCCAQaQOfc/bpfzjfuH7f7AU00SAAECUh6p4pVrEa/e+9Lfc/v/M73HIBTp07x1a9+9Zd+LuMXHR1y38HPNUqlEmNjY8RiMZLJ5B1tid4uJEnCsqz3VYkTBIH5+XkMw2Dfvn1bcpff6TFT61Venlhv0+6GafPTK2v0xfzsvs4ec/3OgO3AS5PrdIV1/vrULKem8zRMT8E/M1fkf3t2Lw8ORjnYE8HFJarLXE8dXNclpKn8X9+/Ss3wjnl5qcTYUpE/+vRBJMFlenqG5eVl9h88zFjO4uEdSX58xfNbK5KIZbs8tjPFq1Pr7eP2xHQCbzfs8/aMflgsewpvvmbQE9V4dGcC1xVoWTYRXeGFq1lem84zk6sxm4OrK1UqTYujAxGeOz5ArWUhiQLdEY2/fGkCGwnH9Ro/P3u4l5n1KjPZMi9cWuJ3n9hJ0RQpVFv0JcP8/fksvTE/+bo3cDu+WqXWMvn1h/r4P3/9MG/OFvniyUEGEwEsxyEV0rzdlIDKRLZGvmYymPDzxK4UluPwz//Tm/y3j4/wJ88d4dJSGb8qEQ+oOK7LixM54rZKttLkZ+Pr/MZD/SSCKi9eXWOx2KRhOZQbJn6fRLXpYjuejefiUoWrq1WiuoTruqxXW5TqBr96JIMmi+RqTa6sVAhpCiulJiFNxicL2K5DrtYirKt85eVp/tHjwywV6rxwJcvnHxrgiyeGeHFinWrLpDfq5+RIgqgmslA2+etTM6zmStiyzrcuZgnrCr9yuIeZXIPPHh1gvdriN44N8O0LSwCoskjUr/CJgxnmclUEXOJ+mQcGohzoCXOkL0RQufV41XsZ77d57pup+5tkf9O7v0n2360g6VbhOM49IbbXn/9eK8L36hruds79H//xH/Pcc8/xR3/0Rxw9epTf//3fv63Hb/cdOHnyJCdPnrxbl9jBfYoOue/g5xKmaTI+Pk61WmX//v2EQiEuXbp0y0VW7wW3U5h1J8jlciwsLBCPxzlx4sQtkxTHdTm/UOLt+xwucGGhxN7uYJscXU+Qyy0bw3RYKNSZXPeGTTcf17IcfnQly0gqQDKloEo3XovlwIXFIsmQD61lYzsutaaBYblky00ayxNomsbJkyepGC5n5+Y42Btmd3eIiWyVoCbTH/OzXGq0NwO6whonRxJslzx6vXK/VDb4/07NUm1ZiIJnF/rJ1RwvT6xTN2xOjMTpDmtkoj7Wqy0KdYOaYfHC1TUO9kVwsZnKVig3THyKxFO7kgx3x1gsNhAFgflchWa9zicO9lCzHEK6yvHhOD94axlREFAkAUkUWC03SQR9KLLIRLbGpeUKe7tDFBoGr89UOdwfBcHlUG+U6bUqb8wWEAQBnyxyZaXCU6NpXp3Kc3w4zlq1xf/9/DhLpSar5SZPjab5nUcG+Z8+upv//MY8by2UkAQQgdlcDVEU0RSRbMXBcV1UUSSiK1iOy45UgP/3pTUCPvn6jRYWi3UaFjRtF58qsb8nwg8vZ3l+bJWGaVNpWCDAh/em+dj+br5xfolLyxVGUwHOFPL89GqWX3+wl994sAfTBr8igOtFhL42nadu2CCrzOVqmA5MVWsUGiar5SZfO7OAbTtYtsNnj/TSMh1EAdJhjbcWC0yteRawdMjHR/d1EVG9Rdzt2HHejdwLgoCNQM1wkASBoPru0a23ipsRy+vV/d7eXmCrd39hYQHTNLck89yJun+vyfW9Pj94iUfvly3znbA5UPte8NRTT/HUU08BMDIywmuvvXYXrqyDXzZ0yH0H9zXebsFwXZeVlRWmpqYYGhpi79697Zu4LMtYlnWzQ901vF/k3jAMLl++jGVZDAwMoKrqbamPAh7Z3A6KtPV1vP64ogCqLFFpWTTNrc9LwLPN1Fo2LcvdltybtkupbiIKIoZlIggCA4kgKhYLqzk+fWRn+0brk100VeaVyRzxgEoqqGJYDqem1vncA30kAwqaIjGYCBD3y9tasjYXJq4Lb8zmqW7EMR7ojbBQaPCvvn8VnyLhV0Salo0qiTyzt4tywyIeUCk3TYKqzM50kKlslaCmkK8ZtBom2BJ7ugLMrNd4/twMDwwlqIga//7FaYp1i3RI5VOHenhwOIHruDyyM8X4aqXdcrszHUQWBWJ+hdOzBfyKxPOXV7m4XGa52OB/+Mgoj+1M0hvTyddM0iEfezMhGobNswe7SYc1/tX3r5CvmfhVmahfxbAdvndxlT1dIT62r4tnD2QQgFzNIOJX2ZkKEtZkvvLyDEGfQqHeQhSgO+QjFlAp1evEA5F2m2xIkzk+nOK7F1dZLtawXZEP7U2zUm60C668z4XAUrEJgkDQJ+O47obNSmCl1KBuOIRVUAS3HT9q2C6FuoGAQNN0MGwHURDxqxL5mmdzmshW+fThHn50eZXvnF/iY/u7OTkcxbYd4nqK3ekwiiyQCvqI+O6MdL8TuRcEgbW6w4sT6yzka/gUiQcG4hzqDaEI753g386uwTup+wsLC1SrVWRZ3pLM825xvveaXN8Ptpx7NdRbrVbfM7nvoIO7gQ657+DnBrVajUuXLqHr+raZ9Zt2mfcbd5vcu67L0tISMzMz7Ny5k66urrZH9zaPxNEBL23Gdq6RYkUSODwQvensQlSX6Qr7SAsaUb9CdsNHLQgCgiBwsCdMWJfb3nevfMrFcV10RUQSQVdlptc9xdVxHNYKZXamwxze0bNFQdMVkYdHEiwU6uRrBvmaAcBIKsDOZICQJrVfk5td72Zajum45GvXXqOBRID/8LNpNFXigf4oyZAPXJeXJ3L4ZJGgT0KRPOId1hVWy00uLJWI6iqfOtJLKqhSXp2n2WiQ9HklT7o/wLdOT+ICkgghXeHcQpEvbDTWjqQCfPW0zXi2StO06FU1vnBikOGEn6m1GpNrNb78yDDz+Tof2duFX5XwqxJ1w6IvppMJaywWm/zt6XlCusJjO5Msl5oMJgLoishgws/5hSJXViuEdZlcpcW+3gg/uLhCqWkxtValO6zxj58Y4R89PsKp6RwXl8qENZknR9PYts0jO7vI1Vq4rhdZujcT5afja9RbFooEIFCsm5QbFrGAirnR7NsV0lBlkULNIBXUGEkGeXFiDUUS+fhIivWawXzBJh3yEdclRBx8ikhPRGc+dy0v38VFFgUG4n7Ozxfw+yQifoXeqI7teA25+zMh/LJL1AextNb+DDjOO8/b3AzvRLArJvzXNxdYKjbaP1suLQK9HBsIvWcFf7uc+VuZHdr8ve3U/c3c/evV/U0rz9vV/XvdOHqvFxdw7xYYd9uW00EHd4oOue/gvodt20xNTbG+vv6OmfWbg67vN+7meTYXLIFAgBMnTrS9spIk0Ww2b/t4A3GNZw9mODWVo1g3iQdUHt6ZoC9y85ZN13U52BdmsdDkSycG+crLM2TLTQRB4PFdSU6MxInqniWnablcWa1wdq6I7brs6wkznAwwkgywIxXk0kIe13VQZYVd3WHSoa0e4nLLRldEPnEgw4VFrxxrtCvE4b4IQd+NEZfbYTMtxyeLDMT9bRuHJIokAyrP7B3h0nIZSRQ4NpTgU4d7EAUB23E4N1/EduFgb5QzM3lylRa5aotqy+QLxwcoyj5em8ySSsT58JEhvntxFRdPxU4EfUR1L3lobLHEvrROT0jmy48M8ujOJMW6iSTC2bkcq6U6+3oi/OfX5vjdx4bp35lgpdQkqMk0TYd0WEdTRDIRjT/+3lVapo1hOciSyHKpSVRXqQigqRKG5ZCJati2y8G+KF89PU9YU6g2zY2oTIO/P7vI//6p/XzpeB+Vlo0giOSrBiulOk+Meir95FqVhmHxzN5u1qoGMV3eGE51qTZNhpMBKk2LqF9FEgQvP992ieoKz+xLU2marBQbfOHkEK/N5CnXvZIpTRF5YrSbfZkQ2XKd/b1Rrmar1FsWmuK11h4bim/4+w0+fbiHSt1AEjbSd2RvgXj95/F2IW5Es15ra715AdrKRkvw9XBdODOXZ18miP4eOeFmzrwgCDQsWK+bVBoWyaBKXJfaz/tWoSgKyWSSZDLZPv47qfumad5z5f6Xmdx3lPsO7gd0yH0H9zVyuRxjY2P09PS8a2a9LMt3oHbfPu7GDoHjOExPT5PNZtm7dy/RaPSGc9zJAkIVBY4NRtjVFcS0HVRJJKpvb2+5HgFFJBZQ2dMd5o9/7RDLpQYRXSUZUAnpEgldxkXgtZk8z19abdu3Z9Zr7MmECSkCH98T50O7k7RsGE4G0GSQr1MQi027XV4V1hT6Yjo7U0EeHIiiK7euNEqSRDCWYL1mMdoVoiuiUWqYBFSRZw9l+JffuMRQMsAjOxL8m+fHKdZMHtkZZ28mwhdPDlGsG5yezrFc2lg8ubCYr3FhocDFmRILNRd1us7H9nfTG9PZmQ7ik0V8sogkeLns8oY9yXEcwgrszwSYXKtzYbFEV1jnoeEYIhL/4pN7mVyr8eevzPA7Dw/x71+YIltpEdEVTgzHeXRXkh2pAHO5OsvlJrIocHwozuRalUxUZ7HQIKTJPLojyVqlyYHeCMulJookEtYUgj6FoE8m4JORRM8iE1BEfnRlnVen1zFaJomwzpH+OI/tSlKqGyRDKsW6ga6I+Db4z0S2zId2p3hrqUS21MQFgrqfj+7rJhX0cXahyA8vrvCpwz0Uagbz+TrhjV2Whuny1dPz/PpDfbx4dZWTIymePdjDeLbCkbpJzK/iUwTOzhb5+IFudqYCvDSxWV4FDwzE0GXhjlR6QRAotFwuLRVZKjbpj+vsy3iDqTdTr+vG9t+rlmljOS7cJXJfNuCb55eZyJZxXPDJIo/sTPHwcNSzMt0h3kndL5VK5HI5CoXCFitPIBD4wAj3L3POvWEYH1jkaQcdvBM65L6D+xqCIPDAAw+gaTdXnjchyzKNRuNdf++94r3acgqFAmNjY3R3d990wXKrJVM3Q1ST2GQpt6KEFho2f3t6npVSE58s0J8IsFZu0jOaIrHRQltomLwxU9gysGs7Nm/OrPOx/d3M5Fvs7o5ydqHAC1dWkAR47sQwxwaj6IrIpaUyk1lPZS83TS4tm1xaLhP0yRztD99wnaYDpYaFIEBEk9tZ/VVbZKIkUC+scWa2yNGBKAGfV7o0vV6nL6bz4b1p/vSnU+SqBhFdIV8zGV8tkw75sG2b+YL3OXFxqTYMWpbL5WwNQ/RhOi1s0+HliXX290UZX6lsmTWQRYHD/dEt9g1NdDmYCbA/E/By7VsOqxWDnqjODy5l+eyRXk7PFFgqNREFyJZbiKLX1jqc9Iq64kGVc/NFPnqgm/PzRaotC78i8aE9aRTJ26VIBn2MJP3YLiwWG9iOSyKokg6HWas0ubLcYigVYqXUAEQkWeb8QomfXM3xyUM9DCd11spNdEWkWDfpDqu4rku5YZItN/jCsX4qLQvLdgjrKq9MrvONN+cZSgZ4aneK3pifv3ltzisz23i7mpbLzHqN6fUaRwcTzOQb/JsfTXCoJ8SDQ3HCmkQm6udoXxTLdfnhpRU0WSIQkHlgMMa+6wa9bxf5psvfvDbbjuS8uFTk/EKRj4x0b0vuXdclE9HxyWK7cG0TffEAQVWEW4jafCe4rouu67w5X+Tqarn985bl8MKVVXakgvSF333BfTu4Xt03DIPe3l4EQaBcLjM/P99W96/P3X837/6d4pdVub+b72cHHbxXdMh9B/c1ksnkLavk73eKzfXnabVa7/6Lb4Npmly9epVGo8GRI0fw+/03/d3tGmrfD7iuiyRJrFUb7OoKMRAPsFCoM7FRWuWTJT57uBtR8CIzqy1vZ8TFpV6vU6lUiEbjREI6o6rK//PCJIW6gQj0RX08P7ZCSJM51BNicq227TVczVZ4YCCy5eaYq1v86HKWqbUahu0wkAjw9O4UyYDKufkyddPlL16b4bljA/xgLMuZ2TxfODHAbK5OzK9iOV4UZDrkQxaFdgLQ+GqFx3YmgTymaVJpmszkGyTDfrrDGt+bKzCda6DJItmKSm/czzN7u7i0XKRUM4n6VR7ZmaAnpNyYtb+ZPmQJfO2NRU5N53nu+AAty6Y7ovOTK1lc10uFefZghlrL4q3FEsmgj9GuEH/24hSzuTqnZ/L88w+PEvUrqLKITxKYWKvzwtU1Lq9U2JeJcGGpRFCTqTQtclWDwXiAxfUy3z47h6oHeXJPF+fmS5RaFgFNxXAEzi0U2ZEKIODyu4+N8J3zy3gM3W2nE4VVCCnewlAUBUZSQc7M5bm6Wtn8xKCrEqrt4m4UmNVaJrbrElBlYn6Vv3hlFtN2ODuXx3RcFEkgFfLx2ycHSfoEPnekh7rhIEkQVt853lIURZo2WLbbTuS5/t8uLBS2ZO0DLBUbTKyLDGe2J9AJXeLpPV38+PJqm+D3RHUe35lEbI8d3zkcx8EWFCazpRv/zfUKwPojsfeNDG4q54FAgFAo1Fb3DcNoJ/PMz89jWVbbu3831f1fZuV+c06pgw7uNTrkvoNfGHxQaTmyLFOv19/9Fzfgui6rq6tMTk4yPDzMvn373vUG8EEsVERRxHZhbLnC195YZGy5RCrk4+ndXUR0hbHlMvlaC8N20GQRn+z5zhfyVQqFApIkkU6lCesqO5JB3pjNIwoC6bBG3K9gt2o4Drw5W2B/T4iIvv2fm7hf3UJ0GpbLd95a4cpKhVzNZKFQ5+XJHLO5Op9/sJeqYXN6tsBQIsBqpcXrM3kAVstN0mEfk9kqrgsB1WvQBQiqMpbj0jBtYgGVp0YTvD6xylzRwUbkkR1JREHg5Yl1JEki6ldpmDbffWuZ/+PXDrGnO4QD+GSB7qCCYRg3fU2nczUurVSYydW5sFDCcWFqrYooCJi2zWeP9PC1NxZYLTcRBNibCeNXZX77kSH+06vz7MmEKDZMvnp6jn/y9E5kWeYbF5awbZe6YfHswW50VaJYNyjUvZz8vrhOrWUSDEfIVg2+f3GF4aSfb59fxBUlBuJ+uoIqvTGdvz+7jGXbPL0njSwJpEMaXUGF8NuSaRzHYWdS5zceGuD16QI1wyQV0vj0oR6+fX6xrd5LokBvVCcd9lEzbEoNs53zbjpego5bMbEc77ugCC4R37Vz3Aw2ImMrdV6fydMwLQbiAR4eSRDTPG+97cJCcfvv4Xy+ftPdLxGHYwMRRlJBlksN/KpEd1gjpNwd9dV1XSTRS6DaDn7feyOdgiDgCgJslMO9/TW82UCrqqpbvPuO47S9+3Nzc9RqNRRFaSv74XD4jtT9e52zDx3lvoMOOuS+g18YfJBRmJvnKTVtaoaNLIpEdAnf26IiG40Gly5dQlXVLQk/hYZNtWWhSAJRXdnSHrt5jrtN7l0ECg1vwFGWBJLpLpZKLf7rmQWy5RbFukmxbjKfr/OHT+4kpMlkojo+WQJcwj6J4wNBrs4sEgqF0XyeKv7YaJKYLiGLAiOpAODi2A7r5RYI0LQdKk2LIxtJPtd7nkOa7OXNX3djLDctprJVmpaXvW87LrIocm6uyOce6KVu2NQNh3REZ3zD5hPWZAp1k8dH01xdqRLRFeIBlXytRVhT0FWJiWyNp0Z1/uG1CZquzIcPD7ErV0WVJfZ0h3htOo8sCriA5bj4ZJHPHOnj5YkcXz+zSLlpsq8nzHPHBoj7ZSzbIe5XCFxHCoWN+Mhc1aDcMPnx5Sx/8OQOXpta5+RIglTQx2KxwVKpieN4w6hN06FpGgRUmT/5zSPMFWrMrtf50sNDLBTq1FoOgzE/kgjFusVfvDzLrq4gz+ztQldEvnl+maZpkQlr1A0b13EpNSySIR1ZVnBcl1rL4dnDPfzgrSWev7BAPKRRrTeJaBKHh5IM701vS7QlHEaTGsPxHkzHS0cybRfcHt6YzVM3LHZ0hRhJBbmwkGd/T4zQRrRozO9jLlenbliMpIOYzq2nxoiiyFvLNf7uzYV28tNKqclyqcFvPtRPQAZJgO6Iznh7V+EaemL6O55HxCGlC6T9Ae+74brcLW7mOA6CY/HQUJyptQrXjxKEdYUdqTu3IVmIzOaavDlXwHbgYF+EnUkdVbx2kltNqxFFkVAoRCgUoq+vD/DU/U3v/tzcHJZlEQwGtyTzvJswcT9EYd6LxKBGo3FPsvU76GA7dMh9B/c1bucP9Adpy1F9Pq5k6/zw0iqrZW/AcXd3iI/u7yKqeZaa2dlZlpeX2b17dzvH2nHh0kqV5y+tkqsZyKLAaHeIj+/vJnpdTMfdtuWYDpyZK/DyRI5C3SDok3lsJEx2rkjLdAj6ZCK6Qqlh0jQd3pwrcqAnzEODMQRcqtUqFy9eJNPbx//8K8e4slrFdlx2d4foCqkIuOxMB/nZ1TUsBxAEFMVHsW4x2qXw48tZIrrKZ472cXauQLFmkAz5OLkjQVdQ2XKtrgu249I0bcK6giZLWI43HKzJIkNJP4qY4uJyiZiueIuQiM6lpQqmvcCnD2fQFJHff2yYlybWKTdN8jWDo/1RhhJ+Tk3WQHD42fgaH9qT5vlLq4R8MsWGyR88uYNXpwsIwNN7ugD4zlsrlBoGluPy2nSelXKLf/rUCN88u0BPVOczR3pJ6mI7vjMV1trWjpVyk6+fnuc3jw+QieqcHEnw7QsrxPwKkigQD6isVVqU6ibnF4rEdIlizeDhkTjORm/Bty8sky03iQdV/KqILAmMLXuNsoOJAEulJp88lOHVyRzFhsVcvsaHdqfpjmqMdocI+SSe3tvNQDzAn63WOTCYwq+IOGaLtWKDF85X2B0VMUrZtkVD1/X2d891XWTBRZYAx0YR4PhgmD3dAUDAp0jUWja1pkmlYfCRfV28cGUNx7GptUwkSeTEcJzLS0XiI3HkWxgmrZsur0/ntkS6gqfIr5Rb7Ez4cByHo/1RLi6WyNcMBDy7WFdYYzTpv6Xvz/uhtrqui2VZ7EjqfOZIH69N56i2LNIhjSdHU8S1W1vgvB2CKPLmbJnvvrXUXjBcWSnx2Giap3bFETcsS9tFcd4qVFUllUqRSqXax3ondT8SiaAoW7+/90MUJtzeveNuoFKpEAwGP9BzdtDBzdAh9x38wuCDVO4dNcQ3zy1RangedNN2eGvR80E/MRzkrfPnSSQSnDhxYouKlatbfPv8crt4yXJcLi2ViegKH9+XZvN+tLlQEQSBpuXgOOBXby0qcjsslVp8760VLw0EqLYsLszlETU/CKBKAkMJLwqx0jIJaRJPjqYIqgITExOsra2xb98+IpEIAL3Ra1nkm+gKqXxobxc/u7pG3bQQZJV0yEcmovHi+Bq249Ib0/n8g32okoCmSCjbcICAKpKJ6lQMm1rLYiJbRZFE/unTO8jVDFbKBrGAj2f2dtMybZZLTcazVRzXZaFQ5+/PLvLPnhnFdhx+77FhinWDq6sVLiwU+atXF8hEfARUyNdaiIKA5bhUWhZXVys0Gi2ODyfpjuiENIWvvbGA43r+fduBWstiZq3KcqlJKqQxl6/zkytr/OrhbkS8GMaBuJ+93SHm8nXqLYv5YoNvnFvkM0d68SsSO9PeXIMqi9QNm5hfoTvsoy+q89ZCiWLDwMXz7R/IBNmRDrNcalBsmIx2hRhOBpkv1NnXE0YA/vDJHVxeLrNcbiEIYNkOQ8kAE6sV9mXChFQR07JwHJfusJf97+KC5MOn+bzdo1gMyS9RKpWYnJykXq+j63qbwIVCoS2fY9u28ckiU7kGZ2YL2I7L3p4IO1JBLNulP67zs8ur7O1ROToQxbBsXp/Jc6gvQtT37qRrM2t/O5QbJoKgUbNgudTg8dE0Y8sV6qbFaCrIwb4wC1ffgkz8Vr8edxWbqrEsOBztDbIrFcB0XPyKiCrevL/h3VBuObw6tb5lJ8AFTk/nONIXIa5dW4zdLXL9Tup+sVi8Qd2PRCL3xUB1ZTeSAAAgAElEQVTtvcDdaKftoIO7hQ657+AXBh9UiZWiKGSrlTax34TjOrwxvshIIM2BAwe2VXGWig3ydRPbcREEUGURERhfrfLwSKKt3ouiSDTZxZn5MmfnCjiuy95MmEN9ES/R4zYgCAIXl0ptYr+J1VKdEz1JZtfrXkutJJAIKiSCCo/uTIJR49SZt+jq6roh1Wc7gqJKAg8PxxjtCpKvm+QqDXI1i9emrimwi4UGFxbLPD2auCnJCWsyHz/QzdqpOV4p5bAclz98cphi3eR//dp5T42WXdLxEJ9/oJ//8SOj/PWrs6zXDHoiOg+PxHl9OsfllSqfOZJhpdjgz16eJV83kUSBlmUz2hVEFQXCmsznHuhhbKXCZw738vXTs4xnqywWGhzqi5IIqFxdreACtuNguyCIQvv9A5jL1ai0nLaPPCALfOJAN/t7I5ydK9IX1xlKBPjx2CqCIPChvWlM22apWGcwESCgShiWTSai8ZPLRfb3RnllMkelabJWNdjV5fLsoQw/GstSapjYjsNH93Ux2hVkeq2KLEAs4GMgrqMpEr/5UD9n5wv89OoajuNwuC/CP3l6F0GfSDqkslre2p/QG/MTUEV8WqS9eHNdl2azSbFYZHV1lfHxcURRbCu2iUSCs8tVvnPhmoo8ka1wfCjBsZEE48tlhqIy/oDOleUShbpB1K9iuQL5lmepiWgSONsT+IAqkonorFdbCEBf3E9XWEeVRYaTQQzb5XsXs5xfKKBKIn1xPzFdpj/uJ66JjH8AqVk3w/XKteM4+Nt3Wec9WX9Mx73hbw5Aw7RpWg6b6Vi3q5xvDoHe6k7hdup+tVqlXC4zOztLLpejVCqRSCTan5e3q/u/iKhUKp0Cqw7uG3TIfQf3NW5na/WDUos8BXPrXbrRaFAqlxjpSdLd1UXwJsOjDdNhPOuVCUmiSDyg0hPVkESB65+qIEqM5WzOLC+0zzSbq5OttHj2QDfqXbC0ZstNQj6JR3d6GeqiIFBpmWQiGnGxydjYNIcOHSIQCNzyMSUR0kGFmC7zw7OTZJvXXoeQJjOUDBDSZOqGc9OdCNd1GUzofPxANw8OxkCAsKbwL//hIg7eroPPMak2NM7NF3hqT5qj/VEUSaRl2pybK7JS9rz6hZrBQqnJyR0Jvn9xFdtxKTYMTNthOBEk5pcJ+XwMxHRsF0aiEtmajStIDCf9njI66w3sbn4WB+N+MhGNs7M5XNcrYRKve+/8irfQMU2Tj+xLkQ5p/N3ZJXLVJomgj4uLBX7n4SHOLxQxbc/OFNUVLiwUsBwXRRIp1A3Cupdd//pMjv29Ef67Z3ahyhKyCPO5Gn//5jyW7SKJAkPJEJ863MPlpRIhn8RQIoAx4hDVZT60O0V/VMV1HD6+P8O3LiyxXvUSZjIRnY/u60aT2EI8BUFA13V0XSeTyQBgWVbbj12zBL5zeoFczUBVfaiqiiLLnF0o8MBQHL8qcfrKIrF4zGsURqAvHuT0dI7XZnKENYXjwwmODUZRhO28/i6P70qyUm4wmAzSMh1eGF8noEo0DJuHhuMsbQzTGrbTLjLLVVt88UT/LX9e3w+8X35vRRSIB3xkK1sXZ0GfjK5sXXjfyvkFQaBqwmqlRaVpkYloJPwyErdnB9xc9IXDYfr6+rh06RLd3d3Ytn1TdT8QCLxvtpl7NdharVY7tpwO7ht0yH0HHdwmRFEkoUvE/CprlQbFYhGAVDLFkeEuItr2X6u1moVpO4Q0ibphYTkO2UoTRRL48N7Nx3k3pkLD5NXJLLI/suUYby2UODmSoDt060qY67rs74lwejq/Rb0XBYGEX2E4FeLMbJGVcpPD/TGGYj4Uu8GxY8fu3LsriwzEA2SXPBLZFdbYnQnz8mSOsZUKb84WODESZ09XEEXa5hyuy8x6lclsla6IzlyujigK+CUJWby2KFguNXAch0rLG8J1XJdSqYggqRwcSpGvm0yt1XhmbxeuC+fmvfdqtCvEpw5lCGteXKIiCShANTvPgZERAgEve/6RnUlWyk1+emWNpuWwpzvIZ4/0Mr5axrA81XlfJoxfdrHta17n0ZROPKAyka1SqhscH4qxtF4EXFZKTQo1L+/8mX3dRDSJH42tsbZB2motk32ZCDO5ajsV5vx8EUUq8dyJYb59doFc/Vpaj+24zOaqPDgU543ZPAKQiersSgcwDAufYOHYNo4gkgypfP6hAcpNC00WiPkVwuqtqbayLJNIJEgkEhRaoOglgkILwzCoVitYpoUoiRTLXXxsfwqjWaNgivhkheG0l5b00lUvDrTUMHl+bIVE0MeelHYDIXNdl66gzO8+Oszp2RJfvTCH3yfjOi5n5wpM5+oc7I3ys6urWx5XbpiYd2vzThSptBxEAULvEtn59mt/P4hrxCfy6K4k3zy36A0146XlPLwj6fVaXBcTeivnzzcdvn5mkYWCt0hSJZHHR9M8PBy9bYJ/PWzbRtM0/H7/Dep+qVRidna27d3fJPvhcPiuqfv3yvPfUe47uJ/QIfcddHCbEEWRVjHLh3f38Y0zJdxwiFBAZ38mzCMjcba7rwqCwHy+zqXFEs8dG+C7F1eoGzbxgMLurhBH+qNcvxtgO1BuGMTfFoVv2A5N0wZu70bYE/Hx8YMZXhpfbw/UPv7gILIs8R9+NsX5+SK1RgvTdnhqb4bfeGiAuCuwHe++Fbiuy56UTtFWmM5W2dcT5q9OzWLYLn0xnbl8naVig88/1M/+TPAGcicARwdijC2VKTcM+uIBRMFbkMQDKo1yFQTQVJmG6fDwSBLHdZlZLbG3L8VoT4zR7hB/89o8r88UGFup8OiOJF88OchA3M/uriC1lsX4WgNNFekKqm3/v2ma7evpi6j87iND/MqRXizbQVMkLi+XkEWRTx8dIBPR6A6pYLXa5G9zqDumijw8FPZaVGtNVofjXM21aFoufp/Crq4wUZ+I6No8OZpgf2+EpVKDdNCH5lP4+ul5ViuN9utxqC9GOqigKjdu2/TH/GTCPnpjXqvtUrHBeqXFIzviKFaNlhPhxck8b8zmaBg23WGN48MJesLqHQ1u+ySBRMhHy3ZQFKW9u+OTBMK6Qn1tkUczIkowSsDvZ7Fs8d0Ly9gI3pPBGy4/O1dkNNWDwI32HMdxwBW4ulIiFdqMS/Vy6FdKTfb1hNtD4JtIhnxot9F2vB0EQaDYcnlxYo2ZdW/e42BflAf6I/jEWxvSfT/IpeM4HOgOkgyMcH6xhO04HOiN0BP2Idxu8ZYg8vLkWpvYg/e35YUrq+xKB+kK3Pn1b5dzf726v4lWq0W5XKZQKDA7O4tlWYRCobaV507V/XuV1lOtVjvkvoP7Bh1y38F9jTv54/5+x6BVq1Wy2SwHMr38zpMHsDcGLkVR8LK8ERC2qcJx8VpFdVXi9x4dpty0KNS8FlPpbfdSTRboigR4u8M2pMkEb7Iz8E5QRDg+GGVXOkDL8pRqu1Hle+NrXFwosl6qoigKiqry8mSOY8MJ4gGFZGD7czmuRwZ8krjtYgaglF3gc0cfpFA3ubJaJayr6IqIuPEAy3F5fSbPjqQfn3zjQQZjGp88lOHUVJ54QOFgb4RqyyagSjQFr9ioP+bnX//gKou5Cp9/aIDfeXwXq6Umpu1g2C5PjKZYKDQ4M1fgJ1dWqRkWO1IB5gt1/u2PJ1nI10kEffzKkR4+cSCDnshQshSsukVc93ZSAgpoQRHXFRBFEX0wxmKxyYXFEqdn8uzviXCkP0JMF9tJJbZtY1lWOw/fqJZ4bCjIyT29FGoGjgsxXcZ0wCeCKrr0hmX6Iteaer94op/pXJ1CzWAwGSATVtFlgY8f6ObbF5bJVVskQz7SIY1HdiSJawK/dWyA5XKTWsuiJ+JHF03WV5YYX6vx4ngWAF2RCGgy67UWubqfrpCMe5spU0FV4InRNH/35jxN0yOWogBPjHbRHfUjRIY5c+YMI4P9OI7DqVyFtXzBI16yhKqqqKqK7TpeGdZNzmNuZPu/ffHnkwUUUUSVr31x/KrEE7vS6Nt8lm4HDQu+cW6ZybVrEZvLpQam7fDkzhjuuyyG2sqxIFJueT77kCa202zeC0QcesMy/dFrirh7B8etmQ7z+RtL5SzHZS5fJxMK33Fa160q5z6fb1vvfqlUYmZmhnq9fkfq/r0i952B2g7uJ3TIfQf3PW41Gxuupcy8HyUqtm0zOTlJPp+nu7ePy0VoGEWmc157aL1l8+BgjGcPdnO0P7IlCcZ1XfpjnlWjPx7gP/5sqp3R3hf382tHe3lqNNnOuw9rMs/s6+KF6Xq7XVUWBR7fldognXcCl9h1j50uGsznqpTrLUIBPxFdQVdlZFGgZTmYls12fyKWygavz+RZr7ToCms8NBQnE1ZveI9M0ySoioR8GpPZKooksFRq0jBs/KpEMuij2rKomTY+WeHtcwyKJPDgQIShZIBqy+b3HhvmjVmvkXRnaoDhdISLi0UuLeToiQWQFJn/5W/P4fcp1A2LkCbzsf3d/NoDvXzugV4kQaBQN3h1Kocii8znPdVyvdpitdLiB2OrnJlYomkLpCJBPrq/i96YjmU7XqKMJlE1XCayNb7y0jRT6x45+u7FFZ4aTfPlRwZZrxnM5upkIhp9UZ2YJrCyssLy8jKhvl1849Vpig2zPWB8YiTJM7uT+KTNrPVrr0FIhSO9wfbnx3VdHMdlMKry3zw8RK5uMZevo8oCqix6xV2yy+6UH9MBWXApl2soPp3zC0VEwVP/h1MhWrYXd3pldZHd6QAPDEZJ+W89ftVxHHandL786AiXlsqYtsPeTJhMWG2ryJvtx5qmcWyHwsXVOrbtYtneoqdRrzMcTnHl0sUtJO764qSgKtIb9ZOvbS0N02SBfT0hRlI6Y0sV/KrE3kz4tp7DzZBvWEyv35idf36+wJG+COF36XUSBAFL0vje2DpXVsrYrstwMshTo6l2Add7geu67znyVxa9tKrtENS2b/a9Vdwpub5e3e/v9+YmWq0WpVKJQqHAzMwMjuNs8e77/f4bhJx7qdwPDg5+4OftoIPt0CH3HfxCYTMO826T+1wux5UrV+jp6eHEiROMzWWZXKgR0hW+cW6p/XsXl7zt8u6wxkDMt+UYyYDCrz7QyzfPrzCfbxBQZXRVIuFXOTW5zp5MiKHYtYjJqFPhy4/uYCJbxbAdRrtCdId9W4Y3bwWCIOC4LuLGIsl1XRYWFqhWaiRDOuGAF/uYq7VYLrcwLYdfOdpLzXRZqXgE3SP9sFw2+JtX59pWiNlcnfHVKl96eIBUYHtVzXVdemJ+ptZq1IyN8q+GyUq5yYNDMS4tVQhrCsmQD78qEtOvkQvTdnl9psArE+soksiu7iAf2pPGr0r8u59McnG5RDIa4ul9XXznwgpXslUO9npzCpbt8penZvnvn9nFK+NruHiJI2FdQblOWeyJaIQ0mb94ZYbBqIIsCvQndMZWKnzrwjKGZRP1qzy8I8GOVIg35wttYu89P+91eP7yGm/OFbiwWEIAHt6R4LeP9+KaJv27D3F5tcqRoRSu41KotSg1DS4sFDjSHyWtX7NzCIKwJW1lmzeU6Vydb51fbMdFBn0ynz3aR3dY4/WJHEulOrvSYXanA4TDAr5sjYN9Meqmw1y+zp++MEmuZpAI+lgshBlfrfDFkwNEbqeQ1HXo8ov07E60r/XtKvLmwjwTVvnovgwvTaxRbkLY7+OhoQQPjsQQ7Djlcpliscj8/PyW4ctYLMbjo0nWqk1WSt5MgiqJPLm7i4QuIftFhmOe8mvb9h2p2G+/3qZh41dluiNeIdFyqUGtZdEwbRzXhZvuM3jwB8P8dCLPmxvzHQDn5gvUWhafP9rDeyyovSvQZYGHhuLM52tbojW7wt6i9L2Q+7u5c+rz+Uin06TTaWCruj89PU29XkdV1bayvxnFea/IfWegtoP7BR1y38EvFO52HKZhGFy+fBnLsjh69Gi7gbDctAjpMmevu4EDNE2HluVwabnMYDy95SYpCpAIqBiWQ3dEw3ZdwpqCLAk4Lsyu1xiOX7uxViplRv0iPbs88nS7N1zHhcVSizOzBaoti309YYZiPqavXETXdWLhEE9EQ0zmGry1WKbYMCk3TE6OJJjJ1XlpYp0nRlMoksCFhRInR+IsFpuUmlvNQoW6wVuLZT60O7ntYKQoijRNm6f2pPn+xRVM26Fh2uxMBznUG2Wl3ODPXp6m2rQ40BvmxHCCBwYi+CSBfMPizEweF88GFFQVvvr6LLu7Q2QrLdarJsulFs891M/4agVNkYjoCobtEFBl1qot6oaFtpGyoikiJ0cS/OUrM+1rHEoGOLdQwtkgbSHNG5b+tz+ZJKTJDCf8rFVafOf8Ml96VGO5uDWtxK9KNCybM3MFQj7vT6ppWTx/YYG93SEe3dnDV16e5dRUjrCusDcT5vFdSWqrDo+MdmPYLoqiYts2juPgOE5bmd2MKbye8JdbDj+6vLolB77asvjexVWODkR5YzbP8eEEU+s1fjSWZU8mxIODcSotz/detxxyNcOzDzkuy6UmPlngykqFk0OR21a+b0VFlnA4Phhmd3eQhungk0Wimuip/KLSHtSFa8VJxWKRqakpXNfl4zt7KbRCtByBoVSIZEBB2oiWvJvFdZslZCd2pLzhawEeGkpSqDWpG54ljHcZNnUUP2NLazf8fHq9Sqllk/bf+wx4x3HY0xXgs0f6OT2bo96y6In6eXw0RcQn4DjvbXfh/bJFvpO6n8/nmZmZwTAMRFFkaWnppur++4Fqtdqx5XRw36BD7ju473E7thxZlu/Kzd51XZaWlpiZmWHHjh10dXVtuUEENRmxZt+g4amy5ym/mbpebdkUGwYzOU/5lUWRvphOMqgS1rcq3+/FYiQIAlezNb5+ep6m5eC6Lq+PL3EgE+RzD+0hHvazsLBAXDb4/cdG+C9vLDBfqDOSCtId1vjzV2YJqBIHeyNIosALV9eI6ArTuboXY6mIOK6LYXsq3XyhgfO2WQNR3EgYEQQuLhbpCfv4Zx/exUqpRTKkEvcruLi8MVOgZdqslltE/A2KtWVCmszBnhC11maGt+ezFrA5NZ7FtmyOjyS4mq3SMF0KdYPusIavbpCrtpharxPRZfZlwgzGAxRqLXRF5vhwHFEUKNa32jxkQSAVVBFxiKour0+tU2u0CF6XOWo6LtNrNYaSAV6ZyrV/rikSpbrJ3u4Qq+UmYQUOjCSJB3UcFyotm1LDJB7wUW6afP3MIrmqQVfYx7cuLPMvPrEXSZK2qPWeBcdp/wfXSGzNEFmvtLZGpwoCYytljg3FOToQ50eX1zgzV8BxHC4vF5lcq/GR/d2kIxpvzBaQRAH/hgWrZlhYjku20rqrJOiG76zrEFEhstnTcBOV/fripE0C12w2CdZq1Go1lq5OM2/bW1pSNU27K9csCCJXVyt8/c0FZtZrWLbLj8ay/PbDgzx7ML1RQvWuB8G0Hd6u8HtFaO9vTOPtCAAyDkd6A+xK+7Fsl4BPRNooYvt5wtvV/dXVVfL5PLZt36Dubyr874dts5OW08H9hA657+AXCnejpbZWq3Hp0iUCgQDHjx/fdojL57QI6xoPDMR4c67A5mZ9KqgS2PD/vv1G27BcTk3leHAgxptzRUzbwXIcFosNdqQCDCYCWx6zSe7v6DkYNi+Pr9O0HAzToFAooPk05qtQtSXieCSq0ajjj0bIRLys/YV8g797cxHb8fzdjgv1loVPFlktN+mNakytVclENFbLTURRZL3aYn9PmG+cX+FAT4ihuO7FB27c6ERBoCfm5/lLq/THdPb3Rnl+bJXhVICRVBC/T+bYcIIP71OYWa9SqpucmSsw2hUg6JPQFYl8uYrlGpSbQRSfj7HlEo/s7uZLJwf56dU15vM1Pnkow3feWmE2V0MUvN0VvyrTFfbxW8cGvLhL0Xsf/unTu/gvbyywXGrSsmyePZjhu28tIQoQi4Yp5eoIgkBEk1hbW0cAVJ/KainIIztSvDlXYLXcomXZiILASCrASDLIcq7M544N8ep0nrHVHIf7IuzriTAY16kbNmPLZVzg4nKZBwcHODMn8upUjtGUjn8j5WWT5F9vLbie6MuigCpBa2PRIyAgSALSBjE3bJczc4WNRwo0TIdK0+LcfJGRpDdQPbVWQxK9RbMiCoiCwHAy8J7sGNvhbi0WNE1D0zQSiQQDAwPYtk25XKZUKrGyskKr1cLv97cJ3J0+j1LL4aXxNSI+mT3dYWotG0kUmMpWeWZ36taOazToiweYzde3/DgV8hH0XYu7fT9wOzGQm6KJLrle/5XrvI9X9sHBdV0CgQD9/f1bFoflcplcLsf09DSO42xJ5rkb6n5Hue/gfkKH3HfwC4X3Qogdx2F6eppsNsuePXuIxWI3/d1GpcixnXvIVi2+cGKQlybWkSWBkWSAJ3alyIR9Nzym1rK5vFxmJBXkD5/cwZm5ArWmTSam8eyBbhJ++a6R+6blkK+3KJZKGK0WsVgcVVFwgVytxWBcQxRFLMsi4pMo1L0h2XLTaif/7OoKslRsMJgMUG5azObqPDgYY3KtRrFuIAoiE9kqO9JBfLLEn780zYHeCP/4yRFUSWTRCrIynmdXl9esO7ZUYiQd4isvz5AIKBi2y5/8cJxLy2UiukJXWOMPHh/hynIJy3JxHAirAvtSKt9frZGOJ+iK+PErEjFd5aXxdRRJ4PMP9tET0RhI+LEcl5cm1nEc2N8bZl8mxKmpdT53pKe9m6LLAs/sSXKgN0zDdNAVkZAqAi6vTuXIVlo8OBRnvtggFlARgj5sy8Y0DHZ3hTAMg997bJiLi2VWyi2Gk34O90X4q1emeGxPD//up1PkawaqJBDWFP70hUk+90AvE2s1Aj6ZhmkTVGU0RWYgrlOot2hYDv6bDDiCR/g3SVtCETjYH+fMbN5TkTcGbR/flaLSbLW96ZsI+iREwaXStFBEkX2ZEGNLQSbXqggCpEIaezMhhpP+nxvVVpIkYrFY+zvqui71ep1SqcTi4iL1ep0zZ860yf6ttqQatrcLBOCTQNuol600DVqWQ+gWYjYblQIf3b+Lb5xfar8XMb/Kx/ZnCKu8u/L/HnAr5F4QBEotl9VKE9eFdMhHTBffNQXo5wXbee43F4fXe/crlQqlUompqSkajQY+n69N9u9E3e+k5XRwP6FD7ju473E7isqdKveFQoHLly+TTqc5ceLELalfst1iT1eYoYSfj+33SpIUSSCmb0/SpOsSKvyKxN7uEM7G/2uydIMqKEk3T/4wHcjXTUp1k6AmE/cr7aQdANdsIRh1JFEklU4hbFgEJEEgGfS1k0wcxyGqSXzyYIZK0+L1mQI+RWIgrvOJg90sFZtMrFYQAE2VODNb4AsnBlgqNJjJ1TkxEifok/mHc4vUTRtBgNl8gxfH15ieXyYai/LieI7Hd6f4wolBzi2UvEjB3Wn+48+m0FXZG2I0PevKC1eznByOkw77aFSKjI2N8cjeA+wdSHNhsUTQJ/HlR4d5/twUAHP5BnXD5rdODPKdt1Z4ZTLH/8/emwfJeZ93fp/37Lfv+5j7wgAzg/sGSImXSMuyZeuwVqZsZZVdp2zZ3iTrVBJXElfFW7VVqUpqnapkK7UpR87aWnud9W3Lki3RlChSJEiRBIhjBoPB3HdP33f3e+WPnmnMAIOTAEHS/fmHxUFPv1f3vN/f836f73Owy4+miFR1kzdmUox2+KnpFlXDwq1KaLKAbVlEnCKWY8vXDk8OBhjr9FFtWEgiWJbN23MZDNPC7XRwfLQDQZSYzdSYT5dZL1TRZJHltEnEJfO1J/fwrYsrBF0qXQEnXk3GrUrMbJS5slpkT9TDxGoBRZJ5am+Ecr2BIjYXAOrNWah3+twJNs+Pxgm6VK6uNa/N/k4f+7v8vDSxRof/Rt+GUxEJuRQsyyLhVTnQ5eP8QpavnuljrVCjbliMJLz0hzTc8qMVno8SQRBwu9243W46OzsplUocOHCgVd1fWFjANM27Jq3IkoDHIVOqN/+G2Ju1bK8m7z5sbReq1Sr73CK/eKqXdFnHtG3CbpWA9ugF9N3EvSiKzGXr/PWFZdKbCUR+p8LPHO5iT9jxUJJ8Hjemad51ISeKYutzsEWtViOfz99S3d8S+3er7rcr920+TLTFfZuPFfcr7nVd59q1a1QqFQ4fPozL5br7L3Gjqm7bNg4JorfJg99O0Cnz5J4Iy/kav/PStU1fLvicCk8Nl/nq6V782o2FwfbklO0xiTXD5kfTad6cTlMzmjGNB7sDPD8aQxMtJicnAfip44P8xfnVzaFXTQ71BIi41db7bz0Z2Btz8S+eHWIlX6fcaFbvvQ6Fum7xymSSvrAbp9zMFQ+6FJL5KjGfg7+/vMp8uoJNMz99b8LH98bXqdQNpM0Fi2nbvH49xWiHD6cisr/TS8O0KNZMZEliX9zDWr6GLAospit86Xg3YbHK4uI6J06cQFVVosBIojnsqmZYDIQcTCQrKJLIaIePxUyFgEvl2nqJi0t5dMsm4lE53B1gKOrhzy+ssJqr0h9x84k9USSxKUS8DhFN2jy/wFKmwt9fWaO82YD86QMJAi6VoEsBQWAxXcajKbxybYNUoYwsCvRFA3z3aor/7YuHmg2elsVG0SBdqhPzNgdLCUDIo3Kk28+euJeeoIsfXd9AEgXODIbwazcWcroF+ZqJYdm4FAm/Jt4imjwKPDMc4nhfAIFmZCTAT4wlSJV1nh9LcHUlT0ATiXkdOB0ypweCaDR4ot9H1bA53OlBU8TW5/hh67JHPW/ibqiqSiQSIRKJALsnrTgcjh1ebL9D4cRAmFeurrcsKgJwoj+M3yHetk9gO1vX0aOAN9j8rtm2/YFUxu8m7su6zffG11vCHprJVd+9skb8TC+e9zkk9nFfc2iK+63gg/thq7ofj8db77P1eZmZmaFSqaBp2m2r+0co0QoAACAASURBVNVq9Z7vH23aPGra4r7NxwpJkqjX63d9nW3brK+vMz09zcDAAGNjY/d1U3owy4zNvoSXV6aasY6yKODVFOI+B+v5Gsu5Gv6Eu/VqdyjOxfU6y1NLxH0a+zu9xDwqS7kqr17baEXY6abNu/NZYm4Jd3mFSCRCZ2cnNgL/9AmV9xbzFGo6Y50+hiJunNu83dufDIRcMiGXTM2A1UKNUt3gaG+AkFtlo1inO6gx1uEj5Fa4utrMyp5LN33FAaeKpoiIAixlKxiWjSZpbDUVNgyLjWKN3pAL2wKPKqPKAsVqg4BL5VB3AFkS6PJrxNwKlinSM7wfVZVa3uCtfc2Udb733jxen4+aCaokcGWlgCyJ/LMn+/n+1Q3mM2U6/RpPDUepNQyurRVxOyTCHgff+NEstYaJIgn0R1z8xFiCuEcmW7WY3ijRG3KTKtW5slxgfKXAC2MJdMPiby+tYtk2lYaJZdl0BDzYQjO7H6BsWDgdEovZZnSiQxJZzVcZ6/DxpWNdDIY1Dnb5yFcMJFHgmX0x4j6NgbCzdWzFhs13J5JMbGbHR70Onh+LMxJ1cbNX27ZtPJvXckv8+1QBtyTxlSNhcmMRapaAJEh4nTKaIoLcnIiriRa2Uaei2zsSeban8twPliCSq5oU6wYeh0xAkx7ofR4Gt6se35y0Ytt2q1qbTCaZnp5GEAT29g4QOJTgyloJURQ50hNgMOK6J2G/tf2tvyUfdCX7buK6qtus5Cq3/DxZrFHRLTzK+7tmjyuGcjv303dwJyRJ2lHdt227lcyTSqWYmZmhXC7ze7/3e5w+fRpZlu97YVOr1Xjqqaeo1+sYhsGXvvQl/tW/+lfMzs7y4osvkslkOHbsGN/85jd3zH9o0+ZutMV9mw8992vLKZdvnby4nWq1ysTEBIqicPLkyQf6o/mgfnjbttFkgeGYB5umMN06unxVbwnZQt3iry+tM5eu4HK6uLKc58JClv/8yQHGVwo7sqlN0ySby/HOjMV/8fQYTkez/CYA3X4HPYH4ju1vP4abbT+6Cd++tNqK+BQF6Ag4Od4X5HhvoJmGY9sc6216758abjaX+pwKc+kykiDQE3Tx5mwavdFgf3cIRW7ui1dTCLoUnhyOUKwZHOkO8u5CBt2ycKsilqEzEHHyg8kk85kaZ/eEifuclOs6EY9KwqehySKvXU9xbS1HsC4yFPeiyhJz6QrvLebZE3Pz3GiM3pCTumHhUSUsG54eiRFyKfzd5XV+dD1FX9hN1K0wtV7CMFf5yqkeVgs1ppJlarrJWKefPTEP7y3msIE/fWeJYt1AE21Eqbm9fR0+XIrEar7G4e4AFxdzxLwaB7p8vDmToZmXIrA34SXhd/I3l9e4tpzGMExUSeTZ0QQut05yNUMgEMDpcnFuNsPFbfGqG8U633pvheiTA4SduwuWLSFj2zYbGxtMTU2xd3Q/uZpFzbC4upbh7bkMoiDw5FCET++P0enTWp/f7RGcW+Jwy7Z1L2Jft0Ven85wbiZNTTdxyCLH+kIMx7vu+HuPinutHguCgNPpxOl0kkgkADAMg0KhQFQs8UTMwjAaCIVlkvVmhd/tdt+TcHxc1eu7CVtBAIcstQbjbaFIAtL9DtB4gO1/EDyqBYYgCLdU92u1Go1Gg9dff521tTWOHDlCb28vZ8+e5ezZs5w5c+aO1XyHw8HLL7+Mx+NB13U+8YlP8JnPfIbf+Z3f4Td+4zd48cUX+frXv843vvENfvVXf/WhH1Objy9tcd/mY8WdbDm2bTM/P8/Kygr79u1r5Wo/6HYeRNx7VIkOv5NkYefTBUkQ6NocHiMIArOpMtPrRcRtXuxMucHV1UKzAkvTD1wpVyiWivh9fiKhAA711ufqd6pk3nwM2arO+Gqh9f+WDcvZKtlyg/6Im8hmg2HQKXGo20dPyMVKLsH5xRzP7ItyuCeAV1OYWC2wUW+QKTfwaDJDUQ9Rj4oiwun+INmqznDcy/GlIHPJPIJlsK+nC8O0GV8pcnowzLvzOV6/fp3huBeXIjIU8/CTBzpYy1eb4sm26Qu7+falVY73Bbm8lGc5V+U/vrVAuW7wqZEYP3e8m//z5SmShTr/8oW9XEsWCblVzG2LGt20GF8t8Z/eXuTSch6AN2czfGokzunBMFPJEuVaA71WpSaqPDcW5YfX06RKdQYjbryajM8pI4gC37m0yqn+EGcHI6SKNXpCLvbEPLx6PcW1tTKaw4m92Wv95nyBfZ29UNzg2rVrOAMR3rqaolQ1UFUHqtK0AhVrBiu5KhGXZ8e1FASBbNXk6nqBtVyViGbT5ZU4fvw46brIXDpJsWbytxdXW7/zytQGNd3gF0714nOIrc9B63pvS+W5OYJzS+hvf70gCCxmarxyLdmy9dQNizemU3j3+Bi6Q9/Io+L9WENkWSYUCu14r3K5TD6fZ3FxkVKptGOars/nu6dG3Q+Ku4lrv0NkrNPPO/OZHT/fm/Dhc9w9w/9umKb5sRX3u6FpGi+88ALPP/88L730EufPn2dhYYE33niDv/7rvyYSiXDo0KHb/r4gCK3BV7quo+vNAs/LL7/MH/3RHwHwta99jd/+7d9ui/s290Vb3Lf5WHG7inqhUGB8fJxwOMzp06ff9x//Bx2WJYnwxFCY5WyVVKkp8EUBTg+FiHlvPEFYzlV3zfefWi/y9EiMc9MbrCZTSJJMLBpDkSWO9wbva3rtzbYcAMOyaRi33uCrDZObNVquovPn7yzRG3bjUkUCToW/u7xGX9jN157ob1bFLYt9CS9nBkItX7gsQtSt4JMtJH+VQyEfgXCEl68mubyUpy/sIuxxIAoCw3EvEa9Kh1+j0jAZXy3QH/EwMScgCDYOWWJitcCx3iD/5suHWcpVW773Q91+fjSVYnqzGl9tmEytl9gT8+B3Ki2TS1fQxavXNxAFcKkylc0puq9e3+DJPWFcskCjVsHjcoEkky3X+ZWnBnlnPotfk+kJuTjSE8CtyrwymeQbr83SG3TRGXRyZaXAP3uyn0tLOYo1g7os4tVkBJre+pV8g7MDffT29lKoW3hnauRqBcrlEjldRxREVIeKrscxDGPH5zZVMfjjtxZZSjenu3rcHvb3Rfn5WJTFbIGQ28EPJpd3XLNizWAhW2Wj1MDnuDXR6WbxfnPmvmmaO8S+oihcWs7v6te/tJzn7Ejnrp+9R8nDtMJsiS+Px0NXV/NJRKPRIJ/Pk81mmZub29F4+X5iOB8Gd22oxeLZfVE0RWJyrYBl2+yJefnEnjCK8P4XYZZlPXZbzuOwBum63rLl9PX10dfXx4svvnhPv2uaJsePH+f69ev8+q//OkNDQwQCgZafv7u7m+Xl5bu8S5s2O2mL+zYfet5PWo5hGFy/fp18Ps+BAwce2nhwSZLQdf3uL9yFTr+Dr57tZS5VoVDT6Q+7SfgcO9JuEj4NQRBuEd9Oh4xTtPn0/gTnXA4qdROfU+H0YIj+yP01c+1my3EqEhGPo7Xw2CLu19C2xQBaNrwxk2Y1X0OVBY72hfk3373Gar5Kb8hF3Odgf6efQ90BLixkeWUqxYm+IB0+FZEbA8L27dtHJBKhUDfZKNbZ1+GjO+jiH64myVd1fvZoF4WqzuXlPHGfhkuVGIy6+d55AdNqLkY+MRzBBn731dnmNFpF5GBXgJEOHwGXyr/+wgG+N77OfLrMwW4/G8U6XYEbQ49CbgflxRyKKDAQcbGSa/Yb2JZFXdf55N4o6UqDZKGOZcNCukxn0MmvPDVI2K0iibBRarBeqPPU3ijFukG2rFNaKzLS4cXtkJu2pVSFqm4wEPEQdMrNJuRtQ7KCbpVnRjt4yRbJV5ufLcsycUkQ1uDdd9/Fsix8Ph/xRIK3V3WmltbRdZ1wKIwky8ymKixkq7gdMvmqQcPceX0lUUCk+bTiXobD3Zy5v72avyX4FXHzc2qDIApAM31IkcSbxpp9cDxKW4yqqkSjUaLRKLAzVnF6eppKpcLFixdbYt/r9X5gYvNebDFu2eb5kQinB4Jgg/ceG4Uf1vYfNY9D3BeLxQe+t0iSxIULF8jlcnzhC19gYmLiltc87iblNh892uK+zceK7eJ+Y6Npd+jp6WHfvn0P9Q+kJEnUarW7v3AXbNsm5JQJ995o1Lr53wejbnqjHmbX8je2KQpE3TL/+99fpisa5lB3gP6Im5BLIeCU77tiuJstJ+iUeG40xrfeW6HSaP6bV5N5biSOX7uxjXLDZD1X43BPgIFoswHVqUgMRjysbTbj/t3ldT53pJNsRWc+XWZytcDPn+yhsjaDbdutJjQAvyZzciDEaq7G7746TcO0+amDHfzbl6fIlHV8msxKvsZTeyP8ty/s5efPDHB5tUq1bvCT+zv4X//uKoWajleTaRg2/+mdJTZKdRqGxVSyyL94dpjVfJXPH+nitakUhmWhiNAZcNIfcRH1OFjMVtBkkf6wk0qt0YyIjPpYydcY6QhwqFtkar3ASq5KZ8BJp9/BpksJf1ijy+fgYKePTw5HmNkoI4oCDcNCFASO9QZ5cyaNZcN6oYbX4SHqVekNuTBtWC3oXFpKUdUtntgTJVup8/ZsmljAw/OjcfoiGn2xk63hTaVqncuz663El2qthkNVURSFybUiT++LspStsi/hY6O40bq+UY+DoFsh5n2w2MMt4bZdwB3pC/LeYo76ZuoO2IgIHO0NYuiNB27SfVA+6MSWm2MV33rrLYaHh8nn86yvrzM1NdVq5t16nWOXpyYPg3sW15aJd8tN9JCEPXw4Gmofxz4UCoX3PZ02EAjwzDPPcO7cOXK5HIZhIMsyS0tLdHZ+8E/A2ny0aYv7Nh8rJEmi0Whw4cIFAI4fP/7QRtPfvJ0HHTC1xZ3EVUCT+PLxHs4vuElVwaMKxHwaFxayyA4364Ua64UaMa+Dr57teyChdrsc/bGEh5h3gIVMBVEQ6Am5iLp3Lh4USeRQb5AfXU8xvlqkbljMp8t0BV10+DWmkiWgmZzTHXRxfqHB1bUCs8kCh7v7CPncO7Zp2zbDMS/fHU9SqBmc7A+RKjW4ulps5oR7/Fi2Tb6ic2Exx6GgxFhHB4qqcT1VJuxxEHKrnBoIIQgChZrOQrrCQNTN315a5T+8Oc/Xnxri371yna8/swcRCLgUYh4Vr0PkyeEIf/HuEqVKjWwux1hvlH1dIf7w3DzVTZuOKkt89nAnnz/qxKOKyIK9Y/81GTTAqyoMhMLMpKqcm01zcTHL3oSPF0/18fp0irpuMhz38KnRGBG3wnvLBf7qwkrLDvXeUpYn90T5Lz+1F0VsxlxunXtBECiVStTqDfZ0xynYDgzdoN6oUyqV0A0dZ7+HRj7F/riHmFejVNO5niwRdqubMxkSBF3SQwu17/I7+Nlj3bx+PUWm3MCtChxKOOnwSC1Lz3Yrz/tJ5bkXHncc4+0adfP5fGvIlq7ruN1u/H4/gUAAt9v9UPbZsqzHeuz/WCv3pVLpgcT9xsYGiqIQCASoVqu89NJL/OZv/ibPPvssf/qnf8qLL77I7//+7/O5z33uEex1m48zbXHf5kPPvd6sbLtp9yiVSgwPD7cemz8KHrSh9n5QrRp73XUGnTbBeCfffHOZQm3nNpPFOoWqQUC7/5vZbrYfaPYAxDwKcW8A2H0R4lIlSnWDl68mGYi4Gev0ka/pGOkyY50+dNPCsCx6Qy4Sfo1feXqQuFfD73agOXe3D9m2hVMW2Rf3kvBpzKUrKLKIZVjYls1IwkuxZjCTKnMw6ASjjsfrQpNFRhMeRjp8vHJtgx/PZfFqMp/en2BPrJnlfmkpj6ZIHOsLki7WeXMmRWfAyYsnu7Esi70RJ186EuPt2Q2kvhBnhmL85YXllrAHaBgmb1zfYPB07w5hvxumaSJLApNrRUzL5u25NH0hF5892EHU62B/hxenDLmqwWtTqR19DrYN78xnONrjJ+C4MdysXGuwlCpQsVx0Jjo4acP1VIUCArKi4HZD3KdxfCiBXkxTz6wQVBz88tku6paAZduEPQ7CbgXhIfrCRWyOdHkZCDnJl6uUi3l6ooFWSshuVp6t/7+fVJ575XGL+92QZZlwONxq4rdtu5WhPj8/T7lcbom8B52QCo9fXD/u7W/xQV//BxX3q6urfO1rX2t9N7785S/z2c9+lrGxMV588UV+67d+i6NHj/JLv/RLj2Cv23ycaYv7Nh8LSqUS4+Pj+Hw+3G73IxX28OANtfeKYRgsLi6SzWY5deoUVRyYuzw9FwUQ7zPCzrIhW20Oqgp19tMMqtwp9rZu0LdLOqnqJvOpMk8OR+jwaRzqbubhv3Y9Ta1hEnCqBJwynznYwfevJvnWxVUqDZNn9sWQxE4Odnpuaf51qRJRr0q20mA1X8PtkJAE8GlNz/pMqoTHodAZcGKZBpbVfHoS9agc7wvxez+aZT5TYb1QZzlnY1gWXznVx+mBEGv5Og5FpC/k5M3ZFIZls5StUNUtrFqFiYkJEokEv/jEMIIgsFrUyZTqiEKz4danKdR0k+Vshbph4VV3X0wJgkBFt6noFm5H02r01kwaG5hPV8hWdL58sgen3BR4DcMms22gUOv8NkxKdZOIqynuk9kCP5xKc2W9ii3KyFKaZ/ZGefFUL+8tZNkoNegJOjncEyDskrBd8R3DeAqFAoVCgYXlNFP1Oh6Ph0AgQCAQwOPxvG8xVKvVmLnajJfdu3fvDmG6m5VneyLPbtX93X7nXvkwivubEQQBr9eL1+ulu7sboJWhvn1C6nYrj6Zpdz0u27Yfq7j+MNhyHgfFYvGBxP2hQ4c4f/78LT8fHBzkrbfeehi71uYfKW1x3+YjjWmazMzMkE6nGR0dxe/38/rrrz/y7T4MW87tSKVSTE5OEovFEAQBt9uNw4LhuLeVP79FT8hFwHnvX+OaYfPmbJa35zLkqzpWtcDnBDdHe/zIYjPBZSlX4+JSnoZhcaQ3SF/QgSrtFBUCQrNx1q/xD1eTvHw1yd64h58+mEAAaoZF0KVweTnPn59fxqXKxH0a+UqDv724QszbT4dv53wBhyTwyeEoa/kaa4U6Z4cihD1Nb3K63KBUMzg9EOJ4b5DS2iyGYeD1egk6ZQSxmQSjSiJht4rHIaPJMi+Nr3Owy89oh0CtYXJ5Kc/BrgBRr4YsChiGwcrcLGNjY3g8ntYkYEmAsFvlUE+QpWyVq+tFwm6VT+6LgSCwVjJAEHApIn6H2BKUC7k6L40nWclXcaoST++N8/ljPVxcytEVcHKo249XFWlYoIoCsiTgcypsFHc2MDtkEbfa7OuYnp5G9yS4nKyD2LzWhmnz0kSSn3Op/PTBOPrmpOLtk4y3kCSJYDBIMBikr6+vVTXO5XLMzc1RKpVwOBwtse/3++9LoKVSKaamphgaGiIWi93T79xLKg80v9/3a+X5KIj73XA4HMRisdY5NE2z1ag7NTXVmoC6vVH35vPxuCvnj3v7j4tSqYTP53vcu9GmTYu2uG/zoed2N+p0Os3k5CSdnZ2cOnXqA72pPApx32g0mJycRNd1jh8/jmVZTE5OAs34yOdGYiDA1HoJy7LpCbt4fjTeipi8G4IgML1R4uWJ9VadPlmo8p1Lq8R8Gr1BjXcX8/wfL10jU2lWk51vLfDLTw3xE6MRRKG5OKg0moOK+iJu/qe/uESp3nyCkS7VubZe4n/4zAhxzaQhOnlrNstIwofHIeNySDhlkZpuMpUs0ekP3yJE+0Ma//SJfsZXipTqBv/LFw/y47kMk2tFPn+kk2dHYsTdIo5wmMXFRRYWFgiFQlTsAH1BB7bgxLShVDdYyFQIuBTODIZxOyTemUtzaijCa1Mpzs/P4ZSbsaSfOzaK16vu2JeAU+aZ0Th/eG6BKyvNpmbLhvHVIv/1p4b5/mSS9xZzdPo1nt4X40Svn4Zp8pfnV0iV6gg0F0nfv3qZr5zqpT/kwqMp/GByg2y5gUuVODUQYjDi4sxgmG9fWsXcNpnsaG8QsVHi/NVxDh06xEszJRBuvc6XlvIc6PAgi/ceAbm9atzT0wM0K++5XI5kMsnU1BSCILT84IFAYNcGUNM0W6Lz2LFj76tJ9E6pPNunEz8qK8/D4mHGYEqS1Dr/W+9drVbJ5/Osrq5y7dq1Hc28fr8fy7IeyM7zsPjHXLl/WElsbdo8DNrivs1Hgu2xfVsiuNFocPToUZxO5y2vfdQVpIfpubdtm/X1daanpxkcHCSRSCAIAvV6fYctJuCU+NlDHeSqBjbg0yQc0r1XKA3L5sJS7pZoQsOyGV/J41Il/uydxZawh6b95v/78QKHu/3UDJNXrm2QLtbZl/BSapjE/Rp6popuWjhkEZcMb88k+YUzfWRyRTxqM9bTxkYWmxGJ2OwQsjcT9ygk9oUBgULNoD/swrbBr0mYho5lWS3hCU1hqpYM9No85WodWZbxeTyMJrw8uSfCWKcHy7Kp1P18f3KDt2Y28Gky3SE36yWDb11c5RdO9eBVb5xLRWzuZ6Zcx7U55dbnVFBlkb+6sIJpWSxkKixmKmSrOoZpsTfhaQn7qmEjIJDwOzm/mONgt5//cG6eybUSIwkvDllgZqPEzx3v4ViPn4BL4d35LLpps7/TS1Cssr66zIkTJ3A4HDiV3ZOZHLLYOqfvB03TSCQStzSA5nI5lpaWaDQaeL3eltg0TZOJiQm6u7sfehIVvD8rz+Os3D/KbQuCgMvlwuVy0dHRATTz1QuFAvl8nqWlJcrlMi6Xq/UdeViNuveKZVkPNPH7YW7/cVz7YrF4z0+t2rT5IGiL+zYfGWzbZnV1ldnZWYaGhojH47v+Id+Kw3xUNxlBEFoDS94vtVqNiYkJJEni5MmTO/Z5t6hKWYSI+8G/trcrLIqiQLluMpsq3/JvyWKdTKXB27NpFtIV6qbFWrFOslDDtmFv3ENdN6nrJrIkogsSgiChqA7iPhffvryBZdm4HBJdQRdhl8xw3HPHKudWpKJ30/JiWdZtYxU1TaPHAZ8/vZcfXtug2jDQG3UGgiqHoxKX33kLn89HIt5DrlhmOOrG47xRZV7NVcnXDLyq0lpECoJAslAl7tPQFImablGo6aiyyFy6zIHO5iN4G8hXdC4t5emPNBOA0hWd5VyNlVwVhyxyrDeAy6FwZaU5+de0bUDAsGzenEkzGHYyHHEyFHGSzxdZmJvBmUjQNzYGNAXLwa7mVNGafmOxJ0sCx/qCbK5DHio3N4BalkWpVCKbzXLp0iUqlQp+vx9d18nlcvh8vkdesb2Tlce27dZ3pV6vt/7/4x7DqSjKjus0MzODLMtYlsXc3FwrKnXLu/+gjbr3yuOeUPu4bEHlcvl9R2G2afMwaYv7Nh8JyuUy4+PjOJ1OTp06dceR74/KDy8IArmqQbrSoKbbuLv3ka4YhF33/zWybZulpSUWFhbYu3fvrg3AD/s4ZFFgf6ef8ws5TMtGlUVUVUUUYSTuRZFFQm61ZbOx7WZl+KcOdiAKAm6nypPDbhayVZKFKoe7A5xfmMXjUFjKlJHlprh7YX+CUsPku1fTSJLA03ujvDGTQpFEQi6FnzzYQcJ7bxaOLZG2JZpuJ5xkAc4MBNkb97JeqOHRZGIeFY8q0hv1bw4yK1CvlqhUGjSqCg5VRXWoaA4VRRJZLxmkyw18zmZEZkfAyWKmQrrcwKbZ5CqLIgMRN1XDwrSaE4clUaBmGGiKhCyJzKcr6Ka9aWOyCLpUCpUG/RE3S5kq0rZjKNWbg6ZkLK5du0aj0WBsbOyWhWnCq/CFo928OpUiU67jd6qcHQrRG9Q+kImooiiiKAobGxvEYjH6+/tpNBrkcjnW1tZ2WES2qvuPuoK7m5Unm81y7do1BgcH79nKI4oidbP5pEWVbt9Efi88br+/bdu43W7C4fAOy1U+nyeVSjEzMwNwS+b+w9rnxz2h9nHZgh60obZNm0dFW9y3+UiwsLDA0NAQwWDwrq+9eUrtw2K1UGcmVeFvL65ycSlHuVLlyX2d/OKZXkZi9z4ddmuh4vF4dgxyuhlRFB+qcEuVdaq6yXDcyw8mk1R1k71RDz8xlqAj4EBE5DMHOvjdV6cxbWiYFj9/soe1Qo1/+/J1Xr2e4rOHOvjpgx14VJm6YfHf/MQ+Xh5PMp+p4JREPjkcYW/cy7mZDN+5vEoyX+Ezh3r4n39mP3XDYi1fw6lI1AwLr6N5E97tGLc3V95r9VUWbGJuibjnxlOBrXPt9/vp745zcgQuL+ea2fD1OsViiU8M9fLm9SSvX1tBEBUcDoWxziDPjcboDblIlxuYlo1lQ8SjcqwvyA8mkxRqOmG3it+pkPA7Cbuaw74WMhVW8zW8msJIh5eD3X7OzabZG/OgGxaO7ZOI/RoYNd4+f5He3l46Ojp2FVoCNqNxF73BbmqGjSoL+By7zyl42Gw9MVtYWGBkZIRQKAQ0q8ZOp3OHRWTLyrOwsNBqeN4S+y6X65EJX8uymJmZIZ/Pc+zYsdZsi7tZeQxBYWK5yOWVPAJwqCfASMyNKj7Y9+5xN5TutrjQNA1N025JT8rn86ytrVGv13c06no8ngc+hsd9/I9L3D9oFGabNo+Ktrhv85FgbGzsnoXMoxD3ugUzGxV+cG2DdxeyAFi2zbsLWZyqyNefHiLkvPNNZetR+draGmNjYy3P+AeBZcOrUykuLGTZl/Dx9aeHNn39NUY6vKhi07j93EiUoFvl+5NJOnwO3A6Z1VyNYs3gULefvrCb//HPL9EZdJIpNxjt8PHFY10c6wuiKiJOReTliSQhj4plQ92w8LsUvnlunktLeQQB9sQ8HOr2c6o/xFKmwomBEJosIQg2PoeMbVt3rNbXTCjUTBqGhUOVCGoSyraG0q3EmMXFRVZWVlopStBsSq4bFrMbJRRF4VjcS388yB+/NYtpizSqFXJ5neRGis6Axs8cSjDa4ePySoGox8HJgRDpUg0R8DhkekIueoJOnhuJ4VREQrd5wgAAIABJREFUNFngK6d6sW1QZJFUsc73J5PsiXo5OxQmXbqRihN0yRxOaMzPTHP48OFbekduxrZtXIqAS2mej92+D7YgkKuamDa4FRGXIryvBWKj0WBiohlxeerUKXJ1eGUqw3KuSm/IxViHl6CzGdepKAqRSIRIJNLav2KxSC6X4/r161QqFZxOZ0vs+3y+hyIEK5UKV65cIRKJcOzYsR2flzum8ggir0ymeP16c4KvgMB0ssjTI3Ge2RN6oMmtj7tyfy/ient6EjT3uVKptAZslUolJEna0ah7pyel23ncDbXtyn2bNk3a4r7Nx45HYcsp1Q0My2ZyLc+BTh97Ez4kLDbKBplSnVSpTug2w5mgOZ58fHycSCTCmTNnPvDqVr5mMpcqY9kwsVpgYrWAJAqsJZNEfS5CPU3x63OIjCa8eDUZEfiz88us5mv4nDLPjyb45rl5clUdjyYTcqvopsW/+8E0XznVyzvTKQ52+Rlfy7Mn5uVAlx9FsGmYFt+f3MCtSgRcCiDwl+8u43Eo9IVdfOviKuv5GgGnwt64h+P9AbzK7tX6bM3iO5fXubCYYylbJeJV+eLRbvbGPXT4msOZqtUq4+PjeL1eTp48ueNmH3aKfOlYJ8WaiQUEnTIvT24gCBIul4to0Idt2xRqOldXCwwEHaxlCnhVWMuW+L/n0/zkgU7+q+eHmUmW8WgyowkPAa3ZHxH3a3z70iyy1PTUm6ZNyKXwiT1hDnf7iHodzGyUcMkCYcXAJ9aJHTjwUARhxYAfXU9xaTlPVTdJ+DQ+NRpnIOR4IIGfy+VYT6Xp6h8iGvCSqZj80VsLrdjOidUCF5fyvHiym4B267XanuSyFcFZrVbJ5XKsrKxw9erVHYkwgUDgnkUk7HyisH0Bdye2W3nSVYuLS3kkUcLa7PGwbJu3Z9Mc7vLhFo37TuV53JXrB9n+Vtyu2+2ms7MTaC7qtqr7CwsLmKaJx+NpXc/bPYV53Mf/OCv39/L5a9Pmg6It7tt8JLgf8fMoKveKJCKLAi+MJchWdL4/mSRfrjPaGeCFsTiasvsNxTRNpqenyWaz7N+//7FVd0Sh6bnfsW9Ws8qobEvcSZZ0fv/1OdaLdTr9zer8tfUiY50+Ai6FUs3A51QIulUCLhVZFMhWmj51XbcwTZuAUyVZqPH8SIxOn4PLqyUAPFqz0p0q1rBoXtPxlQJ/9u4SmiIxHPOwkquQq+r8zKEE8k1tohYCr06lOL+Y49p6c/JroabzB2/M8c+fHEDAhVVIsri4yMjIyG2fjGgSaO7m9RJFAZcq0RnQ2BPzUTctDMum1jBpGCZjvRHWygbfm9hgPV8h7FLwqgJ/+to4ugVfPjOEbFSwrKbYSXhVPn0gwWtTKWpVHU2RONoTYCThQcKiz6/gqNbJZrN0xvpwu90P5foKgsA781leu55q/WwhU+Evzy/ztSf6CO4ivm+HaZqsrCXZaEiMZxzUklmGojodASc1feeieTVfZWKtyJODwbs+Wdue9rIlIrcacrcy903TxOfztcS+0+nc9buv63qrEf3EiRMP1CSqmxZV3QJBQNy2jWrDwkZEkqT7HrD1Uajc3wuqqt7yFGZrou7s7GyrUXdL7G81VP9jFvftyn2bDxNtcd/mY8ejmB7rdUj0R1wsZKv8wbnr2DaYhkGmYmDbNkd7bxWSmUyGq1evtnL4b3fTb5g26bJOtqLjc8pE3Cqa/HAFgl+T2d/lJ3k1uePnCb+bmKfZ+NjMwS+zUWpg2VBumJwdDPPmTJqVXBXDsnE7ZCoNE0UUKNZ01vI1xjr9BJwKnUEnLofMF493860Ly5y7vsGZwQCyHCBTaeDTFMAmV9UB6Ak5+fevrQFQ181m9VQQmVgt8MnhCBHXzpt0oWYykyqRLtV3RGkmi3Uy5QbZfJ7jcZlTp09TbNgkyyayKBBwSoi3yZNpJtEE0C34j28tcGWlgGXDvriHf/n8Xt6Y3uDZkURrmJVTlZheL1IWNA72+PApNjMzM61BUMFgkNFYjD3RPqq6jUMWCDhlZMFu+f/D4TAjIyO7iyBBIF9rWkd8WnNC771U3Yt1q5XHv51spcFyrkaow31P71MoFFhYWED3dvLtqxtsafnFTAVFljg9EOK1qY0dv7OYrQJ374XZDUVRiEajrYZyy7IoFArkcjmuXbvWGty0Jfa9Xi/5fJ7JyUkGBgZaPvIHQZNFAi6FXEXf8fOIT8MhC62G4Nul8mz9VxTFluj/uIj7mxFFEZ/Ph8/no6enB9u2W426GxsbTE9PA82komw2iyiK72vuwYPyuMR9Oy2nzYeNtrhv87HjUVTubdvGqUgsZir0BF2kinVsWSDmc2ID2XKD2GZEpWEYTE5OUq1WOXLkCC5X065TamyJNrkltKq6zQ+vp/nxTJqGaSGJAge7/bwwGsfrEO87sz9bNSjUTBRJIOhUcCpbQsPmZH8Q3bS5vJyjVDfpDjo52xfHJRqb+2KxkKlyfjFPqW40s+CDTn792T28OZehVNN5YSzGhcU8FlCpm1QaJjGPSqpU49P7E8S8DvI1g+dG47wzl2F6o8KT+xJMJUvopklVt5BFEU0RCbpUksUaHlWmK+gk4nFQqunUjWb1/GYEAURBoGHe+DfLBsO0sGwbA4VEdzfvzud4bSpFttLA45A50R/iicEgjtvc82VJ4PxCllSpgSqJOGQREPjO5VWOdAeYXMvz/EiU/pCTa8kSvSEXT+2N0hXQcMkQDx0Emqkk2WyWpYUFCoVCM2/f58Py+ymXyySTSUZHR287ybKk25ybSTOxWkS3LLoCLk4OhFBFCLkVXHdY8NncfnaAYd7dO27bNnNzc2xsbDBy8Ah/diHJ9iK9AMynyxzvC+J3KuSrNwRxd+DOvQL3gyiKtwxuqlQq5HI5FhcXSafTWJZFR0cHkiSh6/p9WXm2E3BKPLU3yncuraJvfqYcsshTw1H8mtT6jt7PgK16vf6BzNm4HR/UdgVBwOl04nQ6d8xGeOedd6hWq0xMTNBoNHC73a3qvtvtfuT79rjEvWEYD/w5bNPmUdAW920+EtyvLader9/9hfeNjSI1q7qDETemYaAqEg5ZotIwEASB9fV1pqam6O/vZ2xsDEEQKDYszi/kGN/MOR/p8HK8N4jXIbKUq/L61EarrmxaNhcWcnQHXZzuDyCK4q437K3zsSVALBuurJZ4aXydbKWBJAoMx7185kCC4Gajr0cVOTMYJOHXqDRMDMtCN01Mmk2rF5dz+J0KlYZBvaGTrDf4xmuzfPFYN7/29BBRj4otCOyJZfiHiSS2DV881kXYrfJ//WCaf/2Fg/z9lXXeW8wS82qcHQpjA2v5Ks+Pxrm0nKPSMOjwa4wmfOTLdb50vAeXQyZdqhPzarhUiaVsBdcuNie/JjGS8HFtvUyu0sCwbAqVBl1BJ3XTwiFJjK+WmEqWyG4O4irVDV6ZTBL1Ojh4U/XaFgTWCg2SRZ3ZjTJRr4Owu9kIXKkbvD2X5Zm9MZZzFWzbxqsKnOjxtbLwb66Ea5pGR0fHjvSYjY0NJicnsW0bVVVZWloiGAwSCATQNO3Gddy0HJ2bSWPZsFao8YPJpn/+icEw6XKdnz3cid+x+/fA6xDZE/OQLO783LtUie6g845V+2q1ypUrVwgEApw4cYKSTisOtXWugLDbQalu4FAk2BT3cZ/GWIf3kaX2bPnBBUFgZWWF3t5eOjs7yefzZDIZZmdnsSxrh5Vn+3m9I7bNkS4fHX6NybUSggD74l7iXuWO5+t2A7bK5TIzMzN0dXVhmuY9W3keJo/TFiPLMqIoMjAw0HqKUS6XyefzLC4uUio1m9i3W3ketiB+HOL+g4iibdPmfmmL+zYfOx6FLQfA65DpDjjJlBoookCmVMIhe9BkhbhX5cKFC9i23ZoqCmBY8IPJDX48m2m9z0quSr6i8zOHOphYLexqGJlYLXC0x99qDt7yFNcMm8VsjSsreRyKxMFOHx1+B5mKzrcvrVLeFGWmZXN1tYDfqfCZAzFEmtaNP3lniflUpbWdUrHAVz8xzIjDxfmFHHXd4p8c7+LvL69TrBtIooBl2/idCp1+BxvlBqv5KiMJL7YNs6kSL0+s43cqFKsG52bT+B0yXSEn0xtl/t8fXqMr4udgp4/+iJsnhsL0h10sZytUdAtFFvn91+fQN6vLIwkvv/bsHvzarTGgAnCiL0SxbvLS+BrXkyX2d/n5zIEOstUGkbCL/+eHM3zxWDdTm558aArT95ZyjCbcSNBK4JnN1PiTHy9yZk+UfFXHtGEpW202EwvNNBzLttkTvbEo2E3U74Zt22xsbLCwsMD+/fsJhUIty0k2m+Xq1avUajXcbjfBYBBHIMaVlRwAFd1kNd+cSHtpOc8LY3HeXcjwznyWT+0L77592+bMYJhsVWdqrYhh2QRcCs+PxjfnMOweN7o94nKrWu5SIe5zkCrtXCi4VZETfUGWsmUWMwrdAScHu/2ttJxHwe2aZmOxWGsi6Fa0Yy6X23Fet8T+naIdJcGm06vQ7b8xrOtBRoKtra2xsLDA2NhY82nNfVh5HiaP2/MONxYwgiDg8XjweDx0dXUBzUbdfD5PNptlbm4Oy7Lwer0twX+7Hot7xTTNx1ZBf5x2rDZtbqYt7tt8JLjfyv2jGGIli/DU3iipUoOVXBVBEHBIAk8M+smuzNPR0XHLCPJc1eDy0q1e6CsrBZ4ZieFUd68yOZSm33p78o9uwY+m0/xw8kal/525DD9/sodKw2wJ++1MJ0sUamECmsRaobZD2ENzEXBuOsVAzEe11uDthSynBmP88tNDFGo6IZdKb9iFgE2xbuLXFASEppVhW7X2YLeflXyFzxzoIOBSqenN/fn88T5en8mQKjfIVhpMrxf5558cYDjqYiHf4MpKnsGoh3yl0RLdEysFBoJRtl9xG4Era2VevbqOV7X57z69j7phk6/q2LaNpoj8eCZNsW6wUawR9qgkCzc9vRFgo2QykypjWjYOpTkxN1epMxj1sFao4XZIGKaFKoubfRQ2w7E7T9O9ma34SFmWdzR77mY5KZfL5HI5SuUyyyvr2JJMwVQwTROHIiOLIqZlI4ki0xslzg4Gcd7GnuN3CHz+cAe5vVEahoVXkwlqEruJ1Ts1pCqiwCf2RFjOVVt+dEGAk/0hunwqQ2EN07aRNu0nj0rY32vT7G7RjlvndWFhoVUx3jr3fr//lvd60CcPhmFw9epVgFuu9da+bX//3aw8dxqwdb88bs//3VBV9ZYei2KxSD6fZ3p6mmq1iqZpLbHv9XrvqxL/OCr3H4YFVZs2N9MW920+djyqIVYAMY/CV071kCzWWU/7CTklXIJOV8fgrhUj07ap6rcuNGq6SU03Gev08/ZsdsdrZFHgeG8A8SZxn6novDmT2SHVGobFj+eyjHXu7uOWRAGB5uIoe1PjIFs/L1aolgoc6w8xmaryh28tIAjQF3IR86p87kg33760QtCpcmowxOeOdJIs1jm/kEMWBcIelaf2REkEnfzNeyv8xbtL9IXdTKwW+PzRLn76UCeXl3Jg2xTqOg2jeQTz6QqLmSrJQq1p9QDS5Todfo1n9oZxKzdEykbZ4E/emGY9ncXn9xP0F3ntegqnIqEpIqWagQCE3M3Jsr1hN+uFKldXC5RqBmcHw0ytV/irC8ubliSbxWyVzx3upFo3+NRojLdmM3g1BUUU2Jfw8txIjIhb3rEfdyOZTDI9Pc2ePXt2nTp887nfqmzWTDiyr4/r60Ua5QZxX9NaEnCpJHwag2GNhmkjiXfeF02ChGf7n/VbhXc6nebatWsMDQ3dshiFpkDs8qt87Ww/1zdKFKo6QzEPnT5Ha4KrAJvxkY+GracbAwMDLU/3vbL9vHZ3dwPNRs9cLkcqlWJ6ehrbtndM090aenU/bMXb9vX1taxYt+N2Vp7tE5jvJ5XndnzUhOb2uFRgR6PulsVxq5l3+0Td22Ga5gd+/OVy+aGlXrVp87Boi/s2HzselS1nC59DpJjKIOeWcChR+oeGgE2hXDUwLRuPQ8IhCWiySNynsV6o7XiPiMeBQxaJOiS+cLyb165tkC43E2XODIXoD7tax7JV5StU9VuiCAHmU2We2Rcj5FbJlBs7/m20w4d/s4G3068hi0KrWdWyLZyyzXMHulmpSgTcKk8MRZBFsCw40R/C71L47vga1YZJtW7yNxdWePFUL7/2zBDzmSqlmkFPyIlPk7mwmOdP3l6kqlvopoVlwx+dm+e3P3eAoKvZhOnVZNTNyrMmSyQLNWzYcVyVhrmj+mgYBhdn18gUS0SiUURRZKNQY1/cy7cvrbEn5kERm1n+lm1TqOr8+9dnGY77+Mn9CWzbJuRW+eO3Fqg0mtuRRAHbhm9fXuOrZ3o5N51iKObhRF+Q4biXbr8DgXuvSm81URuGwfHjx1tJK/eKUxb41GicUs2gYQksZPMIgsDzYwn+4PVZekNOnh+JcPXyRWRZ3tW3fzdM02wNkzp27NgdRZJt24ScIqf7/LftMXgUbE2azeVyHDly5K6Dve4Vh8NBPB7fMaV1a5ruysoK9Xodj8ezw8pzu/Nq2zYLCwusr69z6NChVsP8/bJbdf9hWHk+zJX7u3G7Rt18Pt8asqXreqtRNxAItHoyoHkOHyQW9f1QLBbxeDwf6DbbtLkbbXHf5iPBh8GWA808463mw1gs1hJIFd3m7fkM7y3mqOkmHQEnz+6L0h3QeGYkyrcurFDeFJYuVeK5kRgBrXlTH4276QloVA0LhyTsSNMRRbF1LB5NRpHElj99C59TwatJ/OzRLl4eX2c1X0NTRPZ3+Tk1EGSrehv1qpweDDOxmsch2eRyBZ44MsjrMxnyVbMZAZnw8PWn9zCbKrNWqKEbFgc6/ZxfyBLyqCiiyDvzWbqCTt6dy3C8P8xfXVimJ+hicr2ELIk4bMhXDXpDLmZTZa4s5wlpAou5HE8d78clmoBExKPSE3KxkKlg2c1KsCqJHOsLshX8sr6+TjKZxKElCAZDrWOeS5c5OxTls4c7SBbqVBoGAZfAJ4ejXFsrMBDx0DBMFrMVvnSsk2xF37HAEoCY18H0RhlRENBNi0tLeQ50+Tk9EATbvGfndSaTYXJykv7+fhKJxAOJK9u26fGr/Gdn+5hYK3F6MIwiCbw1m2Y2XaVm2Pz0YY3hgwdbovRmf3kwGLytKC0Wi4yPj9PZ2cnevXvveR8/KFEPNybNhsNhjh8//khFqiRJhEIhQqHmZ8q2bUqlUitvfyvadLuVR5IkGo0GV65cwe12c+LEiYdaJb5Z7AOtav6jtPJ82JFlmXA4TDjc7IvYulb5fJ75+XnK5XLLdlWpVD7wBtf2dNo2H0ba4r7Nx45HYcvZqihubGywf/9+fD4fS0tLGEYzJefCUpZ/GF9vCcLiWpFsucFXz/QyGvcQ/cQA8+kKNtAXdhF137Dw2LaNWxVxq2Lr/7fYbssJuRQO9wZ4e1tzriQKnB4M4VVFfGEn8dM9VBoWotCcvrrdxeGQBM4MBol4FC4v5xkZ6yfm1+gt1jACEnXDwu9S+e//7BIRj0rDsJhKlvjyiW6+9sQAs6kSlgXDcQ9Bl0rM5+AfxpP83eU1Pne4E0USKNcMnKpMTTdxqxKHuv30ht0kfA4+OdJBh8tm/MplTNMk0D3Ml4518sZsjvGVPB5N4RN7IuhGs/J/6eoVAEZGRiibMpMbNZazWwsBODeT4p8c76ErqJGtGLy3lOXiYo5spYEqCYDAcqZMsWogiwJuh0ypZlDRLdaLNWwbhqJuQm4HIwkf+xJeBiNuXPf4V3GrEl4ulzl69OgDWTu2Y9s2igjTySLJQo1MRccwbfbE3CiSwMxGiYFQuCU4+/v7d/jL5+fnKRaLLVEaDAbxer0sLS2RTCb/f/beNDiu+zz3/J2t970bjX0nCALcF5GyFku2LNm+duQkjq0kSqI78ZSn7twaexKnblzzabZKJZXrmWQqSU0W34nHE2diO07seCLHm2RLsrXRIkUCBEDse+/7erb50OgmQII7QEJy/6r4gSDYffqc7j7P//0/7/Ny6NChPWsfWFtbY2Fh4YbDx3YTQRBwu9243W66u7uBWrRpOp0mGo0yPT2NpmlUq1U6Ozvp6em5J6JaFMXbsvLUf3Y/BP+9EtWbr9Vm21Umk2F9fZ2pqSmALVae29nhul2aA6ya7EWa4r7Ju456DNtOkU6nuXTpEq2trZw5c2ZLha1arZIt65xfzGACPodCi9uGbpispoqspsv4212EXQqt7iuNlLfKZnFvlQQ+cKCF0XYPqaKKIECb20arx9J4XKci4lS2v7HHMwW++bMFJiIl7A4HmgGqblLURZaSBY73+Bo7DrIokquopIpVeoNOvvrmEpFMmXSplkP/kcNtDIZdDIRdtHvtvLmQ4rmH+vjXi2sk81W8dgv5qobfqTDS7ubSaoaFRJ6BkIuHjhzHKppkilVeX44jmTqP7QtQVnXenIvx3qEWzs9FON7Tg8vjI1ZQWUzmONDu4VCnj/NLSVJFdWPhYMcpQ7JgcHY+dc1rFkUBQQCvXeJYt58fTESZjRcaNqDj3X4ypQoP7QvR5bXccmNlJpPh0qVLdHZ2bqmEGwgUqjoWScQmC7f9PlQkEcM0a9cXNhYpYJq19J5rEoS28ZfXRenKygrRaBRZlmltbW1UOG/XMrSbqKrKxMQEgiDc8aTZ3cJms9HW1kY4HGZmZoZMJsPAwACFQoGLFy9SrVZxu92NxdZme8hucT0rj67rrK2tNb4vdjuVZzvup9/farUSDodZXV1ldHQUSZIajbqXL19uDEPb3Ki7U8farNw32YvsnW/SJk1uwP3wkWqaxuXLl8nlchw5cuSaqmfjRmqYGKbJA/0BREFgbC2LQ5F4cF8IRRbvyrO82XMPkC1rvDabYD5RRJEEBsNuHuwPopu14VBtboWrey5VVa1NUZV9zKYN7A4HPodCV8DJn/zgMrIAIbeN0U4v55fSZIoq+arOgwMB3jcc5t8urrGYrKXsLCaLWBWJL748x//0sUP888+W+dDhNr5zYZ21TIn/9NQB/uncMrmSxuEuLwc7vLy1mOaHlyIMtDhZTpYIuqwc63Thd9l4eF+YqhZlOVViqNXNw/tbSecKnJ2NMBTq4dzUOj+ciGKKMqIg4rIp/OLxLlpcCi6riLSxVxJwWLbtbdjf6sZrlxBMk0f2BRA2eg4M0+RIp5cuv4MfTcaYjRX41VOdWG5yvzcMg7m5OZLJJIcPH268JwRBIJLXeGUmzmq6jNMicaY/yFDYgSzc+nVXxFqvw2yssGUolc+hMNhya1NmrVZrw7pw/PhxXC5Xw8qzuLiIpmmNXHi/37+rVc0bUbcW1e1Me5FSqcTFixcJhULXWIUMw2hYeWZnZykUCthstobY93g8u57cIopio99DkqTGMd4PK8/9aGbd7hgkSUKSpGuSqUqlEplMhrW1NaamprY083q93jte9DYr9032Ik1x36TJNsTjcSYnJ+np6eHAgQPbip+6uPfYJB7oD/D6fIp/fXu1Yc15eTrO5z88clfHsdlzn6vo/Mv5NZaSRRAgWdC5tJ5nKpIDTCbXcnz0aCfvHQqRK6v4HAqiWmD84kUOHDjAStxsiI3eoJNXpuNUNZ2yafKxfUHOLaZ4/uI6B9rcRHMVXp9N8NkPDNUGCVlk4vkKVlmkWNEoazqr6RIl1cAiiaykS/zFi7N89HA7v/fUMHZFoljVmY8XCbms/Paj/Swni8zHC5xfSnOwfaMJtljF65AZbg/zk+k4L0/H6PTZeexQHza3mx+9NkGholEqZjFNg4zFyovj8OsP9jWEPYDXJvL0sQ6+NxZhJV1EkUSG29w8OhRC2BDEdlkg7LLw6L4gIJAoVHhlOoZumBQrOlXdxHKDNJp8Ps/4+DgtLS2cPHlyi5BJlw2+9ubSliFSS8kin3igm5Gw45YXdqZpsi/k4BePd/L6XJJ8WaPVa+XRoRaCjpvnyV8vPjIUChEKhQAaefu369vfKeoLpFQqxdGjR3esaXaniUQizM3NXdcqVE9x8Xg89PT0ALXFQDqdbthD6gKyLjR3etfkeok921l5DMNo/KmL/c22n7sV5oZh3JfpsFcfw3avQxAEHA4HDodjy5C5bDZLJpNheXkZVVVxuVxbJureyuegWblvshdpivsm7xjqFfBb5U4yn6vVKpOTk6iqysmTJ2/oo6437opCbVLn63OJK/8mivjsCq/NJhhqcTT89LeLJEmoai3CsqgadPod7G9zU9EMJtdzvHF2mbVMiX//UB8vTMb5m5dnEQVI5cvMrEZ56mAnDz5wGrvVQrCcbTyuRa7l3ksChF12uvwO/u61RVxWmapmoukGFd3g5ek4TxwI893xCOmSiqqbqIaJRZIIOi185Eg7XT47v/fBYeZjBboDdr5zcZ2Rdg/fG4/w7bdXEQSBobCLZx7oZn+rG80wMc3a9Ql7baxkSvzP37pIsqjVAtVJkShU+fR7B5mMlZAlgVZvAJtc89bORzIsr8dJrc7h8XgayTHdXju/9kAXeVVHQMBrk7ZUzU3TxOew8NbiMsZVb6OegH3DzrT9sKfFxUXW19cZHR295kYuCAILycI102E1w+SNuSQDQftNdwQ2IwsmRztc7GtxUNVMnBYJi3RzO1c94nJgYKCRCrMdm/P2b+bb39xMuhPUp+EGAgFOnDhx3yu926Hr+pbko9sZilRPetksIOu7JktLS6iqusXK43A47rgBe3l5mbW1tS07SNuxW6k8V7NXYjhv9XwqirKlUbc+ZTiTyTA/P0+xWMRqtTa8+x6PZ1vbWFPcN9mLNMV9k3cloijeViXJNE0ikQgzMzMMDAzcUurJ5sjNYlWj02cn6LSiGQY2RcIqi6SLVcqqcVfivlwuY5iwnCrz/IZFxm1VeKA/wC+f6OTb59eQRYFStZYNOJctAAAgAElEQVSdP76aYcAnUzUlfrpYYKQXbBaT3qCDdm/NumKaJm0eG0uJAt3+2jFbJBFBgKDLgmYYKJJIVTPob3FRqKxgUyTKqo5ZNXlytJXugIMfTkSZiuQoawan+wIMtLiwKRLPX1xnOVUb9KUbJsupEv/wxhKfeWIIj1VCFGE1W0EzBKK5KhW9NqhKEQVCLiuZksqltQx9ISfnl9LkyxpDYTc2m52Qy0pruIXBjmBj4ms9zrAunPx+PxIOYOs1bHUpPLQvxE9nEg3bS6fPzpmBAMI2wr5UKjE+Po7H4+GBBx64blUwW7p2hgBAoVLL1LdK1zZL3wjTNHHIAg5ZAGqLoetxt429N/LtR6PRRtb43Vag73fT7K1QTxXq7Oyks7PzrncwFEW5Ztckl8uRTqcbsaR2u32LledmArm+O6MoCidPnrzthddupfLcjwFSO4koitc06tYz9+PxOLOzs8CVRt164lI+n7/pnIOrWVpa4rd+67dYX19HFEU+/elP89nPfpZkMskzzzzD/Pw8fX19fPWrX20MZ2vS5HZoivsm7xhup3JfT8y5lZtNuVxmfHwcWZZ54IEHblm4bG529dgU7LKIIgpA7TldVpnTA0GyFY1kUcVrV3BZJRzXqRBf7zkMwyBWUPn/3l5lOVWiqhnE1Apfe3OJTz7QwwcPtbGYLFHVdDAMdMPE6XTgsFcpqTrLyRJtbgs+m8Qzp3tYTpUoVnV+88FeDnZ4KJZK+OwKXoeCJAjIooBFFrEpFgJOC5qm89Ej7UxH88zFiwScCp98oJu/eWmOqmYwtpolW1b5waUo/817B2hx1fzvF1YyDLe5WU4W6Q856Q06cVpl9rc6+eHFFbxOOxVTZHw1Q6vHRp9FQhIFcmWNSLZCsqByqKPWB1DRDLJllVaXhRM9Pnw2EdMUtkwmrXugU6lUQzg5HI5GZd/tdmORBB4bCjLa7mE5VcRrt9Dhs+G1bm3CNk2T1dVVlpaWGB4e3vYGKwhCYwegN+hAFNiyIyBLAu8ZDBIvaLyVLOCzW+gO2K95rruhLkbb29tvK+LyZtSbSete+M0V6Nv17V9viuteol4JX11d5eDBg7uWW77Z593b29vwgtfz9icmJrb4xa/2gtdtODvdp3C9VJ7bsfLslcr9TmKz2bDZbFvmI9StPH/8x3/Myy+/jN/v5/jx4wwNDXH8+PFb2umRZZkvfOELnDhxglwux8mTJ3nyySf527/9W5544gk+//nP84d/+If84R/+IX/0R3+02y+zybuQvfct26TJDlAX9zcb1LO8vMzi4iLDw8ON6tqtslncB50Kh7t8vLVYS2yxySKnB4K8vVxrJk2VVKq6wb871E63305fyEnAfvOFhyiKKIrCdDSPpht0+x0spYqUVB1FFpmO5Hj2wR7+9+9NYhgGLR4bw20eUgUViywyFHbjdShE81VsisTkeo6fTMc51OVjLVMimqtgmgJeu4XffrifH0/FqGg62ZKKyyrx0GCQV2cSSBJ84mQXDqtcq8QnSxzu9GJXJI71+BhfzXJ2IcVrcwkeHgwx0OIk95aG167w+Q+PkC2rrKRKZEtVknkJXZT5weUkDw+GaPXY+d54FN00USQRh0XCZ7fQE3DQ4bORKakspYoEnVYe3RfkSJdnW3G82QNdF07FYpFUKsXi4iL5fB6LxYLP5yMQCHC6x9tIVtr8eJVKhUuXLmGxWLYVo/Xm2bcWU8TzVXqCDkY7PLz/QCsvTEYbOwKP7Q8znyjyz29dsQG1e+184lQnwVu49jdCNyEaT5JLJ+5JxOV2Fei6b39ycpJSqbStb7/u67+VKa73C1VVGR8fb1zve1l93uwF7+joaBxPOp1u2KQ0TcPtdjcWr9s19+80d2Ll0TTtvor7exHFKUlSo6Dw53/+5+i6zmc+8xn8fj9/+Zd/yblz53C73XzqU5/iN3/zN6/7OO3t7Y3Pg9vtZmRkhJWVFb75zW/y4osvAvDcc8/x+OOPN8V9kzuiKe6bvCvZLLy3o1AoMD4+jsvl4syZM3dUTdw8LMsmCzw5GqY74GB8NUN/i4uVdE08JwsqC4kChmny/7y6wH983z6+/fYqv3S8E/dN7DqSVGuirAtGn13Gbql57gG6/XaWEnmyxQqjHT4e3R+izWvj1ek47xkMMbaa5e9fX0QCfC4rw61ujnb7eW0+ycuXY4gC9AftZEsqzz7Yy0i7m8n1HI8Nh/E7FC5HsqSKVY50eWlxKciiwGS0gNsukyqqfOvyKqYJx3t8fPq9A7w6k0CWRDBrQ7M+eLCNb55bYS5ewGOTiaXzhLxOfuFoB3/xo3mqmsGHD7dzfjnN2YUUYOB32Hl4XxBV1ylUNB4eDLKacWCaEMtXmIsXGWpxIN0khUYQBJxOJ06n8xq7ydraGpOTk1v85z6fj1QqxczMDPv27aOlpWXbx13LqXzltUUyG1acqUiOS6tZPvlAdy0RKFUk6LJiIvK98TmkTQXttUyJ12aTfPhgeFsb0M0wBYHFRIkfXFggXzUZ6gwSEGy4brMf5W7ZfN6AbX379Urvvn37CIfD9+zYbodUKsXExASDg4N75hgVRaGlpaXx/qtUKly4cAHDMLDZbLz99ts4HI7G+d/JWMfrcTMrj67rpFIpRFFEVdV7GsG5+Xju9eKivrP6zDPPcOrUKaAWxpDJZG75Mebn53nrrbc4c+YMkUikIfrb29uJRqO7ctxN3v00xX2Tdwy3O6V2u0FWhmEwPz9PJBJhZGTkrry/giBsudm5LCIP9Ho51uUhV9F5ayGFbpjEcmWMDeGVr2isZcpkCiqRbAV36MZJIZIkUSgUGOzu5EdSDFWvTbG1SCKlUpnDbQFO9Ldwqj+EVRZJ5iu8OBFjX6uby9E8c/ECbpuECrw6m2AuVuDXzvTw6kwcoJHH73PY+N7YOs8+2MNTIy2YCCQLVXoDdiySiM8moUgC0/Eya+kSFb02RGo1XaKiGSwmizw52soHD7ZRqGrsb/Px371/H6vpMuvZMmG3lU6fDadVZj1b5uJKhpEOD8+PrdMXcvLff2A/K+kSiXyFA21uMiWVb55b4XefHOY/f3eSUlWny2dHEuFnC0meeaCH/S322xa029lN0uk0iUSCsbExTNOktbUVwzCoVCrX7vwIAm/OpxrCvs5KusRcLM+Jbg9Bl4WFRJHX5+NMRvK12QcuK/KG7lhK1VKGHMrtW2hm1rP8n9+/iGJzYrFaScwmmYsX+fXT3fhs969qutm3HwwGG02zHo+HTCbD4uIigiBsWUjdz7x90zQbkabHjh3bs4k99XkK/f39DWtIfUcqnU6zvLxMLpdDluUtVp7baQK+U+r2HF3XmZqaqjXOD9XStXYzled63C9bUC6Xw+PxNP6+eYfrZuTzeT7+8Y/zJ3/yJ1seo0mTu6Up7pu8K9lO3Nf9qqFQaMswqjshXzVIFlXwdZIsavgdMgK1G68sgiSCItU82aq+VYBaZRG7RULcqGzdaHBSPS2n1W3hg4fa+PFkjGShTCmX5fhAiNODLfjsCqGNXfo2l0KHz05R1fm7VxdwW6WavV8AURCI5Suki2rtB9TShEShFlKTKamUVQOHXLs5uxUTtyI1zpMgCOQrOmGvnfNLGX774X6WkkW+/OoiubLKpbUs//Wj/UytZWnzWJBFmFzP0eG1IUu1m3tJNbDKEolCladGW3l5Os4PJ6L8zUtznBnwc6DNjWYYPH9hnXavnYVkEZssEXJaGvn9qm7yxnyS/mAH8i3q4+v1ayiKgiiKpFIpDhw4QEtLC5lMhlQq1YjH29Kka7UTy1e2eQZYzZQ50ePjhYkIs7E8Ya+dYlWjWNWoagY9ATsC4LTKKNKt911AbRGyshbh5YUSdrdvy3s3ki0zEy9wqnt7u9K9ZLum2e2SY6727ft8Pux2+z3J2y+Xy4yNjeHz+fZsYk89oSkSiXDkyBEcDkfj3zbvSHV2dgK1lK90Ok0ymWRubg7DMLac292aZZDP5xkbG6O7u7thK6qzW6k810PTtPvSz5HP5+9ImKuqysc//nGeffZZfvmXfxmA1tZW1tbWaG9vZ21tbc/sJjV559EU903elWxOstF1nZmZGVKpFAcPHrzr2LJoXuVf3l5jMV5gLRLhQH+Rpw62MdrmaghQj1XicJeP5XQZu0WiotVuaJ0+O/vCLiI2mTcXUqxnSrT7HNgVEadFwmPbmmVe3/aVRTjZ46HdKRLP5gl6+2j1ObFKW2/YggB+u1TbKdis80wIOC0sJYv4HQo2RSRfMRCAkMuCAPSHXCDAXLJMSTXwOxSCDhlx44FyVZOXp+NMRLJcXM4iiQKP7W/h9z80zLcvrKGIAook8v6RMC9cWmdqKYLf72cxWSJX0Taq93YWErUGW49d4X95+hD/9NYyfSEnFdWgP+ji++MRLkdzPLSvj7V0iZlYnt6gg6BTaaTGFCo6qm4i30Tdl3WI5VUyJZWQy0LIqTTiMXVd5/LlyxSLxS0pM4FAgEAgAGxNN5mamsJms+MULOTzeaxWy0aFtHYMXX47mbLO28tpSlWdQ10+eoMOFhJFkoUqrR4rbqvEmf7ARrTlrb3fkskkk5OT7Bs5hC6ZiGLxmt+J5yq3HRV7qwiCQL5qkCnXqrBum3xNU3C9adY0zevGR97Itz81NbXFt1+3m+y0II3FYkxPTzM8PNy4xnsNVVUZGxvDbrdz6tSpWxK/FouFcDjcEIP1xs/tZhn4fD5cLtddi+r6Qu5636nXs/Js/lM/1vrv343Y3yuV+1vBNE0+9alPMTIywu/+7u82fv7000/zpS99ic9//vN86Utf4mMf+9hOH26TnxOa4r7JO4bbteXouk4ymWRiYoKOjg5Onz5912JBNeDFqRgL8ULjZ+miyncurNPm6SPkvPKROtHjpaobOGcTXFrL0uV38PGTXUxH83xvPMLHjnVwdjHNqy/M4LErHO3y8sRIK0MtVywC9W3verXRZrNxeHj4phUqj1WiN+RkbOWK99NhkTjY4cGuSLx3fwsvTkRpcVuxGBXCbitHun185dVFXp1Nki2rhD02nj3Ty7FuD36bxOR6hoVEAVGoucVj+QrfeGuFnqCDLr8dn12hzaWQSGX4yfgiFoeLU2EPQWecQlUjWajid1iQJYF2n53/9/VFDnZ4+b0PDlPVTBQZ4rkq7x0K89RoG3ZFIl6oAibrmTJum7yRRgQDIScWWSJeVMmUNFw2Gb9dwSKZjR2UXNXkXy+sM7GexTBBkUQeHAjy6FCAcq5md+jq6mJ4ePi674vt0k08uSoz8SKriSyaqiJJEiPdIbo8MhVNp1TVMYG3l9J84mQ3ZxdSzCcKDIVdvHd/CwPBW7MTGYbB9PR0Y9Ksze6gK1BpTAveTF/o1qbX3i6CILCe0/j226sspYqYJrR5bfy7w+30+WuTcO+0afZGvv3FxcUdzds3DIOpqSnK5TInT568r5agG5FOp7l06dJd9wBsbvyEa89tPp9HUZQt5/ZWq96GYTRmgdxu+tHtpvIIgnDLgv9+RXFua9+7Ca+88gpf/vKXOXz4MMeOHQPgD/7gD/j85z/PJz/5Sb74xS/S09PD1772td045CY/Bwj3exu3ybY0L8o26Lq+rY9+O5aWllheXkZRFEZHR7dsa98NiaLG//XyPNlyzXO9HonQ2hpGQODXzvQw0ro1xcJEIF1SKVYNChUNw4RvvLVC2G1BEES+fnYJqAnPA21uWt1WnnvoyiJB13XOnj2L2xcgGG4j6HUji7eWDBHNq3z7Qm2HwTBr+fUfPNROt99GSTVYz1S4HMlRLWZ5/Eg/37mwznfGI6SLVaA2iKvLb+O3H+nnUIebr59dZSqSYzVTpqwaTEdzFFWDp0Zb6Q7YONHtp1vOIjj8fPmNVSqaidehMNDiZiaWZyqS58H+ACG3le+NryMJIslChc9+YD8dPhsz0QIvT8eYjxcJuiw8vC+E36HwxnyKFyYi7G91Y5VFeoO1RdL5xTQ/mUlQVHUqmslAi5PTfQGsssBAyMnZxRTfHYtsOSciJp843ooaX2RoaOiO3heCIBAv6oytZohkK/T4bfT5FLKxVUyLk6+fixIrVLFYrPhddvpa3IQ9Nk72+vFZb626vjnisru7u7H4SJR0/uGNZSLZcuN3D3Z4+eiR1o1M/J2lYsBX31xhOprf8vN2r41nT/cQW5kjkUhw8ODBHfuMbabeAJ1Op8lkMnfk2y8UCoyNjdHW1rblXO4lTNNkYWGBWCzGoUOH7kkPQKVS2XJuTdPcMstgu3kJpVKJCxcu7Nq5vNrKc7Vl8UbV/VQqRTweZ2hoaEeP6WY88sgjnDt3bk++r+4xP/cnYC/RrNw3edcRjUaZnZ3F7XZz/PjxHf3SFQUBeZMVRtywQgiCsOGjrtF4TtPEb5fx20EQrERyKlVVo9Xj4Ttj643fV3UDzTDJlFSWU0VaXF6KxSILS0vg7WDVsJFMavxwZhWfQ+FAm5uw23JDQRd2KTxzsotUSUPTDbx2Bf9GBKNTEQk5ZA62OZiaypBOJplcT5PKl7HIEgFXrRKVLWuAgG6Axy4DJhW1NizrYIcX1TA53u3jvUNBcqUqPl8nDotMlz/NTKxANFvltdlFHt4X5Hi3l8EWF3/+4jSlqsZIu5feoIMWt4VYtsz3xtcoqQarmRJLqSKz8QL/4bFBBoIOjj91AAFocSt0ee3EClVemIximpAoVFlMFBlbzeCxyRTKGkupUmNwVB1NVUmlUlyOuvjFE8fuuNJtmiZBu8jjQ8HG303TpNVXayb8JZuPr78+RyKbZzGbJZFM8okHBzEKKSo4byhIN3utt8tbDzlkfv1MN7OxAvF8ld6gg26/bVeEPUC+YrC0zU7BciLPciyJFTh58uSu2SGuboDWNK0hSG/m2zdNk7W1NRYXF3fEjrdbVKtVxsbGcDqdu3our8ZqtdLa2rolwz2TyZDJZBpD4VwuV+PclkolZmdnGRkZwev17sox3Y2V535U7pvF0SZ7laa4b/KO4WYivZ5PDjA8PEw6nd7xaorPJjHS4eWVy7HGMZmGSXvARovLgm5CvKCynCwhSwJdfjtBR+1jZpomiiTgc1gAYevAI1FE2rCc6AZEkxnSyTiuli5ePDtLSyjEf/n+ZaK5ClZZ5ESPn48eaefhwcANp986LSLObcRkvcHNNE0GBwdJlnSsShLT0Ak4bSwlC6QKVSRRYCVVolTVONbtZ2w5g89hIRHLU9HK7A+76Q3Y+f2vn8dtt9AfcvH+kVYeO9BKvrLGWqaERRZZSZd5dCjEP721TKGi4XNYsEgCDouEyyLxw0tpTLPWbBxyWUkVq6iawXy8QEXTaPfZONXjbSykfjxds9qohkkkW0E3a2Wji6tZTvb4OLuQ5EOHOjYGS5nkc3lKpRI+v4+Qz7Uj/vTtGqFN02Skzcl/eGKExWQB04TugB2rXiYeizA7O4uqqtsOgKpbrzwez3W91qZp4rOKnOz2NF7DbgoMgdoCdjOlYpFCIY/TMUBPV3DXnns7ZFm+Jd++x+MhmUxed1bBXqEexXmj6NV7RX3iaiAQoL+/H9M0G0PhLly4QLlcxuPxkEgk0HX9rmxSt8OtWnnK5XJtIrau37MYzvpnr1m1b7LX2JvfeE2a3Ab1aaLz8/MMDQ0RDocbI8N3GkGAhwYCmKbJ+EqGssPCUKuLJw914LHJnFvO8p0La5TUWmXJ77DwsROdDG74rN1WiVP9AeZjBQ51eJmPFxCAsMeKVRIRBBAFg9cWcrQFQ8TTKg+P9nFuKU1ZMzCBsmawlinz/MV1uvx2Rtu2WoEqukmuXHt+r11G2XSP23xThCuVsoBD4T2DIXRTYClVJFPSEASBw51eEvkyP760xGef2M/Hjrby+kKWoMtCyGnhYKeX5y+uoZkCugk/vhxjMVng18708uj+EC6LRFU3MYFcWSNdrDW2tnvtKJLAo/tqEZ6avnE8QNhtxaZI5CsaHofCI4PteKxC45gFQcAqSwjUknMShSpVzailD22I+apqYJVFrCIsrcexWC20tLTgsikMt7pvmFB015gmIYdIi9PTOOfgwuupVeE3C9J6s6MkSZTLZQYHB+no6LipWNhtUV/HY5PYF3ZxYSWDaRik02lM4MRwLyH3tbaNe812vv1IJMLU1BR2u518Ps/bb7+9I779ncQ0Tebn54nH43s2ilMQBCwWC9FolPb2dvr6+qhUKmQyGaLRKNPT0wANK4/X693WyrPTbDdgKxqNsrq6yoEDBxqfbU3TGiJ/t8R+pVK5J6+5SZPbpSnum7yjKZVKjWSJ06dPN1I6Nqfl7DRuq8hToy2c7vOTSHlp9bnxOC0kiio/vBRpCHuAVLHKC5citJ3uYT1b5tXZJK0eGwMtTiyKhGYYLCaKWGURh1LzT6cLKvGywN8+P8FapsyJbj+iCM+e6eGvX5pF1U0qms56pkyxqlNSTewbmemRvMr3xiMsJopIAgy2unniQBi/XWoIwnqT2mYBWVQNuvwOjvXoHO/x80BfkMVEgQcHg1xYzuCw2ZmJlzjabuNUt4sun5Ww18FX31wiXzEIua0N+8Z8okihovH6TIxnH+yj3ylhIFBUTf7XXzrEQrxARTMYanXR5q5V8A+0e1hMFilrJnOJAoWKhiKJ+OwKYytpHuz3N47VMAxG2t38eCpOplRBEgQ0oxZBOtruZTpawKZI+KwCD/e5eNsuU9YFeoMOHtnXQotLwdxNcb/B9cT3ZkHa2dnJwsICLo8fi92GWszz5ptvIssyfr8fn8+Hx+O5b3GNsmDygdEwuq7xxtQSXreDka4WPjDaitMi7ilbwmbf+qlTpxo9AOVyuSFIL1++fN/z9qvVKhcvXsTtdt9TG87tUk9p2pwsZLPZsNlsDSuPpmmNeNOVlRUqlUojOtbn8+F0One1qm2aJjMzM+TzeU6dOtX4/r+Rlacu9HfivOdyuWusc02a7AWa4r7JO4bNN4n6jbxerbk61m7z9NjdQAQCDpn1+RiGUwFsZMvaNcONoDa4aild5h/fXKKs6kysZfE5FA52ePj37+kjX6mSzuSw2Oy8OpvCZbfyjZ8tU9EMRAHyVY3paJ6gy8pou4dzSxm8dqUWuylAvRBZqBp8+/waC4krST5vL6UxTZNfONJGJFdhfDWLYcBoh4cOrxVZMMlWTL55fpWZaJ6KAelilRM9fj52vIN/vbBeE8KCQLvfyXSixJdeHMfl8fLEaDsz0Vwt+95tpVgsI0piraJmmqgGtQhIlwPRNHEp4FJkur01oV676ZoYhsnRLi/RbIXvjEU2bDsKHzrYRixXZmylNvG323slXrHNY+FDh9v4l/OriIJArqzy2HCYeL7C+FqGZ463YRRSPHKgg5P7oKLXvPkzsTzJooWw24pTEXAo90+gJpNJssUygr+LF+fTLCTTDIRcPDJygqBFJx6Ps7q6ysTExB0nm9wtpmmSWV/iZMDkfU+fRJJk3FYRq3RvJ+LejEqlwtjYGC6X6xrBfD1BmkqlWFpa2mKT2u28/b1kw7ke9QFfqVSKEydO3DAJRpZlgsEgweCV/pN8Pk86nWZ2dpZCoYDNZmucW4/Hs2M7J/VFktfr5dixY1uu2c2sPPVwhs2Fjjup7udyuT3by9Hk55umuG/yjqM+OMXn83HmzJltbxbXm1C702xOCLHKNd+8bmwVPf0hJ2/MJShvquiniyovT8exCDohLU5/fz8XYxolVWc5VSLgtCCKAjZZQhYFuv12Lq1lOdThYTVTRpYEjvf4aXHV7DymaZIuaczE8qi6iSCARRIRBJPLkRyLKT9//9p8Y6DW6/NJnjrYxnv6/by9mmQ6mqeimSymiiiSyH95eY7/+P59RDK1nP4urwWnpPPdmRiBYAgEgVRJ42h3gBenYmimgNtpo1Cu0u5WUKsVItEYViFMNpvF5XI1br7bWWJcCjw5GqbDbyeWq6BIIsupIpPrOUQB5uMFev2BK3aijcd5aDBIu682RXd8Nct0tsRvnO7mgV4vfseViuwPJta5sJwhWaxNBt7f6uJjxzrx2iSGWhwI9yigShAEVN1kPRpFNDQ0W4D/44fTzMZqC7KXLsd5bS7B7z01TEdHB+H2DrJlA80wENQy2VSMmZkZoGaHqFf3d6r6LAgCqmFS1UxkdC6N1xo9BwYGtgifvSTsE4kEU1NTDA0N3dJk0KsFaX2WQSqVavj2HQ5H49zuRCb85om4m2cq7DXqzb0ul4vjx4/f9usWBAG3243b7aa7uxuo7a6m0+mGXWondk4ymQzj4+O3fM23s/Jc7duH28/cb4r7JnuVprhv8o6hnvsdi8U4ePDgDQeHSJK0q5V7w4RYQWW+aMGiqozYdIJOhfePtHJ2PkmyUG38bk/AyZsLyS3/X91Ibom32Xnf6cMAtKsVllJFbIpMuqhit0jMbXjy94XdDIScHOz0oBkmI+0eTvb46PBYG0IrV9GYjOQpVTUkSSTgtNDusQIGiXxly6Rc3TB55XKMkTY3M9E8ggDpkkq2pBJyWdkXdrGSKjHc5kLVdD56qAWHzYJssUGpFsM4F8vz4SMdnOrzoxu1LPrVTIkTPX5molkeP9yHU9JZWFgml8ths9ka2dtut3tLjrVpmqiGwQ8vRbmwkiZf0TFMk6DLQrfPgdsubxGUpmkScln5zsVa4pBDEfEqBiGbgMumEHBaGxakuUSBi8sZchWd+XgBEzi3lGZfiwtFEnDbFNrdu/9VqJkwHcnxwsUlECUO9bbixiCa3TrxdjKS4+xiivfuC/H9SxEm13OoukGn38FTB7s4MzDQmPaaSqW2pMbUBemdeLhNBOaSZV6dSRDL5PFaTJ44tJ/ugGNPifk6hmEwMzNDLpe7aYX5RmyeZQC191axWGyc282Z8H6//7Z9+/VdBY/Hs2cn4sKVjP2d3lWw2+3Y7fbGDITNVp76zonb7Wvu0D0AACAASURBVG4sVh0Ox3V3TkzTZGVlhdXV1bvqVbjdVJ7rWXma4r7JXqUp7pu8YxAEAavVypkzZ256g9zt9IKLqzn+9cIa8VSGR0c6iE7GWc+WcVplTvUFsMoi6ZJKh8/OvhYX0VyZxUSxNlwpn6NULOH3+znc29p4zLDLwlDYTaJYpStg5+XLcaq6gSKJlJbTfPRIO8e7fJzs8eN3yFssJemyTrak0uG1MR3Lo+kG0UwZiyRwpj9IJFu65jWkSxq6CR5b7WugutHUGs9XcFplBlqcdLgkilWd3hYvVlmgJ+hkLVNGEOBwl4+fLaa5uJLBIom0em08dbANVTN4ZCjMYMiBXYZwoNbsWCqVyGQyrKysUCqV8Ld2kdZkogWNrqAbmyJxss/P6/OJ2oRdIJGvcqDVTV/QeY247/LZONHr57WpNeYjGdweNwe6Qoy2e7akWEzHcpjUprhulqjTsTzHur1cWMnQOdqyq022giBwdjbGV16axOXxIisiyxfXcdtlnhhp5ZvnVja9NpAlkRemYpxbSjd+vpAo8K1zq/zGmR481murz9lstmH7KJfLuFyuhti/mfdZEAQWUhX+/tV51uPJWiqPz0/2fITfeLAHn00iWdJJFqrYLRJ+u4LLcv+sOaVSiYsXL9LS0rLjcbeCIOB0OnE6nXR1dQFbffubG0lvtnNS963v37+/ca32GqZpsrS0xPr6+j1p7r3ezkkmk2FmZqZh5bm650TXdS5duoQgCJw8eXLHG6PvxMqTz+eb4r7JnqQp7pu8YxBFkZ6envteRUyVdH44EaVY1TnSG2I+WeI7Y1EcFpmhsIvx1QwP7QshAueXUljlmoVmai3Dhdnlmv83HOZIt49u/5XtebsiEHBaWEgUePpoB8WKzlyigNcm89hwmGxJ5ceXY3ziZBeKWGuONRGI5qu8uZCmVNX4rYf6+PqbS7y9nEEUIeSy8YHRVv7twlrjeQpVg1iujCKJJAoVjnT7mVzP4rUpRDeGI0kY+Czwb2PrfOBgJza5JuQe7A+wnilhmJAt63zj7BKtXjutbiuJXJkfjK/zXz3cT8ghbblOVUMga1jQnUG6A23owLfeWmZscY1ypYIgSjiddn7lZDe/9Z5efjqTIFvW6A06+ciRdoKOa3soLILOiRaBAU8nGa2HkNtGl8+G27rVR1+352hXvW98Dguqbm5pgN4NyuUy8WyJn86lsXsDiJK4ETEJU5E8Tx/14LLK5Cs1G5ksCrR7bfxg0xyEOpFsmWSxiueqKvV2qTH1GMO699lutzeqz/Wdkzq6Ca9cjrC0FsHlcm3YzQTi+QrRXIWFhM73L0XIlTUEoTYR9+kjHQTs974Kvb6+zvz8/K7mrV/NjXz79Z0Tt9vdEKQ2m425uTnS6fRd7SrsNpqmMT4+3ogMvR+7Cpt3Turf7/XhZaurq0xOTgK1z1E4HGZgYOCeRXDCVisP1Kr5dcH//PPPUyxeOweiSZP7TVPcN2lymxSqtQqmKEBHwMk3fnAZEChVNTTDZClVJJot88wDPUys5ZiO5PjFI2E+csDLQ/tCZMo6XX4HHV7rNRn1iiRQ0Qzmk1lO9vp5crSVQkVjYj3LcrK4sSsgYBg1y8lkpMA/vbXMfKLIQqLIUIuLX3+wl1883kVR1ZAEgZBD5kC7m5V0kVylNllWM0w+cridc4tpHBaJXzrRzauzCewWEdk0eHiohflkgaF2P0e7fQ2xHHRI/MLRTtIlla+dXaavxYXTIrMR0U+urDEbK9DS572yq7DRsFsoa7T77CTyFfwuCy9NJ2lx2/DZ7ZjAbLzA+eUMVkHjeLsdiyxRrGhUy2UqFcuWJtJ6jGRXVxeHe7u2DC7aLOwNw2C03cPZ+RQBh4XsRsOzyypzosfP2fkEv3Sic9cWjOvr6+SrBkXZw9h6gWiugtMi0+mz47SKBBwWVN3AuSHuBeBot4+Qy9rYvbga/RY2GDZ7n+uCqe59Xl6u2aQ2N+mWNZO1WIpAILjlPEuigGrAd8fXKVRqiyDThLlYgZem4zx9uPWe9Svous7ExAS6rnPy5MlGMsr94Ea+/YmJCdLpNDabjc7OTiqVCoqi7Dk7Ti6XY2xsjL6+vsaQsL2AIAhbrDyxWIzLly/T19eHqqpcuHBhy/Ayr9d7QyvPTlG/fqIoUqlU+P3f/31yuRx/9md/tqvP26TJndAU903e1RiGseM3VUUSsMi16qtuQK6sYrFYkCUR3TDJllSKFR2rIqKqKpFkkhct8KnHhxmw1qpA1xOTHruM3Srhd1j4+9cXgVqTW6vPSX/IxUibe5PHXufFyQj5sobLKiOLApdjef70B1P8t48N8tPLMd4zGMShCJzu8xHY8Kg7LBJHu304LRKvzyXQDZOBkJNfOODl8X4nVruTWL7KUJuHVrcFp3LlphnJa3zl9UVG2z1kiypzsQIOq8RAyIki1gZz5SoaU7ESVlmkxW3h5ek4b8wleM9giP/83Qke6AsiizUvfKpYZbjVjSJCq9vGYrLMSJubi6u1XHWvVeQxp8S5c+caN/RKpYKqqhw9erRhIbiROG9xyjzzQDcvTSfwOywoksCDg0EimSLHenx0+2w7Lu5VVWVychKX281ixUW+UqTFbWMtUyZbVlETBvvCLnx2mdP9QfwOC9PRPIMtTk70+PDbZfpbXJzfZMsBCLosBF23L2oFQcDhcOBwOOjo6ABqXvBoNMrY2BiBQIDekIulyzGsVisWi7W2G+BQyJU1ipVrdzcW4gWyFR2vdfdFa12Idnd339IcgHtNvfqsaRpra2scPXoUm83WmKR7t779nWZlZYXl5WUOHz6M0+m8+X+4D5imyezsLJlMhlOnTm2xPtUXU+l0munpaYrFIg6Ho7FYvXpnaidZX1/nueee4+mnn+Zzn/vcnlu0NWkCTXHf5B3G7UwWrcdh7vSXr9+uMNLu4e2lNLJYmzhb1MDvtKAaBpmSRk/QwWqqRKasEgwGMSQLVd3ANG98LE5F5ExfgMVUidP9Ad5aSONx2+j02jjS5WFf+Ir3vKTqNRuNaWKTRXoCDlbSJdJFlUJV50i3j5O9fkzTxCLCvpCDY90+koUKkUyJaLaMCeiaxsX5ddydFrra25EkiXZP7Ua65VwLAm/OJ0kVqqxlSuwLu7i4miFf1siVNbx2hXi+giQKfPWNRVTdYKjNQ9Bp5WCnlx9NxchXdPIVld6AE02vNeF2+uz47TIuq8TBTi8hp0JH0U6Hz8aZ/iCtLpmellNkMhnGxsaw2WzIssz58+e3WCGuF2FomiYdHoVfOdHOhw62ohom+ZKGo8uLTZEoVg0km4Qi3l4KTP19dbVXP5VKMTk5SV9fH45AmG//dBHThMeHW1hMFkgXVUqqTkXTeeJAmE6PwkAgxIc2fP+1xzN433ALxarObCyPbpi0emx86GAbfpu0I4uRehX/8OHD+P1+2osaa3mD2bUkuVwOTDja3otLEdB0DUnaeruQJRFpl0W2aZosLy+ztrbGoUOH9mymuGEYDSG62YbjdDrp7OwEaoupdDpNLBa7Ld/+TlLf/TBNk1OnTu2JgV7bUa/QezyebXsqNlt5ent7t+xMrayskM1mkWW5Udn3+Xw7stPz05/+lN/5nd/hC1/4Ak8++eRdP16TJrtFU9w3eddSH2S109v3FgmeHG3F77BQKFd5+lgnP51N47LJpEsqkmjy/v0hXp2JsZRWcdistHhsOC1bP26brSSbCbsUfHaZ3sA+orkK68kcHT47PSF3Y3CQYRiImLgsMhWtlszjdyg4rTKCAAfaPbS5FBTxymNbJCirOheWMxs/MSkUChQKBUZOH6O784qVZjvxWNFMIhvJLouJIo8MtfDwvhCvzyWpaAaSKPD4cJj1TImKVhO855cyHOr00h1w8G9jEQAurGT4wEgrF1czLKdK5CsaAYeCyyrxyGCQoFPiVK8Pu0VE3Hit8/PzxGIxjhw50hB4m6t39QhDl8vVqI5e3UQqYeK11oYDuCxWXp9Pcn4pQ0XT6fLbed9wmDa3clPhbAoC0bzG5HqKUlVnuN1Dh8eCTC29JZvNcuzYsVrltmKg6SapYhVnssCnHx1gOpanUNF5fLiF4bATWTC3TXby20R+5UQH2bKOZpi4rBJe693n8muaxuTkJLqubxn8E3IqPPfIIIvJdhL5Kl1+O36rSamiYkNlJZJAURQsVgtWq5UjXa3X9DfsJKqqNhZzu9FAuVOUy2UuXrxIIBDgxIkT191VsFqttLa2bjsAanNqzM0Wq3dKoVDg4sWLdHZ20tnZued2P+pks1nGx8cZHBy85dSe7XamqtVq4/zOz89jGMY1fRG3eg5M0+SLX/wiX/nKV/jWt75FX1/fnb68Jk3uCcL9bk5ssi3Ni3IdVFW95VST8+fPMzg4uGvVPkEQiGWLVCpVyoKF1+ZSyKLAoU4vyXyF/+17k5RUg9EOD599/xA9/itNdbGCymysQFnVGWhx0boxqXU7lpeXsVqttLS0YJo1IWiaJqIk8ZPZNN8dX2fzx/hYt4+PHGrFso0WihV1/uGNJdbTBVKpNLIs856Rbp4+0oHjZmsgQeDbF6K8MV+L9bQpEgc7vPgcFkJuK5pu8PZymsuRXOO/aAb4HAqj7R6+Ox7hxakYACd6fPzamV5emorhtSsMhByc6Q/Q6bWw+cUUi0XGx8fx+XzXZK1fjWnWFiupVIpUKkWhUGhs1debSOspFz+eTvH9S5Et/7/da+PXT/fgsd44VeZyrMQ//myZYnUjE1uAhwcD9Cp5LHKt6buR6Y/At95e563Fmr1GkQTavXbcNoWnDrbiu8Fz7QaZTIZLly7d0N6yOZ60Hie6kq3ywkSMhWgGAYND7W5GghLFTKJRfd6JPPg6de/64OAg4XB4Rx5zN4jH41y+fHnLFNc7ZfNiNZ1ON6wmO5G3H4lEmJubY3R09IYRwvebzXahzTNEdgJd17ec31KphNPpbFT3r2flKZfLfO5zn0NVVf7qr/5qx4/rXcTeXC3+nNIU93uT5kW5Dpqm3XJ+/djYGJ2dnY0Ekd2gWCySSGdYrtj5yUycuXiBkmrwvgNhTvb4eWM+yaFOL6d6vBiGgSAIzMSL/OPZZXLlK+ko7zvQynsG/Mjb3Lvn5+dRFIW2trbGwqYuwKoGTEQKvLWYRtMNDrR7ONrl5Ua27NVUgYtLCUqmwv52Hz1+O85b3NyI5DX+7rXFLZN4e4MOnj3Tw0uXE7x0Obbl9wVgX5uHvqCDVFHlL16coVzV2d9a24V48mAbp3r9BB0S4qa3fT3Penl5+Y5TUepb9XWxn8/nsVgshLsH+Pr5KKlSbfLuZn7zPb0MhezXrUaXNPjK60ssJusJGSb5fAFDrfCfnj5Ju/faVJRkyeBf3l5lPl7AMMFrrwn7g+0uhHv0/WuaJvPz88TjcUZHR2/bZy0IAhXdJF8xkETw2WUMXadYLJJOp7ec382TdG+32l73WadSKQ4dOrRnhz1tztg/dOjQrlhq6nn7dTG6uQn6Vn37hmEwNTVFpVJhdHT0vjYh3whd15mcnMQwDEZGRu7JLs31zq+iKExPT/P4449TKpV47rnn+OQnP8lnPvOZpr/+xjTF/R6iKe73Js2Lch1uR9xPTEwQCoVuaYLh7ZCrGKSKKhVNx2eTiOcr/A/fuECqbJDfqOYqIvyPTx/CoQh4bBaOddey10uqwd+/ucz8xkTSOnZF4lOP9hO+SpUbJizHMmi6TmvAjW0b9V8XXoZZG+Z0vZ2NSqXC+Pg4NpuN4eFhZFm+Jl3mZggblpS3ltLEchX6Qk4OdXhwWUVieY3/+ydzFKpXro9FFnn2wV48VplEUUUQBGZjeWK5Cq0eG31BB30BG7Jw5RjK5TKXLl3CbrczNDS0ozf6crlMsqjx1z+aZj2VRRRELFYrVqsFi8XCr57u5WCb87rnJFky+OuXZilWdXRdI5VKoygyXo+XZx/sYzh87cJAEAQKqkGqqFHRDPwOBZ9duq6wv9quJQgC2YrRsD557RLibVyzcrncGKI0ODi4qwKlHmGYTqfJZDJbIjpv5nuu21v8fj8DAwN71jZSP85gMEhfX989Pc66b79+foGGp9zn822J3KzPAgiHw1t2k/YapVKJCxcu0N7eTldX1309zkqlwszMDH/6p3/Km2++STQa5ZFHHuFXf/VXefjhh+np6blvx/YOYG++wX5OaXrum7xrqTfU7iTRvMq/nF9lMVGkXKlwotuNJlopGwJWRUI1TKqagWrACxNRfvuRPjq8V9JYClWDWKZ8zeOWVJ1EobpF3GfLOj+6HOPiUopMNkeHz8aTI63YjCIul6tRuas3zMK1zZ1QE4n1bPDNw3TuZGiTaZq0OCU+OBLCADRDYHw1y4WVLB1+Ox8+0sHZhRSpQgWvXeGhwRA9PisiJiGnHQOBNreVqm4gi+C2iltE7vr6OnNzc7s29Mdms9FqEzg00I6xZMcwdCqVCuVSGb1SQtEKTE2t4PP5CAQClHQRwwSPTQTTbEy0TWRy5LI5vD4vVqsNSRRw2+RtFwWmaeKQBRweZfMPr/m9il57f0WyJQJOK20eC3ZFZCJS5MXJKJGNIWnHuv08vC+A/RbWPJFIhNnZ2R2xjdwKNpuNtra2RrSiqqoNMVr3PdfFqN/vb4jRaDTKzMwMBw4cwO/37/px3in1Ztj7dZw38u0vLy83fPuyLDd2afby+UwkEkxNTTEyMrKrO6y3itVq5cCBAxw/fpyZmRn++Z//mfX1dV555RW+/OUvs7y8zPPPP99okm7SZK/SFPdN3lHcTlVHlmU0Tdux59YMeGEyxlw8TyaTQVU1RCVEqayjGyCJtfx0TTHRDRObItHisuLalGWvSAJOq7ylug0137bLeuXjqP//7L1plFx3ee7727vmeeh57lar1epuybKsyTa2bCAMySU+l3NNDiQBh3FBCAvnwsEri3udsJIvZOXkHK9wGMLlwuEQsM0JXMDYGIyxHUM8ybKknrvV8zzUPO6qvff9UNpbVT1IPVS3ynb91tKX6lLVrr1reP/v/3mfR4FnR1Z48fIqCGC0OVlKwzPjMf7D0SqWFqYZGRnBaDRSV1eHxenBYrXhshQ6qaTTaQYHBzEajQXDk7tFVVVkVeCpgSVeGFsFYGQpSpXLwrt66qhymbAYROymq+5G2mLCaQI0e80rf5MkicHBQURRLOpxboQBlbs6KgnFJaYDCWw2O1VeN+/orqG11kYoABIGnrw0w2uTAUTRSFeTnzsP1eCzGThW72B8KYLLX4UoCghAT4MHv33nX6fJLPxqYJnzUwEUNdcC66x1cffhXHqtpu+PprL828gybpuRW1u9my7QNJlDJpPZ8/N5LUwmE1VVVfpgpCzLejE6OzuLJEm6Xe1+hlJtF0VRGB0dJR6Pc+LEiX1xttkKa/32NTeclZUV7HY7g4OD+2YRuR1UVWV8fJxgMFhS5zOZTPLZz34Wo9HIr3/9a2w2G52dndx1110AurSyTJlSp1zcl3nDUuziPpzKMjIfYGlpWXdlmVhNcvfhan56YZ6MrCAKYDYICAaBOzoq8dqNBV1aj9XAiVY/T/bOo+Q1bw/VuqlwmHQnnGBSZnAufGWj8+qPyVw4jSSYOHz4MACxtMy5yQCvXpokk8nQ0+Dl5kYPVlEmlUoxNTXFwYMHt+Q6IYoiGUXFILBhZ3kt8YyCKArc1l5JKCExvpKT2zzRO8d9t7VgM27NWlIbSjxw4IDekdxrct73jazGM6QyMpVOC36bAQGVyqpqnh5e5cWZFKpoRZIknumfY3YlxB/eVI/RIHD34RpengwCInd0VHJrmw/LDtVDgiAwFUxwbjKg36YCkVSWC9PhDRN0L82EOVrvwmZcX2hobiOl6AlvMBjw+/34/X5isRi9vb1UVVVhNpuZnJxcN0SqDUHfSDR5S1VVFR0dHTf8eDYjnU7T29uL1+vl9ttv14eiNV35RuFlHo+nILRsP9BckBwOB8ePHy+JxQbA9PQ09913Hx/84Af51Kc+teFxlcqxlilzPcrFfZnXFdv5YTUYDEiSVJTnzWazhIJBpGSSyspKjFc8v6OpLEuBCJ9+60H+9dwM08EEfruZ3z9axy3Nvg111cebPNjMBs5PBpFkhY5qF7e0eLEZBd0JRwWyKqyVMapARr7SCUfg+bEg/345ABgAAy9MJ0gqBo64UyzMzmCz2QgGgwDX1DxHJZWB+RBDi1GcFiMnWnw0XpHTbEQyCy9PhPj5xXkkWeFwrZvbD1bx8vgqkWRGP8brndPh4WEkSSrwBt8PVFXFaRJwes152Qm5Yw6nZC7O5NxtBEHEYraQTqVzycExgcdfm8NthhaPDRGVkelFOvwmBJuwo6RMQRAYWYytvx2IpDLIKqwtKTZKr1VVlcnJSZaWlko+nGhubo7p6Wl6enpwuVwFf4vH43r4UzQaxWKx6MW+x+PZ1wJLkwuVimxkMzR3obVyNkEQcDgcG/rtr6yscPnyZWBz3X6x0cLI2tra9m0hvxWeffZZHnjgAb7yla9w9uzZG304ZcrsmnJxX+YNS7E69ysrKwwNDdHZeZi39LTy29GVgr+fH1/i//zD4xxr9JCQZMxGkUqHGesmny6rUeB4o5vDNQ5kBRxmEVmW9WMVBAGXxUBzhZ2BuUjB//U5zPiveFaGUjIXr3jWi4IAAkhSlt/1T3H7H97M4Y6DuUXJFTcTTfOs6Z214JxkFh6/tED//NXnGpiP8L6TTRyq2sA5RhB4YTzAU/2LTAUSZBWVmWCSlVia2w9UMBmIY9nI9icPLeipubmZurq6G9oNXfv6ZFUleUUGk81kCAaD2Gw2DtRXc3EmjNlsJgVMhGVAJSNlGFmMkbHEWF5exuFwFNgXbuW1uW3r3yxL0TSnDlQwuhjRcwM0uurc2E1XPebzh2ZPnjxZsh3GTCbDwMAABoOBU6dOrRuWFgQBp9OJ0+mksbERyL22YDDIwsICQ0NDejiR9m8vOs+KojAyMkIymbyhsqbroS3oVlZWOH78+Jbchbaq29fO704WrBuhLehKaeGpKApf/epX+dnPfsbjjz+uv+fKlHm9Uy7uy7xh2W1xn8lkGBoaIp1Oc+LECaxWK7e7FFSgfy6MlFVo8Nk4XuPDYTJgN4rA1ooA9UqqrKqqZLNZ3U9c+xE1CSpv66wmlsoyE0igAn6Hmd8/UovXZgBVRVFU4ukskbTMUjRNJitT5bLg8fgRjWb9HOQ7BsmyrBf7U1NTiKKIpbqVV8cXMJssiFeKrXRW4beXV2j2Na6Tm4RTChdmQliMAjVuK7OhJJALp3pndw13HarCYzVsqAfXLAQjkQjHjh3DZrNt/8LsMXaTgTqPld7JJRKJBD6fF5PJjNVkxLiuaBYwmc3Y7FYOtdfT3t6ue+1PTEwQi8Ww2WwFXvtrC29FUeiqc/PSeEC3R4XcHEa9x8of3FTHM4PLBBMSFqNIT4OHm5uuBo5p3eX9GprdKZrHfmtrqz5wuxWsVit1dXXU1dUBufmMUChEIBBgbGwMVVULkl5323nOd5k5dOhQycpwNHmL3W7nlltu2fGCbq1uP99vf3R0VJdK7VS3ryiKPv9x4sSJfZcBbUY8Huczn/kMLpeLp556al93DsuU2WtK41NWpswW2a4sZ6duOcvLywwPD9PW1lbQWXZZRN7ZXcXpNh+KAi6rgb4L55Gbq7f1g6dp69f61udT4zTy/lONhJJZMrKC127GJOasON0WEYtJxGQ08MrQMsl0BlEUmQokufNgJaZN7CMNBkPBD7mqqrw0EUSSsjm/dkXBbDJjtlgIRA2ksgoWQ+HrkhWF1BUdeLXLgtNqJBiXEAWBSpeFrmr7hoV9NBqlv7+f2traayZ53mgMisTJeiuzS0YcjkoEQcRuNtBR7cTrMDEXShTc32IUOVjl1IfttM5zU1MTqqrqnWdN86x5wft8PtxuNwaDgWqHkf90qonfjq6wHJXw2Ezc1l5BlcNIjdNFq99OIqNgNgh4bUY91Ta/aNqToURBICYpGEUBm1HY0gzFWvI99m+66aZdhwCZzWaqq6v1cCttwbq287yTpFct7KmUh3vh6kLpwIEDRQ/5EkURj8eDx+OhpaVlV7r9VCrFpUuXSs6Oc2Jigj/7sz/jox/9KJ/4xCdK5rjKlCkW5eK+zBuWnXTuNdmALMucPHlyw26OCPjzZBSyLCPL8qbF/VVNd478lNmNivr8+zlNAi6zmZWEzNODS0yvJjAbRY41eTja4OGmBg8XpgJ4a12ksgrLkTTHW3ykszKCYLpuMSYIAg0VLrwed27AV1WRMhLpdBorGRZnpphJxfH5fPh8Pmw2Gw6zgXqvjctLsZzLj9mAy2LDbTVS57YgrNHpK4rC5OQky8vL9PT07FlicDHQrCO7u7v59LuOMhdOkckqNPps+O1Gaj0WgjGJ12ZCJCWZCoeZuzqrqHKa2CieQhAEbDYbNpuN+vp64KoX/MLCAsPDwxgMBrxeL5WVlbz3WC1pObdgsGrFtApeq4jXqr2/VH1otrGxkYaGhqIXJ4IgEEjKvDgWYGI1jsVo4ESrj8M1DszbaBCn02n6+vpwuVycOHFiT+RCaxesWuc5GAwyPDysJ5FeSyqVH/Z04sSJkpbhzMzMMD8/X5SF0lbYqW4/EAgwNDRUcvamv/71r/niF7/I1772Nd7ylrfc6MMpU2ZPKIdYlSbli7IJqqpueUhWkiQuXLjAqVOntnT/paUl3bWltrZ2XQEgyRBMZpCyKk6LiO9Kgf/qq6/S1dVVIDERBIFAIsNcKEVckmn02ah0mDBw7W59/v/PKDl5jKIKPPzyNDPBqx3jVCrFh892MBtK4XNY6J+P4LIa6ah28eL4CscafZxscm3Jyz4lwy/7lzg3GdRvc1iM/NHJJlq8JiKRiJ7ymkqlqKioQPTU8bPeRULJLCBgMxt4d08tRYjtmgAAIABJREFUNzW4CoZw4/E4/f39+P1+2tratlTcCYJAKKWwGs9d5wqHCa/VsKOu8VbJZrMMDQ0hyzJdXV16cbc2UApARiCUlMkqOWmV54oH/nYRBAFZBVXOEggECAQCejBR/lxEfqGZPzTb09OzZ9rlWEblh+dmmVi5GrYmCvCHx+r1tOXrobkg7VVmwVbRhnSDwSChUIhYLIbVatXPsdFo1HeUmpqaSraLm81m9XmFzs7OfUlx3Sr5uv1QKEQ8HkdVVVpbW6msrCyabn83KIrCQw89xC9/+UsefvhhXeZVpmiU5gfnTUq5c1/mdcVeyHIkSWJgYABVVTl16tSG8oZoWuGpgSX6rmjtfXYzdx+u5kidC1dVI2EJjJZcMi3AbDjNj1+dZTGSC6wyGgTe2lnNqRY3RkHAYMhJXgwIGMW1Q50Ck8E0L4wFSGVlDtW6GVmK6emz0VgUVVERBJHR5QTf/v/6MF8ZYK1yWvjE2QNUOLbeebQa4B1d1RyqcTG4EMFjM9Fd76HGmbPx1LboW1tbUVWVaDRKIhHnP3R5WYxKyAg0V7qodJgQVAWu7FTMzMwwOzu7LYmDIORe+88uzLEUTQNQ6bTwnmN1HKy0E0llSWdVbGYR+w5lImvRhntbWlrWLeo2enwDKhW2vEXKNo9BEAQWY1nOTwVZjkk0+23c1FDB4epqfQZjo+Anh8PB4uIiHo9nz4dml2MSkyuFKcqKCq9MBOmsceIwbf451DzhY7HYvrsgbcRaqRTkdPXBYJDR0VHC4TAOhwNJkggEAjfEHvJ6aLahLS0tJVmUarp9j8dDLBajurqa2tpawuFwUXT7uyUWi/GpT32KmpoafvWrX5WMr36ZMntFaX2DlSlTRERRvG6HcWFhgcuXL9Pe3n7NIb9zk0HOT13tbAcTEj95bQ6o56cvT2O02DhY5+WOg5WYDQZ+NbDEXDiFQcjZFkoZhWcGFzlY7cBhMdI/HaJ/PorRIHBLs4/2Spsud5iLSvzgpSkSkozXbmIhnGJ0KUZ7lZ1sRiIjWrHajETTuQJRyiq6F3o4meHSbJjbDvi3lUBrM0JXjZ3u2lwnWFGUDYtWQRBwu9243W4ADgCJRILV1VVGZgIkEgmsVivJZBKXy8XJkye3VSjFJJUn+xb1wh5gJZbmF72LvPtILb/qWyCVkTne7KPRn5MkuK0GfDbjpradm6EoCmNjY4TD4X0d7l2IZviXF6cIJzMAjCxGGZiP8v5TTXgswoZD0JOTk4yNjWGxWFhdXSWTyeid52IftyAIRJPZDc9mXMqSUVQ2a9IlEgn6+vqoqqri+PHjN7xbuxlms5lwOOd8dPbsWVRV3VBmku8qdaOYn59ncnKSI0eOlLSkTVuA5A9Me73eXev2d8vly5f58Ic/zKc+9Sk+8pGPlOx7skyZYlIu7su87lirYb/W/TYjnU4zMDCAIAibdus1YpLC4EK04DZZVbm8FOXiTBib2cRySubnF+a5NBPhD47W8pvBZexmA0ZRyAVbGUUEWWUunCYQi/Ds8LL+WKOLUd59pI5bWz2oCLwyESpIJPXbTRiQWY1lyKoQiCdprbAzshjDahI52epnNpRAUaDSaWYlliYuybjN29u2V1V1R51wu92O3W6nsbGR+fl5xsfHqaysJJ1O8/LLL+sdO80t5lrXJZlR1g2sCuQK4GONXhYjSW4/WM0L4wFe+c0oLRUOKp1m7jxYyalWL0Zha8cfj8fp6+ujurpadxpRcvJ2DMLWwrd2hCDwymRQL+w15kJJRpZinGp2F0qAZFnPArjtttswm80FmvLBwUFSqZRuXejz+XA6naSv5AxYDNvf3VBVlRq3BbNRRFpjwVnnteE0G9hIOagVoaU+jKpd+7q6OhobG/X3Y/6QriYz0VylstksbrdbP8dWq3XPi0RtYFqb/ym13YR8FhYWmJiY2HQBcqP89n/5y1/y13/913zzm9/k9OnTu368MmVeL5Tut0WZMnuAqqosLCwwNjZGR0fHlpwmBHJ643wkWSWSzCKKYDAaCawmWYymmQwkuPdEI00+G8+NrAAqoiDgtZto9tmxGA1cmg0VPLaiwkvjAQ7XunBaRMLJqzMFyWSK3vE57j3ZwpMDS8iZXLFV4bTQVungW8+PcajGhd9uRgWSkoyqgsmwv90pTdpkNBo5c+aMXohoHTutSIpGo1itVn1Ad+32vCComAxiga+7CoRSWRSg0edgLpzUF0ehZAa7WeSpgUWaK+w0uK8tR9LkQnNzc3R3d+NyuZAUGF9K8NpUCFlVuanRy8EqO9Y9kDRLsspy3q5EPgvhFIJw1eJSC/xZOzSb72aiSaVisRihUIjlUIyXJ4L0zkUxmkycaKvkSIOXDWz0r0mFw8idHZU8O7RM9kqUcoXTzNmOSkxrRgy0eQVFUfa8CBUEgWRWJZbOLX5dFsPVweMtoC1A1oZnrWUje0ht9kRbUGkp1T6fD4fDUdRiP5FI0Nvbu24BUmpoeQCpVGrb134v/fYVReEf/uEfeO655/jlL3+5J4FZ2qKroaGBxx57rOBv6XSaD33oQ5w7d46KigoeeeQRWltbi34MZcpsRrm4L/O6Y6udew3NlSadTtPf34/RaOT06dNbdsRwWgwcafAwE0wW3O62Gmn02bk0uUIgISErKomMjNEg0FHj4icXZnFYTBgElYQkY7cYcFqNBOK5rm06q5LKyoiCgChKZBUVkyhwoNrJyGKESDhCJptlzuDF4VL4y987xORqHIMoYDIIZBWFzloX0VSWhJQlnVUwGQTe1VOD324EdeuynLVkVYFwUiarqthMIh6LuOk5X15eZnR0lPb29nWLpfyOXWNj4zpryEgkoieQ+nw+XG4Ph2pcXJoN64+hAN11blZjaXwOE/8+dlUelZFzrzGrqAwtRGnyVm4qR9KcWxwOBydPnsRgMCAIAuemQvyyb4ErNSzDC1He0lHJ2zorMRR5tt1sEKnzWhlfo2cHaPTZ9N2TqakpFhcXtxT4IwgCLpcLp8vN08OrPDceJJuVSacT9E8tc2dnNSdqcrKlzbz212IU4PYDfg5WO5lcTWA3G2j22/HbCgebtQXIfoSRabMKT/Yt5AZ9BWirdPCunlqqHcZrfifstgsuiqJeZAIFC6qxsTHi8fh18wy2ytLSEmNjYyW/A5JOp7l06RKVlZVFyQMolt9+JBLhk5/8JM3NzTz55JN75nz00EMP0dXVRSQSWfe3b33rW/h8PkZHR3n44Yd54IEHeOSRR/bkOMqU2YhycV/mDY0gCMiyzOLiIhMTExw6dIiqqqptPYaqqhxr9BBNZbkwHSIhZWny2Xhndw3jyzFkRUUgJ9XpqHKSziiEEhIfP9vOa1O5+x9r8nCmzY9RBLvZwGwozXQwoRenNzV6MIgCqwmZJq+Nd/fUMDRvYSokoagCJoNIPJ3h8lKUrKISSWbw2Ey870QTS9E0F6ZDWE0ibztcw7FGN+IuCvuopPLUwCID87lk1Bq3ld/rquFQta2gZat1bLPZ7JZ91jezhgwGg8zNzZEYHeWW+hbkjJXx1SQmk4n2ahdHGjz86NUZDlY7C3YlfHazrhAxXyMVVwt6WuvcEkzKvDQe0At7yD3cuckgx5t9VNqKPPSnKtzS7GNoIcpq7OoOTUuFnfYqh54063Q6tz00G0zKVxyPBIxGE0ajCYcDxsIqdx+pJxNeZnZ2lkgkgslk0vXkHo9nQ+cVo6BS7zLR4L5a0GoFtKqqTE9Ps7CwsG+Jo4msyhO9C4wtx/TbRhZjqOoC7zvRsOlOSywWo6+vj4aGhqLZhmoLKpfLpecZJJPJTTXlXq/3uu422iByIpEoaTtOQN/B2MvgtI389rVB6PxzLMsys7OzvPWtb2VpaYmPfOQj3H///Xzwgx/cs8XmzMwMP//5z/niF7/IP/7jP677+09+8hP+5m/+BoB7772Xv/iLv9CbTGXK7Afl4r7MGxpBEDh//jw2m21b3fq1OMwi7+iq4mSrD1lRsZpEkpLCyEIUUVDx2U14HBbe3VPLajzN473zNHhtnO2o0jv6j1+a53Pv6OT0gUr++bnLmAwiogAmg8j7TjbzzPAKlyaWWAqGqa+u4nirn9s67BhEgQavlYH5KMG4pPeSw8kMEysx/tOpJt7RVYXJIOK2GlB2GNx15YTxwvgqr01flQ4tRlL87MIsH35LG/4rxW4gEGB4eLgoHdu1CaTpdBqf3Uwo5UKSJAzZOE6sNHnNLEWSnGjx0TsbodJpxmHO9dbtZgOHatZbf+ZbXJ48eXLd9c8q6jr9O+TkTemMDMUu7oFqh5E/PdNC31yEhUiKtkoHh2qcZGJBeoeGdmwdKau5xOK1JCQZBZH6+np9QaXpnTX7V60zrRX8+Z3ttR1xSZLo7+/HarXuuWtPPrG0zOTq+h2P8ZU4sbSC1b7+OObm5piamrquDGe3CIKgz56sPcfazpYgCAXFfv5iOJVK0dvbS2VlJR0dHSVbBGq7SktLSxw/fhyr1bpvz51/jjXdviRJ9Pb28sQTT/C3f/u3LC0t8e53vxu73c7i4uK2kpC3w/3338/f//3fE41GN/z77Oys7sxkNBrxeDysrq7qQ/Jlyuw15eK+zOuOrfzwqarK7OwssViMrq4u/cdgN4gCVNivfmTcFgMfONPEcrQKSZLJqPBU/wJjSyp3Hqykdy7Cj87PUOGwsBpLc7DaSTorc6TezRfedZjz0yEcFiPHm7wsRZJ88+lBuhq8+P1+UlmFly6vcLCqlUNVNhRF4Ui9C1Wt59WpIOmsQrPfzh0dFdiNYLuSIrtZYS8IAomMikG89pBlJCUztLD+ByuSyjIXTuG1WBkZGSEej3PzzTcX9cddu64WiyUn1blyuyRJhEIh7miyEJUETBYrn3nrAV6bCpDKKFR7bNx1qJqqK9adGqFQiMHBwQ0tLjVMBhG/w7xOB++0GLGa9sZHXFVV/DaRuzpyHU9ZlhkcHNQDlHbqzGI2iFQ4zHo+gIbfYcZiLHzta/XOmUyGUChEIBBgbGwM2NgtRgsmOnjw4LZ3wIrBVuV42jlVVfWGDaNupCnXLE61IV3NJWZpaYnu7u6SCntaSzabpb+/H7PZvGeBZNvFbDZz7NgxWlpaWFxc5IknnmBsbIznn3+eb3zjG6ysrPD0008X9bw+9thjVFdXc+LECZ555pkN77PR+7RUF2xl3piUi/sybziSySR9fX3Y7Xaqqqr2pGOnpcw6jSouf845QxEE6r02UlmZjAI1A0v0zoVYjqQ42uDirYdrCcQlhhbj/Oy1WRr9NmRZJZlKk5bBarNe6SznfgRkFV6bDtJRlbM6tBjgdKuHrjonsqLishj0NNi0nJOTaF74+UQllQszIQYXohhFgVtaNk8aFQRh0ySSbDbLpUuX8Pl8RdHY5j9nKKUwFYgTTmZorrBT4zJjuXJ8ZrO5QMufzWaptqkc9FaQzsgIsoQlucTKkk2XP4yNjREKha5rcem1itxxsJLHLs6RueIwIwpwe3sFPtvGrjDFQtMU9/f3U19fz+HDh3d1Tj1WkbOHqvj5pXnd5cZkELmzoxLfdULATCYTVVVVesG+kVuMdsw9PT03RAvuNBto9juYWNO9b/bbcViuvpk1GU5jYyP19fUlU1SttTjVdpUWFxexWq0MDg4WDJAWe0h3N8TjcXp7e/WdulIhHA7z8Y9/nM7OTh5//HFMJhNtbW28/e1vB3LnuNhhX7/97W/56U9/yuOPP04qlSISifCnf/qnfO9739Pv09jYyPT0NI2Njfpnaa/kS2XKbEQ5obY0KV+Ua5DNZjcMp9J0wNPT0xw+fJiKigr6+/upq6srWudGVVUU5dops6Io8tpslPlQCpNBQEDFgILZYMBsNfPIS1OMryaIJiUkSeLOzlp8Tisji7F14VPHGr3ce0vdpkOiKRn652NcnAkhq3CoxsXBaici4HcYUVSVxy4ucHHm6oCqIMC7umu5/YB3fcEniDw1uMy/jazkv2iMisQHTtZR57UXXV+9HJf54bkZPfBLFOC29kru7qhgK26eWtc5FAqxurpKIpHA5XLR3NyM3++/rhRLVgXmoxKXZsLIisqRBjcNHoseSLYX5A/Ndnd3F82/XEFgMZphaDGKqkJnrZMapxnDFi1CNyKZTNLb24vdbsdmsxEKhUin07jdbr2zb7PZ9rwQFQSB+WiGxy8tMB3IFfjNfgd/cLSOWpcRRVGYm5tjZmaGnp6ekvaE1+Qkbreb9vZ23SRAGyANBoMkEglsNpt+jvc7+EljcXGR8fHxPZc2bZeBgQE+/vGP8/nPf54PfOADN2Qh9Mwzz/AP//AP69xy/vt//+9cunSJr3/96zz88MP86Ec/4tFHH93349tnSmMlWgYod+7LvA7Z6EtcC89xOp0FVoxGo1HvOu4WrVuvDUZt9mOiKApVLgu/7F0gLmW5rb2KoZUUq/E0DrORqUCCeo+VBVUhosjMBBPc3lHNUjRV8DiiAMeaNijAteNB4N/HAjw7tERWgdlwih+8NM1dnVXUuq3UuC0cb/YxMJdzc9AOV1Xhlckg3fUuPJY1xYKqcKbNT1yS6Z+LEE+mcIhZ/rdbWmiv9xb9B1RF4HeXV/XCHnLWoC+MrdJd76bxOtaWkOs6V1ZWkkqlEASB48ePoyiK3nXWEl41Rx6z2VzguGQQVBrdJpqP5HYHdur3v1XWuvYUs2ATUalzGWnwXHUc2U2vQBtEPnz4cMECOd9rf3h4mGQyuafWkJC7LnUuE+8/1UA0rSCQs8K0mwQymQyDg4MIgqA7IZUqoVCIgYEBOjo6CjTY+QFxzc3NmwY/XW8QuliU6oCvqqr89Kc/5ctf/jLf+c53uPnmm2/0IQHw4IMPcvLkSe655x4++tGP8sEPfpCDBw/i9/t5+OGHb/ThlXmTUe7clybli3INZFnWC3atAzo7O0tXV9e6Dv3Y2Bg2m21XW8lb6dbrx6YKBBJZFqNpJFlFyirMh5N86/lxbmrw0l3v5jvPDeO0melq8KGoKoosc8+xejIZiWf6ZohKKnV+F3d21nKyxbtpBzmYUvh/nx8nmsoSTGaJpjKkswqLkRSf/b1D/PjVGf78rQf58flZMlmFUDKDoqh47Cb8djOfvOsAFZsMjGZUWAzESKUlfG4HlS7rnhS8MUnluy9MFRT3Gu8+UsvtbZsvbjQ0i1ObzUZHR8e6gkeWZcLhMPF4nLRgIiIJyIg0Vrjw2s1YTVflTXuNNly506HZ/SI/PKu7u/u6hV2+NWQwGCQej28rvGw35NtxasOspUj+MOqRI0d2lCycSqX0XapwOFxg0en1eotWgKfTaXp7e/H7/bS2tpaMPEiWZf7u7/6OCxcu8C//8i8l/Rl6E1Iab5IyQLlzX+Z1jJY06Xa7OXPmzIZdLIPBsKGEZ6tstVsPOW/4lydDPDO0TCojowINPjtHGzwca/SQzaSpcYi01XqYD6cJJzNUO80YLUYa/A7aKys51lxJSsqQkVKIUpSJ6TgWqw2fw4LdZi14jbICsXSW9montV4bc6EUBjEXmFXntqKSs+g0GQQG5mJIV2w3FyMp7uyowrrJqiGZTNLf34/L5aK9vR2D4dp67d1gNgg4LBt3Hz226xcqm1lc5mMwGPD7/WTMHh5/bZa5UIJYSiaZmeNdPTU4DQotfhs+s4LD4dgTBxBZlvWwn90Mze4Hmma9vr5+ywFKm1lDarsnsVgMs9ms75643e5d71hoQ/Ozs7ObJqOWCplMhv7+fiwWy66GUa1WK7W1tboLTL4kbWJiAkVRCuRSO3kvazsLpbYADQaDfOxjH+PYsWM89thjJZ3YW6bMjabcuS9NyhflGsiyzOjoqJ4yqgXLbMTs7CyZTGbb6YBb7tYLAqGkTDKrkJFVvvu7Cd27HnJWi00+O0ImwdBClIMNVRxv8RNIZHBZjNjNBuq9VmpdJsS8y56WBS7ORnh1YpVYMkWj18ItDQ6SgQUcDgc+nw+Dw8MzQ6tkVfjGc2PMhXIhW2cO+Pm//6CLmWAKh9VAMC7xLy9MsnjFEcZrN/Gh21o50+rHk5fwrqoq8/PzTE1N0dnZue05BVEUkVUVka27mgiCQP9igv/1yrSehApQ77Xxx6ebcJk3LizzPfa7urquWyxnVYF/PT9H/1yEZEZhZClGRlbw2U187M4DvHx5kT+6pY7V6cukUincbrfedd6tnjx/aLaU00bzi2UtvbeYaHkGwWCQSCSC0WjUi/3tSkyy2SwDAwMYDAY6Ozu3LU/R3qsGQdh0nqVYaDsLbW1te5KUmo+2S6UV/JIk4XQ69WL/WimvWnrz/Pw8R48e3dHOwl7R19fHJz7xCf7qr/6K973vfSX7GXqTU74oJUR56VvmdYf2o3XrrbdeP2nTaCSZTF7zPmvZsrYegb65GE8PLmE2inhsZkaX4zR4beTylFQyqSSzAYXTByqYSwikZQVZVRlejBJJSNjNBrrq3Nx1qFIvZPNTU3PlrkD/skQKE//x5m4SoVWWlpaQ5Xl66g/wf/2kn0BcymmQrUbuOdbAP//bOLF0FimrcLajkk/efZDR5RiZrEJrhZ2x5RirVQ48llx1L0kSAwMDmEymbVsHCoJAMClzaS7EdCBBjcvKTU2e66aGaue6o8rO+0418fJ4gHg6S5Pfzpk2P+5NUnE1i8vteOxHUjIzgQSCACuxtL4ACyYyRJJZRIOJybDMXSdOkM1mdT350NAQqVRKL5B8Pt81C6S1r00Leir1Ac9MJsPAwABGo3HPNOtr8wwkSSIYDLK8vFzgtX89iUkkEqG/v5+WlpZty+0EQWAlIXNhJsB8KEWjz8bRBg8V9uLvTuUvlvYr6EvbpdKcWRRFIRaLEQwGC1Je84d0taC/gYEBRFHkxIkTJTOzoKoqP/rRj/jHf/xHvvvd73L06NEbfUhlyrwuKBf3ZV53+Hy+LRdK25HlbEdbD7ASz/Lzi/MkMzKVTgsWk4GVWBqLUaTaYWQ1sIrVYqW9voJ6nwPLbITDtW6e6F0gkszgt5uIpbO8PBFAQeU9R2oQUQmnZc5PBddt34yvxAmmFJrzvLPHAhIem4n2KgeyAm85WMGlmRBP9C5w16EqXFYjT/Qu8Kv+Re7/vQ7OTQZ5dmgJQQCHOffx16QtO/UuDyZlHn55mvlwTjc/tBCldy7MH59ppsp+/SLBKKh0Vds5UGFDklUcJhGB9UOtiqJs2eJyLQZRwGw0oKaypDJXO7UCYDMbyMgKS9GcP3x+MmZra6uuJ88vkLTdk82GR7U5ALvdvq9BTztBWyztR2c5H7PZvKHXfr7ERCv0fT4fJpOpoLO8k2J5JSHz/RenWInldrGGF6P0zkX449NN+KzFu0ZasXyjB3xFUdSHdLWU10QioculotGo3gCpra3VZXilQDab5Utf+hJDQ0NF96ovU+aNTrm4L/O6Yztbslt1y9mOtl47honVOMlMbuGwGktztNFLg9fGciSJTTDi9XgxW8ycavVzc5Ob1gob4aTMv42uUGE3FRTvwwtRwgcr8VlFZBn9cQuPERJpucDpxWwUiaQyoIJRBLfVyItjAZwWIwYRkJJYjQKzwQSXZsOkMjJZReWWFi8ei0BfXx+yLO9YBy6KIkOLEb2w1wjEJc5PhXhXV+WWOqKqqmIWwSwKbKRKi8fj9Pf3U1lZyYkTJ7a9Le+1GTja4OGZoSVcVmPunAGdtS4URSUhyRyocmwaPqPpyTUXk3g8TjAYZGxsTB8e1Yr9ZDLJ5cuX17mhlBqqqjIxMcHKysq2F0t7wVqvfU1ikq/bt1gstLS07KgAFUWRCzMBvbDXWIqk6JuPcrbdVxSJjuYJ39jYWJTwvGIiCAIOhwOHw0FjYyNLS0uMjo7S2NhIOp3mlVdeKUgr9ng8N8QlZ3V1lY985CPceuut/OQnPymZBUeZMq8XysV9mTc01yvut9utz0fMkxiqQO90gPceq+HcVBiT0YDTYuREi48j9U5EVaHCZiCeljGJ671ZMrKKJjl3WETqPDaiqcKkWLvZQKXTUlCAVjpNvOVgJb/oXQAgK6vYLQYafDZsJgOC0UqFScEkCngsImo6Q3e7h85qGxdefZnm5uZN01u3iqb1X8tCJEVGAeM2HlrrcGt2lJoOeG5ujq6uLtxu946OUVUUTrf5EAR4dSqE0SDQ4LNx+4FKXpsKcrDaycFNivu1CIKA0+nE6XTqw6OJRILV1VUuXrxIOp3G6/USj8cxm8176hSzUzQ7TrfbXTJpo2vRJCZGo5Hl5WU6Ozux2+0Eg0H6+/tJp9O4XK4t6ckBZFVlPrTelQlgNphAEHYfMrSwsMDExETJecKvRVVVLl++TDQa5dSpUwUFfH5a8fj4eIGVrNfrxWKxXOORd8/Fixf55Cc/yYMPPsh73/vekvvslCnzeqBc3Jd5Q2MwGDYt7rfbrV/7f1sq7TgsRuLpDLFYnKWlOPF0De871UK1y4JBYJ1u3GMzUem0rOseNvvtuK54zlsMAnd3VhGIS/r9LEaRtx6uxm8vTE21iiofON2Mz2bmpYkA6azMfzzewPMjKxgFUAUDFrOBwz4Ht3dUYxUrmJua4PLAKAaDgdnZWeLxOH6/f0e+2aqq0uiz8dp0aN3f6j02TGJux+F6CILASlxmZCmsOwBVO4yM9F/CarUWRdpgN8LdHX6ON3tJZRTiUpbJ1QTv6K6hpcKG07SzIkLbSZmfn6e5uZmGhgZ9eFSTPlitVr2zf6PCiDRWVlYYGRkpOTeUteTPLNx0003Y7XYAvF4vbW1teujTRnIpr9eL0+ks+EwbBIFGn43hxei652ry23eluVcUheHhYdLp9LZnVvYbLUDL4/Fw8803r/veu9YOyszMDJlMpmBRVawAM1VVefTRR/nKV77C97//fbq7u3f9mGXKvFkpu+WUJuWLch3S6fT170SuC3X+/HlOnz6t37abbn0mYrI1AAAgAElEQVQ+qiAwMB/jJy+PsByVqK30cazZz50HK7Bv8tsuCAIjKymeuDTPSiyNQM4u855j9dS6rg6gCoJAMCUzG0qRkmQa/XYq7cbNk0ZFkeVYFgFwWg2cmwhyfjpEQpKp81i5u7MKvzHDpUuXaGho0F1bJEnSu3Sab7bP59tWsR9OqzzyyjSzwasd/EqnhT8+3UyF/fpFrCAITIXS/PCVGcLJnFwmnUpyS6OLd3TX4nPvzSCidt21HQJBEEjLKnEpt9PhthquK9NQVZXl1SDBhESF30eV24a6wf/RbCGDwWBBGJEmfdiPYl9RFEZGRkgkEvT09JS0HWe+deRG2QUbkS+XCoVCxGIxbDZbgdd+KA0/eGmKpejV7486j5X3n2rCu0PNfTKZ5NKlS9TW1tLU1FTSneZwOEx/f/+O52ugMMAsFAqRTCZxOBz6eV67qNoKmUyGBx98kImJCf7H//gf13RA2y6pVIqzZ8+STqfJZrPce++9fOlLXyq4z3e+8x3+83/+z7qM6i/+4i/42Mc+VrRjeJNQum/8NyHl4r40KV+U6yBJ0pY6bYqi8MILL3D77bcDu+vW56M5YYRCIWpbDoLBjEEEr/X6gUiCIBBKKYQSGQwGAa/NiMu8sTOMdnzb/ZwKgkA4pZBVVOwmgaW5aRYXF+nu7r7mIOJmxb7WpduoyNJez8BClMnVOHUeK0fqt+5AIinwv16dY2ghiqoohMNhFEWhssLHx+86SL1r7zW/2gLj2eFlFiO5JOFTrT6ONrgwb1LzZTIZJpYj/G4syEoKzEYjRxs8nG7zbbq401hrC2kymQp0zsXWGCcSCXp7e6mpqaG5ubnkC9CBgYFdD/hqXvtasFY0GqWyshJbVSN983FmQ0naq5zc1OjBZ93483c9tFCyrq6uohakxUb7vpqbm+PIkSP6LkixHjsej+vnORaLYbVa9ffz9TINlpeX+fCHP8zdd9/NF7/4xaK/97XjczqdZDIZ7rjjDh566CFuvfVW/T7f+c53eOWVV/jKV75S1Od+k1G6XypvQkp377BMmSKQr+EuRrcecp26gYEBHA4HXV1da36MtjY86rEIeCzmgts2u+9OUFUVt0UgFotz8VI/VVVVnDx5ct1rzv/RVRQFs9lMdXU11dXVwNVif2VlhdHR0Q2Lfe313N7m5fY2r/5YWz32ZEZhKZpGktKEQiGcTid2ux0FgcVImga3edPHyqo5t56EJOMwG/DZrrG7cQ1WEjL/69wMoURu5yCSzPDYxSRmYyM31TvXPf/KygqBWJpfT6RYiGnDzzK/GVpCVlXe3llxTT3SWlvIdDr32peWlhgZGcFgMOjF0WaLqq0yPz/P5OQk3d3dO55Z2A+0BNfFxcUCGc5OEQQBu92O3W7Xk2vnIxI/eXUKKSNhFgWGU3HcBhnRb8Jut29ZTqMoCpcvXyYWi5V8KJksywwODgLsic1l/gxKY2MjgL6ompubY3BwEKPRqL+f84PiXn31VT796U/zd3/3d7znPe/Zk0WndnyQW5BnMpmSXtyWKVMMysV9mTcFsiyjKAqiKO6qWz83N8f09DSHDh3SvaRLEa1QWlhY2DCQSPOm75sPMxtK0uSz0VXrwmcr7LbvpNjfLiYRxGyaSDiC31+hF1gCucCtzQr7lAzPjaxybjJAKqNgNxs401bBbe0+LNtQWAiCwORqXC/sNRQVzk0GOVTtQAvR1ZJmM5kMroYOFmOT6x6vdzbMyRYvnm0chMViKbCFzD/Ply9fRhCEgmJ/K0VoNpvVi7pS14FnMhn6+vqw2Wx7Zh2aUQSeGV5hPiYDuQuqKDKr/Uu8/2Qtg4OvAhSc542cYtLpNL29vfh8vg0166WEJhmqr6+noaFh347VZrNhs9kKMg209/NHPvIR5ubmaG5uZnh4mP/5P/+nvrO6V2iOYKOjo3z605/mzJkz6+7zr//6rzz33HMcOnSI//pf/ytNTU17ekxlyuwlZVlOaVK+KNchk8lsSQ+tKAqvvfYa2Wy2QEu+3eIhlUoxMDCA1Wqlo6OjpAulZDJJf38/breb9vb2DV9rIKnw8MvTLEauuofUeWy8/3QTXsvWC4BMJqPLS0Kh0LaL/Xg8zuTkJHFbNb8YWCWbd0kPVjv5P25p2FDiIggCl+bj/PCV6YLbRQH+5NYWOiptW945EEWR50YD/Kp/cd3fGrw2/vRME3aTQCwWo6+vj7q6OpqbmxlZSfG9F9YX906LkY/e0YbfVrwCVXMw0c4zXLsI1YKempub9a51qRIKhRgYGKC9vV1fRO4FgaTC//P8OPH0+gH7+25v5YDfQjabLTjPa51i4vE4Q0NDdHZ2lvTiHq4OTnd3d+PxeG704ehIksQDDzzA5OQkN998My+99BKhUIgTJ07w/ve/n7e//e179tyhUIj3vve9/NM//RNHjhzRb19dXcXpdGKxWPj617/Oo48+ytNPP71nx/EGpXRXuW9CSrdCKVNmF2jaekVROHbsmJ6GOT8/z9DQECaTCb/ff11NqOaCMjU1RUdHR8m7i2g7C52dnZuGvoiiyKXZYEFhDzAfTtI/H+GOA1v3+zaZTAWdfa3Yv15nPz+9s6urC7vLg9fl5JWJAMmMTHuVk5ubvJtq1xVyHfKOaie1Xjsmg0AslWViNU7vbJhDVVt3P1EUhbZKBwZRQFYK/8+BKid2s8j01BRzc3MFuyAVDjNuq5FIqrBYbPTZcFkKXY12y1oHk/widG3gUzweZ2VlZd9SUXeKqqpMTk6yvLzMzTffvOc++6IARnF9/SEAJkPu8280GqmsrNTzCfKdYkZGRpAkiaqqKlKpFMlk8oZnA2yEqqqMjY0RDodLTjK0uLjIhz/8Yd75znfyta99Tf/elSSJc+fO7fnze71e7r77bn7xi18UFPf53+sf//jHeeCBB/b8WMqU2UvKnfvSpHxRrsNmnfutauu1gcZAIEA0GsVsNuvFvmZVmE6nGRgYwGw2c+jQoZLu1mvHqrmLXOtYVUHgBy/PMrSw3hKwu97NB042FCXMBzbu7LvdbsLhME6nk87OTr3gFwSBjAJZRcVmFK5dnAsCr05HuTgb5nejK4STGZr8Dt7RXY3bauQt21igAGRVgX8fC/JvI8ukswqiAO1VTv7gaA3Tw30buraogkDvXIwn+xaIXinw6702/sPN9dS5NpcT7QWyLOudWm12Qus4+3y+Pfcm3y6SJNHX14fD4eDgwYP74xYkCDzZt8wLY6sFtzf57bz/VOOmVqjasTqdTtra2nRHnmAwSCqVwuVy6bso1/Pa32symQy9vb24XC7a29tLSjL08ssv85nPfIYvf/nL/P7v//6+Pe/y8rI+sJ5MJnnnO9/JAw88wHve8x79PvPz87qE6Mc//jFf/vKXeeGFF/btGN8glM6brUy5uC9RyhflOmSzWWS5MMU1v1u/XW19vlVhJBLJ2SKm07S1tZW8vd3i4iJjY2NbTkQVRZGnhlZ4dmh53d/e0V3D2YP+ohX3a5mfn2d0dBSXy4UkSbqWXJNLbXUBJYoiz10O8vdPDJJ/pJVOM3/7vx+lyb39hVhWFQgksqzE0jgtRuxilonhflpaWja1DVTIzS4sR9OYjSJVTvO6bIP9IBAIMDQ0pFsc5necQ6EQmUwGt9utF/vaQOONIBgMMjg4uCs7xp0Sy6g8P7JK/3yErKLS7LfxtsM11DiNG14zzbmnvb19w2NVVZVYLKZ/dyQSiYK04p3YQu4UTYp14MCBPZU3bRdVVfnud7/Lt7/9bb7//e9z8ODBfX3+ixcvct999+m/DX/0R3/Egw8+yIMPPsjJkye55557+Ku/+it++tOfYjQa8fv9fO1rX+Pw4cP7epxvAEr3R/JNSLm4L03KF+U65Bf3xXTCkSSJgYEBAHw+H5FIhGg0is1m0zv7+/mDfS0ymQxDQ0Ooqsrhw4e3FRO/mlD4/ktTBWFaVa6cN30xteIasiwzNDREJpOhq6tLlwrkp2GGw+EtD46mZPiXl2Z4dTLIYjSNlFWwGEVqPVY+eGsLZ1rcu3Ia0tI7u7u7t9T11vzy9xtFUXQJRk9Pz6ZFu6IoRCIRvQhNp9MFxf5+yEtUVWViYkKXDO3lAkMVBFKZ3HvCIBS6TimCQCQpo6jgshgwG9a7UuUHaB09enTL50dLK9YWVfsVYDY7O8vMzAxHjhwpKSlWOp3mC1/4AtFolG9961sldWxlis6N/1Eso1Mu7kuT8kW5Dlpxv5tu/Vq0DvjajqLmlx0IBHQfZ6075/f7cTgc+17sr66uMjw8TFtbG7W1tdv+/1oi7IWZEPPhFPVeK8cavVv2pt8OWvezqamJ+vr6a56rrRb7iYzK916cZi6UJC2rKCoYRDAbBN7ZXcNbDvh29DpisRj9/f2vi0CiZDJJX18fFRUVtLa2butY84OI8uUl+cV+MV+7loqqyUX2SoYjCAKLsSwvjQeYDyfxOcycaaugyWNmq1+r2WyW/v5+TCYTnZ2duz5WbVcwFArpmQba+3m3mQaKojA4OIgsy3R3dxfd5nI3LCwscN9993HPPffwuc997oamMpfZF0r3y/JNSLm4L03KF+U6ZLNZJEkqWrd+aGgIYEsdcK07pxX78Xhcj733+/17qrvVrBiTySRdXV277n4KgoCi5oYNi/1doCgK4+PjBAIBenp6duRbvtYlRiv2a2pqOLeY5Zk10iKTQeSjd7RR59qeLEdVVWZmZtYNzZYq2kK0WOFJqqoWFPvJZBKn06kXobtZwGqSoa3KxnZDIKnwvRcmWY1L+m12s4E/vbWFBvf1d7ai0Sh9fX20tLToGuxio2UaBIPBgqA4bSB6q9K0ZDJJb28vtbW1euJ0qfDCCy9w//3381/+y3/hHe94x40+nDL7Q+m8AcuUi/sSpXxRrsPnP/95zGYzd911F6dPn97xwODS0hKXL1/elQ2floCoFfuJREIvjPx+f9G6oFoHvLGxcV89q3dCIpEo6CoXq2unFfuhUAjR5ubpkSDDizFEo4kKl523dtVwS5NnW0FWkiTR39+v25yWUvdzLfnypu7u7m1JsbZDMbTkqqrqi7sjR47suc5fFEWevxzkyb6FdX+7pcXHPUdrrpkerTlN9fT06KFH+8FGNqf59psbud1oO3elloyrqirf+ta3+P73v8/DDz9Ma2vrjT6kMvtH6f4gvQkpF/elSfmiXIe5uTl+85vf8Mwzz/Dyyy9TUVHBnXfeydmzZzlx4sR1ix5Nr64oCocPHy6qXVx+YRQIBEgmk7rkQSv2t4Omqw6FQnR3dxc1Or7YaBaXMzMzdHV17bm/diKjsBJNEUmksCCjJEOIsOWwJ61IuhHDndtF89lvaGjY98WdtoDViv14PI7NZivQkucfTzqdpq+vD7fbzYEDB/ZFkiGKIj+5uMArE8F1f2urdPAnZ5owbbDo0xJcVVXdIHF6/8lmswXD0NlstqDYn5+f1xdMpeSClEql+NznPkcmk+Gf//mfS/p7qsyeUC7uS4hycV+alC/KNtCG355++ml+85vfcP78eerq6jh79ixnz57l2LFjBUXe+fPnSaVSO9ar7+T4otGo3tlPpVL6MKPf779mR1PTgFdXV9PS0lLS3XqtA26xWDh06NANKZI26oLm++xr7wNFURgZGSEej9PT01NSRdJa8jMB9rurfK1j2mxw1GAw6FkL+5kLIQgCvfNxHl0TbAbw1sPVvLXDv056Fo/H6evr2/cE1+0gyzKRSITV1VVmZmYAqKysxO/34/V6iz4fsRPm5ua47777uPfee/nsZz9b1te/OSm9D8+bmHJxX5qUL8ou0EJcnn76aZ555hkuXLhAa2srp06d4ne/+x2yLPPoo4/eMDvA/GHGQCCgO5dobjxWq1UP+FlaWqKrq6vkNeDLy8uMjo6WXAd8o2Lf4XAQCoWoq6ujra3thhdG1yKTydDf369nLdzorvK1SCQSDA8PEw6HMZlMBS4x1wqKKybxDDx2aZ7+uYh+W6PPxvtONOK1Fj7/4uIi4+PjdHd343a79/zYdoM2C9DW1kZVVZX+/REKhfT5CG3wfL8H/J9//nk+//nP89/+23/jbW972749b5mSo3S/SN+ElIv70qR8UYqIoih8+9vf5q//+q/p6upidnaWzs5OvbNfDEeM3R6fZlOoFfuZTAaPx0NnZ2dJpmBqyLLM8PAw6XSa7u7ukkrDXIu2YJqZmcHr9ZJIJAD0osjn8+1JUJkqCERSMgICHpsBdYsZAqFQiMHBQdra2qipqSn6cRWTdDpNb28vXq+XAwcOIAiCHhSnZUdoLjFasb9XC5VEVmUmmGI6mKTaZaGlwo4nL3tA27VJJpP09PTs2dxCsdASsjezudRkgNoiNh6PY7fb9ff1WslUsVAUhW9+85v88Ic/5OGHH6a5ubnoz1HmdUW5uC8hysV9aVK+KEUimUzyl3/5l8zNzfGNb3yDuro6FEWhr69P7+wPDw/T3d3NXXfdxdmzZ/dNI7wWTX4xPT1NY2Ojnu6qFfpaZ79UCujtWFzeaDaTDGWzWb0oCgZzWu18683dFn6htMJvR1e5vBzDIAh017s53erDsUkaKhQOovb09JT04g6uzi10dnbi9/s3vV86ndbPczgcxmg0FpzrYhb7mnuWqqoFUpxUKsWlS5eoqqoqeZmboigMDw8jSRLd3d1bXnhqkintfR2NRrFYLEXdRUkmk9x///2IosjXv/71or5HU6kUZ8+eJZ1Ok81muffee/nSl75UcJ90Os2HPvQhzp07R0VFBY888kh5ePfGU7ofpjch5eK+NClflCIhyzKPPfYY99xzz6Y/5LIsc+HCBV2zPzk5ybFjx7jzzju5++6798XvPJ1O09/fj81mW+fYkp82GgwGkWW5oNjf786joihMTEywurq6Y4vL/UQrPrfiiFTMYj8lw4/PzzG4EC24/dYDFby7u2pD55ZUKkVfXx8ej+eGLTK3Sn6A1k6GOyVJKij2RVHcUoDZTllZWWFkZITDhw/j8/mK+tjFRluEVFdX09zcvOvvn2Qyqb+vI5EIRqNRL/a367U/PT3Nn/3Zn/Enf/In/Pmf/3nR36Pa8LbT6SSTyXDHHXfw0EMPceutt+r3+epXv8rFixf5+te/zsMPP8yPf/xjHnnkkaIeR5ltUy7uS4hycV+alC/KDSSbzXLu3Dm92F9YWOCWW27Ri/3a2tqiFvsLCwuMj49z6NChLQ0gyrJcUIAqioLX69UH7Pay2N8ri8u9QFEURkdHicViOx6a3U2xvxiT+cZzl5GVwo+zy2rko3e04VujAdfmFq7XAS8FUqkUvb29+P3+os0taDtVWrEPxdlF0RKHI5EIR44cKZmdr83QcgH2chGiLax0W9krCyvt32bn+rnnnuMLX/gCX/nKVzh79uyeHFs+iUSCO+64g6997WucOXNGv/1d73oXf/M3f8Ntt91GNpultraW5eXlkt6JeRNQPvklRLm4L03KF6WEkCSJF198kaeffppnn32WQCDAqVOnOHv2LHfeeSdVVVU7+lHJZDIMDg4iCAKdnZ07Ll60Yl9z44Hi68hVVdV9wPfD4nK3aC4oNTU1Rel8auQX+6FQCFVVNy1AZyIZvvnc2LrHMBkEPnlXO5X2XLdU04AnEgl6enpKvvjUOuB7vQjJzzQIhUL6IlZ7X2/l8yJJEpcuXSqYBShVtJmQlZWVfckFyGezc3358mWOHTtGXV0dX/3qV/nZz37GD37wAxobG/f0eGRZ5sSJE4yOjvLpT3+aL3/5ywV/P3LkCL/4xS/042hvb+fFF1/cVkiaqqol/X54HVI+mSVEubgvTcoXpYRJpVL87ne/4+mnn+a5554jFotx5swZ7rrrLu644w58Pt91fzS0AunAgQNFH5bMZrN6BzTfDlLr7G9X2yxJEgMDA68Lx5Z8n/2enp49dxla29nXin2/34/B4eN7L04XpKUCNPvt/PHpJmzGq4uQ2trafZF/7QZFUbh8+TLRaPSGdMDz/d+1HSvN/32jWZRgMMjg4OCWd8RuJNlslr6+Pj1I7UbviGlSwIceeoinnnqKlZUVnE4nDzzwAG9/+9tpamral+MIhUK8973v5Z/+6Z84cuSIfntPTw9PPvlkQXH/0ksvbfk6K4qC+P+zd+dxUdbr/8dfw6pssihEgoEiIqCJIK6A6TGDbLHMrY5rmpWlqZnHzLTFLfVommampabyVcusxKVYXDAhTVIWFRVkERFZZGdg5v794W/uAybmAswNfp6Px3mc0zDHuWfukd73574+12VgIP+3UCeU+8vrISTCvTKJk9KIFBcXEx0dTUREBEeOHKGyspKePXsSFBRE7969a3SruHHjBufOncPY2BhPT88G6a+uW5XLy8uTb8FX7/1+p7CuuwhRWovL21HCRYgugObl5VGl0VBgYM2+M1lUSCpMTExpZdmM53xa42JjSmZmJmlpaY2iFaOuDEdXjqWEi5Bb96Lohj1ZW1vLA7c6deqkt5a3d6u4uJj4+HhcXFwaZO7GvUhNTWXMmDGMGjUKHx8fjhw5wuHDh8nMzGTcuHFMmTKl3o9h/vz5mJubM2PGDPmxuijL+fPPP5k7dy59+vShTZs2jBw5sj4O/2Gi/18KgkyEe2USJ6WRkiSJwsJCjhw5QkREBNHR0ahUKvr06UOrVq34+uuv+eCDD3jxxRf1FpB0tc15eXncuHEDQ0PDv22ua0wtLuHeNs02pCqtlpyiCtKuFyFpNNg1N8CoqpTc3FxMTEzuqQuKvuj2Aih9I6pWq5W/B1qtFiMjI3lYnI2NjSK7Dl29epXU1FS8vb0VMZysuoiICGbPns3atWvp3bt3jZ9pNBry8/PvqQzmbuXk5GBsbIy1tTVlZWU8+eSTvPfeewwaNEh+zhdffMGZM2fkDbU//PADO3bsuOvX+PXXX5k1axZLliwhLi6Ozz77jPj4+Hp5Pw8REe4VRIR7ZRInpYmQJImrV68yefJkTpw4gZOTkxz2g4KC8Pf313vo0G2u04V9XY9yR0dH2rVrp+gynOqbZj09PRW7Squ7kMvLyyMhIUHuBHI/deQNRffZ6qb4Kv0Cr7CwkMTERHkuQPX5Efn5+VRUVGBpaVkj7OvrAlu3z6K8vBwvLy9FXeBptVo+//xzDhw4QGhoKI6Ojg36+qdPn2b06NFoNBq0Wi1Dhw5l7ty5zJ07Fz8/P5599lnKy8v597//zalTp7C1tSU0NJS2bdve9s+TJIns7Owad0V+/PFHnJ2dyc3NZc6cObz//vs899xzog7/wYgPTkFEuFcmcVKaiDNnzvDqq68ybNgwpk6dikqlIicnh8jISCIjI4mJiaFFixby5lw/P78GKdW5HV1/9ZycHB555BFKSkrk4UO6tpsNNWn0bujq1e3t7RXfs1ySJNLS0sjOzsbb21tuH1pb5yN9h/2ysjLi4+MbRT94SZLIyMggKyurxmd7q+qTofPz8ykvL8fCwkL+rM3MzBrkfVZUVHDmzBlatmypuM+2uLiYN954A3t7e1asWKH4C7q78ddff/H777/j5ubGV199xYcffsj333/PunXr8PHxYfny5bi7u5OUlERCQgJDhgzR9yE3Vsr5Igsi3CuUOClNRHR0NDY2Nnh6et7255IkkZWVJbfdPHHiBPb29gQEBBAYGIiPj0+DBLyysjISEhKwsbHB1dW1RoDXTRrNy8ujqKgIExMTOexbWlo2eNivvmm2MdSrq9VqEhISMDc3x83N7Y6flz7CfpWkorC8CgMDFdbNDMm5do0LFy7QsWNHrK2t6/z16lJVVRVJSUkYGhrSoUOHe7rLpJvsqvusS0tLMTc3lz9rc3PzOg/euk2+Smx3eunSJcaMGcPrr7/OuHHjFHXR8aCGDx9OeHg4zz//POvXrwfA19eX5557jrlz5/LXX38xduxY3njjDV599VU9H22j1XS+ME2ACPfKJE7KQ0rXDk8X9uPi4nByciIwMJCgoCA6depUp2Uy99PisqysTA5EhYWFNGvWTO7GU1+j7nUqKytJTEzE2Nj4nsOcPuhqwNu3b39f9bz1GfZVKhVXiyuJPJdDRl4pJgYqXGyM6exgSutWyioRup3i4mISEhLkCckPSjc8SfdZl5SUYGZmJn/WFhYW9/3dliSJ9PR0srOzFbnJ9+DBg8ydO5evv/4af39/fR/OA9NoNDV+N2zcuJGtW7fy8ssv8/TTT+Pg4EBSUhJDhw7Fz8+PuLg4ZsyYwcsvv6zHo270RLhXEBHulUmcFAH4X/vBiIgIoqKiOHPmDG3btpXDfseOHe975VzXXcbY2Bh3d/f7rvstKyuTe+wXFRXRvHlzeWX/QQLRrXTDfZS2afZ2qreNvN8BWrdzu7Cvm1ZsbW19T2UUxWqJbX+kk5lfhqaqirz8PJo3b05I17b079ASJG2dHHN9yMrK4vLly/W6EVWSJEpLS+XPuri4WL6QvZe7VlVVVTUuSJVS1gY3v6fLli0jKiqK0NDQOm/Lqw/V21t+/fXXdO/eHWdnZ65du8a8efMIDg7mlVdeQaVSkZGRgbGxMWq1usHaezZhItwriAj3yiROinBbWq2Ws2fPymFfd4s/MDCQwMDAu+6RXV8tLiVJqhH2i4uL5dVPW1vb+yp10AXlwsJCvLy8FLfqeStdvXrLli3rvW3kg4T91PwKvolOpayslKKiIqytb/aKtzM3YXSvx2hhqpwQqqPRaDh37hxVVVUN3mlI993WfdZFRUWYmprKYf92+1FKSkqIj4+vs7sLdamwsJDXX38dZ2dnli1bpvg7NfeioKCAoUOH0qpVKxwcHNBoNHzyySccOnSIHTt20K9fP7Zt28Y777xDcHCwvg+3qRDhXkFEuFcmcVKEu6LVajlz5owc9i9evIi3t7cc9l1dXWuEy+LiYtLS0uQWl/W9eVe3+qkL+yUlJXJds62t7T9uYg97SvkAACAASURBVGxMm2YBsrOzSUlJwcPDQy/16rre77rPu7ZBTyqVirPZpaw5eBqtRoO1jY0cTK2aGTGujys2zZQV7ktLS4mPj8fR0VHuOqVvt4Z9Y2Nj+bOuqKggNTW1QYap3avz588zbtw4pk6dyr///W9FfJYPorCwEDMzM4yMjCgpKWHFihU4Ojoybtw4evfuTfPmzenSpQtLly7lhx9+ICIiAnNz879NvhUeSOP+EjUxItwrkzgp92H//v1MmTIFjUbDq6++yqxZs/R9SA2uqqqKuLg4uWY/IyODLl26EBAQgJWVFR9++CFLly6lX79+evkXuq6uWRc+S0tL5Y4ltra2cnvC6nsBGsOm2eoryh07dlTMKmj1sF9QUIBGo6FFixaYm5tTrjJl28mraA1NqP7vZe/WLRjcxREjlXJ+DV27do1Lly7d9b4QfSkvLycvL4/Lly9TVlaGpaWlXKKmmyGhT5IksW/fPj7++GM2btyIr6+vXo+nLkydOpWUlBQqKir46aefMDExITU1FUtLS1544QUGDBhA//79mTFjBi+88ALTp0+nqqpKUe1HmwgR7hVEhHtlEiflHmk0Gtzd3fn1119xcnKiW7dubN++vdYuNQ+LyspKjh8/zscff8xff/2Fq6srnp6eBAQEEBQUhIODg15X7ap3LMnLy6OsrAxzc3PKyspo3rw5Xl5eeg9E/6SoqIjExERat25N69atFb0KqtFoSElJITMzk1b2DuSprDiYmE25BpqZmuJsZ8nzXR6llbkyPvPqvfa9vb0Vc9FUG7VazZkzZ+SuU7qBcfn5+dy4ceOepkPXNY1Gw+LFi4mJiWHbtm2Knzj9T0pKShgyZAh2dnasXLmSESNGYGVlxa5du4CbbYiXLVvGt99+S1lZGS+++CIeHh7MmzdP8YsFjZRyf/E9hMSlq9AkxMbG4ubmJg8yGT58OHv27Hnow31GRgZz584lMDCQvXv3otVqOX78OBEREWzatIkbN27g7+8v99m3s7Nr0HCqUqmwtLTE0tKSNm3akJubS1JSEtbW1lRWVhIbG6vYKaO6/upXrlzBy8tLcRNGb1X97kKvXr0wNjZGo5Vwc7Ql+0YpKklDcyopyr6M9v+XTumzz3l5eTnx8fHY2dnRpUsXRV80wc0676SkJNzd3bGzswPAxMQEBwcHeaOqWq2moKCA69evc+HCBQwMDOTuR9bW1vW2mnzjxg0mTpyIu7s7YWFhir9Iuhs7d+6kqKiITz75BDs7Ow4ePEi/fv3IyMjAyckJOzs7tm/fjo+PDwcPHsTX15f58+crakOzINQXEe6FJiEzM7NGtwMnJydiYmL0eETKEBUVxYIFC+jZs6f8WFBQEEFBQcDNmuFjx44RHh7O2rVrKS8vp3v37gQFBdGnTx9atGjRIKGq+qZZPz8/edNs9cFDSUlJVFRUYGVlJZc66Gtzra4lp4mJCX5+foq/u6Dbu3BrvbqhgYpHWjTD0bo5kiTJZTz5+fmkp6dTVVVVo/VmQ4V9XQtRJfaDv1X1IVpdunS54wWoiYkJ9vb2crenyspKCgoKyMvL49KlSwA19kjURQhPSkpiwoQJzJgxgxEjRtTp3+f09HRGjRrF1atXMTAwYOLEiUyZMqXGc6KionjuuedwdXUF4IUXXmDu3LkP/NpPPvkkly9f5ueff8bJyQlJksjJyWHLli34+Pjw1FNPERMTw//93//Rs2dP5syZ88CvKQiNhSjLUSZxUu7Rzp07OXDgAF9//TUAW7ZsITY2llWrVun5yBqXoqIijh49SkREBEePHkWj0dCrVy+CgoLo1atXvWwMLC0tJSEh4a66y2i1WgoLC+UynsrKyhphvyGm++oGETWGlpzwv7aR97Oxs3rYz8/Pr/ewr5uSnJ+fj7e3t96mNd8tjUZDUlISBgYGdTJ3oaqqSu5+VFBQUOuG6LshSRI///wzixYt4ttvv6VLly4PdGy3k5WVRVZWFl27dqWoqAhfX19+/PHHGndMo6KiWLp0Kb/88kudv/6pU6fYtm0bhYWFREdH88QTT2Bpacn+/ftp1aoVM2fOpH///nX+usJtKfvW2kNGrNwLTYKTkxPp6enyP2dkZCiu9VxjYGlpSXBwMMHBwUiSxI0bNzh8+DAREREsXLgQQ0ND+vTpQ1BQED169MDMzOy+X+t+Bmjpyhisra1xdXVFq9XK4TMzM5PKykq5FWRdh09d8MzLy/vHFVol0JXhaDQa/Pz87qvkw9DQEFtbW3n1/HYr+3X1eavVauLj47GyssLHx0fx5RO67j26vRZ1wcjIiJYtW8oDz6q3Oq3+eevCfm0XP7rWj3/99Rfh4eFymVBdc3R0xNHREbj5u6Njx45kZmbWeTmk7q7Srd9hHx8f8vLyWL16NQMHDmTZsmUATJgwgT///FO+WyAIDxuxcq9M4qTco6qqKtzd3QkPD6d169Z069aNbdu24eXlpe9DazIkSSI3N5eoqCgiIyM5fvw45ubm9OnTh8DAQPz9/e+6TKayspKkpCQMDQ3p0KFDndUa37rSrOsOowuf91vmUF5eTkJCgnxRofTgqZveWt+bfG+3sn8/YV9Xr17XcxfqS05ODhcvXmzwTk4ajUa+c5Wfny/fuTIwMMDY2Jh27dqRn5/Pq6++yuOPP84nn3zSYF1hUlNTCQwMlC/QdKKionjxxRdxcnLi0UcfZenSpff0e7msrIw//vhDHgCWmppKt27dkCRJ/l5v3ryZhIQEfH19GTp0aJ2/N+GuiJV7BRHhXpnESbkPYWFhTJ06FY1Gw7hx43j//ff1fUhNmiRJZGdnExkZSWRkJLGxsdja2sqdeLp27XrbcKcra2nbtm29T8S83ZAna2trecjT3YT9nJwcLly40CjqvwGuXLlCWlqaXvqr68KnrtXpP600S5JEWloa165dw9vbW/F3QyRJkveGeHt763XDMfyvTO3YsWN89tln5OTkUFVVxTPPPMOMGTMabOW6uLiYoKAg3n//fV544YUaPyssLMTAwAALCwvCwsKYMmUKycnJd/1nazQali9fzokTJzhx4gRz5sxh7NixAHLALysr47///S9ZWVm89957ODk51en7E+6KCPcKIsK9MomToiC1bRqbN28e69evl1caFyxYQEhICAALFy5kw4YNGBoa8vnnnzNw4ECgaffilySJzMxMIiIiiIiI4NSpUzzyyCNy2Pfw8OD999+nZcuWzJgxQy+bYXVhXxc+gRo15NVXOTUaDcnJyZSXl+Pp6an3IPdPNBoNZ8+eRZIkPDw8FNHH+3Yrzbqwb2FhwcWLFzE1NcXd3V3xd0N0ZUMtWrSgbdu2iureI0kSu3fvZvny5cyYMYMrV65w6NAh0tPT6dy5MwsWLKi3wFtZWcmgQYMYOHAg06ZN+8fnu7i4cOLECbn0qDYajUbew/Drr78yYsQIhgwZwpdfflnjebqAn5mZSUlJCe7u7vf/ZoQHoZy/EIII9wolToqC1LZpbMeOHVhYWDBjxowaz09MTGTEiBHExsZy5coV/vWvf3H+/HmAh6oXvyRJpKamEhERwU8//cTRo0fp3LkzAwYMoF+/foroYV9VVSUHz4KCAgC55WZmZiaOjo44OzsrKsjdjq4MR1f6oNTj1e2RuHr1KllZWRgbG9OyZct/rCHXtxs3bpCUlES7du0UVzZUVVXFRx99RFJSEt999x02NjbyzzQaDWfOnKF9+/aYm5vX+WtLksTo0aOxtbVlxYoVt33O1atX5XkasbGxDBkyhMuXL9/xO1q95CY3NxdjY2P++usvNm/ezBNPPMFTTz3VKO6iPWSU+UvnIaX/pR1BULjaNo3VZs+ePQwfPhxTU1NcXV1xc3MjNjYW4KHqxa9SqXBxccHIyIiMjAzCwsKwsbEhPDyc5cuXk5iYiJubG4GBgQQGBuLh4dHgq7dGRka0atVKDmxqtZqUlBQuXLiAiYkJ2dnZqNVqvQwduhu6TckZGRmNote+SqWitLSUwsJCunfvTvPmzf+2Ibqhux/9E90sg86dOz/QBvL6kJuby/jx4/H39+enn3762/fT0NCwXrrk6ERHR7NlyxY6deokv86CBQtIS0sDYNKkSezatYu1a9diZGRE8+bNCQ0Nvatgn5OTwzPPPIONjQ09e/ZkxowZ5OXl8f333+Pi4oKzszPx8fEEBwfX2/sThMZKrNwrkzgpClV909jy5cv59ttvsbKyws/Pj2XLlmFjY8PkyZPp0aMHr7zyCgDjx4+X/wW0f//+Gu06Y2JiWL16td7eT30LDw9n27ZtrFy58m/BU6vVkpiYSEREBFFRUZw/f56OHTvKYb9du3YNGvarqqrktoYeHh4YGhrKE0bz8vK4ceMGhoaG8ipzixYt9Br2q6qqOHv2LCqVSj5eJdO1jbzT8VbvflR9w6g+wn71MqeOHTsq7vM9ffo0kyZNYu7cuQwePFixd2vux+HDhzl06BDOzs54eHiwe/dutFotS5YsYeXKlSQkJPDjjz8yZ86cv/XVF/Sm6XwBmwAR7pVJnBQFunXTWHZ2Ni1btkSlUvHBBx+QlZXFxo0befPNN+nZs2eNcB8SEoJWqxW9+O9Ao9Fw+vRpOeynpKTQuXNnOew/9thj9RZgdGUXjz32mHyX5nbUanWNsG9sbFwj7DfUxUhRUREJCQm0adOmUbR8LSkpIT4+Hicnp3tqG6nbMKrbI6EL+7rPvL72bZSVlXHmzBkeffTReu02dD8kSWLHjh2sXr2aLVu2NIk7f9Xr6wEGDx5MeHg4V69exczMjJiYGDZv3kzHjh2ZPHkyly9fJj8/v17vSgj3TDl/SQRRliMId6OyspIXX3yRl19+We4GUb3Ty4QJExg0aBBw5577ohd/7QwNDfHx8cHHx4fp06dTVVXFn3/+SUREBO+88w5ZWVn4+PgQEBBA3759cXR0fODQJUkSly9fJicn567KLkxMTHBwcJDPfUVFBfn5+WRlZXHu3DmMjY3lVWZdi8K6pNu0nJmZibe3t+LLcOBmzXVqaup9de+pPtcAaob9K1eu1EvYv379OsnJyXh6et7V7IWGVFlZydy5c0lNTSU8PFz+XBozrVYrB/uTJ0/i6+vL7t276dSpE5988gkLFizAx8eHgoIC1qxZQ9u2bQkJCeGxxx7T85ELgnKJlXtlEidFQWrbNJaVlSWv8v73v/8lJiaG0NBQEhISGDlypLyhtn///iQnJyNJkujF/wDUajWxsbFERkYSFRVFbm4ufn5+BAYGEhAQgL29/T2F/YqKChISErCwsMDNza1Ognh5ebm8sl9UVISJiYkc9i0tLR/oNXRlQ7rZAEorE7mVVqvl/PnzVFRU4Onped8zBv7pNW43sfh+wr5uSFlBQYEi2lzeKicnh7FjxxIUFMScOXMUf/7vRUpKCq+++iqPPvooNjY2jBgxAmdnZ7y9vdmyZQvPPPMMubm5nD9/np49e+r7cIXbEyv3CiLCvTKJk6IgR48eJSAggE6dOsnhbMGCBWzfvp24uDh54+i6devksP/pp5+yceNGjIyMWLFihVxzL3rx153y8nKOHz9OeHg4hw4dori4GH9/f4KCgggICMDGxqbWsK8LCu3bt//HlnwPoqysTK4fLywspFmzZtjY2GBra4ulpeVdX4wUFhaSmJj4j2VDSlFWVkZ8fDz29va0adOmwcpa7jfsV1ZWEh8fL1/oKakMB+DUqVO88cYbfPzxxzzzzDOKO74HUVRUxJtvvsnYsWNxdHQkJCSEOXPmMG7cOH766Seef/55Ll68KKbNKl/T+VI2ASLcK5M4KUINLi4uWFpaYmhoiJGRESdOnCAvL49hw4aRmpqKi4sLO3bswMbGBkmSmDJlCmFhYZiZmfHtt9/StWtXADZt2sQnn3wCwJw5cxg9erQ+31adKikp4dixY4SHh3PkyBHUajU9e/YkMDCQ3r17Y2VlhVqt5tNPPyU4OJiuXbs2eDeWsrIyuX68qKiI5s2byyv7FhYWfwttkiSRkZFBVlYWXl5e9dLOsK7phn517NhR72Uj1cN+fn4+FRUVNYZqNWvWTL5watu2Lfb29no93ltJksS2bdtYt24dW7dupUOHDvo+pAd2a319QUEBM2fOpGfPnmzYsIERI0bw5ptvcv36dVq2bMmCBQt44oknxIq98olwryAi3CuTOClCDbcb/DJz5kxsbW2ZNWsWixYtIj8/n8WLFxMWFsaqVasICwsjJiaGKVOmEBMTQ15eHn5+fpw4cQKVSoWvry8nT56s0Re7KSksLOTIkSNEREQQHR1NeXk5JSUl9OrVi0WLFum9nlqSpBphv7i4GDMzM3ll38TEhLNnz2JsbIy7u7viyzC0Wi0XL16kqKhIkWUt8PewX1JSgkajwcXFhUceeUQvg9Vqo1armT17NtnZ2Xz77bcNPm24PlTvX79//368vLxwdnZm2rRp/PDDD2zatImgoCDUajXvvfceb7zxBu3bt9fzUQt3SYR7BRHhXpnESRFquF2479ChA1FRUTg6OpKVlUXfvn05d+4cr732Gn379mXEiBE1nqf7z7p16wD+9rymLDQ0lIULF/Lyyy+TkZHB77//jomJCX369CEoKEjuua5PkiRRWlpKXl4e165do6CgAEtLSxwdHbG1tcXMzEyx5RgVFRXEx8djY2ODq6urYo9TR6vVcu7cOSorK3FycpLbb1ZUVNRovamvsJ+dnc3YsWN58sknmTVrluKn9/6T6qH++vXrBAcHY2try+XLl9m1axfnzp3jyJEjWFpaMmLECN566y1cXFzYsGGDno9cuAfK/kv/kBHdcgShEVCpVDz55JOoVCpee+01Jk6cSHZ2tlx/7ejoyLVr1wDIzMzE2dlZ/v86OTnJHVZu93hT9+6775KRkcHhw4fl1XpJkrh+/TqRkZH8/PPPzJkzB0tLS3lzbrdu3Rq8ZEelUmFmZkZubi4ajYYePXogSRJ5eXlcuHCB0tJSLCws5JX95s2bKyJE5+Xlce7cOdzd3bGzs9P34fyj8vJyzpw5g4ODgzx92NbWFldX1xor+4mJiXoJ+ydOnGDy5MksWrSIkJCQen+9+lY92F+8eJHCwkLGjx/PpEmT+PDDD1m6dCnvvvsurq6ufPHFFyxbtoxBgwbxzjvv6PnIBaHxEuFeEBqB6OhoHn30Ua5du8aAAQPw8PCo9bm3uxunUqlqfbype/vtt3FycqrxXlUqFa1atWLo0KEMHToUSZK4evUqERER7Ny5k5kzZ9KyZUsCAgIICgrCx8enXrq9VFdZWUliYiKmpqb4+fnJq7UWFha0adMGSZIoLi4mPz+f8+fPU1ZWhqWlZY2w35AkSSI1NZXc3Fx8fHwUVdJSG91G6tr2A1RvvakL+0VFReTl5dUI+7rPvC7fsyRJbNmyhY0bN/L999/XaTlKeno6o0aN4urVqxgYGDBx4sS/DX+6016dB6H7e7dixQo+++wz2rRpg5OTE5MmTWL+/PmMHz+ejRs38sEHH7BhwwbUarUiS7oEoTER4V4QGgFdP3x7e3sGDx5MbGwsDg4OcjvOrKwseTNgbX32nZyciIqKqvF43759G/Jt6EX1uxW1UalUODo68vLLL/Pyyy8jSRJpaWlERkayceNG4uLiePTRRwkMDCQoKIjOnTvXaQ28boiWq6trjfkJtx6jpaUllpaWctjXBc+zZ89SXl5eozNMfYZ9XXcZc3NzunbtqviyEd2FSF5e3j1tpDYwMKBFixa0aNHiH8P+g3zmFRUVvPfee9y4cYPIyMg63zhtZGTEsmXL6Nq1K0VFRfj6+jJgwIAaA7D27dtHcnIyycnJxMTE8PrrrxMTE3Nfr1dQUMDJkyfp378/cDPYX7p0ib/++ov4+HhWrlzJpk2bGD16NMuXL2fgwIH88ssvvPLKKyLYC0IdEDX3yiROiiArKSlBq9ViaWlJSUkJAwYMYO7cuYSHh2NnZydvqM3Ly2PJkiXs3buX1atXyxtq3377bWJjY8nLy8PX15c///wTgK5du3Ly5ElsbW31/A6VT5IkLl68KE/PPX36NK6urnLY9/T0vK+Aq7uIuHbtGl5eXv84ROtOdMFT1wayvkpKdBciSuwuczuVlZUkJCRgZmZWZ/MMdGr7zO8l7F+9epUxY8bwzDPPMH369Aa5UHruueeYPHkyAwYMkB+rba/OvbZelSSJNWvWcPr0aUaNGoWHhwdPPPEEnp6ebNmyBY1GQ1hYGJs2bWLatGkEBQWRnp5+VxfhgqI1/dvAjYhYuRcEhcvOzmbw4MHAzUFGI0eO5KmnnqJbt24MHTqUDRs20KZNG3bu3AlASEgIYWFhuLm5YWZmxjfffAOAra0tH3zwAd26dQNg7ty5ItjfJZVKhZubG25ubkycOFEe0BQeHs5nn31GUlISHTp0ICAggMDAQNzd3f8xpOlCZ/PmzfH19X3gUFd9ldnFxaVG/XhCQoLc810X9u91T4EkSaSnp3P16tW7muarBEVFRSQkJNzxjsiDuN1nrgv7d3M35fjx40ydOpWlS5fy5JNP1vnx3U5qaiqnTp2ie/fuNR6vbU/OvYR7XX39Cy+8QG5uLnv37sXDw4P169czc+ZM/vjjD3r16sWAAQO4ePEin332Gd26daN169Z19v4EQRAr90olToqgGOfOnWPYsGHyP1+6dImPPvqIgoIC1q9fT6tWrYCbg710GwAXLlzIhg0bMDQ05PPPP2fgwIHAzfZ3U6ZMQaPR8OqrrzJr1qyGf0P1QKvVkpCQQHh4OFFRUVy4cAEvLy95Zf/WDjLnz58nNzeXdu3aNdjqt1arlbvC5OfnU1lZSYsWLeSwf6dyiKqqKhITExtNW064OUE6LS0NLy8vLCws9HIM1cN+fn4+sbGxRERE0KdPH9RqNXv37mX79u0NNqCpuLiYoKAg3n//fV544YUaP3v66af5z3/+Q58+fQDo378/S5YswdfX967+7Fv718fFxbF582YcHByYOXMm69atY//+/Xz22We0b9+ea9euUVpaiouLS529P0GvxMq9gohwr0zipAiKpNFoaN26NTExMXzzzTdYWFgwY8aMGs9JTExkxIgRxMbGcuXKFf71r39x/vx5ANzd3fn1119xcnKiW7dubN++vUbdb1Oh0WiIi4uTy3jS0tJ4/PHH6d27N/Hx8Rw7doz9+/frtXe5RqOpEfarqqqwtraWw75uA3FxcTHx8fGNZjqu7q6KWq3G09MTIyPl3KBWq9UcOXKENWvWkJSUhJWVFb6+vvTt25egoKB6DbqVlZUMGjSIgQMHMm3atL/9/EHKcrRaLQYGBlRWVrJlyxaeeOIJ2rRpQ0xMDFu2bKF79+6MGTOGadOmkZGRwfr16/U+Z0KocyLcK4hyfusJgqB44eHhtGvXjscee6zW5+zZs4fhw4djamqKq6srbm5uxMbGAuDm5kbbtm0BGD58OHv27GmS4d7Q0BBfX198fX159913qaysJDw8nKlTp2JqaopKpWLWrFkEBATQt29fHBwcGrxzkaGhIba2tnJplkajoaCggPz8fC5fvoxWq8XY2JjS0lK8vb31Pm32bpSXlxMfH0+rVq3o0KGD4rpBXb9+nUWLFjFkyBD27t2LJEnExcURFRXFW2+9xfz58+ukQ82tJEli/PjxdOzY8bbBHuDZZ59l9erVDB8+nJiYGFq0aHHXF3MGBgYkJSUxZswYHn/8cf744w/atm3LjBkzyMrKIjIyEkdHRxYvXsw333wjgr0g1DMR7gVBuGuhoaE1hl6tXr2azZs34+fnx7Jly7CxsSEzM5MePXrIz6neT//Wmt777cbR2MTExDBr1iyWLl3KoEGDUKvVHD9+nMjISMaPH09BQQHdunUjKCiIPn360LJlS72EfTs7O+zs7NBoNCQlJVFeXk6rVq3kOy/W1tZy/biSVsThf/32PTw8FDl1OTo6munTp7NixQr69esnP667CJw+fXq9vvaWLVvo1KkTXbp0AW6W0aWlpQEwadKkWvfq1Ka4uFgud7p+/TpffPEFCxcuxMvLi6CgINLS0mjevDmTJ0/m4sWL7Nmzh+7duzNx4sR6e5+CINwkynKUSZwUQXHUajWPPvooCQkJODg4kJ2dLYfQDz74gKysLDZu3Mibb75Jz549eeWVVwAYP348ISEhaLVaDhw4wNdffw3Ali1biI2NZdWqVfp8Ww1i8eLFjBw5staOIGVlZfz++++Eh4dz+PBhysrK6N69O4GBgfTp0wdra+sGC/ulpaXEx8fj6OhYYz5AVVWVXMJTUFAAIAd9GxsbvdXhS5LE5cuXuX79Ot7e3orrt6/Valm/fj27du1i+/bttGnTRt+H9MASExO5ePEi3bt3Z/fu3fKeGrVazUsvvcTUqVMxMzNjzZo1zJw5k379+pGbm4uTk5Oej1yoR8q6TfaQU9bSiyAIirVv3z66du0qdx2p3n1kwoQJDBo0CKi9zz5Q6+NN3XvvvXfHnzdv3px+/frJK7rFxcVER0cTHh7OypUrqaqqolevXgQGBtKrVy8sLS3rJexnZ2eTkpKCp6cnVlZWNX5mZGREq1at5A3UlZWVFBQUkJuby8WLFzEwMJCDvrW1dYOE/aqqKhISEjA1NVVkv/2ysjKmTp2KgYEBv/32W4MPGqsv9vb2fPrpp8yZM4cWLVowceJEVCoVP/30E8899xxjx47l999/x9TUlJMnTxIcHCyCvSA0IBHuBUG4K9u3b69RkqMboAWwe/duvL29gZu1uyNHjmTatGlcuXKF5ORk/P39kSSJ5ORkUlJSaN26NaGhoWzbtk0v70XpLCwsGDhwIAMHDkSSJAoLCzl8+DAREREsXrwYAwMD+vTpQ2BgID169HjgoUdarZbk5GTKysrw9fW9q2m8xsbGfwv7+fn55OTkcOHCBQwNDeWw36JFizoP+0rf6Juens6Yh3G6FQAAHCRJREFUMWN4+eWXeeONNxR34XGvdJtmAVq2bCnX1s+fP1++0DQxMWHx4sU8/vjjrFixgjFjxjB+/Hh9HrYgPJREWY4yiZMiKEppaSnOzs5cunRJ3gz373//m7i4OFQqFS4uLqxbt04OWZ9++ikbN27EyMiIFStWEBwcDEBYWBhTp05Fo9Ewbtw43n//fb29p8ZKkiTy8vI4dOgQERERHD9+nObNm9OnTx+CgoLw9/e/p9KU8vJyzpw5Q6tWrXjsscfq7I6AWq2WhzvduHEDY2PjGmH/QcKu7g6Dt7e33tpc3snhw4eZOXMmq1evJjAwUN+H88B0/esBdu7cyR9//MGsWbM4cOAAkZGRDB48WP47/tVXX5GYmEjHjh157bXX9HnYQsMSZTkKIsK9MomTIgjCXZEkiWvXrhEZGUlUVBQxMTHY2NjIA7V8fX1r7WGfm5vL+fPnG2QTakVFhRz2CwsLMTY2lttuWllZ3VXY12q1XLhwgbKyMjw9Pe/qDkND0mq1rF27lp9++ont27c3uVKUefPmERYWxqpVq+QhWEuWLCE9PZ1Ro0Zx6NAh+vfvT+fOnRvFLAShTolwryAi3CuTOCnCQ2/cuHH88ssv2NvbEx8fD9zsiDJs2DBSU1NxcXFhx44d2NjYIEkSU6ZMISwsDDMzM7799lu5peCmTZv45JNPAJgzZw6jR48G4OTJk4wZM4aysjJCQkJYuXKl4lon3g9Jkrhy5QoRERFERkZy8uRJHBwcCAwMJDAwUO6W8p///IeePXsyaNCgOw6wqi/l5eU1wr6pqakc9i0tLf8W9isqKoiPj8fW1hYXFxfFnavS0lLeeustLCwsWL169T1PAFai6iv2mZmZTJw4kR07dqBWqzl58iRXrlxh1KhRfPLJJ5w8eZJr165x8ODBBy4TExolZf2FfMiJcK9M4qQIDUpXT1tYWPi3jZT6cvjwYSwsLBg1apQc7mfOnImtrS2zZs1i0aJF5Ofns3jxYnk1MSwsjJiYGKZMmUJMTAx5eXn4+flx4sQJVCoVvr6+nDx5EhsbG/z9/Vm5ciU9evQgJCSEt99+Wy4taEp03WR0Yf/EiRNUVFTg5eXFzJkz6dKliyJWWcvKyuRuPIWFhTRr1gwbGxtsbW3RaDScPXsWd3d37Ozs9H2of5OamsqYMWMYN24cr732muIuPO5V9VB//fp1WrZsCdzcOP/bb7/Rr18/VCoVBw8eZPLkycycOZMrV648NBvkhdtq3F/6JkaEe2USJ0VocFu3bmXXrl0kJyfTtWtXXnzxRQYNGqTX4JeamsqgQYPkcF99amZWVhZ9+/bl3LlztU7X1P1n3bp1wP+mcPbt25cnnniCs2fPAjc3C1d/XlMVHR3Nm2++yeTJk9FoNERERJCQkICbmxsBAQEEBQXh4eGhiM2fZWVl5ObmkpmZSXFxMTY2NtjZ2WFra4uFhYViAnRERASzZ89m7dq19O7dW9+H88CqqqrkGQZbtmxhxYoVdOrUienTp9OpUycOHDhAz549sbKyYuvWrWRlZTF9+nTFnA9Bb8QXQEFEtxxBeMhJksT69ev5+uuvWbNmDc7OzuzevZvffvsNX19fnJyc0Gg0GBgY6P1f4NnZ2fKmXUdHR65duwbcLBm4dUBWZmbmHR+vXg9dfdBWU/Xjjz+yfPlyfv75Z/kzee2119BqtZw9e5bw8HAWLlwoD4LSlfG4ubnpJeybmJiQn5+PlZUVfn5+VFRUkJeXR2pqKsXFxZiZmckr++bm5g3+3dRqtXz++eccOHCA/fv3N/pVa7VajYmJiRzsv/zyS44ePcqmTZtYv349GzZsYMSIEQwcOJCqqiqmTZtGREQEW7du1fvvBUEQahLhXhAecqmpqRw7dozZs2fj5+cH3JxYmZKSQmlpKcDfVu8lSUKSJEWs8MLN47mVSqW658ebsn79+hESEvK3+noDAwM8PT3x9PTkrbfeQqPREB8fT3h4OHPnzpW70gQGBhIUFFSnHXVqU1JSQnx8PM7OznJoNjMzw8zMDCcnJyRJorS0lLy8PC5dukRJSQnm5uZy2DczM6vXYywuLuaNN96gVatWHDx4sEnU148dO5aCggL27t3LmTNnWLVqFU8++STe3t7MnTuXxYsXc/DgQczMzFCr1Vy9epXo6GhRXy8ICqSMfzMLgqA3p06dwtraGn9/fwA0Gg0Arq6uWFhYsGTJEtavX8/+/fuprKwEbgZhfQR7BwcHsrKygJt99u3t7YHaB2fd6fGMjIy/Pd6UWVlZ3dXGWUNDQx5//HGmTZvGTz/9xKlTp5g2bRqFhYVMnz6dnj17MmnSJLZu3UpmZuZtL5QexLVr1zhz5gyenp61nhOVSoW5uTnOzs507tyZHj160LZtWyRJ4sKFCxw/fpwzZ86QkZFBaWlpnR7jpUuXePrppwkODmbNmjV1HuzHjRuHvb29PDfiVlFRUbRo0YIuXbrQpUsXPvroozp53W+//ZakpCRWrVpFp06deOedd4iOjubPP//Ezs6OMWPGkJKSwt69e/Hw8GDbtm0i2AuCQomVe0F4yBUVFVFYWCgHKUNDQ3lDXX5+PgsXLmTixImEhoby2GOPYW1tza5du3B3d6d///4YGRnV2IAHNzfh5eXl1XlJx7PPPsumTZuYNWsWmzZt4rnnnpMfX716NcOHDycmJoYWLVrg6OjIwIEDmT17Nvn5+QAcPHiQhQsXYmtri6WlJcePH6d79+5s3ryZt956q86OsykxMjKiW7dudOvWjffee4/KykpiY2OJjIzk9ddf5/r16/j6+splPPb29ve1aq7Varl48SLFxcV3PUhLR6VSYWFhgYWFBW3atEGSJIqLi8nPz+f8+fOUlZVhaWkpr+zf76TYX3/9lblz5/LVV1/JrSDr2pgxY5g8eTKjRo2q9TkBAQH88ssvdfq6xsbG7N69m4CAAHx8fHj11VfJzc1l0aJFfPnll3h6ejJs2DBsbGxEqBcEhRMbapVJnBShwWzfvp2vv/6a8PBwNBpNjVX577//nuXLlxMdHQ3A+fPnWbJkCf7+/kRGRqJSqdi4cSPNmjVDrVZjaGiIoaEhsbGxLFu2jA8++ABJkrh06RIDBgzAzMzsro9rxIgRREVFcf36dRwcHJg/fz7PP/88Q4cOJS0tjTZt2rBz505sbW2RJInJkyezf/9+zMzM+Oabb+QSo40bN7JgwQIA3n//fcaOHQvAiRMn5FaYwcHBrFq1qsmX5tSHiooKfv/9d7nPflFREf7+/gQGBhIQEICtre0/fq5qtZozZ85gY2ODq6trnZ8HSZIoKioiLy+P/Px8ysvLsbKykodq/VPY12q1LF++nMjISEJDQ3FwcKjT47vVrRvJq4uKimLp0qV1Hu51vvnmG/7zn/9w9uxZTExM+M9//sP169fZunVrvbye0GSIX54KIsK9MomTIjSYixcv8vHHHzNs2DC5FWReXh4Aa9eupaSkhAULFpCRkcHy5csJDQ3lq6++4rHHHmPVqlVMmDCBjh078vHHH7Nnzx58fHzw9/fnwoULLFy4EFNTU/766y/8/f3Zu3cvbm5utGvXTt64p1S367P/7rvv8vPPP2NiYkK7du345ptvsLa2JjU1lY4dO9KhQwcAevTowZdffgnU3k+/tp79jV1paSnHjh0jIiKCQ4cOoVar6dGjB4GBgfTp0wcrK6sa4T09PZ2MjAzat28vt1ysb1qtlqKiIrnPfkVFBVZWVnKf/eoTfouKipg0aRLOzs4sW7asQQZn/VO4f/HFF3FycuLRRx9l6dKleHl51enrv/HGG/z1119ER0dz9uxZfvrpJ6ZPn66IlqmCYolwryAi3CuTOClCg9q5cyeLFy9GrVbTrl07XnnlFdq3b8/q1asJDg5m8ODBnD59mg0bNtC6dWuys7OJjo7G0NCQSZMmUV5eztatW4mKimLbtm2sW7eOJ598ktmzZ/Pdd9/h7++PjY0NI0eOZP369bi6uur7Lf+j2/XZP3jwIP369cPIyIj33nsPgMWLF98xjNXWT7+2nv1NTVFREUePHiU8PJzo6Gi0Wi29e/cmICCAkydPsn//fvmOi75otVoKCwvlsD9v3jysra15/PHH2bVrF9OnT2fUqFENdmfnTt+nwsJCDAwMsLCwICwsjClTppCcnFznx9C1a1e5tl4Q7oII9woiNtQKgsBLL73EiRMnOHjwIBMnTqRPnz4YGBhw4cIFfH19gZu942NiYnjuuedYtmwZx48f58iRI7i7u3Px4kWmTJkCgJubGyqVCj8/PzIzM4mIiOCPP/5g27ZtHDp0iNdee43NmzfXeiyDBw/m6NGjDfK+7yQwMBBbW9sajz355JPyHYcePXrU2JR7O1lZWRQWFtKzZ09UKhWjRo3ixx9/BGDPnj3ytNzRo0fLjzc1lpaWBAcHs3TpUo4dO8b+/fvp0aMHs2fPZufOnTRr1owlS5YQGRkpd2dqaAYGBlhbW+Pq6oqvry87duzAx8eH3bt306xZMz7//HMmT57Mrl27yMnJ0csx6lhZWWFhYQFASEgIlZWVXL9+vc5f59ixY4wcObLO/1xBEOqfCPeCIMgeeeQRgoODcXBwwNvbm127dtGmTRsATE1NGT16NB9++CHfffcdiYmJ5Obm4uLiQkxMDJ06dQKgvLwcd3d32rVrR3JyMubm5gwaNAgbGxtGjx7N2LFj5bIHrVYrdzIpKChgw4YN/Pzzz39bLdR18FGSjRs31phom5KSgo+PD0FBQRw5cgTgjv30a+vZ35TpypGWL1/O9OnTOXv2LD/++CPdunVj7969DBw4kODgYBYsWMDRo0epqKho8GPUaDSsWLGCqKgoDh48SGxsLNHR0QwdOpT4+Hg++eSTBj+m6q5evSr/nYmNjUWr1dbL1N5mzZoxaNCgOv9zBUGof8ouehUEQa9uXbkeNWoUhoaG7Nu3j++++45ly5bh4uJSI4Rt3LgRExMTXF1dOXDgAC1atMDa2pqoqCgCAwPlKbJwc8VUo9FgaGjI9u3buXbtGs8++6zcslE3LbN6ra9Wq9V7f/1PP/0UIyMjXn75ZeBmOE9LS8POzo6TJ0/y/PPPk5CQ8FD20/8nX331FV988YV8R6hly5a89NJLvPTSS0iSRHZ2NhEREezatYuZM2diZ2cnd+Lp2rVrvda837hxg9dee4327duzb98++bWaNWtGUFAQQUFB9fbaOtU3kjs5OTF//ny5Be2kSZPYtWsXa9euxcjIiObNmxMaGvrQf6cEQahJhHtBEO6aubk5EydOZOLEiQByC8zJkyczcOBAfHx8yM7OZsCAAcDN/vFt27ZFq9WSkpLCpEmTqKioqNEb3NDQkMrKSjZs2MDChQtxcnIiLS2NgoICrK2t2bRpE61bt8bHxwc7Ozu9B/tNmzbxyy+/EB4eLocqU1NT+T35+vrSrl07zp8/f8d++rqe/Y6OjjV69jd1d9pXoFKpeOSRRxg5ciQjR45EkiTS09OJjIxk06ZNTJkyBUdHR4KCgggMDKRz5851tjE7KSmJCRMmMGPGDEaMGKG3wLx9+/Y7/nzy5MlMnjy5gY5GEITGSIR7QRDumy4A6cJYRkYGly5dwtjYmKysLDIzM+nduzdXrlzBzs6ODh06/G3oT2VlJcuWLaN9+/YMGDAAtVrNkCFD5BC/cOFCOnXqhEql4ty5c+zZs4cbN27QsmVLWrduDfzvIkOSJHllvz7C2f79+1m8eDGHDh2qsQE0JycHW1tbDA0NuXTpEsnJybRt2/aO/fRr69kv/I9KpaJNmzaMHj2a0aNHI0kSKSkpREREsHbtWk6fPo2LiwsBAQEEBQXh6el5zx1dJEni559/ZtGiRXz77bd06dKlnt6NIAhCwxDdcpRJnBShSbhy5QrNmjXD3Nyct956i4MHD7JmzRpCQkLkQB4eHs5bb71FSUkJrVu3xsbGhlOnTnHlyhWuXbtGr169+Oqrr+jXrx/jx49Hq9VibW3Nnj17mD17NmPGjMHIyIjKysrblmzcbxnP7frsL1y4kIqKCrnGWdfy8vvvv2fu3LlyCdH8+fN55plngNr76efm5t62Z79w97RaLcnJyURERBAREcHZs2dp3749AQEBBAYG0qFDhzuee41Gw6effkpcXBxbt26tl9p1QXhIiNowBRHhXpnESRGapOjoaB555BHatWsH3ByKNXnyZEJCQpg6dSqnT59m/fr1pKens2vXLn744Qe++uorfvvtN3kjZkpKClu3biU+Pp5XXnmFuLg4rly5woIFC4iLi8PHx4d58+bVGtRunaYrNB1arZbExETCw8OJioriwoULdOzYUa7Zb9u2rRz28/PzmTBhAp06dZL3UAiCcN/EL1UFEd1yBEFoML1795aDPYCNjQ2vv/46w4YNA6Bz586sWrWK8vJysrKy+P333+nduzdws7tMeXm5vCL++++/y5t5Fy5ciFarZefOnRgaGrJp0yYAIiIi2LBhQ41jqB7sdWU8Wq22Xt/3gxg3bhz29vZ4e3vLj82bN4/WrVvTpUsXunTpQlhYmPyzhQsX4ubmRocOHThw4ID8+P79++nQoQNubm4sWrRIfjwlJYXu3bvTvn17hg0bhlqtbpg3Vg8MDAzw9vZmypQp7N69m7i4OGbNmkV5eTmzZ8+mV69eTJgwgcWLFxMcHMyYMWNYtGiRCPaCIDQpItwLgqA3rVq1YvDgwXJLSN2dxLKyMiwtLfnxxx8JCAgAbpb4FBcX4+npCdwcKDVo0CD279+Pubk5b775Jo6Ojnh5eXHs2DEADh06xB9//AHcvEvw4YcfcunSJdRqNWq1GpVKhYGBQY3SjbS0NAoLCxvsM/gnY8aMYf/+/X97/J133iEuLo64uDhCQkIASExMJDQ0lISEBPbv388bb7yBRqNBo9Hw5ptvsm/fPhITE9m+fTuJiYkAvPfee7zzzjskJydjY2Pzt4uhxszQ0JCuXbsyY8YMfvnlF/7880/efvttLly4wJIlSxg6dKi4iyMIQpMjwr0gCIqhC1qHDh3C2tqa7777jr59+wI3Q3dxcTHu7u4AxMfH06NHD0pKStBqtTzyyCPAzQsDPz8/EhMTuXr1Kv369QNuBt/09HSaN2/O9u3b8fPzY9myZbz77rty0L1x4wYHDhyQ/1kJbjdMqzZ79uxh+PDhmJqa4urqipubG7GxscTGxuLm5kbbtm0xMTFh+PDh7NmzB0mSiIiIYMiQIUDTHqYFYGRkRPfu3dm6dStPPfWUvg9HEAShXohwLwiCYvXu3VsumXjllVeYP38+zZo1Iz09nZKSEjw8PPD09OTPP/+Ua+xDQ0Px8PCgoqKC3NxcuZ/62bNncXZ2xtHRkaNHj+Ls7Iy/vz/p6enyavWaNWtYsWIFMTEx+nnD92D16tV07tyZcePGkZ+fD9wcmuXs7Cw/Rzc0q7bHc3Nzsba2lj/j6kO2BEEQhMZJhHtBEBoFY2Nj2rVrh1arxdnZmaSkJIyMjGjfvj1eXl507NiR559/nvbt2/P8889TUVHB+fPn5Rr/y5cvy9N2jxw5wty5cwkICOCjjz7iwoUL5OTk0KxZM8zMzNi8eTNLly6lrKxMn2+5Vq+//joXL14kLi4OR0dHpk+fDlDr0Kx7fVwQBEFovMQuIkEQGhVdfby5ubn83ytXriQ7O5u0tDS8vLwAMDExwczMjD179pCfn89vv/3GhAkTSElJQaVS0b17dzQaDTk5OZSVldGqVSvgZmmK0ocEOTg4yP97woQJDBo0CLi58p6eni7/rPrQrNs93rJlSwoKCuRJwNWfLwiCIDROYuVeEIQmwcHBgW7dusnDpTp37sz7779PWFgYUVFR9OzZE1dXV37++Wfat28P3FzpPnXqFO7u7lRVVZGamlpjOJVSZWVlyf979+7dciedZ599ltDQUCoqKkhJSSE5ORl/f3+6detGcnIyKSkpqNVqQkNDefbZZ1GpVDzxxBPs2rULQAzTuku362BUnSRJvP3227i5udG5c2f+/PPPBj5CQRAeZmLlXhCEJsnIyIhnnnlGbp1ZVFSEpaUllZWVPP/88wCUlpZy5MgRnn76adLS0tBoNPj4+ADK6YdffZiWk5MT8+fPJyoqiri4OFQqFS4uLqxbtw4ALy8vhg4diqenJ0ZGRnzxxRfyxNbVq1czcOBANBoN48aNk+9wLF68mOHDhzNnzhx8fHwYP3683t5rYzFmzBgmT57MqFGjbvvzffv2kZycTHJyMjExMbz++uuNYh+HIAhNgxhipUzipAhCAygpKeGzzz5j+PDhuLi4MHz4cHr37s27776r70NTlHHjxvHLL79gb29PfHw8AMOGDePcuXMAFBQUYG1tTVxcHKmpqXTs2JEOHToA/5viC3Dy5El5Wm9ISAgrV65EpVKRl5fHsGHDSE1NxcXFhR07dmBjY6OfN3uXUlNTGTRokPx5VPfaa6/Rt29fRowYAUCHDh2IioqSW74KQhOk/5UQQSbKcgRBeGiZm5szb948PDw8MDU1ZciQIfz+++/s27fvtptNH1a367X/f//3f3Kf/RdffJEXXnhB/lm7du3kn+mCPdzcCPzVV1/Jq9q6P3PRokX079+f5ORk+vfvX2PIVmNUW3ciQRCEhiDCvSAIAje7xLzyyiv88MMPBAcHK6IkRynu1GtfkiR27Nghr1LXJisri8LCQnr27IlKpWLUqFFyT/09e/YwevRooGn02hddiARB0CcR7gVBEIT7duTIERwcHORNygApKSn4+PgQFBTEkSNHgJur2U5OTvJzqq9mZ2dnyyUrjo6OXLt2rQHfQd27U9ciQRCE+ibCvSAIgnDftm/fXmPV3tHRkbS0NE6dOsXy5csZOXIkhYWFD9Vq9rPPPsvmzZuRJInjx4/TokULUW8vCEKDEd1yBEEQhPtSVVXFDz/8wMmTJ+XHTE1NMTU1BcDX15d27dpx/vx5nJycyMjIkJ9XfTXbwcGBrKwsHB0dycrKwt7evmHfyD26XQejyspKACZNmkRISAhhYWG4ublhZmbGN998o+cjFgThYSLCvSAIgnBffvvtNzw8PGqU2+Tk5GBra4uhoSGXLl0iOTmZtm3bYmtri6WlJcePH6d79+5s3ryZt956C7i50r1p0yZmzZrVKHrtb9++/Y4/V6lUfPHFFw10NIIgCDWJshxBEAThjkaMGEHPnj05d+4cTk5ObNiwAYDQ0NC/baQ9fPgwnTt35vHHH2fIkCF8+eWX8mbctWvX8uqrr+Lm5ka7du0IDg4GYNasWfz666+0b9+eX3/9lVmzZjXsGxQEQWhCRJ97ZRInRRAEQRCExqJpbqBppMTKvSAIgiAIgiA0ESLcC4IgCIIgCEITIcK9IAiCIAiCIDQRItwLgiAIgiAIQhMhwr0gCIIgCIIgNBEi3AuCIAiCIAhCEyHCvSAIgiAIgiA0ESLcC4IgCIIgCEITIcK9IAiCIAiCIDQRItwLgiAIgiAIQhMhwr0gCIIgCIIgNBFG+j4A4bZU+j4AQRAEQRAEofERK/eCIAiCIAiC0ESIcC8IgiAIgiAITcT/a7cOZAAAAAAG+Vvf4yuK5B4AACbkHgAAJuQeAAAm5B4AACYCnthte69TWGkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 720x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(10,8))\n",
    "ax = fig.add_subplot(111, projection='3d')\n",
    "plt.title(\"Figure 9: Multivariate Exploration of Credit Amount, Instalment Rate and Age of Customers in Year\", fontsize =15)\n",
    "\n",
    "xs = dataset['CR_Amount']\n",
    "ys = dataset['Inst_Rate']\n",
    "zs = dataset['Age_Years']\n",
    "ax.scatter(xs, ys, zs, s=50, alpha=0.6, edgecolors='w')\n",
    "\n",
    "ax.set_xlabel('Credit Amount')\n",
    "ax.set_ylabel('Instalment Rate')\n",
    "ax.set_zlabel('Age in years')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Encoding Categorical Descriptive Feature\n",
    "\n",
    "Prior to modelling, we must encode all the Categorical features including target feature into the numeric values. For this purpose we can use two types of standard encoding techniques namely 'One-hot-encoding' and 'Integer-encoding'. Since we dont actually know the class value difference between values of ordinal features hence we are going to use One-hot-encoding for all the categorical values.\n",
    "\n",
    "For encoding the Categorical features, first we will separate out the target feature from rest of the data and then find out categorical features from the dataset.\n",
    "\n",
    "- Pandas series 'target' contain the target feature `Credit_Risk_Status` of the original dataset.\n",
    "- After the above step, we drop this feature from the dataset and save rest of the data in 'data' dataframe.\n",
    "\n",
    "we can check the shape of both 'data' and 'target'. Both have equal number of observation instances(rows)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "target = dataset['Credit_Risk_Status']\n",
    "data =dataset.drop(columns = 'Credit_Risk_Status')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1000, 20)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1000,)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "target.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Encoding the Target Feature.\n",
    "\n",
    "The target feature contains 2 class values 1 which represents Good Credit_Risk_Status and 2 which represents Bad Credit_Risk_Status. Generally, Positive class is represented by 1 and negative class is represented by 0. So we are replacing value 2 with 0 for representing Bad Credit_Risk_Status."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The values of the target class before encoding are:\n",
      "1    700\n",
      "2    300\n",
      "Name: Credit_Risk_Status, dtype: int64\n",
      "_________________________________________________________\n",
      "\n",
      "The values of the target class after encoding are: \n",
      "1    700\n",
      "0    300\n",
      "Name: Credit_Risk_Status, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "target = target.astype(str)\n",
    "print(\"The values of the target class before encoding are:\")\n",
    "print(target.value_counts())\n",
    "print(\"_________________________________________________________\\n\")\n",
    "print(\"The values of the target class after encoding are: \")\n",
    "target = target.replace({'2': '0'})\n",
    "print(target.value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['1', '0'], dtype=object)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "target.unique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Encoding all other Categorical Descriptive Features\n",
    "\n",
    "In the next step we have to encode the categorical features into Numeric values using One-hot-encoding technique. For this purpose we have to find out the Categorical Features from the dataset 'data'. We are storing all the features with their datatype as `np.object` into the Categorical feature list."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Existing_CA_Status',\n",
       " 'Credit_History',\n",
       " 'Purpose',\n",
       " 'Saving_Acc',\n",
       " 'Present_Employment',\n",
       " 'Gender_Status',\n",
       " 'Other_Debtors',\n",
       " 'Property',\n",
       " 'Inst_Plans',\n",
       " 'Housing',\n",
       " 'Job',\n",
       " 'Telephone_Status',\n",
       " 'Foreign_Worker']"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "categorical_cols = data.columns[data.dtypes==object].tolist()\n",
    "categorical_cols"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we are planning to perform feature selection, we are going to define ***q*** dummy variables for Categorical Descriptive Features with ***q*** different levels. Only the excception is, when a certain Categorical feature is binary i.e 2 different levels, in that case we are going to define ***q-1*** i.e. only one dummy variables. \n",
    "\n",
    "The coding operation is shown below. For each two-level categorical variable, we set the drop_first option to True to encode the variable into a single column of 0 or 1. Next, we apply the `get_dummies()` function for the regular one-hot-encoding for categorical features with more than 2 levels. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "for col in categorical_cols:\n",
    "    n = len(data[col].unique())\n",
    "    if (n == 2):\n",
    "        data[col] = pd.get_dummies(data[col], drop_first=True)\n",
    "\n",
    "data = pd.get_dummies(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After encoding, we have few new dummy features in our dataset 'data'. we can checck the shape of dataset as well as columns in it. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1000, 59)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Duration_Months', 'CR_Amount', 'Inst_Rate', 'Present_Residence_Time',\n",
       "       'Age_Years', 'Exsisting_CR_Count', 'Liability_Count',\n",
       "       'Telephone_Status', 'Foreign_Worker', 'Existing_CA_Status_A11',\n",
       "       'Existing_CA_Status_A12', 'Existing_CA_Status_A13',\n",
       "       'Existing_CA_Status_A14', 'Credit_History_A30', 'Credit_History_A31',\n",
       "       'Credit_History_A32', 'Credit_History_A33', 'Credit_History_A34',\n",
       "       'Purpose_A40', 'Purpose_A41', 'Purpose_A410', 'Purpose_A42',\n",
       "       'Purpose_A43', 'Purpose_A44', 'Purpose_A45', 'Purpose_A46',\n",
       "       'Purpose_A48', 'Purpose_A49', 'Saving_Acc_A61', 'Saving_Acc_A62',\n",
       "       'Saving_Acc_A63', 'Saving_Acc_A64', 'Saving_Acc_A65',\n",
       "       'Present_Employment_A71', 'Present_Employment_A72',\n",
       "       'Present_Employment_A73', 'Present_Employment_A74',\n",
       "       'Present_Employment_A75', 'Gender_Status_A91', 'Gender_Status_A92',\n",
       "       'Gender_Status_A93', 'Gender_Status_A94', 'Other_Debtors_A101',\n",
       "       'Other_Debtors_A102', 'Other_Debtors_A103', 'Property_A121',\n",
       "       'Property_A122', 'Property_A123', 'Property_A124', 'Inst_Plans_A141',\n",
       "       'Inst_Plans_A142', 'Inst_Plans_A143', 'Housing_A151', 'Housing_A152',\n",
       "       'Housing_A153', 'Job_A171', 'Job_A172', 'Job_A173', 'Job_A174'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Scaling Of Features\n",
    "- Once we have encoded all the categorical features, we perform `Min-Max scaling` of all the descriptive features. \n",
    "- It is convienient to choose Min Max Scaling for this Data Set as after integer encoding it is observed that the values are 0 and 1.\n",
    "- Hence after Min-Max Scaling we can obtain all the values in the range 0 - 1. \n",
    "- Before scaling the data set we copy Data to keep track of all the Feature Names."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import preprocessing\n",
    "\n",
    "scaledData = data.copy()\n",
    "Data_Scaler = preprocessing.MinMaxScaler()\n",
    "Data_Scaler.fit(data)\n",
    "data = Data_Scaler.fit_transform(data)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the below chunk the descriptive features are displayed after scaling. Since the output of the Scaler function is a `Numpy array`, all column names are lost, hence we add those from the previously stored copy of the data set.\n",
    "The following observations can be made from the scaled data: \n",
    "- The `Binary features` are kept binary even after scaling.\n",
    "- All the `Descriptive features` are in the range of 0-1."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Duration_Months</th>\n",
       "      <th>CR_Amount</th>\n",
       "      <th>Inst_Rate</th>\n",
       "      <th>Present_Residence_Time</th>\n",
       "      <th>Age_Years</th>\n",
       "      <th>Exsisting_CR_Count</th>\n",
       "      <th>Liability_Count</th>\n",
       "      <th>Telephone_Status</th>\n",
       "      <th>Foreign_Worker</th>\n",
       "      <th>Existing_CA_Status_A11</th>\n",
       "      <th>Existing_CA_Status_A12</th>\n",
       "      <th>Existing_CA_Status_A13</th>\n",
       "      <th>Existing_CA_Status_A14</th>\n",
       "      <th>Credit_History_A30</th>\n",
       "      <th>Credit_History_A31</th>\n",
       "      <th>Credit_History_A32</th>\n",
       "      <th>Credit_History_A33</th>\n",
       "      <th>Credit_History_A34</th>\n",
       "      <th>Purpose_A40</th>\n",
       "      <th>Purpose_A41</th>\n",
       "      <th>Purpose_A410</th>\n",
       "      <th>Purpose_A42</th>\n",
       "      <th>Purpose_A43</th>\n",
       "      <th>Purpose_A44</th>\n",
       "      <th>Purpose_A45</th>\n",
       "      <th>Purpose_A46</th>\n",
       "      <th>Purpose_A48</th>\n",
       "      <th>Purpose_A49</th>\n",
       "      <th>Saving_Acc_A61</th>\n",
       "      <th>Saving_Acc_A62</th>\n",
       "      <th>Saving_Acc_A63</th>\n",
       "      <th>Saving_Acc_A64</th>\n",
       "      <th>Saving_Acc_A65</th>\n",
       "      <th>Present_Employment_A71</th>\n",
       "      <th>Present_Employment_A72</th>\n",
       "      <th>Present_Employment_A73</th>\n",
       "      <th>Present_Employment_A74</th>\n",
       "      <th>Present_Employment_A75</th>\n",
       "      <th>Gender_Status_A91</th>\n",
       "      <th>Gender_Status_A92</th>\n",
       "      <th>Gender_Status_A93</th>\n",
       "      <th>Gender_Status_A94</th>\n",
       "      <th>Other_Debtors_A101</th>\n",
       "      <th>Other_Debtors_A102</th>\n",
       "      <th>Other_Debtors_A103</th>\n",
       "      <th>Property_A121</th>\n",
       "      <th>Property_A122</th>\n",
       "      <th>Property_A123</th>\n",
       "      <th>Property_A124</th>\n",
       "      <th>Inst_Plans_A141</th>\n",
       "      <th>Inst_Plans_A142</th>\n",
       "      <th>Inst_Plans_A143</th>\n",
       "      <th>Housing_A151</th>\n",
       "      <th>Housing_A152</th>\n",
       "      <th>Housing_A153</th>\n",
       "      <th>Job_A171</th>\n",
       "      <th>Job_A172</th>\n",
       "      <th>Job_A173</th>\n",
       "      <th>Job_A174</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>842</th>\n",
       "      <td>0.205882</td>\n",
       "      <td>0.093155</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.071429</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>0.470588</td>\n",
       "      <td>0.086332</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.321429</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>308</th>\n",
       "      <td>0.058824</td>\n",
       "      <td>0.054308</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.089286</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>881</th>\n",
       "      <td>0.294118</td>\n",
       "      <td>0.496699</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.517857</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>350</th>\n",
       "      <td>0.073529</td>\n",
       "      <td>0.054253</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.071429</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Duration_Months  CR_Amount  Inst_Rate  Present_Residence_Time  Age_Years  \\\n",
       "842         0.205882   0.093155   1.000000                     1.0   0.071429   \n",
       "68          0.470588   0.086332   1.000000                     1.0   0.321429   \n",
       "308         0.058824   0.054308   0.666667                     1.0   0.089286   \n",
       "881         0.294118   0.496699   0.333333                     1.0   0.517857   \n",
       "350         0.073529   0.054253   0.000000                     1.0   0.071429   \n",
       "\n",
       "     Exsisting_CR_Count  Liability_Count  Telephone_Status  Foreign_Worker  \\\n",
       "842                 0.0              0.0               0.0             0.0   \n",
       "68                  0.0              0.0               1.0             0.0   \n",
       "308                 0.0              0.0               0.0             0.0   \n",
       "881                 0.0              0.0               1.0             0.0   \n",
       "350                 0.0              0.0               1.0             0.0   \n",
       "\n",
       "     Existing_CA_Status_A11  Existing_CA_Status_A12  Existing_CA_Status_A13  \\\n",
       "842                     0.0                     0.0                     0.0   \n",
       "68                      0.0                     0.0                     0.0   \n",
       "308                     0.0                     1.0                     0.0   \n",
       "881                     0.0                     0.0                     0.0   \n",
       "350                     0.0                     0.0                     0.0   \n",
       "\n",
       "     Existing_CA_Status_A14  Credit_History_A30  Credit_History_A31  \\\n",
       "842                     1.0                 0.0                 0.0   \n",
       "68                      1.0                 0.0                 0.0   \n",
       "308                     0.0                 0.0                 0.0   \n",
       "881                     1.0                 0.0                 0.0   \n",
       "350                     1.0                 0.0                 0.0   \n",
       "\n",
       "     Credit_History_A32  Credit_History_A33  Credit_History_A34  Purpose_A40  \\\n",
       "842                 1.0                 0.0                 0.0          0.0   \n",
       "68                  1.0                 0.0                 0.0          0.0   \n",
       "308                 1.0                 0.0                 0.0          0.0   \n",
       "881                 1.0                 0.0                 0.0          0.0   \n",
       "350                 1.0                 0.0                 0.0          0.0   \n",
       "\n",
       "     Purpose_A41  Purpose_A410  Purpose_A42  Purpose_A43  Purpose_A44  \\\n",
       "842          0.0           0.0          0.0          0.0          0.0   \n",
       "68           0.0           0.0          0.0          0.0          0.0   \n",
       "308          0.0           0.0          1.0          0.0          0.0   \n",
       "881          1.0           0.0          0.0          0.0          0.0   \n",
       "350          0.0           0.0          0.0          0.0          1.0   \n",
       "\n",
       "     Purpose_A45  Purpose_A46  Purpose_A48  Purpose_A49  Saving_Acc_A61  \\\n",
       "842          1.0          0.0          0.0          0.0             1.0   \n",
       "68           0.0          1.0          0.0          0.0             1.0   \n",
       "308          0.0          0.0          0.0          0.0             1.0   \n",
       "881          0.0          0.0          0.0          0.0             0.0   \n",
       "350          0.0          0.0          0.0          0.0             1.0   \n",
       "\n",
       "     Saving_Acc_A62  Saving_Acc_A63  Saving_Acc_A64  Saving_Acc_A65  \\\n",
       "842             0.0             0.0             0.0             0.0   \n",
       "68              0.0             0.0             0.0             0.0   \n",
       "308             0.0             0.0             0.0             0.0   \n",
       "881             0.0             0.0             0.0             1.0   \n",
       "350             0.0             0.0             0.0             0.0   \n",
       "\n",
       "     Present_Employment_A71  Present_Employment_A72  Present_Employment_A73  \\\n",
       "842                     0.0                     1.0                     0.0   \n",
       "68                      0.0                     0.0                     1.0   \n",
       "308                     0.0                     0.0                     1.0   \n",
       "881                     0.0                     0.0                     1.0   \n",
       "350                     0.0                     1.0                     0.0   \n",
       "\n",
       "     Present_Employment_A74  Present_Employment_A75  Gender_Status_A91  \\\n",
       "842                     0.0                     0.0                0.0   \n",
       "68                      0.0                     0.0                0.0   \n",
       "308                     0.0                     0.0                0.0   \n",
       "881                     0.0                     0.0                1.0   \n",
       "350                     0.0                     0.0                0.0   \n",
       "\n",
       "     Gender_Status_A92  Gender_Status_A93  Gender_Status_A94  \\\n",
       "842                1.0                0.0                0.0   \n",
       "68                 0.0                1.0                0.0   \n",
       "308                1.0                0.0                0.0   \n",
       "881                0.0                0.0                0.0   \n",
       "350                1.0                0.0                0.0   \n",
       "\n",
       "     Other_Debtors_A101  Other_Debtors_A102  Other_Debtors_A103  \\\n",
       "842                 1.0                 0.0                 0.0   \n",
       "68                  1.0                 0.0                 0.0   \n",
       "308                 1.0                 0.0                 0.0   \n",
       "881                 1.0                 0.0                 0.0   \n",
       "350                 1.0                 0.0                 0.0   \n",
       "\n",
       "     Property_A121  Property_A122  Property_A123  Property_A124  \\\n",
       "842            1.0            0.0            0.0            0.0   \n",
       "68             0.0            0.0            0.0            1.0   \n",
       "308            1.0            0.0            0.0            0.0   \n",
       "881            0.0            0.0            0.0            1.0   \n",
       "350            1.0            0.0            0.0            0.0   \n",
       "\n",
       "     Inst_Plans_A141  Inst_Plans_A142  Inst_Plans_A143  Housing_A151  \\\n",
       "842              0.0              0.0              1.0           0.0   \n",
       "68               0.0              1.0              0.0           0.0   \n",
       "308              0.0              0.0              1.0           0.0   \n",
       "881              0.0              0.0              1.0           0.0   \n",
       "350              0.0              0.0              1.0           1.0   \n",
       "\n",
       "     Housing_A152  Housing_A153  Job_A171  Job_A172  Job_A173  Job_A174  \n",
       "842           1.0           0.0       0.0       0.0       1.0       0.0  \n",
       "68            0.0           1.0       0.0       0.0       1.0       0.0  \n",
       "308           1.0           0.0       0.0       0.0       1.0       0.0  \n",
       "881           0.0           1.0       0.0       0.0       1.0       0.0  \n",
       "350           0.0           0.0       0.0       0.0       1.0       0.0  "
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(data, columns=scaledData.columns).sample(5, random_state = 999)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Feature Selection and Ranking <a class=\"anchor\" id=\"5\"></a>\n",
    "\n",
    "\n",
    "\n",
    "Most important features are selected by using `Random Forest Feature Importance(RFI)` in the full data set. Random Forests are amongst the most popular machine learning methods known for thier relatively good accuracy, robustness and ease of use.\n",
    "This is for the quick ranking of the most relevant 10 features to gain some insights of the data. During the hyper-parameter tuning phase, RFI will be included as a part of the pipelining process and we will search over 10, 20 and full set of 59 features to determine which set of features work best with each classifier.\n",
    "\n",
    "We use `np.argsort()` method to sort the list of the features in the descending order of their RFI scores and then we choose first 10 features as the highly important 10 features for the modelling purpose.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['CR_Amount', 'Age_Years', 'Duration_Months',\n",
       "       'Existing_CA_Status_A14', 'Inst_Rate', 'Existing_CA_Status_A11',\n",
       "       'Present_Residence_Time', 'Saving_Acc_A61', 'Credit_History_A34',\n",
       "       'Purpose_A40'], dtype=object)"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "num_features = 10\n",
    "\n",
    "model_rfi = RandomForestClassifier(n_estimators=100, random_state= 999)\n",
    "model_rfi.fit(data, target)\n",
    "fs_indices_rfi = np.argsort(model_rfi.feature_importances_)[::-1][0:num_features]\n",
    "\n",
    "best_features_rfi = scaledData.columns[fs_indices_rfi].values\n",
    "best_features_rfi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.10278397, 0.08216811, 0.08195924, 0.05667467, 0.03575125,\n",
       "       0.03390708, 0.03344957, 0.020316  , 0.01960295, 0.01928142])"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feature_importances_rfi = model_rfi.feature_importances_[fs_indices_rfi]\n",
    "feature_importances_rfi"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the following chunks the important features are plotted with respect to thier importance using Altair Graphs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "%config InlineBackend.figure_format = 'retina'\n",
    "\n",
    "import altair as alt\n",
    "alt.renderers.enable('html') # required in Jupyter notebooks\n",
    "\n",
    "def plot_imp(best_features, scores, method_name, color):\n",
    "    \n",
    "    df = pd.DataFrame({'features': best_features, \n",
    "                       'importances': scores})\n",
    "    \n",
    "    chart = alt.Chart(df, \n",
    "                      width=500, \n",
    "                      title=method_name + ' Feature Importances'\n",
    "                     ).mark_bar(opacity=0.85, \n",
    "                                color=color).encode(\n",
    "        alt.X('features', title='Feature', sort=None, axis=alt.AxisConfig(labelAngle=45)),\n",
    "        alt.Y('importances', title='Importance')\n",
    "    )\n",
    "    \n",
    "    return chart"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "<div id=\"altair-viz-73377d3eff9449dba79cdb6a86a75b24\"></div>\n",
       "<script type=\"text/javascript\">\n",
       "  (function(spec, embedOpt){\n",
       "    let outputDiv = document.currentScript.previousElementSibling;\n",
       "    if (outputDiv.id !== \"altair-viz-73377d3eff9449dba79cdb6a86a75b24\") {\n",
       "      outputDiv = document.getElementById(\"altair-viz-73377d3eff9449dba79cdb6a86a75b24\");\n",
       "    }\n",
       "    const paths = {\n",
       "      \"vega\": \"https://cdn.jsdelivr.net/npm//vega@5?noext\",\n",
       "      \"vega-lib\": \"https://cdn.jsdelivr.net/npm//vega-lib?noext\",\n",
       "      \"vega-lite\": \"https://cdn.jsdelivr.net/npm//vega-lite@4.8.1?noext\",\n",
       "      \"vega-embed\": \"https://cdn.jsdelivr.net/npm//vega-embed@6?noext\",\n",
       "    };\n",
       "\n",
       "    function loadScript(lib) {\n",
       "      return new Promise(function(resolve, reject) {\n",
       "        var s = document.createElement('script');\n",
       "        s.src = paths[lib];\n",
       "        s.async = true;\n",
       "        s.onload = () => resolve(paths[lib]);\n",
       "        s.onerror = () => reject(`Error loading script: ${paths[lib]}`);\n",
       "        document.getElementsByTagName(\"head\")[0].appendChild(s);\n",
       "      });\n",
       "    }\n",
       "\n",
       "    function showError(err) {\n",
       "      outputDiv.innerHTML = `<div class=\"error\" style=\"color:red;\">${err}</div>`;\n",
       "      throw err;\n",
       "    }\n",
       "\n",
       "    function displayChart(vegaEmbed) {\n",
       "      vegaEmbed(outputDiv, spec, embedOpt)\n",
       "        .catch(err => showError(`Javascript Error: ${err.message}<br>This usually means there's a typo in your chart specification. See the javascript console for the full traceback.`));\n",
       "    }\n",
       "\n",
       "    if(typeof define === \"function\" && define.amd) {\n",
       "      requirejs.config({paths});\n",
       "      require([\"vega-embed\"], displayChart, err => showError(`Error loading script: ${err.message}`));\n",
       "    } else if (typeof vegaEmbed === \"function\") {\n",
       "      displayChart(vegaEmbed);\n",
       "    } else {\n",
       "      loadScript(\"vega\")\n",
       "        .then(() => loadScript(\"vega-lite\"))\n",
       "        .then(() => loadScript(\"vega-embed\"))\n",
       "        .catch(showError)\n",
       "        .then(() => displayChart(vegaEmbed));\n",
       "    }\n",
       "  })({\"config\": {\"view\": {\"continuousWidth\": 400, \"continuousHeight\": 300}}, \"data\": {\"name\": \"data-0cc4bf0dfc73750b2c1cea4b7f660241\"}, \"mark\": {\"type\": \"bar\", \"color\": \"Blue\", \"opacity\": 0.85}, \"encoding\": {\"x\": {\"type\": \"nominal\", \"axis\": {\"labelAngle\": 45}, \"field\": \"features\", \"sort\": null, \"title\": \"Feature\"}, \"y\": {\"type\": \"quantitative\", \"field\": \"importances\", \"title\": \"Importance\"}}, \"title\": \"Random Forest Feature Importances\", \"width\": 500, \"$schema\": \"https://vega.github.io/schema/vega-lite/v4.8.1.json\", \"datasets\": {\"data-0cc4bf0dfc73750b2c1cea4b7f660241\": [{\"features\": \"CR_Amount\", \"importances\": 0.10278396944469094}, {\"features\": \"Age_Years\", \"importances\": 0.08216810717910855}, {\"features\": \"Duration_Months\", \"importances\": 0.08195924103505021}, {\"features\": \"Existing_CA_Status_A14\", \"importances\": 0.0566746708597807}, {\"features\": \"Inst_Rate\", \"importances\": 0.03575124897913127}, {\"features\": \"Existing_CA_Status_A11\", \"importances\": 0.03390707588464699}, {\"features\": \"Present_Residence_Time\", \"importances\": 0.03344957444796026}, {\"features\": \"Saving_Acc_A61\", \"importances\": 0.02031599521857861}, {\"features\": \"Credit_History_A34\", \"importances\": 0.01960295013055181}, {\"features\": \"Purpose_A40\", \"importances\": 0.019281419549935175}]}}, {\"mode\": \"vega-lite\"});\n",
       "</script>"
      ],
      "text/plain": [
       "alt.Chart(...)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot_imp(best_features_rfi, feature_importances_rfi, 'Random Forest', 'Blue')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here,  we can check out from the above graph that, `CR_Amount` is the most important feature along with `Duration-months` and `Age_Years` while considering top 10 RFI features. We include this in the below pipeline execution to check the Model performance for top 10, 20 and all the features(59) in the encoded dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Data Modelling <a class=\"anchor\" id=\"6\"></a>\n",
    "\n",
    "\n",
    "## Train-Test Splitting of the Dataset.\n",
    "\n",
    "\n",
    "Before starting the deifferent modelling process using different algorithms, first we have to split the dataset into `Train` and `Test` datasets. We are going to train our model  and Tune it only on the Training dataset and then we are going to predict the performance of the model on the Test dataset.\n",
    "\n",
    "For the simplification of the process, as we have comparitively less amount of data, we are going to perfrom 70:30 splitting in which 70% of the total data set is used for the Training and HP tuning while we are going to predict the Test dataset to check the performance of the model on unseen data.\n",
    "\n",
    "To achieve this, we are using `train_test_split()` function from the `sklearn.model_selection` package."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(700, 59)\n",
      "(300, 59)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "Data_sample_train, Data_sample_test,target_sample_train, target_sample_test = train_test_split(data, target, \n",
    "                                                    test_size = 0.3, random_state=999,\n",
    "                                                    stratify = target)\n",
    "\n",
    "print(Data_sample_train.shape)\n",
    "print(Data_sample_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model Evaluation Strategy\n",
    "\n",
    "From above step, we can see that we are going to train and tune our classification models on the 700 observation instances and will test the performance of each of the model on remianing 300 unseen instances.\n",
    "\n",
    "For each model, we are going to use Stratified K-Fold Cross-Validation evaluation method with **5 folds** with **2 repititions** for the Hyper Parameter Tuning.\n",
    "\n",
    "we are using `StratifiedKFold()` method for this purpose."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import StratifiedKFold, GridSearchCV\n",
    "\n",
    "cv_method = StratifiedKFold(n_splits=5,random_state=999)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Hyperparameter Tuning <a class=\"anchor\" id=\"7\"></a>\n",
    "\n",
    "\n",
    "\n",
    "# 1. K-Nearest Neighbours (KNN)\n",
    "\n",
    "- Using Pipeline, we stack `feature selection` and `grid search` for KNN hyperparameter tuning via cross-validation. We will use the same Pipeline methodology for NB and DT and Random Forest Ensemble.\n",
    "\n",
    "\n",
    "- KNN is a distance based classification algorithm technique.  we are using Euclidean and Manhattan distances to calculate the performance of the algorithm on our data.\n",
    "\n",
    "\n",
    "- For KNN hyperparameters are: \n",
    "\n",
    "\n",
    "    - Number of Neighbours(`n-neighbors`)i.e. value of K and\n",
    "    - The distance metric `p` (1= Manhattan and 2 = Euclidean).\n",
    "    \n",
    "    \n",
    "- For feature selection, we are using  Random Forest Importance (RFI) method with 100 estimators.we define the custom `RFIFeatureSelector()` class below to pass in RFI as a \"step\" to the pipeline."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.base import BaseEstimator, TransformerMixin\n",
    "\n",
    "# custom function for RFI feature selection inside a pipeline\n",
    "# here we use n_estimators=100\n",
    "class RFIFeatureSelector(BaseEstimator, TransformerMixin):\n",
    "    \n",
    "    # class constructor \n",
    "    # make sure class attributes end with a \"_\"\n",
    "    # per scikit-learn convention to avoid errors\n",
    "    def __init__(self, n_features_=10):\n",
    "        self.n_features_ = n_features_\n",
    "        self.fs_indices_ = None\n",
    "\n",
    "    # override the fit function\n",
    "    def fit(self, X, y):\n",
    "        from sklearn.ensemble import RandomForestClassifier\n",
    "        from numpy import argsort\n",
    "        model_rfi = RandomForestClassifier(n_estimators=100, random_state =999)\n",
    "        model_rfi.fit(X, y)\n",
    "        self.fs_indices_ = argsort(model_rfi.feature_importances_)[::-1][0:self.n_features_] \n",
    "        return self \n",
    "    \n",
    "    # override the transform function\n",
    "    def transform(self, X, y=None):\n",
    "        return X[:, self.fs_indices_]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following code chunk represents a `pipeline()` function which is responsible for following tasks:\n",
    "\n",
    "- The pipeline function implements KNN algorithm on the data returned by RFI FeatureSelector.\n",
    "- RFIFeatureSelector computes RFI scores for each feature in the dataset and returns specified number of TOP RFI scored features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "\n",
    "pipe_KNN = Pipeline(steps=[('rfi_fs', RFIFeatureSelector()), \n",
    "                           ('knn', KNeighborsClassifier())])\n",
    "\n",
    "params_pipe_KNN = {'rfi_fs__n_features_': [10, 20, data.shape[1]],\n",
    "                   'knn__n_neighbors': [10, 15, 20, 25, 30],\n",
    "                   'knn__p': [1, 2]}\n",
    "\n",
    "gs_pipe_KNN = GridSearchCV(estimator=pipe_KNN, \n",
    "                           param_grid=params_pipe_KNN, \n",
    "                           cv=cv_method,\n",
    "                           refit=True,\n",
    "                           n_jobs=-2,\n",
    "                           scoring='roc_auc',\n",
    "                           verbose=1) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 30 candidates, totalling 150 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-2)]: Using backend LokyBackend with 3 concurrent workers.\n",
      "[Parallel(n_jobs=-2)]: Done  44 tasks      | elapsed:    9.2s\n",
      "[Parallel(n_jobs=-2)]: Done 150 out of 150 | elapsed:   25.8s finished\n"
     ]
    }
   ],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "gs_pipe_KNN = gs_pipe_KNN.fit(Data_sample_train, target_sample_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'knn__n_neighbors': 20, 'knn__p': 1, 'rfi_fs__n_features_': 59}"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gs_pipe_KNN.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7348153547133137"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gs_pipe_KNN.best_score_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here,  we observe that, Optimal KNN model has mean AUC score of 0.7348. The best performing KNN selected 20 nearest neighbours and all the 59 features with value of p = 1 i.e. Manhattan distance.\n",
    "\n",
    "Even though, we get the optimal score and conditions for a optimal KNN model, we have to see the difference between performance of KNN with different combinations of the hyperparameters. we can see the difference between their `mean_score` values. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "# custom function to format the search results as a Pandas data frame\n",
    "def get_search_results(gs):\n",
    "\n",
    "    def model_result(scores, params):\n",
    "        scores = {'mean_score': np.mean(scores),\n",
    "             'std_score': np.std(scores),\n",
    "             'min_score': np.min(scores),\n",
    "             'max_score': np.max(scores)}\n",
    "        return pd.Series({**params,**scores})\n",
    "\n",
    "    models = []\n",
    "    scores = []\n",
    "\n",
    "    for i in range(gs.n_splits_):\n",
    "        key = f\"split{i}_test_score\"\n",
    "        r = gs.cv_results_[key]        \n",
    "        scores.append(r.reshape(-1,1))\n",
    "\n",
    "    all_scores = np.hstack(scores)\n",
    "    for p, s in zip(gs.cv_results_['params'], all_scores):\n",
    "        models.append((model_result(s, p)))\n",
    "\n",
    "    pipe_results = pd.concat(models, axis=1).T.sort_values(['mean_score'], ascending=False)\n",
    "\n",
    "    columns_first = ['mean_score', 'std_score', 'max_score', 'min_score']\n",
    "    columns = columns_first + [c for c in pipe_results.columns if c not in columns_first]\n",
    "\n",
    "    return pipe_results[columns]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean_score</th>\n",
       "      <th>std_score</th>\n",
       "      <th>max_score</th>\n",
       "      <th>min_score</th>\n",
       "      <th>knn__n_neighbors</th>\n",
       "      <th>knn__p</th>\n",
       "      <th>rfi_fs__n_features_</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.734815</td>\n",
       "      <td>0.025957</td>\n",
       "      <td>0.779883</td>\n",
       "      <td>0.708819</td>\n",
       "      <td>20.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>59.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.733503</td>\n",
       "      <td>0.018448</td>\n",
       "      <td>0.766885</td>\n",
       "      <td>0.715500</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>59.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>0.730224</td>\n",
       "      <td>0.033539</td>\n",
       "      <td>0.775389</td>\n",
       "      <td>0.684038</td>\n",
       "      <td>30.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>20.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.729106</td>\n",
       "      <td>0.029015</td>\n",
       "      <td>0.776968</td>\n",
       "      <td>0.690355</td>\n",
       "      <td>20.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>59.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>0.728110</td>\n",
       "      <td>0.037205</td>\n",
       "      <td>0.777211</td>\n",
       "      <td>0.672133</td>\n",
       "      <td>30.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>20.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    mean_score  std_score  max_score  min_score  knn__n_neighbors  knn__p  \\\n",
       "14    0.734815   0.025957   0.779883   0.708819              20.0     1.0   \n",
       "8     0.733503   0.018448   0.766885   0.715500              15.0     1.0   \n",
       "25    0.730224   0.033539   0.775389   0.684038              30.0     1.0   \n",
       "17    0.729106   0.029015   0.776968   0.690355              20.0     2.0   \n",
       "28    0.728110   0.037205   0.777211   0.672133              30.0     2.0   \n",
       "\n",
       "    rfi_fs__n_features_  \n",
       "14                 59.0  \n",
       "8                  59.0  \n",
       "25                 20.0  \n",
       "17                 59.0  \n",
       "28                 20.0  "
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_KNN = get_search_results(gs_pipe_KNN)\n",
    "results_KNN.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- A minimal difference of 0.001 can be observed in the `mean_score` between 1st optimal and 2nd optimal models of KNN hyperparameter values.\n",
    "\n",
    "- It can also be observed that most of the top 5 mean scores are generated when calculating scores with all the descriptive features i.e all the 59 features taken into consideration.\n",
    "\n",
    "- Hence, we have plotted a graph for results of a Grid Search corresponding to all the 59 features that got selected."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "<div id=\"altair-viz-a3705cd6c16b44fea7ee091d8840edae\"></div>\n",
       "<script type=\"text/javascript\">\n",
       "  (function(spec, embedOpt){\n",
       "    let outputDiv = document.currentScript.previousElementSibling;\n",
       "    if (outputDiv.id !== \"altair-viz-a3705cd6c16b44fea7ee091d8840edae\") {\n",
       "      outputDiv = document.getElementById(\"altair-viz-a3705cd6c16b44fea7ee091d8840edae\");\n",
       "    }\n",
       "    const paths = {\n",
       "      \"vega\": \"https://cdn.jsdelivr.net/npm//vega@5?noext\",\n",
       "      \"vega-lib\": \"https://cdn.jsdelivr.net/npm//vega-lib?noext\",\n",
       "      \"vega-lite\": \"https://cdn.jsdelivr.net/npm//vega-lite@4.8.1?noext\",\n",
       "      \"vega-embed\": \"https://cdn.jsdelivr.net/npm//vega-embed@6?noext\",\n",
       "    };\n",
       "\n",
       "    function loadScript(lib) {\n",
       "      return new Promise(function(resolve, reject) {\n",
       "        var s = document.createElement('script');\n",
       "        s.src = paths[lib];\n",
       "        s.async = true;\n",
       "        s.onload = () => resolve(paths[lib]);\n",
       "        s.onerror = () => reject(`Error loading script: ${paths[lib]}`);\n",
       "        document.getElementsByTagName(\"head\")[0].appendChild(s);\n",
       "      });\n",
       "    }\n",
       "\n",
       "    function showError(err) {\n",
       "      outputDiv.innerHTML = `<div class=\"error\" style=\"color:red;\">${err}</div>`;\n",
       "      throw err;\n",
       "    }\n",
       "\n",
       "    function displayChart(vegaEmbed) {\n",
       "      vegaEmbed(outputDiv, spec, embedOpt)\n",
       "        .catch(err => showError(`Javascript Error: ${err.message}<br>This usually means there's a typo in your chart specification. See the javascript console for the full traceback.`));\n",
       "    }\n",
       "\n",
       "    if(typeof define === \"function\" && define.amd) {\n",
       "      requirejs.config({paths});\n",
       "      require([\"vega-embed\"], displayChart, err => showError(`Error loading script: ${err.message}`));\n",
       "    } else if (typeof vegaEmbed === \"function\") {\n",
       "      displayChart(vegaEmbed);\n",
       "    } else {\n",
       "      loadScript(\"vega\")\n",
       "        .then(() => loadScript(\"vega-lite\"))\n",
       "        .then(() => loadScript(\"vega-embed\"))\n",
       "        .catch(showError)\n",
       "        .then(() => displayChart(vegaEmbed));\n",
       "    }\n",
       "  })({\"config\": {\"view\": {\"continuousWidth\": 400, \"continuousHeight\": 300}}, \"data\": {\"name\": \"data-010640ade6cbf7376e2027a5f1f56039\"}, \"mark\": {\"type\": \"line\", \"point\": true}, \"encoding\": {\"color\": {\"type\": \"nominal\", \"field\": \"knn__p\", \"title\": \"Values of Distance Metric P\"}, \"x\": {\"type\": \"quantitative\", \"field\": \"knn__n_neighbors\", \"title\": \"Number of Neighbors\"}, \"y\": {\"type\": \"quantitative\", \"field\": \"mean_score\", \"scale\": {\"zero\": false}, \"title\": \"AUC Score\"}}, \"title\": \"KNN Performance Comparison with 59 features\", \"$schema\": \"https://vega.github.io/schema/vega-lite/v4.8.1.json\", \"datasets\": {\"data-010640ade6cbf7376e2027a5f1f56039\": [{\"mean_score\": 0.7348153547133137, \"std_score\": 0.025957256411678557, \"max_score\": 0.779883381924198, \"min_score\": 0.7088192419825072, \"knn__n_neighbors\": 20.0, \"knn__p\": 1.0, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7335034013605441, \"std_score\": 0.018448441920860023, \"max_score\": 0.7668853255587949, \"min_score\": 0.7155004859086491, \"knn__n_neighbors\": 15.0, \"knn__p\": 1.0, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7291059280855199, \"std_score\": 0.02901489078083252, \"max_score\": 0.7769679300291544, \"min_score\": 0.690354713313897, \"knn__n_neighbors\": 20.0, \"knn__p\": 2.0, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7267978620019435, \"std_score\": 0.02159192915730419, \"max_score\": 0.7642128279883381, \"min_score\": 0.7056608357628765, \"knn__n_neighbors\": 15.0, \"knn__p\": 2.0, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7206511175898932, \"std_score\": 0.03123551795961527, \"max_score\": 0.7752672497570456, \"min_score\": 0.6870748299319728, \"knn__n_neighbors\": 25.0, \"knn__p\": 2.0, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7196064139941691, \"std_score\": 0.03354066945265218, \"max_score\": 0.777089407191448, \"min_score\": 0.6793002915451894, \"knn__n_neighbors\": 30.0, \"knn__p\": 1.0, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7192176870748299, \"std_score\": 0.028229954354761097, \"max_score\": 0.7670068027210885, \"min_score\": 0.68488824101069, \"knn__n_neighbors\": 25.0, \"knn__p\": 1.0, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7173712342079689, \"std_score\": 0.029850535410072716, \"max_score\": 0.7684645286686103, \"min_score\": 0.6863459669582119, \"knn__n_neighbors\": 30.0, \"knn__p\": 2.0, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7051263362487853, \"std_score\": 0.02721170865057192, \"max_score\": 0.7419825072886298, \"min_score\": 0.6720116618075802, \"knn__n_neighbors\": 10.0, \"knn__p\": 1.0, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.6993197278911566, \"std_score\": 0.02189440425544276, \"max_score\": 0.7322643343051507, \"min_score\": 0.6763848396501458, \"knn__n_neighbors\": 10.0, \"knn__p\": 2.0, \"rfi_fs__n_features_\": 59.0}]}}, {\"mode\": \"vega-lite\"});\n",
       "</script>"
      ],
      "text/plain": [
       "alt.Chart(...)"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_KNN_59_Features = results_KNN[results_KNN['rfi_fs__n_features_'] == 59.0]\n",
    "\n",
    "alt.Chart(results_KNN_59_Features, \n",
    "          title='KNN Performance Comparison with 59 features'\n",
    "         ).mark_line(point=True).encode(\n",
    "    alt.X('knn__n_neighbors', title='Number of Neighbors'),\n",
    "    alt.Y('mean_score', title='AUC Score', scale=alt.Scale(zero=False)),\n",
    "    alt.Color('knn__p:N', title='Values of Distance Metric P')\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see from the above graph that, AUC_ROC scores are not increasing after the number of neighbors 20.  Hence,  we can say that, `K =20`  is a optimal hyperparameter value which gives us optimal performance of the model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. (Gaussian) Naive Bayes (NB) \n",
    "\n",
    "\n",
    "The Next model is Gaussian Naive Bayes (NB). It is a probabilty based classification algorithm. We are using `var_smoothing` (variant of Laplace Smoothing) as a hyperparameter to tune our model. By default, the `var_smoothing` parameter's value is $10^{-9}$ . We conduct the grid search in the `logspace` (over the powers of 10) sourced from `NumPy`. We will start with $10^{0}$  and end with  $10^{-9}$  and we will try 100 different values.\n",
    "\n",
    "As Gaussian Naive Bayes considers probability is noramlly distributed, Before fitting our data into the Naive Bayes model we need to normalise the probability of each value. For this purpose we are going to perform pwoer transforamtion of Training data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import PowerTransformer\n",
    "Data_sample_train_transformed = PowerTransformer().fit_transform(Data_sample_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.naive_bayes import GaussianNB\n",
    "\n",
    "np.random.seed(999)\n",
    "\n",
    "pipe_NB = Pipeline([('rfi_fs', RFIFeatureSelector()), \n",
    "                     ('nb', GaussianNB())])\n",
    "\n",
    "params_pipe_NB = {'rfi_fs__n_features_': [10, 20, data.shape[1]],\n",
    "                  'nb__var_smoothing': np.logspace(0,-9, num=100)}\n",
    "\n",
    "\n",
    "gs_pipe_NB = GridSearchCV(estimator=pipe_NB, \n",
    "                          param_grid=params_pipe_NB, \n",
    "                          cv=cv_method,\n",
    "                          refit=True,\n",
    "                          n_jobs=-2,\n",
    "                          scoring='roc_auc',\n",
    "                          verbose=1) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 300 candidates, totalling 1500 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-2)]: Using backend LokyBackend with 3 concurrent workers.\n",
      "[Parallel(n_jobs=-2)]: Done  44 tasks      | elapsed:    6.5s\n",
      "[Parallel(n_jobs=-2)]: Done 194 tasks      | elapsed:   26.8s\n",
      "[Parallel(n_jobs=-2)]: Done 444 tasks      | elapsed:   58.7s\n",
      "[Parallel(n_jobs=-2)]: Done 794 tasks      | elapsed:  1.7min\n",
      "[Parallel(n_jobs=-2)]: Done 1244 tasks      | elapsed:  2.6min\n",
      "[Parallel(n_jobs=-2)]: Done 1500 out of 1500 | elapsed:  3.1min finished\n"
     ]
    }
   ],
   "source": [
    "gs_pipe_NB.fit(Data_sample_train_transformed, target_sample_train);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'nb__var_smoothing': 0.8111308307896871, 'rfi_fs__n_features_': 20}"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gs_pipe_NB.best_params_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see from the above that hyper parameters for optimal NB performance are `var_smoothing` of 0.8111 and with top `20` features selected by RFI method. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7605928085519923"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gs_pipe_NB.best_score_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The performance score of the optimal NB model is 0.7605 which higher than optimal KNN score which is calculated above. That means the model has approximately `76.1% ROC_AUC` score.\n",
    "\n",
    "We try to visualize other results in the grid search by ploting a chart of var_smoothing factor against AUC Score for every number of features we selected (10, 20 and 59). We can visualize other scores with the help of below graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean_score</th>\n",
       "      <th>std_score</th>\n",
       "      <th>max_score</th>\n",
       "      <th>min_score</th>\n",
       "      <th>nb__var_smoothing</th>\n",
       "      <th>rfi_fs__n_features_</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.760593</td>\n",
       "      <td>0.029212</td>\n",
       "      <td>0.793975</td>\n",
       "      <td>0.714529</td>\n",
       "      <td>0.811131</td>\n",
       "      <td>20.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.760398</td>\n",
       "      <td>0.028542</td>\n",
       "      <td>0.791545</td>\n",
       "      <td>0.714529</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>20.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.760398</td>\n",
       "      <td>0.029864</td>\n",
       "      <td>0.797619</td>\n",
       "      <td>0.714286</td>\n",
       "      <td>0.533670</td>\n",
       "      <td>20.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.760350</td>\n",
       "      <td>0.029574</td>\n",
       "      <td>0.795190</td>\n",
       "      <td>0.714043</td>\n",
       "      <td>0.657933</td>\n",
       "      <td>20.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.760253</td>\n",
       "      <td>0.030370</td>\n",
       "      <td>0.801506</td>\n",
       "      <td>0.714286</td>\n",
       "      <td>0.284804</td>\n",
       "      <td>20.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    mean_score  std_score  max_score  min_score  nb__var_smoothing  \\\n",
       "4     0.760593   0.029212   0.793975   0.714529           0.811131   \n",
       "1     0.760398   0.028542   0.791545   0.714529           1.000000   \n",
       "10    0.760398   0.029864   0.797619   0.714286           0.533670   \n",
       "7     0.760350   0.029574   0.795190   0.714043           0.657933   \n",
       "19    0.760253   0.030370   0.801506   0.714286           0.284804   \n",
       "\n",
       "    rfi_fs__n_features_  \n",
       "4                  20.0  \n",
       "1                  20.0  \n",
       "10                 20.0  \n",
       "7                  20.0  \n",
       "19                 20.0  "
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_NB = get_search_results(gs_pipe_NB)\n",
    "results_NB.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "<div id=\"altair-viz-2b1fb273f1db4714ab390c884fbbc36a\"></div>\n",
       "<script type=\"text/javascript\">\n",
       "  (function(spec, embedOpt){\n",
       "    let outputDiv = document.currentScript.previousElementSibling;\n",
       "    if (outputDiv.id !== \"altair-viz-2b1fb273f1db4714ab390c884fbbc36a\") {\n",
       "      outputDiv = document.getElementById(\"altair-viz-2b1fb273f1db4714ab390c884fbbc36a\");\n",
       "    }\n",
       "    const paths = {\n",
       "      \"vega\": \"https://cdn.jsdelivr.net/npm//vega@5?noext\",\n",
       "      \"vega-lib\": \"https://cdn.jsdelivr.net/npm//vega-lib?noext\",\n",
       "      \"vega-lite\": \"https://cdn.jsdelivr.net/npm//vega-lite@4.8.1?noext\",\n",
       "      \"vega-embed\": \"https://cdn.jsdelivr.net/npm//vega-embed@6?noext\",\n",
       "    };\n",
       "\n",
       "    function loadScript(lib) {\n",
       "      return new Promise(function(resolve, reject) {\n",
       "        var s = document.createElement('script');\n",
       "        s.src = paths[lib];\n",
       "        s.async = true;\n",
       "        s.onload = () => resolve(paths[lib]);\n",
       "        s.onerror = () => reject(`Error loading script: ${paths[lib]}`);\n",
       "        document.getElementsByTagName(\"head\")[0].appendChild(s);\n",
       "      });\n",
       "    }\n",
       "\n",
       "    function showError(err) {\n",
       "      outputDiv.innerHTML = `<div class=\"error\" style=\"color:red;\">${err}</div>`;\n",
       "      throw err;\n",
       "    }\n",
       "\n",
       "    function displayChart(vegaEmbed) {\n",
       "      vegaEmbed(outputDiv, spec, embedOpt)\n",
       "        .catch(err => showError(`Javascript Error: ${err.message}<br>This usually means there's a typo in your chart specification. See the javascript console for the full traceback.`));\n",
       "    }\n",
       "\n",
       "    if(typeof define === \"function\" && define.amd) {\n",
       "      requirejs.config({paths});\n",
       "      require([\"vega-embed\"], displayChart, err => showError(`Error loading script: ${err.message}`));\n",
       "    } else if (typeof vegaEmbed === \"function\") {\n",
       "      displayChart(vegaEmbed);\n",
       "    } else {\n",
       "      loadScript(\"vega\")\n",
       "        .then(() => loadScript(\"vega-lite\"))\n",
       "        .then(() => loadScript(\"vega-embed\"))\n",
       "        .catch(showError)\n",
       "        .then(() => displayChart(vegaEmbed));\n",
       "    }\n",
       "  })({\"config\": {\"view\": {\"continuousWidth\": 400, \"continuousHeight\": 300}}, \"data\": {\"name\": \"data-3016f4373954cb4e98c673eb4ed6bcea\"}, \"mark\": {\"type\": \"line\", \"point\": true}, \"encoding\": {\"color\": {\"type\": \"nominal\", \"field\": \"rfi_fs__n_features_\", \"title\": \"Number of Features selected\"}, \"x\": {\"type\": \"quantitative\", \"field\": \"nb__var_smoothing\", \"title\": \"Var. Smoothing\"}, \"y\": {\"type\": \"quantitative\", \"field\": \"mean_score\", \"scale\": {\"zero\": false}, \"title\": \"AUC Score\"}}, \"title\": \"NB Performance Comparison with All the features\", \"$schema\": \"https://vega.github.io/schema/vega-lite/v4.8.1.json\", \"datasets\": {\"data-3016f4373954cb4e98c673eb4ed6bcea\": [{\"mean_score\": 0.7605928085519923, \"std_score\": 0.02921220402712172, \"max_score\": 0.793974732750243, \"min_score\": 0.7145286686103013, \"nb__var_smoothing\": 0.8111308307896871, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7603984450923227, \"std_score\": 0.0285424907831263, \"max_score\": 0.7915451895043731, \"min_score\": 0.7145286686103013, \"nb__var_smoothing\": 1.0, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7603984450923227, \"std_score\": 0.02986437408555057, \"max_score\": 0.7976190476190477, \"min_score\": 0.7142857142857143, \"nb__var_smoothing\": 0.533669923120631, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7603498542274052, \"std_score\": 0.02957367940420961, \"max_score\": 0.7951895043731777, \"min_score\": 0.7140427599611272, \"nb__var_smoothing\": 0.657933224657568, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7602526724975703, \"std_score\": 0.03036964042558449, \"max_score\": 0.8015063168124392, \"min_score\": 0.7142857142857143, \"nb__var_smoothing\": 0.2848035868435802, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7601554907677356, \"std_score\": 0.02977965992272885, \"max_score\": 0.7978620019436345, \"min_score\": 0.7140427599611273, \"nb__var_smoothing\": 0.43287612810830584, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7599611273080662, \"std_score\": 0.03008633979364029, \"max_score\": 0.8000485908649174, \"min_score\": 0.7145286686103013, \"nb__var_smoothing\": 0.3511191734215131, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7595724003887269, \"std_score\": 0.030467905971230707, \"max_score\": 0.8012633624878522, \"min_score\": 0.7137998056365403, \"nb__var_smoothing\": 0.23101297000831597, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7594266277939747, \"std_score\": 0.031216142181999692, \"max_score\": 0.802964042759961, \"min_score\": 0.7133138969873664, \"nb__var_smoothing\": 0.15199110829529336, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7594266277939747, \"std_score\": 0.031196848974352754, \"max_score\": 0.802721088435374, \"min_score\": 0.7133138969873664, \"nb__var_smoothing\": 0.1873817422860384, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7591350826044703, \"std_score\": 0.03160521083079549, \"max_score\": 0.8039358600583091, \"min_score\": 0.7130709426627794, \"nb__var_smoothing\": 0.1, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7590379008746355, \"std_score\": 0.03157935226898841, \"max_score\": 0.8034499514091351, \"min_score\": 0.7125850340136055, \"nb__var_smoothing\": 0.12328467394420659, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.758843537414966, \"std_score\": 0.03161716139611723, \"max_score\": 0.8032069970845481, \"min_score\": 0.7123420796890185, \"nb__var_smoothing\": 0.08111308307896872, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7587949465500486, \"std_score\": 0.03182438286123348, \"max_score\": 0.8034499514091351, \"min_score\": 0.7118561710398446, \"nb__var_smoothing\": 0.0657933224657568, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7587463556851313, \"std_score\": 0.03209867078727207, \"max_score\": 0.804178814382896, \"min_score\": 0.7113702623906705, \"nb__var_smoothing\": 0.04328761281083057, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7587463556851312, \"std_score\": 0.03156349782796281, \"max_score\": 0.8029640427599611, \"min_score\": 0.7120991253644315, \"nb__var_smoothing\": 0.0533669923120631, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7586491739552963, \"std_score\": 0.03219327240553352, \"max_score\": 0.8041788143828961, \"min_score\": 0.7113702623906705, \"nb__var_smoothing\": 0.03511191734215131, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7585034013605442, \"std_score\": 0.032419976554845145, \"max_score\": 0.80466472303207, \"min_score\": 0.7108843537414965, \"nb__var_smoothing\": 0.02848035868435802, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7584062196307094, \"std_score\": 0.03240926911554168, \"max_score\": 0.80466472303207, \"min_score\": 0.7108843537414965, \"nb__var_smoothing\": 0.02310129700083159, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7584062196307094, \"std_score\": 0.032596684753441574, \"max_score\": 0.8051506316812439, \"min_score\": 0.7108843537414966, \"nb__var_smoothing\": 0.006579332246575682, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.758357628765792, \"std_score\": 0.03242966119171644, \"max_score\": 0.80466472303207, \"min_score\": 0.7108843537414965, \"nb__var_smoothing\": 0.01873817422860384, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7583090379008747, \"std_score\": 0.03280635554252133, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 0.0004328761281083057, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7583090379008747, \"std_score\": 0.032665712644667305, \"max_score\": 0.8053935860058309, \"min_score\": 0.7108843537414966, \"nb__var_smoothing\": 0.002848035868435802, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7583090379008747, \"std_score\": 0.032574512747811854, \"max_score\": 0.8051506316812439, \"min_score\": 0.7108843537414966, \"nb__var_smoothing\": 0.005336699231206307, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7583090379008747, \"std_score\": 0.032665712644667305, \"max_score\": 0.8053935860058309, \"min_score\": 0.7108843537414966, \"nb__var_smoothing\": 0.003511191734215131, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7583090379008747, \"std_score\": 0.03280635554252133, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 0.0001873817422860383, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7583090379008747, \"std_score\": 0.03280635554252133, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 0.0003511191734215131, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7583090379008747, \"std_score\": 0.03280635554252133, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 0.0001, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7583090379008747, \"std_score\": 0.03280635554252133, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 0.0002848035868435802, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7583090379008747, \"std_score\": 0.03280635554252133, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 0.0001232846739442066, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7583090379008747, \"std_score\": 0.03280635554252133, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 0.0002310129700083158, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7583090379008747, \"std_score\": 0.03280635554252133, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 0.0001519911082952933, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7583090379008747, \"std_score\": 0.03280635554252133, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 8.111308307896872e-05, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7583090379008747, \"std_score\": 0.032574512747811854, \"max_score\": 0.8051506316812439, \"min_score\": 0.7108843537414966, \"nb__var_smoothing\": 0.004328761281083057, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1.873817422860383e-05, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.03273632599599515, \"max_score\": 0.8053935860058309, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 0.0006579332246575676, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 2.848035868435799e-06, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 5.336699231206302e-08, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 2.310129700083158e-06, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 5.336699231206313e-07, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 4.3287612810830526e-08, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1.873817422860383e-06, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1.519911082952933e-06, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 3.5111917342151277e-08, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.03273632599599515, \"max_score\": 0.8053935860058309, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 0.0005336699231206307, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 2.848035868435799e-08, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 2.310129700083158e-05, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.03273632599599515, \"max_score\": 0.8053935860058309, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 0.0008111308307896872, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 2.310129700083158e-08, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.03273632599599515, \"max_score\": 0.8053935860058309, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 0.0012328467394420659, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 6.579332246575682e-07, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1.873817422860383e-08, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.03273632599599515, \"max_score\": 0.8053935860058309, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 0.0023101297000831605, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1.232846739442066e-06, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1.519911082952933e-08, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1e-06, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 4.3287612810830526e-07, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 3.5111917342151275e-06, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 6.579332246575682e-08, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 4.328761281083053e-06, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1.519911082952933e-07, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1.873817422860383e-07, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 2.848035868435799e-05, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1.5199110829529332e-05, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1.2328467394420658e-05, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 3.511191734215127e-05, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1.232846739442066e-07, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 2.310129700083158e-07, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 4.328761281083062e-05, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 5.3366992312063123e-05, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1e-05, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1e-07, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 6.579332246575683e-05, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 8.111308307896873e-06, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 2.848035868435799e-07, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 8.111308307896873e-08, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 6.579332246575683e-06, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 5.336699231206313e-06, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 3.5111917342151277e-07, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.03273632599599515, \"max_score\": 0.8053935860058309, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 0.001, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 8.111308307896872e-07, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1.519911082952933e-09, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1e-08, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 6.579332246575682e-09, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.032528671745944954, \"max_score\": 0.80466472303207, \"min_score\": 0.7106413994169095, \"nb__var_smoothing\": 0.012328467394420659, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 5.336699231206302e-09, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 3.5111917342151273e-09, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 8.111308307896856e-09, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.032669181892277466, \"max_score\": 0.804907677356657, \"min_score\": 0.7103984450923225, \"nb__var_smoothing\": 0.008111308307896872, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 2.310129700083158e-09, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 4.328761281083061e-09, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1e-09, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1.2328467394420635e-09, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1.873817422860387e-09, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 2.848035868435805e-09, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582604470359573, \"std_score\": 0.0328277955852317, \"max_score\": 0.805636540330418, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 1.232846739442066e-08, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582118561710398, \"std_score\": 0.032542242186745785, \"max_score\": 0.80466472303207, \"min_score\": 0.7106413994169095, \"nb__var_smoothing\": 0.01519911082952933, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582118561710398, \"std_score\": 0.03259987164491228, \"max_score\": 0.80466472303207, \"min_score\": 0.7103984450923225, \"nb__var_smoothing\": 0.01, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582118561710398, \"std_score\": 0.03270074957203267, \"max_score\": 0.8053935860058309, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 0.0015199110829529332, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.7582118561710398, \"std_score\": 0.03270074957203267, \"max_score\": 0.8053935860058309, \"min_score\": 0.7106413994169096, \"nb__var_smoothing\": 0.001873817422860383, \"rfi_fs__n_features_\": 20.0}, {\"mean_score\": 0.745821185617104, \"std_score\": 0.05643163228133805, \"max_score\": 0.8367346938775511, \"min_score\": 0.6588921282798834, \"nb__var_smoothing\": 1.0, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7448007774538388, \"std_score\": 0.05718684469470027, \"max_score\": 0.8364917395529641, \"min_score\": 0.6562196307094266, \"nb__var_smoothing\": 0.8111308307896871, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7448007774538385, \"std_score\": 0.03050546728474383, \"max_score\": 0.7728377065111759, \"min_score\": 0.6868318756073857, \"nb__var_smoothing\": 0.1873817422860384, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7448007774538385, \"std_score\": 0.030072838797785675, \"max_score\": 0.7742954324586977, \"min_score\": 0.6880466472303207, \"nb__var_smoothing\": 0.3511191734215131, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7447035957240038, \"std_score\": 0.030272303830328865, \"max_score\": 0.7735665694849369, \"min_score\": 0.6873177842565598, \"nb__var_smoothing\": 0.2848035868435802, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7447035957240038, \"std_score\": 0.030162913936922102, \"max_score\": 0.773080660835763, \"min_score\": 0.6878036929057336, \"nb__var_smoothing\": 0.23101297000831597, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7445578231292517, \"std_score\": 0.029824105486493974, \"max_score\": 0.7747813411078717, \"min_score\": 0.6885325558794947, \"nb__var_smoothing\": 0.43287612810830584, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7444606413994168, \"std_score\": 0.030923571545755892, \"max_score\": 0.7728377065111759, \"min_score\": 0.6856171039844509, \"nb__var_smoothing\": 0.15199110829529336, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443634596695821, \"std_score\": 0.031377210054782265, \"max_score\": 0.772594752186589, \"min_score\": 0.6846452866861029, \"nb__var_smoothing\": 0.12328467394420659, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443634596695821, \"std_score\": 0.030061373892235996, \"max_score\": 0.7755102040816326, \"min_score\": 0.6880466472303206, \"nb__var_smoothing\": 0.533669923120631, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443634596695821, \"std_score\": 0.03105271826661202, \"max_score\": 0.7704081632653061, \"min_score\": 0.6856171039844509, \"nb__var_smoothing\": 0.004328761281083057, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1.519911082952933e-07, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 2.310129700083158e-05, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1.2328467394420658e-05, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1.873817422860383e-05, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1.873817422860387e-09, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 2.848035868435805e-09, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1.232846739442066e-07, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1.5199110829529332e-05, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1e-05, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 2.310129700083158e-07, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 2.310129700083158e-09, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1.873817422860383e-07, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 5.3366992312063123e-05, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 8.111308307896873e-06, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1.519911082952933e-09, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1e-06, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 6.579332246575682e-07, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1.232846739442066e-06, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1.519911082952933e-06, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 5.336699231206313e-07, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1.873817422860383e-06, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1e-09, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 2.310129700083158e-06, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 4.3287612810830526e-07, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 2.848035868435799e-06, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 3.5111917342151275e-06, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 3.5111917342151277e-07, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1.2328467394420635e-09, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 4.328761281083053e-06, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 5.336699231206313e-06, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 2.848035868435799e-07, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 6.579332246575683e-06, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 3.511191734215127e-05, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 2.848035868435799e-05, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 4.328761281083062e-05, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.0008111308307896872, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 4.3287612810830526e-08, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.0002848035868435802, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 3.5111917342151277e-08, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 6.579332246575682e-09, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.0004328761281083057, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.0005336699231206307, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 2.848035868435799e-08, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.0006579332246575676, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 2.310129700083158e-08, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.0001873817422860383, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 8.111308307896872e-07, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.001, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 8.111308307896856e-09, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.0012328467394420659, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1.873817422860383e-08, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1.519911082952933e-08, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1e-08, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1.232846739442066e-08, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.0002310129700083158, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.0003511191734215131, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 5.336699231206302e-09, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 6.579332246575682e-08, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 1e-07, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 3.5111917342151273e-09, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 6.579332246575683e-05, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 8.111308307896872e-05, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 4.328761281083061e-09, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.0001, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 8.111308307896873e-08, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.0001232846739442066, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 5.336699231206302e-08, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7443148688046647, \"std_score\": 0.031105593133440513, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.0001519911082952933, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7442662779397473, \"std_score\": 0.031104074995038908, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.003511191734215131, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7442662779397472, \"std_score\": 0.03111128549266575, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.0023101297000831605, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7442662779397472, \"std_score\": 0.03111128549266575, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.001873817422860383, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7442662779397472, \"std_score\": 0.03111128549266575, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.0015199110829529332, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7442176870748299, \"std_score\": 0.03107893903484223, \"max_score\": 0.7701652089407192, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.002848035868435802, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7441690962099126, \"std_score\": 0.031087142730019092, \"max_score\": 0.7704081632653061, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.005336699231206307, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7441690962099126, \"std_score\": 0.031087142730019092, \"max_score\": 0.7704081632653061, \"min_score\": 0.6853741496598639, \"nb__var_smoothing\": 0.006579332246575682, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7441690962099125, \"std_score\": 0.031204643353068183, \"max_score\": 0.7704081632653061, \"min_score\": 0.6851311953352769, \"nb__var_smoothing\": 0.008111308307896872, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7441205053449951, \"std_score\": 0.031163909465198723, \"max_score\": 0.7701652089407192, \"min_score\": 0.6851311953352769, \"nb__var_smoothing\": 0.01, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7441205053449951, \"std_score\": 0.03130262589557235, \"max_score\": 0.7704081632653061, \"min_score\": 0.6848882410106899, \"nb__var_smoothing\": 0.02310129700083159, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7441205053449951, \"std_score\": 0.031126763443132827, \"max_score\": 0.7701652089407192, \"min_score\": 0.6851311953352769, \"nb__var_smoothing\": 0.012328467394420659, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7440719144800778, \"std_score\": 0.031132527770999743, \"max_score\": 0.7701652089407192, \"min_score\": 0.6851311953352769, \"nb__var_smoothing\": 0.01873817422860384, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7440719144800778, \"std_score\": 0.031132527770999743, \"max_score\": 0.7701652089407192, \"min_score\": 0.6851311953352769, \"nb__var_smoothing\": 0.01519911082952933, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7440719144800776, \"std_score\": 0.031421274439235196, \"max_score\": 0.771622934888241, \"min_score\": 0.6841593780369291, \"nb__var_smoothing\": 0.1, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7438775510204082, \"std_score\": 0.03143547315787897, \"max_score\": 0.7704081632653061, \"min_score\": 0.684402332361516, \"nb__var_smoothing\": 0.02848035868435802, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7438289601554907, \"std_score\": 0.03149932566682738, \"max_score\": 0.771622934888241, \"min_score\": 0.6839164237123421, \"nb__var_smoothing\": 0.08111308307896872, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7438289601554907, \"std_score\": 0.030225158461692405, \"max_score\": 0.7769679300291545, \"min_score\": 0.6878036929057335, \"nb__var_smoothing\": 0.8111308307896871, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7437803692905733, \"std_score\": 0.03156372223880976, \"max_score\": 0.771622934888241, \"min_score\": 0.6839164237123421, \"nb__var_smoothing\": 0.0657933224657568, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7436831875607386, \"std_score\": 0.031480356016970655, \"max_score\": 0.7708940719144801, \"min_score\": 0.6841593780369291, \"nb__var_smoothing\": 0.03511191734215131, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7436345966958211, \"std_score\": 0.030126492993102997, \"max_score\": 0.7755102040816326, \"min_score\": 0.6873177842565598, \"nb__var_smoothing\": 0.657933224657568, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7435374149659865, \"std_score\": 0.031637690883383175, \"max_score\": 0.7711370262390671, \"min_score\": 0.6836734693877551, \"nb__var_smoothing\": 0.04328761281083057, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.743488824101069, \"std_score\": 0.03171885716072786, \"max_score\": 0.771379980563654, \"min_score\": 0.6834305150631681, \"nb__var_smoothing\": 0.0533669923120631, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7434402332361516, \"std_score\": 0.03036676374327604, \"max_score\": 0.7776967930029155, \"min_score\": 0.6878036929057336, \"nb__var_smoothing\": 1.0, \"rfi_fs__n_features_\": 10.0}, {\"mean_score\": 0.7424198250728864, \"std_score\": 0.05844622669098871, \"max_score\": 0.8352769679300291, \"min_score\": 0.6511175898931001, \"nb__var_smoothing\": 0.657933224657568, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7406705539358601, \"std_score\": 0.05912284378653896, \"max_score\": 0.8352769679300291, \"min_score\": 0.6491739552964042, \"nb__var_smoothing\": 0.533669923120631, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7375121477162292, \"std_score\": 0.060384507327003366, \"max_score\": 0.8330903790087463, \"min_score\": 0.6431000971817298, \"nb__var_smoothing\": 0.43287612810830584, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7358600583090379, \"std_score\": 0.06081751916143204, \"max_score\": 0.8321185617103983, \"min_score\": 0.64067055393586, \"nb__var_smoothing\": 0.3511191734215131, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7339650145772596, \"std_score\": 0.0615504534814407, \"max_score\": 0.8309037900874636, \"min_score\": 0.6372691933916423, \"nb__var_smoothing\": 0.2848035868435802, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7331875607385812, \"std_score\": 0.06151158258074783, \"max_score\": 0.8294460641399417, \"min_score\": 0.6360544217687075, \"nb__var_smoothing\": 0.23101297000831597, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7321185617103984, \"std_score\": 0.06154174515735914, \"max_score\": 0.8282312925170068, \"min_score\": 0.6348396501457726, \"nb__var_smoothing\": 0.1873817422860384, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7310981535471331, \"std_score\": 0.06191883411155677, \"max_score\": 0.8284742468415938, \"min_score\": 0.6338678328474247, \"nb__var_smoothing\": 0.15199110829529336, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7307094266277939, \"std_score\": 0.06187042602576051, \"max_score\": 0.8277453838678328, \"min_score\": 0.6333819241982507, \"nb__var_smoothing\": 0.12328467394420659, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7295432458697765, \"std_score\": 0.06202988850859664, \"max_score\": 0.8267735665694849, \"min_score\": 0.6319241982507288, \"nb__var_smoothing\": 0.1, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7289601554907678, \"std_score\": 0.06189274651385736, \"max_score\": 0.8258017492711369, \"min_score\": 0.6314382896015549, \"nb__var_smoothing\": 0.08111308307896872, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7286443148688047, \"std_score\": 0.06249890327236223, \"max_score\": 0.8267735665694849, \"min_score\": 0.630709426627794, \"nb__var_smoothing\": 0.0657933224657568, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7283041788143828, \"std_score\": 0.06251038667387501, \"max_score\": 0.8265306122448979, \"min_score\": 0.630466472303207, \"nb__var_smoothing\": 0.0533669923120631, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7282069970845481, \"std_score\": 0.06254767429425635, \"max_score\": 0.826287657920311, \"min_score\": 0.6299805636540331, \"nb__var_smoothing\": 0.04328761281083057, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7276239067055393, \"std_score\": 0.06261317115541308, \"max_score\": 0.8258017492711369, \"min_score\": 0.629251700680272, \"nb__var_smoothing\": 0.03511191734215131, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7274781341107872, \"std_score\": 0.06280853770742678, \"max_score\": 0.825072886297376, \"min_score\": 0.6280369290573371, \"nb__var_smoothing\": 0.01873817422860384, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7273566569484936, \"std_score\": 0.06289324959291878, \"max_score\": 0.825072886297376, \"min_score\": 0.6277939747327501, \"nb__var_smoothing\": 0.01519911082952933, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7273566569484936, \"std_score\": 0.06298272111086498, \"max_score\": 0.825072886297376, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.01, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.727332361516035, \"std_score\": 0.06274554021786163, \"max_score\": 0.825315840621963, \"min_score\": 0.6285228377065111, \"nb__var_smoothing\": 0.02310129700083159, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.727332361516035, \"std_score\": 0.06249943215956755, \"max_score\": 0.825315840621963, \"min_score\": 0.629251700680272, \"nb__var_smoothing\": 0.02848035868435802, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7273080660835762, \"std_score\": 0.06297094889881084, \"max_score\": 0.825072886297376, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.012328467394420659, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7273080660835762, \"std_score\": 0.06298744437355479, \"max_score\": 0.825072886297376, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.008111308307896872, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7272108843537415, \"std_score\": 0.06282808228797701, \"max_score\": 0.824829931972789, \"min_score\": 0.6277939747327501, \"nb__var_smoothing\": 0.006579332246575682, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.727162293488824, \"std_score\": 0.06282481275284563, \"max_score\": 0.824586977648202, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.004328761281083057, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.727162293488824, \"std_score\": 0.06282481275284563, \"max_score\": 0.824586977648202, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.005336699231206307, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7271137026239067, \"std_score\": 0.06290995309626078, \"max_score\": 0.8248299319727892, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.002848035868435802, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7271137026239067, \"std_score\": 0.06299096784679935, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.001, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.727089407191448, \"std_score\": 0.06291267403342492, \"max_score\": 0.8248299319727892, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.0023101297000831605, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270651117589895, \"std_score\": 0.06299606528904991, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.0002848035868435802, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270651117589895, \"std_score\": 0.06299606528904991, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.0004328761281083057, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270651117589895, \"std_score\": 0.06299606528904991, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.0003511191734215131, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270651117589895, \"std_score\": 0.06299606528904991, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.0002310129700083158, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270651117589892, \"std_score\": 0.0629154323806372, \"max_score\": 0.8248299319727892, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.0012328467394420659, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270651117589892, \"std_score\": 0.06283450811983236, \"max_score\": 0.824586977648202, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.003511191734215131, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270651117589892, \"std_score\": 0.0629154323806372, \"max_score\": 0.8248299319727892, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.001873817422860383, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1.519911082952933e-08, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1.873817422860383e-08, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 2.310129700083158e-08, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1e-06, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1.232846739442066e-08, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1e-08, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 8.111308307896856e-09, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 6.579332246575682e-09, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 4.328761281083061e-09, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 3.5111917342151273e-09, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 2.848035868435805e-09, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 2.310129700083158e-09, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1.873817422860387e-09, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1.519911082952933e-09, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1.2328467394420635e-09, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 5.336699231206302e-09, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 2.848035868435799e-08, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 8.111308307896872e-07, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1.873817422860383e-07, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1.5199110829529332e-05, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 2.310129700083158e-07, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1.2328467394420658e-05, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1e-05, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 2.848035868435799e-07, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 8.111308307896873e-06, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 6.579332246575683e-06, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 3.5111917342151277e-07, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 5.336699231206313e-06, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 4.328761281083053e-06, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 4.3287612810830526e-07, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 3.5111917342151275e-06, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 2.848035868435799e-06, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 5.336699231206313e-07, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 2.310129700083158e-06, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1.873817422860383e-06, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 6.579332246575682e-07, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1.519911082952933e-06, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1.232846739442066e-06, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1.873817422860383e-05, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 2.310129700083158e-05, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 8.111308307896873e-08, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.0001, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 3.5111917342151277e-08, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 4.3287612810830526e-08, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 5.336699231206302e-08, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.0001873817422860383, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 6.579332246575682e-08, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.0001519911082952933, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.0001232846739442066, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 2.848035868435799e-05, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1e-09, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 8.111308307896872e-05, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1e-07, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 6.579332246575683e-05, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 5.3366992312063123e-05, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1.232846739442066e-07, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 4.328761281083062e-05, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 3.511191734215127e-05, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.7270408163265307, \"std_score\": 0.0629986700727211, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 1.519911082952933e-07, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.727016520894072, \"std_score\": 0.06300131222526398, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.0005336699231206307, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.727016520894072, \"std_score\": 0.06300131222526398, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.0006579332246575676, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.727016520894072, \"std_score\": 0.06300131222526398, \"max_score\": 0.8250728862973762, \"min_score\": 0.6275510204081631, \"nb__var_smoothing\": 0.0008111308307896872, \"rfi_fs__n_features_\": 59.0}, {\"mean_score\": 0.727016520894072, \"std_score\": 0.0629923172098912, \"max_score\": 0.8248299319727892, \"min_score\": 0.6273080660835761, \"nb__var_smoothing\": 0.0015199110829529332, \"rfi_fs__n_features_\": 59.0}]}}, {\"mode\": \"vega-lite\"});\n",
       "</script>"
      ],
      "text/plain": [
       "alt.Chart(...)"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "alt.Chart(results_NB, \n",
    "          title='NB Performance Comparison with All the features'\n",
    "         ).mark_line(point=True).encode(\n",
    "    alt.X('nb__var_smoothing', title='Var. Smoothing'),\n",
    "    alt.Y('mean_score', title='AUC Score', scale=alt.Scale(zero=False)),\n",
    "    alt.Color('rfi_fs__n_features_:N', title='Number of Features selected')\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. Decision Tree(DT)\n",
    "\n",
    "- Decision Tree Learning is one of the predictive modelling approaches which uses Decision Trees to go from observations about an item to conclusions about an item's target value. \n",
    "\n",
    "- Decision Trees hyperparameters are as follows: \n",
    "    - `Max_Depth` : The maximum depth of the tree.\n",
    "    - `Split_criterion`: The tree can be split using 2 splits Gini Index and Entropy.\n",
    "    - `Min_Sample_Spilt`: The minimum number of samples required to split an internal node.\n",
    "\n",
    "- Feature selection is carried out using the Random Forest Importance(RFI) method with 100 estimators as defined in the `RFIFeatureSelector()` class mentioned above. \n",
    "----\n",
    "The following code chunk represents a `pipeline()` function which is responsible for following tasks:\n",
    "\n",
    "- The pipeline function implements Decision Tree algorithm on the data returned by RFI FeatureSelector.\n",
    "- RFIFeatureSelector computes RFI scores for each feature in the dataset and returns specified number of TOP RFI scored features.\n",
    "\n",
    "The `GridSearchCV()` function performs an exhaustive search over specified parameter values for an estimator."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 54 candidates, totalling 270 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-2)]: Using backend LokyBackend with 3 concurrent workers.\n",
      "[Parallel(n_jobs=-2)]: Done  44 tasks      | elapsed:    6.0s\n",
      "[Parallel(n_jobs=-2)]: Done 194 tasks      | elapsed:   27.6s\n",
      "[Parallel(n_jobs=-2)]: Done 270 out of 270 | elapsed:   36.6s finished\n"
     ]
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "\n",
    "pipe_DT = Pipeline([('rfi_fs', RFIFeatureSelector()),\n",
    "                    ('dt', DecisionTreeClassifier(random_state=999))])\n",
    "\n",
    "params_pipe_DT = {'rfi_fs__n_features_': [10, 20, data.shape[1]],\n",
    "                  'dt__max_depth': [3, 4, 5],\n",
    "                  'dt__criterion': ['gini', 'entropy'],\n",
    "                  'dt__min_samples_split': [2, 5, 10]}\n",
    "\n",
    "gs_pipe_DT = GridSearchCV(estimator=pipe_DT, \n",
    "                          param_grid=params_pipe_DT, \n",
    "                          cv=cv_method,\n",
    "                          refit=True,\n",
    "                          n_jobs=-2,\n",
    "                          scoring='roc_auc',\n",
    "                          verbose=1) \n",
    "\n",
    "gs_pipe_DT.fit(Data_sample_train, target_sample_train);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'dt__criterion': 'entropy',\n",
       " 'dt__max_depth': 5,\n",
       " 'dt__min_samples_split': 10,\n",
       " 'rfi_fs__n_features_': 20}"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gs_pipe_DT.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.703206997084548"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gs_pipe_DT.best_score_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- It can be observed that the optimal Decision Tree has mean AUC score of 0.703. The best performing DT selected `entropy criterion, max_depth 5, min_samples_split 10 and all the 20 features`.\n",
    "\n",
    "- Even though we have obtained the best score, we need to find the difference between the performance of DT with combination of different hyperparameters.\n",
    "\n",
    "- The difference can be observed in the `Mean_score` values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean_score</th>\n",
       "      <th>std_score</th>\n",
       "      <th>max_score</th>\n",
       "      <th>min_score</th>\n",
       "      <th>dt__criterion</th>\n",
       "      <th>dt__max_depth</th>\n",
       "      <th>dt__min_samples_split</th>\n",
       "      <th>rfi_fs__n_features_</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>0.703207</td>\n",
       "      <td>0.0386538</td>\n",
       "      <td>0.76725</td>\n",
       "      <td>0.648567</td>\n",
       "      <td>entropy</td>\n",
       "      <td>5</td>\n",
       "      <td>10</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>0.700267</td>\n",
       "      <td>0.0434598</td>\n",
       "      <td>0.760569</td>\n",
       "      <td>0.628887</td>\n",
       "      <td>entropy</td>\n",
       "      <td>5</td>\n",
       "      <td>10</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>0.695554</td>\n",
       "      <td>0.0512508</td>\n",
       "      <td>0.78535</td>\n",
       "      <td>0.63654</td>\n",
       "      <td>gini</td>\n",
       "      <td>5</td>\n",
       "      <td>10</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.694485</td>\n",
       "      <td>0.0345392</td>\n",
       "      <td>0.753158</td>\n",
       "      <td>0.648445</td>\n",
       "      <td>gini</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.694485</td>\n",
       "      <td>0.0345392</td>\n",
       "      <td>0.753158</td>\n",
       "      <td>0.648445</td>\n",
       "      <td>gini</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   mean_score  std_score max_score min_score dt__criterion dt__max_depth  \\\n",
       "52   0.703207  0.0386538   0.76725  0.648567       entropy             5   \n",
       "53   0.700267  0.0434598  0.760569  0.628887       entropy             5   \n",
       "26   0.695554  0.0512508   0.78535   0.63654          gini             5   \n",
       "0    0.694485  0.0345392  0.753158  0.648445          gini             3   \n",
       "3    0.694485  0.0345392  0.753158  0.648445          gini             3   \n",
       "\n",
       "   dt__min_samples_split rfi_fs__n_features_  \n",
       "52                    10                  20  \n",
       "53                    10                  59  \n",
       "26                    10                  59  \n",
       "0                      2                  10  \n",
       "3                      5                  10  "
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_DT = get_search_results(gs_pipe_DT).round(3)\n",
    "results_DT.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- A minimal difference of 0.003 can be observed between the 1st and the 2nd optimal mean score values of the DT hyperparameter values.\n",
    "\n",
    "- We plot a graph for results of the Grid search corresponding to **10**, **20** and  **59** features.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "<div id=\"altair-viz-f8a4378d46e74e8cb489c486f8d22a69\"></div>\n",
       "<script type=\"text/javascript\">\n",
       "  (function(spec, embedOpt){\n",
       "    let outputDiv = document.currentScript.previousElementSibling;\n",
       "    if (outputDiv.id !== \"altair-viz-f8a4378d46e74e8cb489c486f8d22a69\") {\n",
       "      outputDiv = document.getElementById(\"altair-viz-f8a4378d46e74e8cb489c486f8d22a69\");\n",
       "    }\n",
       "    const paths = {\n",
       "      \"vega\": \"https://cdn.jsdelivr.net/npm//vega@5?noext\",\n",
       "      \"vega-lib\": \"https://cdn.jsdelivr.net/npm//vega-lib?noext\",\n",
       "      \"vega-lite\": \"https://cdn.jsdelivr.net/npm//vega-lite@4.8.1?noext\",\n",
       "      \"vega-embed\": \"https://cdn.jsdelivr.net/npm//vega-embed@6?noext\",\n",
       "    };\n",
       "\n",
       "    function loadScript(lib) {\n",
       "      return new Promise(function(resolve, reject) {\n",
       "        var s = document.createElement('script');\n",
       "        s.src = paths[lib];\n",
       "        s.async = true;\n",
       "        s.onload = () => resolve(paths[lib]);\n",
       "        s.onerror = () => reject(`Error loading script: ${paths[lib]}`);\n",
       "        document.getElementsByTagName(\"head\")[0].appendChild(s);\n",
       "      });\n",
       "    }\n",
       "\n",
       "    function showError(err) {\n",
       "      outputDiv.innerHTML = `<div class=\"error\" style=\"color:red;\">${err}</div>`;\n",
       "      throw err;\n",
       "    }\n",
       "\n",
       "    function displayChart(vegaEmbed) {\n",
       "      vegaEmbed(outputDiv, spec, embedOpt)\n",
       "        .catch(err => showError(`Javascript Error: ${err.message}<br>This usually means there's a typo in your chart specification. See the javascript console for the full traceback.`));\n",
       "    }\n",
       "\n",
       "    if(typeof define === \"function\" && define.amd) {\n",
       "      requirejs.config({paths});\n",
       "      require([\"vega-embed\"], displayChart, err => showError(`Error loading script: ${err.message}`));\n",
       "    } else if (typeof vegaEmbed === \"function\") {\n",
       "      displayChart(vegaEmbed);\n",
       "    } else {\n",
       "      loadScript(\"vega\")\n",
       "        .then(() => loadScript(\"vega-lite\"))\n",
       "        .then(() => loadScript(\"vega-embed\"))\n",
       "        .catch(showError)\n",
       "        .then(() => displayChart(vegaEmbed));\n",
       "    }\n",
       "  })({\"config\": {\"view\": {\"continuousWidth\": 400, \"continuousHeight\": 300}}, \"data\": {\"name\": \"data-e0bc352eb45e3a097cea3d953c8fe98f\"}, \"mark\": {\"type\": \"line\", \"point\": true}, \"encoding\": {\"color\": {\"type\": \"nominal\", \"field\": \"dt__max_depth\", \"title\": \"Max Depth\"}, \"x\": {\"type\": \"quantitative\", \"field\": \"dt__min_samples_split\", \"title\": \"Min Samples for Split\"}, \"y\": {\"type\": \"quantitative\", \"field\": \"mean_score\", \"scale\": {\"zero\": false}, \"title\": \"AUC Score\"}}, \"title\": \"DT Performance Comparison with 10 Features\", \"$schema\": \"https://vega.github.io/schema/vega-lite/v4.8.1.json\", \"datasets\": {\"data-e0bc352eb45e3a097cea3d953c8fe98f\": [{\"mean_score\": 0.6944849368318755, \"std_score\": 0.034539185228335684, \"max_score\": 0.7531584062196306, \"min_score\": 0.6484450923226432, \"dt__criterion\": \"gini\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6944849368318755, \"std_score\": 0.034539185228335684, \"max_score\": 0.7531584062196306, \"min_score\": 0.6484450923226432, \"dt__criterion\": \"gini\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6934645286686103, \"std_score\": 0.03438740184319652, \"max_score\": 0.7531584062196306, \"min_score\": 0.6484450923226432, \"dt__criterion\": \"gini\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6921039844509232, \"std_score\": 0.028967410600410137, \"max_score\": 0.7274052478134111, \"min_score\": 0.640913508260447, \"dt__criterion\": \"gini\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6913508260447037, \"std_score\": 0.03202700172296867, \"max_score\": 0.7274052478134111, \"min_score\": 0.6327745383867833, \"dt__criterion\": \"gini\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6913508260447037, \"std_score\": 0.03202700172296867, \"max_score\": 0.7274052478134111, \"min_score\": 0.6327745383867833, \"dt__criterion\": \"gini\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6880952380952381, \"std_score\": 0.03013893197298464, \"max_score\": 0.7253401360544217, \"min_score\": 0.6333819241982507, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6880952380952381, \"std_score\": 0.03013893197298464, \"max_score\": 0.7253401360544217, \"min_score\": 0.6333819241982507, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6876822157434404, \"std_score\": 0.03348682738847577, \"max_score\": 0.7295918367346939, \"min_score\": 0.6270651117589893, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6827016520894073, \"std_score\": 0.039696089923226624, \"max_score\": 0.7608114674441205, \"min_score\": 0.6543974732750243, \"dt__criterion\": \"gini\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6817784256559766, \"std_score\": 0.04414070232080112, \"max_score\": 0.7632410106899903, \"min_score\": 0.6348396501457726, \"dt__criterion\": \"gini\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6810495626822158, \"std_score\": 0.038887311998016025, \"max_score\": 0.7350583090379008, \"min_score\": 0.6198979591836734, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6805636540330418, \"std_score\": 0.045479050856051764, \"max_score\": 0.7632410106899903, \"min_score\": 0.6287657920310982, \"dt__criterion\": \"gini\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6787657920310981, \"std_score\": 0.044626986264125944, \"max_score\": 0.7475704567541301, \"min_score\": 0.6092079689018465, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6787414965986395, \"std_score\": 0.03624326459913626, \"max_score\": 0.7292274052478135, \"min_score\": 0.6163751214771623, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6787414965986395, \"std_score\": 0.03624326459913626, \"max_score\": 0.7292274052478135, \"min_score\": 0.6163751214771623, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6787414965986395, \"std_score\": 0.03624326459913626, \"max_score\": 0.7292274052478135, \"min_score\": 0.6163751214771623, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 10}, {\"mean_score\": 0.6777696793002915, \"std_score\": 0.0461964471750899, \"max_score\": 0.7475704567541301, \"min_score\": 0.6042274052478134, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 10}]}}, {\"mode\": \"vega-lite\"});\n",
       "</script>"
      ],
      "text/plain": [
       "alt.Chart(...)"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_DT_10_features = results_DT[results_DT['rfi_fs__n_features_'] == 10.0]\n",
    "\n",
    "alt.Chart(results_DT_10_features, \n",
    "          title='DT Performance Comparison with 10 Features'\n",
    "         ).mark_line(point=True).encode(\n",
    "    alt.X('dt__min_samples_split', title='Min Samples for Split'),\n",
    "    alt.Y('mean_score', title='AUC Score', scale=alt.Scale(zero=False)),\n",
    "    alt.Color('dt__max_depth:N', title='Max Depth')\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "<div id=\"altair-viz-bc1d13d097e94809b670988811f2f7a5\"></div>\n",
       "<script type=\"text/javascript\">\n",
       "  (function(spec, embedOpt){\n",
       "    let outputDiv = document.currentScript.previousElementSibling;\n",
       "    if (outputDiv.id !== \"altair-viz-bc1d13d097e94809b670988811f2f7a5\") {\n",
       "      outputDiv = document.getElementById(\"altair-viz-bc1d13d097e94809b670988811f2f7a5\");\n",
       "    }\n",
       "    const paths = {\n",
       "      \"vega\": \"https://cdn.jsdelivr.net/npm//vega@5?noext\",\n",
       "      \"vega-lib\": \"https://cdn.jsdelivr.net/npm//vega-lib?noext\",\n",
       "      \"vega-lite\": \"https://cdn.jsdelivr.net/npm//vega-lite@4.8.1?noext\",\n",
       "      \"vega-embed\": \"https://cdn.jsdelivr.net/npm//vega-embed@6?noext\",\n",
       "    };\n",
       "\n",
       "    function loadScript(lib) {\n",
       "      return new Promise(function(resolve, reject) {\n",
       "        var s = document.createElement('script');\n",
       "        s.src = paths[lib];\n",
       "        s.async = true;\n",
       "        s.onload = () => resolve(paths[lib]);\n",
       "        s.onerror = () => reject(`Error loading script: ${paths[lib]}`);\n",
       "        document.getElementsByTagName(\"head\")[0].appendChild(s);\n",
       "      });\n",
       "    }\n",
       "\n",
       "    function showError(err) {\n",
       "      outputDiv.innerHTML = `<div class=\"error\" style=\"color:red;\">${err}</div>`;\n",
       "      throw err;\n",
       "    }\n",
       "\n",
       "    function displayChart(vegaEmbed) {\n",
       "      vegaEmbed(outputDiv, spec, embedOpt)\n",
       "        .catch(err => showError(`Javascript Error: ${err.message}<br>This usually means there's a typo in your chart specification. See the javascript console for the full traceback.`));\n",
       "    }\n",
       "\n",
       "    if(typeof define === \"function\" && define.amd) {\n",
       "      requirejs.config({paths});\n",
       "      require([\"vega-embed\"], displayChart, err => showError(`Error loading script: ${err.message}`));\n",
       "    } else if (typeof vegaEmbed === \"function\") {\n",
       "      displayChart(vegaEmbed);\n",
       "    } else {\n",
       "      loadScript(\"vega\")\n",
       "        .then(() => loadScript(\"vega-lite\"))\n",
       "        .then(() => loadScript(\"vega-embed\"))\n",
       "        .catch(showError)\n",
       "        .then(() => displayChart(vegaEmbed));\n",
       "    }\n",
       "  })({\"config\": {\"view\": {\"continuousWidth\": 400, \"continuousHeight\": 300}}, \"data\": {\"name\": \"data-75598fed29e25c49fa9ef712187c55cf\"}, \"mark\": {\"type\": \"line\", \"point\": true}, \"encoding\": {\"color\": {\"type\": \"nominal\", \"field\": \"dt__max_depth\", \"title\": \"Max Depth\"}, \"x\": {\"type\": \"quantitative\", \"field\": \"dt__min_samples_split\", \"title\": \"Min Samples for Split\"}, \"y\": {\"type\": \"quantitative\", \"field\": \"mean_score\", \"scale\": {\"zero\": false}, \"title\": \"AUC Score\"}}, \"title\": \"DT Performance Comparison with 20 Features\", \"$schema\": \"https://vega.github.io/schema/vega-lite/v4.8.1.json\", \"datasets\": {\"data-75598fed29e25c49fa9ef712187c55cf\": [{\"mean_score\": 0.703206997084548, \"std_score\": 0.03865382688892024, \"max_score\": 0.7672497570456753, \"min_score\": 0.6485665694849367, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6934645286686103, \"std_score\": 0.031175650522846515, \"max_score\": 0.7300777453838678, \"min_score\": 0.6376336248785228, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6927113702623907, \"std_score\": 0.03253135725565082, \"max_score\": 0.7300777453838678, \"min_score\": 0.6338678328474247, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6900145772594752, \"std_score\": 0.043483353891515863, \"max_score\": 0.7588678328474248, \"min_score\": 0.6406705539358601, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6900145772594752, \"std_score\": 0.043483353891515863, \"max_score\": 0.7588678328474248, \"min_score\": 0.6406705539358601, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6890427599611273, \"std_score\": 0.04461486895612519, \"max_score\": 0.7588678328474248, \"min_score\": 0.6358114674441206, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6810495626822156, \"std_score\": 0.023862592332952258, \"max_score\": 0.7164723032069971, \"min_score\": 0.6543974732750242, \"dt__criterion\": \"gini\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6755102040816326, \"std_score\": 0.03225783901390927, \"max_score\": 0.7096695821185617, \"min_score\": 0.6286443148688047, \"dt__criterion\": \"gini\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6755102040816326, \"std_score\": 0.03225783901390927, \"max_score\": 0.7096695821185617, \"min_score\": 0.6286443148688047, \"dt__criterion\": \"gini\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6755102040816326, \"std_score\": 0.03225783901390927, \"max_score\": 0.7096695821185617, \"min_score\": 0.6286443148688047, \"dt__criterion\": \"gini\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6747084548104956, \"std_score\": 0.03901126297625412, \"max_score\": 0.7201166180758017, \"min_score\": 0.6248785228377065, \"dt__criterion\": \"gini\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6741253644314869, \"std_score\": 0.04048297104636125, \"max_score\": 0.7255830903790087, \"min_score\": 0.6213556851311952, \"dt__criterion\": \"gini\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6735422740524781, \"std_score\": 0.03158907039608403, \"max_score\": 0.716958211856171, \"min_score\": 0.6353255587949466, \"dt__criterion\": \"gini\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6730806608357629, \"std_score\": 0.032813353874850355, \"max_score\": 0.716958211856171, \"min_score\": 0.6315597667638484, \"dt__criterion\": \"gini\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6730806608357629, \"std_score\": 0.032813353874850355, \"max_score\": 0.716958211856171, \"min_score\": 0.6315597667638484, \"dt__criterion\": \"gini\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6532069970845481, \"std_score\": 0.03460913134627281, \"max_score\": 0.6944849368318756, \"min_score\": 0.6163751214771623, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6532069970845481, \"std_score\": 0.03460913134627281, \"max_score\": 0.6944849368318756, \"min_score\": 0.6163751214771623, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 20}, {\"mean_score\": 0.6532069970845481, \"std_score\": 0.03460913134627281, \"max_score\": 0.6944849368318756, \"min_score\": 0.6163751214771623, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 20}]}}, {\"mode\": \"vega-lite\"});\n",
       "</script>"
      ],
      "text/plain": [
       "alt.Chart(...)"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_DT_20_features = results_DT[results_DT['rfi_fs__n_features_'] == 20.0]\n",
    "\n",
    "alt.Chart(results_DT_20_features, \n",
    "          title='DT Performance Comparison with 20 Features'\n",
    "         ).mark_line(point=True).encode(\n",
    "    alt.X('dt__min_samples_split', title='Min Samples for Split'),\n",
    "    alt.Y('mean_score', title='AUC Score', scale=alt.Scale(zero=False)),\n",
    "    alt.Color('dt__max_depth:N', title='Max Depth')\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "<div id=\"altair-viz-ed592b3e6aa949f39f4c2516032363ef\"></div>\n",
       "<script type=\"text/javascript\">\n",
       "  (function(spec, embedOpt){\n",
       "    let outputDiv = document.currentScript.previousElementSibling;\n",
       "    if (outputDiv.id !== \"altair-viz-ed592b3e6aa949f39f4c2516032363ef\") {\n",
       "      outputDiv = document.getElementById(\"altair-viz-ed592b3e6aa949f39f4c2516032363ef\");\n",
       "    }\n",
       "    const paths = {\n",
       "      \"vega\": \"https://cdn.jsdelivr.net/npm//vega@5?noext\",\n",
       "      \"vega-lib\": \"https://cdn.jsdelivr.net/npm//vega-lib?noext\",\n",
       "      \"vega-lite\": \"https://cdn.jsdelivr.net/npm//vega-lite@4.8.1?noext\",\n",
       "      \"vega-embed\": \"https://cdn.jsdelivr.net/npm//vega-embed@6?noext\",\n",
       "    };\n",
       "\n",
       "    function loadScript(lib) {\n",
       "      return new Promise(function(resolve, reject) {\n",
       "        var s = document.createElement('script');\n",
       "        s.src = paths[lib];\n",
       "        s.async = true;\n",
       "        s.onload = () => resolve(paths[lib]);\n",
       "        s.onerror = () => reject(`Error loading script: ${paths[lib]}`);\n",
       "        document.getElementsByTagName(\"head\")[0].appendChild(s);\n",
       "      });\n",
       "    }\n",
       "\n",
       "    function showError(err) {\n",
       "      outputDiv.innerHTML = `<div class=\"error\" style=\"color:red;\">${err}</div>`;\n",
       "      throw err;\n",
       "    }\n",
       "\n",
       "    function displayChart(vegaEmbed) {\n",
       "      vegaEmbed(outputDiv, spec, embedOpt)\n",
       "        .catch(err => showError(`Javascript Error: ${err.message}<br>This usually means there's a typo in your chart specification. See the javascript console for the full traceback.`));\n",
       "    }\n",
       "\n",
       "    if(typeof define === \"function\" && define.amd) {\n",
       "      requirejs.config({paths});\n",
       "      require([\"vega-embed\"], displayChart, err => showError(`Error loading script: ${err.message}`));\n",
       "    } else if (typeof vegaEmbed === \"function\") {\n",
       "      displayChart(vegaEmbed);\n",
       "    } else {\n",
       "      loadScript(\"vega\")\n",
       "        .then(() => loadScript(\"vega-lite\"))\n",
       "        .then(() => loadScript(\"vega-embed\"))\n",
       "        .catch(showError)\n",
       "        .then(() => displayChart(vegaEmbed));\n",
       "    }\n",
       "  })({\"config\": {\"view\": {\"continuousWidth\": 400, \"continuousHeight\": 300}}, \"data\": {\"name\": \"data-8a5029a1ccdc821cbee2a737c27861d1\"}, \"mark\": {\"type\": \"line\", \"point\": true}, \"encoding\": {\"color\": {\"type\": \"nominal\", \"field\": \"dt__max_depth\", \"title\": \"Max Depth\"}, \"x\": {\"type\": \"quantitative\", \"field\": \"dt__min_samples_split\", \"title\": \"Min Samples for Split\"}, \"y\": {\"type\": \"quantitative\", \"field\": \"mean_score\", \"scale\": {\"zero\": false}, \"title\": \"AUC Score\"}}, \"title\": \"DT Performance Comparison with 59 Features\", \"$schema\": \"https://vega.github.io/schema/vega-lite/v4.8.1.json\", \"datasets\": {\"data-8a5029a1ccdc821cbee2a737c27861d1\": [{\"mean_score\": 0.7002672497570457, \"std_score\": 0.04345980920457182, \"max_score\": 0.7605685131195336, \"min_score\": 0.6288872691933918, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.6955539358600582, \"std_score\": 0.05125083326453809, \"max_score\": 0.7853498542274053, \"min_score\": 0.6365403304178814, \"dt__criterion\": \"gini\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.6874149659863946, \"std_score\": 0.036169314579613546, \"max_score\": 0.7514577259475219, \"min_score\": 0.6465014577259476, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.6856171039844509, \"std_score\": 0.04503746747189591, \"max_score\": 0.7523080660835763, \"min_score\": 0.6194120505344995, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.6856171039844509, \"std_score\": 0.04503746747189591, \"max_score\": 0.7523080660835763, \"min_score\": 0.6194120505344995, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.682069970845481, \"std_score\": 0.031060491802303362, \"max_score\": 0.7378522837706512, \"min_score\": 0.6560981535471331, \"dt__criterion\": \"gini\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.6803692905733721, \"std_score\": 0.04030577829882875, \"max_score\": 0.7514577259475219, \"min_score\": 0.6438289601554907, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.6803692905733721, \"std_score\": 0.04030577829882875, \"max_score\": 0.7514577259475219, \"min_score\": 0.6438289601554907, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.676773566569485, \"std_score\": 0.04819943162642813, \"max_score\": 0.7580174927113703, \"min_score\": 0.6241496598639457, \"dt__criterion\": \"gini\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.6764820213799806, \"std_score\": 0.03275195512476434, \"max_score\": 0.7378522837706512, \"min_score\": 0.6494169096209912, \"dt__criterion\": \"gini\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.6764820213799806, \"std_score\": 0.03275195512476434, \"max_score\": 0.7378522837706512, \"min_score\": 0.6494169096209912, \"dt__criterion\": \"gini\", \"dt__max_depth\": 4, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.675971817298348, \"std_score\": 0.04877562057173126, \"max_score\": 0.7580174927113703, \"min_score\": 0.6241496598639457, \"dt__criterion\": \"gini\", \"dt__max_depth\": 5, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.6697764820213801, \"std_score\": 0.03484930598797571, \"max_score\": 0.7111273080660837, \"min_score\": 0.6117589893100097, \"dt__criterion\": \"gini\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.6697764820213801, \"std_score\": 0.03484930598797571, \"max_score\": 0.7111273080660837, \"min_score\": 0.6117589893100097, \"dt__criterion\": \"gini\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.6697764820213801, \"std_score\": 0.03484930598797571, \"max_score\": 0.7111273080660837, \"min_score\": 0.6117589893100097, \"dt__criterion\": \"gini\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.6452137998056365, \"std_score\": 0.02988176218480442, \"max_score\": 0.6801506316812439, \"min_score\": 0.6066569484936831, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 10, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.6452137998056365, \"std_score\": 0.02988176218480442, \"max_score\": 0.6801506316812439, \"min_score\": 0.6066569484936831, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 5, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.6452137998056365, \"std_score\": 0.02988176218480442, \"max_score\": 0.6801506316812439, \"min_score\": 0.6066569484936831, \"dt__criterion\": \"entropy\", \"dt__max_depth\": 3, \"dt__min_samples_split\": 2, \"rfi_fs__n_features_\": 59}]}}, {\"mode\": \"vega-lite\"});\n",
       "</script>"
      ],
      "text/plain": [
       "alt.Chart(...)"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_DT_59_features = results_DT[results_DT['rfi_fs__n_features_'] == 59.0]\n",
    "\n",
    "alt.Chart(results_DT_59_features, \n",
    "          title='DT Performance Comparison with 59 Features'\n",
    "         ).mark_line(point=True).encode(\n",
    "    alt.X('dt__min_samples_split', title='Min Samples for Split'),\n",
    "    alt.Y('mean_score', title='AUC Score', scale=alt.Scale(zero=False)),\n",
    "    alt.Color('dt__max_depth:N', title='Max Depth')\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Random Forest Classifier(RF)\n",
    "\n",
    "A random forest is a meta estimator that fits a number of decision tree classifiers on various sub-samples of the dataset and uses averaging to improve the predictive accuracy and control over-fitting.\n",
    "\n",
    "- Since a Random Forest ensembler uses collection of Decision trees to to esitmate the prediction it shares some similar features of the hyperparameters which are as follows: \n",
    "    - `Max_Depth` : The maximum depth of the tree.\n",
    "    - `Split_criterion`: The tree can be split using 2 splits Gini Index and Entropy.\n",
    "    - `n_estimators`: The number of trees in a forest. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 54 candidates, totalling 270 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=-2)]: Using backend LokyBackend with 3 concurrent workers.\n",
      "[Parallel(n_jobs=-2)]: Done  44 tasks      | elapsed:   20.3s\n",
      "[Parallel(n_jobs=-2)]: Done 194 tasks      | elapsed:  1.1min\n",
      "[Parallel(n_jobs=-2)]: Done 270 out of 270 | elapsed:  1.5min finished\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "pipe_RF = Pipeline([('rfi_fs', RFIFeatureSelector()),\n",
    "                    ('rf', RandomForestClassifier(random_state =999))])\n",
    "\n",
    "params_pipe_RF = {'rf__n_estimators': [120, 200, 250],\n",
    "                  'rf__max_depth': [4, 5, 6],\n",
    "                  'rf__criterion': ['gini', 'entropy'],\n",
    "                  'rfi_fs__n_features_': [10, 20, data.shape[1]]\n",
    "                  }\n",
    "\n",
    "gs_pipe_RF = GridSearchCV(estimator=pipe_RF, \n",
    "                          param_grid=params_pipe_RF, \n",
    "                          cv=cv_method,\n",
    "                          refit=True,\n",
    "                          n_jobs=-2,\n",
    "                          scoring='roc_auc',\n",
    "                          verbose=1) \n",
    "\n",
    "gs_pipe_RF.fit(Data_sample_train, target_sample_train);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7866375121477163"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gs_pipe_RF.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'rf__criterion': 'entropy',\n",
       " 'rf__max_depth': 5,\n",
       " 'rf__n_estimators': 250,\n",
       " 'rfi_fs__n_features_': 59}"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gs_pipe_RF.best_params_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- It can be observed that the optimal Random Forest Ensemble has mean AUC score of 0.7866. The best performing RF selected `entropy criterion, max_depth 5, n_estimators 250 and all the 59 features`.\n",
    "\n",
    "- Even though we have obtained the best score, we need to find the difference between the performance of RF with combination of different hyperparameters.\n",
    "\n",
    "- The difference can be observed in the `Mean_score` values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean_score</th>\n",
       "      <th>std_score</th>\n",
       "      <th>max_score</th>\n",
       "      <th>min_score</th>\n",
       "      <th>rf__criterion</th>\n",
       "      <th>rf__max_depth</th>\n",
       "      <th>rf__n_estimators</th>\n",
       "      <th>rfi_fs__n_features_</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>0.786638</td>\n",
       "      <td>0.0311885</td>\n",
       "      <td>0.825559</td>\n",
       "      <td>0.734208</td>\n",
       "      <td>entropy</td>\n",
       "      <td>5</td>\n",
       "      <td>250</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>0.786054</td>\n",
       "      <td>0.0324027</td>\n",
       "      <td>0.82483</td>\n",
       "      <td>0.730564</td>\n",
       "      <td>entropy</td>\n",
       "      <td>5</td>\n",
       "      <td>200</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>0.785374</td>\n",
       "      <td>0.0276988</td>\n",
       "      <td>0.824587</td>\n",
       "      <td>0.745384</td>\n",
       "      <td>entropy</td>\n",
       "      <td>6</td>\n",
       "      <td>250</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>0.784111</td>\n",
       "      <td>0.0279647</td>\n",
       "      <td>0.823129</td>\n",
       "      <td>0.74344</td>\n",
       "      <td>entropy</td>\n",
       "      <td>6</td>\n",
       "      <td>200</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.783965</td>\n",
       "      <td>0.034582</td>\n",
       "      <td>0.834062</td>\n",
       "      <td>0.730321</td>\n",
       "      <td>gini</td>\n",
       "      <td>5</td>\n",
       "      <td>250</td>\n",
       "      <td>59</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   mean_score  std_score max_score min_score rf__criterion rf__max_depth  \\\n",
       "44   0.786638  0.0311885  0.825559  0.734208       entropy             5   \n",
       "41   0.786054  0.0324027   0.82483  0.730564       entropy             5   \n",
       "53   0.785374  0.0276988  0.824587  0.745384       entropy             6   \n",
       "50   0.784111  0.0279647  0.823129   0.74344       entropy             6   \n",
       "17   0.783965   0.034582  0.834062  0.730321          gini             5   \n",
       "\n",
       "   rf__n_estimators rfi_fs__n_features_  \n",
       "44              250                  59  \n",
       "41              200                  59  \n",
       "53              250                  59  \n",
       "50              200                  59  \n",
       "17              250                  59  "
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_RF = get_search_results(gs_pipe_RF)\n",
    "results_RF.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- A minimal difference of 0.001 can be observed between the 1st and the 2nd optimal values of the RF hyperparameter values.\n",
    "\n",
    "- It can be observed from the above values that the top mean values are obtained while considering **59** features. \n",
    "\n",
    "- Hence, we plot a graph for results of the Grid search corresponding to **59** features."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "<div id=\"altair-viz-bb9564ec7cf34ceda89a5de5d423adc1\"></div>\n",
       "<script type=\"text/javascript\">\n",
       "  (function(spec, embedOpt){\n",
       "    let outputDiv = document.currentScript.previousElementSibling;\n",
       "    if (outputDiv.id !== \"altair-viz-bb9564ec7cf34ceda89a5de5d423adc1\") {\n",
       "      outputDiv = document.getElementById(\"altair-viz-bb9564ec7cf34ceda89a5de5d423adc1\");\n",
       "    }\n",
       "    const paths = {\n",
       "      \"vega\": \"https://cdn.jsdelivr.net/npm//vega@5?noext\",\n",
       "      \"vega-lib\": \"https://cdn.jsdelivr.net/npm//vega-lib?noext\",\n",
       "      \"vega-lite\": \"https://cdn.jsdelivr.net/npm//vega-lite@4.8.1?noext\",\n",
       "      \"vega-embed\": \"https://cdn.jsdelivr.net/npm//vega-embed@6?noext\",\n",
       "    };\n",
       "\n",
       "    function loadScript(lib) {\n",
       "      return new Promise(function(resolve, reject) {\n",
       "        var s = document.createElement('script');\n",
       "        s.src = paths[lib];\n",
       "        s.async = true;\n",
       "        s.onload = () => resolve(paths[lib]);\n",
       "        s.onerror = () => reject(`Error loading script: ${paths[lib]}`);\n",
       "        document.getElementsByTagName(\"head\")[0].appendChild(s);\n",
       "      });\n",
       "    }\n",
       "\n",
       "    function showError(err) {\n",
       "      outputDiv.innerHTML = `<div class=\"error\" style=\"color:red;\">${err}</div>`;\n",
       "      throw err;\n",
       "    }\n",
       "\n",
       "    function displayChart(vegaEmbed) {\n",
       "      vegaEmbed(outputDiv, spec, embedOpt)\n",
       "        .catch(err => showError(`Javascript Error: ${err.message}<br>This usually means there's a typo in your chart specification. See the javascript console for the full traceback.`));\n",
       "    }\n",
       "\n",
       "    if(typeof define === \"function\" && define.amd) {\n",
       "      requirejs.config({paths});\n",
       "      require([\"vega-embed\"], displayChart, err => showError(`Error loading script: ${err.message}`));\n",
       "    } else if (typeof vegaEmbed === \"function\") {\n",
       "      displayChart(vegaEmbed);\n",
       "    } else {\n",
       "      loadScript(\"vega\")\n",
       "        .then(() => loadScript(\"vega-lite\"))\n",
       "        .then(() => loadScript(\"vega-embed\"))\n",
       "        .catch(showError)\n",
       "        .then(() => displayChart(vegaEmbed));\n",
       "    }\n",
       "  })({\"config\": {\"view\": {\"continuousWidth\": 400, \"continuousHeight\": 300}}, \"data\": {\"name\": \"data-27e78fd47ef79d03dee21b15cca8a342\"}, \"mark\": {\"type\": \"line\", \"point\": true}, \"encoding\": {\"color\": {\"type\": \"nominal\", \"field\": \"rf__criterion\", \"title\": \"criterion\"}, \"x\": {\"type\": \"quantitative\", \"field\": \"rf__max_depth\", \"title\": \"Max Depth of the Tree\"}, \"y\": {\"type\": \"quantitative\", \"field\": \"mean_score\", \"scale\": {\"zero\": false}, \"title\": \"AUC Score\"}}, \"title\": \"RF Performance Comparison with 59 Features\", \"$schema\": \"https://vega.github.io/schema/vega-lite/v4.8.1.json\", \"datasets\": {\"data-27e78fd47ef79d03dee21b15cca8a342\": [{\"mean_score\": 0.7866375121477163, \"std_score\": 0.03118852272960078, \"max_score\": 0.8255587949465502, \"min_score\": 0.7342079689018465, \"rf__criterion\": \"entropy\", \"rf__max_depth\": 5, \"rf__n_estimators\": 250, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.7860544217687074, \"std_score\": 0.03240271179416913, \"max_score\": 0.8248299319727892, \"min_score\": 0.7305636540330417, \"rf__criterion\": \"entropy\", \"rf__max_depth\": 5, \"rf__n_estimators\": 200, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.7853741496598641, \"std_score\": 0.027698753617543974, \"max_score\": 0.8245869776482022, \"min_score\": 0.7453838678328474, \"rf__criterion\": \"entropy\", \"rf__max_depth\": 6, \"rf__n_estimators\": 250, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.7841107871720118, \"std_score\": 0.02796470765746266, \"max_score\": 0.8231292517006803, \"min_score\": 0.7434402332361516, \"rf__criterion\": \"entropy\", \"rf__max_depth\": 6, \"rf__n_estimators\": 200, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.7839650145772595, \"std_score\": 0.0345820199051136, \"max_score\": 0.8340621963070943, \"min_score\": 0.7303206997084548, \"rf__criterion\": \"gini\", \"rf__max_depth\": 5, \"rf__n_estimators\": 250, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.7839650145772594, \"std_score\": 0.03053014736771598, \"max_score\": 0.825801749271137, \"min_score\": 0.7351797862001944, \"rf__criterion\": \"gini\", \"rf__max_depth\": 6, \"rf__n_estimators\": 250, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.7829446064139941, \"std_score\": 0.031096103552451418, \"max_score\": 0.8129251700680271, \"min_score\": 0.7259475218658893, \"rf__criterion\": \"entropy\", \"rf__max_depth\": 5, \"rf__n_estimators\": 120, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.7815354713313897, \"std_score\": 0.036436474931522515, \"max_score\": 0.8299319727891156, \"min_score\": 0.7220602526724976, \"rf__criterion\": \"gini\", \"rf__max_depth\": 5, \"rf__n_estimators\": 200, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.781438289601555, \"std_score\": 0.02419200531376593, \"max_score\": 0.8083090379008746, \"min_score\": 0.7436831875607386, \"rf__criterion\": \"entropy\", \"rf__max_depth\": 6, \"rf__n_estimators\": 120, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.7812439261418852, \"std_score\": 0.03504866867219131, \"max_score\": 0.8335762876579202, \"min_score\": 0.7264334305150631, \"rf__criterion\": \"gini\", \"rf__max_depth\": 4, \"rf__n_estimators\": 250, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.7810981535471331, \"std_score\": 0.030974151462484355, \"max_score\": 0.8248299319727892, \"min_score\": 0.7342079689018464, \"rf__criterion\": \"entropy\", \"rf__max_depth\": 4, \"rf__n_estimators\": 200, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.7809037900874636, \"std_score\": 0.029174919975841117, \"max_score\": 0.8216715257531584, \"min_score\": 0.7371234207968902, \"rf__criterion\": \"entropy\", \"rf__max_depth\": 4, \"rf__n_estimators\": 250, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.7800291545189503, \"std_score\": 0.03524717168965138, \"max_score\": 0.826530612244898, \"min_score\": 0.7201166180758017, \"rf__criterion\": \"gini\", \"rf__max_depth\": 5, \"rf__n_estimators\": 120, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.7794946550048591, \"std_score\": 0.030357665413467533, \"max_score\": 0.8165694849368319, \"min_score\": 0.7281341107871719, \"rf__criterion\": \"gini\", \"rf__max_depth\": 6, \"rf__n_estimators\": 200, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.7791545189504373, \"std_score\": 0.036555059785888924, \"max_score\": 0.8328474246841594, \"min_score\": 0.7223032069970845, \"rf__criterion\": \"gini\", \"rf__max_depth\": 4, \"rf__n_estimators\": 200, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.7789115646258502, \"std_score\": 0.030833957320101885, \"max_score\": 0.8134110787172011, \"min_score\": 0.7257045675413022, \"rf__criterion\": \"gini\", \"rf__max_depth\": 6, \"rf__n_estimators\": 120, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.7776482021379981, \"std_score\": 0.0347584012070807, \"max_score\": 0.8299319727891157, \"min_score\": 0.7254616132167153, \"rf__criterion\": \"gini\", \"rf__max_depth\": 4, \"rf__n_estimators\": 120, \"rfi_fs__n_features_\": 59}, {\"mean_score\": 0.7768707482993198, \"std_score\": 0.02920735413322771, \"max_score\": 0.8160835762876579, \"min_score\": 0.7300777453838679, \"rf__criterion\": \"entropy\", \"rf__max_depth\": 4, \"rf__n_estimators\": 120, \"rfi_fs__n_features_\": 59}]}}, {\"mode\": \"vega-lite\"});\n",
       "</script>"
      ],
      "text/plain": [
       "alt.Chart(...)"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_RF_59_features = results_RF[results_RF['rfi_fs__n_features_'] == 59.0]\n",
    "\n",
    "alt.Chart(results_RF_59_features, \n",
    "          title='RF Performance Comparison with 59 Features'\n",
    "         ).mark_line(point=True).encode(\n",
    "    alt.X('rf__max_depth', title='Max Depth of the Tree'),\n",
    "    alt.Y('mean_score', title='AUC Score', scale=alt.Scale(zero=False)),\n",
    "    alt.Color('rf__criterion:N', title='criterion')\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Performance Evaluation <a class=\"anchor\" id=\"8\"></a>\n",
    "\n",
    "\n",
    "During Hyperparameter tuning phase we used 700 observation samples as a Training data within a cross-validation framework. We found out the best hyperparameter values for each of the 4 classifiers. For example, for KNN, it turned out that it worked best with value of K as 20,  with P =1 i.e Manhattan distance metric and by taking all 59 features into account.\n",
    "\n",
    "Now it is time to test these best performed tuned classfiers on the Test dataset to check the performance of the models on unseen data. We are going to perform this using 10 fold cross-validation without any repititions. For this purpose we are using `cross_val_score method` from the `sklearn.model_selection` package.\n",
    "\n",
    "1. **KNN** on the test data: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7444444444444445"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "cv_method_ttest = StratifiedKFold(n_splits=10, random_state=999)\n",
    "\n",
    "cv_results_KNN = cross_val_score(estimator=gs_pipe_KNN.best_estimator_,\n",
    "                                 X=Data_sample_test,\n",
    "                                 y=target_sample_test, \n",
    "                                 cv=cv_method_ttest, \n",
    "                                 n_jobs=-2,\n",
    "                                 scoring='roc_auc')\n",
    "cv_results_KNN.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. **NB (Naive Bayes)** on the test data: \n",
    "   \n",
    "   Before calculating the performance we have to normalize each probability value by using power tranformation on the test      dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.752910052910053"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Data_sample_test_transformed = PowerTransformer().fit_transform(Data_sample_test)\n",
    "\n",
    "cv_results_NB = cross_val_score(estimator=gs_pipe_NB.best_estimator_,\n",
    "                                X=Data_sample_test_transformed,\n",
    "                                y=target_sample_test, \n",
    "                                cv=cv_method_ttest, \n",
    "                                n_jobs=-2,\n",
    "                                scoring='roc_auc')\n",
    "cv_results_NB.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3. **Decision Trees (DT)** on the test data:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7113756613756613"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_results_DT = cross_val_score(estimator=gs_pipe_DT.best_estimator_,\n",
    "                                X=Data_sample_test,\n",
    "                                y=target_sample_test, \n",
    "                                cv=cv_method_ttest, \n",
    "                                n_jobs=-2,\n",
    "                                scoring='roc_auc')\n",
    "cv_results_DT.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "4. **Random Forest Ensemble** on the test data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7883597883597883"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cv_results_RF = cross_val_score(estimator=gs_pipe_RF.best_estimator_,\n",
    "                                X=Data_sample_test,\n",
    "                                y=target_sample_test, \n",
    "                                cv=cv_method_ttest, \n",
    "                                n_jobs=-2,\n",
    "                                scoring='roc_auc')\n",
    "cv_results_RF.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see from our results, Random Forest is working with better performance than all other classifiers on the unseen data. The next step is to check the Statistical Significance of these scores using paired statistical t-tests. We can see from the above that, perfromance of Random forest (RF) (0.7883) is slightly higher that that of Naive Bayes (NB)(0.752).\n",
    "\n",
    "Hence we are going to check the statistical significance of the score of these two classification models.\n",
    "\n",
    "To perform the statistical paired T-tests we are using `ttest_rel()` function from the scipy stats library package. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Ttest_relResult(statistic=-1.847148735065825, pvalue=0.09779358355358424)\n",
      "Ttest_relResult(statistic=-2.7038077932704687, pvalue=0.02424205755138577)\n",
      "Ttest_relResult(statistic=-2.1121922481003104, pvalue=0.06384210478804028)\n",
      "Ttest_relResult(statistic=0.7913699829612472, pvalue=0.44908050240877695)\n",
      "Ttest_relResult(statistic=-0.4135166166932492, pvalue=0.6889075434373727)\n",
      "Ttest_relResult(statistic=-0.998874968142931, pvalue=0.34395224491849563)\n"
     ]
    }
   ],
   "source": [
    "from scipy import stats\n",
    "\n",
    "print(stats.ttest_rel(cv_results_NB, cv_results_RF))\n",
    "print(stats.ttest_rel(cv_results_KNN, cv_results_RF))\n",
    "print(stats.ttest_rel(cv_results_DT, cv_results_RF))\n",
    "print(stats.ttest_rel(cv_results_KNN, cv_results_DT))\n",
    "print(stats.ttest_rel(cv_results_KNN, cv_results_NB))\n",
    "print(stats.ttest_rel(cv_results_DT, cv_results_NB))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As P-values smaller than 0.05 represents statistically significant difference. Even though the performance of the RF is better than NB by slightly higher percentages, by looking at the above result we can conclude that at a 95% significance level difference between the scores of Random Forest classifier and naive Bayes classifier is **not statistically significant**.\n",
    "\n",
    "We are using AUC score to check the performance of the models. We can consider other metrics to determine the performance of the classifiers like\n",
    "\n",
    "- Accuracy\n",
    "- Precision\n",
    "- F1 Score\n",
    "- Recall\n",
    "\n",
    "etc.\n",
    "\n",
    "\n",
    "In the last step we are going to show the confusion matrices for each of the classifiers. Confusion matrix is also one of the measures to check the performance of the classifier. For this purpose we are going to store predictions of each classifier on the test data into seperate lists and plot them against the target test data. We are using `skelearn metrics` to draw confusion matrices.\n",
    "\n",
    "These matrices are in the format : \n",
    "Thus in binary classification, the count of true negatives is at position [0,0] , false negatives is at [1,0] , true positives is at [1,1] and false positives is at [0,1] ."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_KNN = gs_pipe_KNN.predict(Data_sample_test)\n",
    "Data_test_transformed = PowerTransformer().fit_transform(Data_sample_test)\n",
    "pred_NB = gs_pipe_NB.predict(Data_test_transformed)\n",
    "pred_DT = gs_pipe_DT.predict(Data_sample_test)\n",
    "pred_RF = gs_pipe_RF.predict(Data_sample_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Confusion matrix for K-Nearest Neighbor\n",
      "[[ 31  59]\n",
      " [ 13 197]]\n",
      "\n",
      "Confusion matrix for Naive Bayes\n",
      "[[ 27  63]\n",
      " [ 16 194]]\n",
      "\n",
      "Confusion matrix for Decision Tree\n",
      "[[ 32  58]\n",
      " [ 33 177]]\n",
      "\n",
      "Confusion matrix for Random Forest\n",
      "[[ 12  78]\n",
      " [  4 206]]\n"
     ]
    }
   ],
   "source": [
    "from sklearn import metrics\n",
    "print(\"\\nConfusion matrix for K-Nearest Neighbor\") \n",
    "print(metrics.confusion_matrix(target_sample_test, pred_KNN))\n",
    "print(\"\\nConfusion matrix for Naive Bayes\") \n",
    "print(metrics.confusion_matrix(target_sample_test, pred_NB))\n",
    "print(\"\\nConfusion matrix for Decision Tree\") \n",
    "print(metrics.confusion_matrix(target_sample_test, pred_DT))\n",
    "print(\"\\nConfusion matrix for Random Forest\") \n",
    "print(metrics.confusion_matrix(target_sample_test, pred_RF))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Limitations and Proposed Solutions<a class=\"anchor\" id=\"9\"></a>\n",
    "\n",
    "- The modelling strategy has some flaws and limitations. First the data collected is skewed and has class imbalance problem hereby preventing us from using different performance measuring techniques such as Accuracy, F1-score, Precision, Recall etc hence we have used mean AUC scores to check the performance of the model as ROC AUC is robust towards class imbalance problems.\n",
    "\n",
    "- Secondly, various different combinations of the RFI Feature selector can be used to test the performance of the models. We have tested the models on 10,20 and 59 features.\n",
    "\n",
    "- We have tuned the model hyperparameters to a limit, we can re-run our experiments and try various different combinations for tuning the data. Since the Random Forest Ensembler method statistically outperforms other methods, different n_estimators can be used to tune the model to increase the mean AUC score.\n",
    "\n",
    "- K-fold cross validation techniques have been used to validate the model, 5 Folds were used to validate the traning data and 10 folds were used to validate the testing data. In future repeated K-fold cross validation can be used as a validation technique.\n",
    "\n",
    "- Other models such as Support Vector Machines and Neural networks can be considered and mean scores can be observed.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Summary and Conclusion <a class=\"anchor\" id=\"10\"></a>\n",
    "\n",
    "The Random Forest Ensembler(RF) with all 59 features selected by Random Forest Importance(RFI) produces the highest cross-validated AUC score for the training data. In addition, when evaluated on the test data in cross validated format, the Random Forest Ensembler(RF) outperforms Naive Bayes with slightly higher percentage with respect to the mean AUC score, however, the difference between the mean scores of Naive Bayes and Random Forest is not statistically significant.\n",
    "\n",
    "Though RF classification model outperformes others like NB, KNN and DT models, It's performance difference is not statistically significant if we consider 95% significance level. If we consider the cost matrix provided with the data, using Random Forest ensembler gives more loss than a Naive Bayes classfier.\n",
    "\n",
    "The cost Matrix provided is : \n",
    "\n",
    "          0        1\n",
    "    ----------------------------\n",
    "     0    0        5\n",
    "    ----------------------------\n",
    "     1    1        0\n",
    "\n",
    "\n",
    "It is worse to class a customer as good when they are bad (5), than it is to class a customer as bad when they are good (1).\n",
    "\n",
    "when we try to multiply the resultant confusion matrices of RF and NB with the cost matrix above, we found out that, `with NB,  the loss cost is much lesser than that of the RF clssifier`.\n",
    "\n",
    "Hence, we can say that, having best performance accuracy is not good enough when it comes to the selection of the ideal classification model for a certain problem. There are many more parameters affecting to the model. Its always a trade-off between the precision and the recall of the model. Only having optimal precision or recall doesnt make it an overall optimal classification model.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# References <a class=\"anchor\" id=\"11\"></a>\n",
    "\n",
    "- UCI Machine Learning Repository Collection of Data bases : https://archive.ics.uci.edu/ml/about.html\n",
    "\n",
    "- Dr Vural Aksakalli, Zeren Yenice, Yong Kai Wong, Imran Ture, Milad Malekipirbazari https://www.featureranking.com/ SK Part 1: Basic Modeling Techniques.\n",
    "    - https://www.featureranking.com/tutorials/machine-learning-tutorials/sk-part-2-feature-selection-and-ranking/ SK Part 2: Feature Selection and Ranking\n",
    "    - https://www.featureranking.com/tutorials/machine-learning-tutorials/sk-part-3-cross-validation-and-hyperparameter-tuning/ SK Part 3: Cross-Validation and Hyperparameter Tuning\n",
    "    - https://www.featureranking.com/tutorials/machine-learning-tutorials/sk-part-4-model-evaluation/ SK Part 4: Model Evaluation\n",
    "    - https://www.featureranking.com/tutorials/machine-learning-tutorials/sk-part-5-advanced-topics-pipelines-statistical-model-comparison-and-model-deployment/ SK Part 5: Advanced Topics: Pipelines, Statistical Model Comparison, and Model Deployment\n",
    "    \n",
    "- Regression Case Study(Predicting Age in Census Data):https://www.featureranking.com/tutorials/statistics-tutorials/regression-case-study/ \n",
    "- Case Study: Predicting Income Status : https://www.featureranking.com/tutorials/machine-learning-tutorials/case-study-predicting-income-status/ \n",
    "\n",
    "- Towards Data Science : A Medium publication sharing concepts, ideas, and codes : https://towardsdatascience.com/ \n",
    " \n",
    "\n",
    "\n",
    "\n",
    "\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}