Twitter Sentiment Analysis Project.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "9d6060d2",
   "metadata": {},
   "source": [
    "<center>\n",
    "    <img src=\"https://weclouddata.s3.amazonaws.com/images/logos/wcd_logo_new_2.png\" width='30%'> \n",
    "</center>\n",
    "\n",
    "----------"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "361124d5",
   "metadata": {},
   "source": [
    "### Twitter Sentiment Analysis Lab\n",
    "\n",
    "#### Dataset (tweets.csv)\n",
    "The dataset contains approximately 2000 different (scrapped) tweets with the following attributes:\n",
    "- 'id'         : unique 19 digit id for each tweet\n",
    "- 'created_at' : date & time of each tweet (or retweet)\n",
    "- 'text'       : tweet details/ description \n",
    "- 'location'   : origin of tweet\n",
    "\n",
    "#### Objective \n",
    "- **Sentiment label** - for each tweet based on it's text, devise a method to assign an appropriate sentiment ('positive', 'negative' or 'neutral'). This is achieved by using TextBlob (https://textblob.readthedocs.io/en/dev/)\n",
    "- **Text Analytics/NLP** - to extract features from tweet texts\n",
    "- **Machine Learning** - Building a robust & optimized ML model to accurately predict the sentiment associated with each tweet & explanation of the built model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "e91f59ed",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Libraries for data preparation & visualisation\n",
    "\n",
    "import re\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "sns.set(color_codes=True)\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "from wordcloud import WordCloud\n",
    "\n",
    "\n",
    "# Library to ignore warnings\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "# Library for assigning sentiment label\n",
    "\n",
    "from textblob import TextBlob\n",
    "\n",
    "# Libraries for text analytics\n",
    "\n",
    "import nltk\n",
    "from nltk.stem.wordnet import WordNetLemmatizer\n",
    "from nltk.stem.porter  import PorterStemmer\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "\n",
    "# Libraries for ML modeling \n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import StratifiedKFold, ParameterGrid\n",
    "from sklearn.model_selection import GridSearchCV, cross_val_score\n",
    "from sklearn.preprocessing import OneHotEncoder, StandardScaler\n",
    "from imblearn.over_sampling import RandomOverSampler\n",
    "from imblearn.pipeline import Pipeline\n",
    "from sklearn.svm import LinearSVC\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from xgboost import XGBClassifier\n",
    "from sklearn.metrics import f1_score, classification_report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "dbe80ec8",
   "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>id</th>\n",
       "      <th>created_at</th>\n",
       "      <th>text</th>\n",
       "      <th>location</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1083193473539420160</td>\n",
       "      <td>2019-01-10 02:47:03</td>\n",
       "      <td>@CIBC please explain to me why I want to remai...</td>\n",
       "      <td>Canada</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1083191479215026176</td>\n",
       "      <td>2019-01-10 02:39:08</td>\n",
       "      <td>RT @CIBCLiveLabs: We are pleased to announce, ...</td>\n",
       "      <td>Oshawa, Ontario</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1083184422709575683</td>\n",
       "      <td>2019-01-10 02:11:05</td>\n",
       "      <td>CIBC World Markets Inc. Decreases Holdings in ...</td>\n",
       "      <td>The Netherlands</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1083182915826126848</td>\n",
       "      <td>2019-01-10 02:05:06</td>\n",
       "      <td>Le patron de la Banque @cibc s’attend à un ral...</td>\n",
       "      <td>Montréal</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1083177871881818112</td>\n",
       "      <td>2019-01-10 01:45:03</td>\n",
       "      <td>Your home is a valuable asset. Use your equity...</td>\n",
       "      <td>Lower Mainland, BC</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    id           created_at  \\\n",
       "0  1083193473539420160  2019-01-10 02:47:03   \n",
       "1  1083191479215026176  2019-01-10 02:39:08   \n",
       "2  1083184422709575683  2019-01-10 02:11:05   \n",
       "3  1083182915826126848  2019-01-10 02:05:06   \n",
       "4  1083177871881818112  2019-01-10 01:45:03   \n",
       "\n",
       "                                                text            location  \n",
       "0  @CIBC please explain to me why I want to remai...              Canada  \n",
       "1  RT @CIBCLiveLabs: We are pleased to announce, ...     Oshawa, Ontario  \n",
       "2  CIBC World Markets Inc. Decreases Holdings in ...     The Netherlands  \n",
       "3  Le patron de la Banque @cibc s’attend à un ral...            Montréal  \n",
       "4  Your home is a valuable asset. Use your equity...  Lower Mainland, BC  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Reading in data \n",
    "\n",
    "data = pd.read_csv('tweets.csv')\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "a6b1f852",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1951, 4)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Shape of data\n",
    "\n",
    "data.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f4316810",
   "metadata": {},
   "source": [
    "- There are 1951 rows & 4 columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "858c8e94",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 1951 entries, 0 to 1950\n",
      "Data columns (total 4 columns):\n",
      " #   Column      Non-Null Count  Dtype \n",
      "---  ------      --------------  ----- \n",
      " 0   id          1951 non-null   int64 \n",
      " 1   created_at  1951 non-null   object\n",
      " 2   text        1951 non-null   object\n",
      " 3   location    1509 non-null   object\n",
      "dtypes: int64(1), object(3)\n",
      "memory usage: 61.1+ KB\n"
     ]
    }
   ],
   "source": [
    "data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1ecd3d7a",
   "metadata": {},
   "source": [
    "- 'id' is a 16 digit random number associated with each tweet. This should have no relevance on final modeling & can be dropped \n",
    "- 'created_at' is of object type & will be converted to datetime format\n",
    "- 'location' has 442 missing records "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "ac5f452b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dropping 'id'\n",
    "\n",
    "data = data.drop('id', axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "ed7435c4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Changing datatype for created_at \n",
    "\n",
    "data['created_at'] = pd.to_datetime(data['created_at'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "1be298e1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Replacing missing records under location with 'unknown'\n",
    "\n",
    "data = data.fillna('Unknown')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "5383ad6f",
   "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>count</th>\n",
       "      <th>unique</th>\n",
       "      <th>top</th>\n",
       "      <th>freq</th>\n",
       "      <th>first</th>\n",
       "      <th>last</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>created_at</th>\n",
       "      <td>1951</td>\n",
       "      <td>1924</td>\n",
       "      <td>2019-01-08 20:15:59</td>\n",
       "      <td>4</td>\n",
       "      <td>2018-12-30 23:26:02</td>\n",
       "      <td>2019-01-10 02:47:03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>text</th>\n",
       "      <td>1951</td>\n",
       "      <td>1768</td>\n",
       "      <td>RT @NHLBlackhawks: Staying warm never looked s...</td>\n",
       "      <td>25</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaT</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>location</th>\n",
       "      <td>1951</td>\n",
       "      <td>429</td>\n",
       "      <td>Unknown</td>\n",
       "      <td>442</td>\n",
       "      <td>NaT</td>\n",
       "      <td>NaT</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           count unique                                                top  \\\n",
       "created_at  1951   1924                                2019-01-08 20:15:59   \n",
       "text        1951   1768  RT @NHLBlackhawks: Staying warm never looked s...   \n",
       "location    1951    429                                            Unknown   \n",
       "\n",
       "           freq               first                last  \n",
       "created_at    4 2018-12-30 23:26:02 2019-01-10 02:47:03  \n",
       "text         25                 NaT                 NaT  \n",
       "location    442                 NaT                 NaT  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.describe(include='all').T"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "35f406eb",
   "metadata": {},
   "source": [
    "- 'text' value is repeated in the dataset, e.g., text 'RT @NHLBlackhawks: Staying warm never looked s...' occurs 25 times in the dataset. This indicates some of the texts have retweets "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "fb49edf0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Checking duplicates \n",
    "\n",
    "data.duplicated().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8efe17dd",
   "metadata": {},
   "source": [
    "- There are no duplicate records "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "160791df",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Checking for missing values \n",
    "\n",
    "data.isnull().sum().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "26e32196",
   "metadata": {},
   "source": [
    "- No further missing values "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dee66a2b",
   "metadata": {},
   "source": [
    "#### Feature Engineering & EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "d273d446",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Language detection\n",
    "\n",
    "from langdetect import detect\n",
    "\n",
    "def language_detection(text):\n",
    "    \n",
    "    try:\n",
    "        language = detect(text)\n",
    "        return(language)\n",
    "    except:\n",
    "        return('error')\n",
    "\n",
    "data['language'] = data['text'].apply(language_detection)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "df858747",
   "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>created_at</th>\n",
       "      <th>text</th>\n",
       "      <th>location</th>\n",
       "      <th>language</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2019-01-10 02:47:03</td>\n",
       "      <td>@CIBC please explain to me why I want to remai...</td>\n",
       "      <td>Canada</td>\n",
       "      <td>en</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2019-01-10 02:39:08</td>\n",
       "      <td>RT @CIBCLiveLabs: We are pleased to announce, ...</td>\n",
       "      <td>Oshawa, Ontario</td>\n",
       "      <td>en</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2019-01-10 02:11:05</td>\n",
       "      <td>CIBC World Markets Inc. Decreases Holdings in ...</td>\n",
       "      <td>The Netherlands</td>\n",
       "      <td>en</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2019-01-10 02:05:06</td>\n",
       "      <td>Le patron de la Banque @cibc s’attend à un ral...</td>\n",
       "      <td>Montréal</td>\n",
       "      <td>fr</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2019-01-10 01:45:03</td>\n",
       "      <td>Your home is a valuable asset. Use your equity...</td>\n",
       "      <td>Lower Mainland, BC</td>\n",
       "      <td>en</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           created_at                                               text  \\\n",
       "0 2019-01-10 02:47:03  @CIBC please explain to me why I want to remai...   \n",
       "1 2019-01-10 02:39:08  RT @CIBCLiveLabs: We are pleased to announce, ...   \n",
       "2 2019-01-10 02:11:05  CIBC World Markets Inc. Decreases Holdings in ...   \n",
       "3 2019-01-10 02:05:06  Le patron de la Banque @cibc s’attend à un ral...   \n",
       "4 2019-01-10 01:45:03  Your home is a valuable asset. Use your equity...   \n",
       "\n",
       "             location language  \n",
       "0              Canada       en  \n",
       "1     Oshawa, Ontario       en  \n",
       "2     The Netherlands       en  \n",
       "3            Montréal       fr  \n",
       "4  Lower Mainland, BC       en  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "33dd756d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "en       1809\n",
       "fr         57\n",
       "error      13\n",
       "ca         12\n",
       "it          6\n",
       "es          6\n",
       "so          6\n",
       "sv          5\n",
       "ro          5\n",
       "vi          4\n",
       "da          4\n",
       "de          3\n",
       "pt          3\n",
       "zh-cn       3\n",
       "tr          2\n",
       "id          2\n",
       "cy          2\n",
       "pl          1\n",
       "ru          1\n",
       "ja          1\n",
       "th          1\n",
       "nl          1\n",
       "no          1\n",
       "fi          1\n",
       "fa          1\n",
       "ar          1\n",
       "Name: language, dtype: int64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Language \n",
    "\n",
    "data['language'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "1ee90959",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Filter the dataset to only contain english texts\n",
    "\n",
    "data = data[data['language']=='en']\n",
    "data.drop('language', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "aa94bc7d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1809, 3)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Shape of data\n",
    "\n",
    "data.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "99fd3704",
   "metadata": {},
   "source": [
    "- There are 1809 rows remaining "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "15566719",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create new features from 'created_at'\n",
    "\n",
    "# Day of the week\n",
    "data['day_of_week'] = data['created_at'].dt.dayofweek\n",
    "data['day_of_week'] = data['day_of_week'].replace({0:'Monday', 1:'Tuesday', 2:'Wednesday', 3:'Thursday',\\\n",
    "                                                   4:'Friday', 5:'Saturday',6:'Sunday'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "6c9e659f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Hour \n",
    "\n",
    "data['hour']        = data['created_at'].dt.hour + data['created_at'].dt.minute/60"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "c096b9a9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Wednesday    441\n",
       "Monday       432\n",
       "Tuesday      335\n",
       "Friday       230\n",
       "Thursday     216\n",
       "Saturday      92\n",
       "Sunday        63\n",
       "Name: day_of_week, dtype: int64"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Value counts for day_of_week\n",
    "\n",
    "data['day_of_week'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c16acf9c",
   "metadata": {},
   "source": [
    "- Majority of the tweets are on Wednesday (middle of the week), followed by Monday\n",
    "- Least number of tweets are on Saturday or Sunday"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "0151f96d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dropping the 'created_at' attribute\n",
    "\n",
    "data = data.drop('created_at', axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "aceadf11",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Let's create target column, i.e., sentiment associated with text\n",
    "\n",
    "# Defining a function to assign sentiments (positive, negative or neutral)\n",
    "\n",
    "def get_sentiment(text):\n",
    "    blob = TextBlob(text)\n",
    "    sentiment = blob.sentiment.polarity\n",
    "    if sentiment > 0:\n",
    "        return 'positive'\n",
    "    elif sentiment < 0:\n",
    "        return 'negative'\n",
    "    else:\n",
    "        return 'neutral'\n",
    "    \n",
    "data['sentiment'] = data['text'].apply(get_sentiment)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "b73473b0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "positive    793\n",
       "neutral     715\n",
       "negative    301\n",
       "Name: sentiment, dtype: int64"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Value counts for sentiment\n",
    "\n",
    "data['sentiment'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b7647e11",
   "metadata": {},
   "source": [
    "- Maximum tweets have a positive sentiment followed by neutral sentiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "710c78cc",
   "metadata": {},
   "outputs": [],
   "source": [
    "# function to create labeled barplots\n",
    "\n",
    "\n",
    "def labeled_barplot(data, feature, target ,perc=False, n=None):\n",
    "\n",
    "    total = len(data[feature])  # length of the column\n",
    "    count = data[feature].nunique()\n",
    "    if n is None:\n",
    "        plt.figure(figsize=(count + 2, 5))\n",
    "    else:\n",
    "        plt.figure(figsize=(n + 2, 5))\n",
    "\n",
    "    plt.xticks(rotation=90, fontsize=15)\n",
    "    ax = sns.countplot(data=data.sort_values(by=target),x=feature,palette=\"Paired\",hue=target,\\\n",
    "                       order=data[feature].value_counts().index[:n].sort_values(),)\n",
    "\n",
    "    for p in ax.patches:\n",
    "        if perc == True:\n",
    "            label = \"{:.1f}%\".format(100 * p.get_height() / total)  # percentage of each class of the category\n",
    "        else:\n",
    "            label = p.get_height()  # count of each level of the category\n",
    "\n",
    "        x = p.get_x() + p.get_width() / 2  # width of the plot\n",
    "        y = p.get_height()  # height of the plot\n",
    "\n",
    "        ax.annotate(label,(x, y),ha=\"center\",va=\"center\",size=12,xytext=(0, 5),textcoords=\"offset points\",)  # annotate the percentage\n",
    "\n",
    "    plt.show()  # show the plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "b7b684a6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# day_of_week\n",
    "\n",
    "df = data.copy() \n",
    "\n",
    "from pandas.api.types import CategoricalDtype\n",
    "cats     = [ 'Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday']\n",
    "cat_type = CategoricalDtype(categories=cats, ordered=True)\n",
    "\n",
    "df['day_of_week']=df['day_of_week'].astype(cat_type)\n",
    "\n",
    "labeled_barplot(df, 'day_of_week', 'sentiment')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "23a9c05f",
   "metadata": {},
   "source": [
    "- All weekdays (except Tuesdays) have majority of tweets as positive tweets, while weekends & Tuesdays have majority of tweets as neutral tweets "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "a7f00fec",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# hour\n",
    "\n",
    "plt.figure(figsize=(9,5))\n",
    "sns.histplot(data=data, x='hour');"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c29159b8",
   "metadata": {},
   "source": [
    "- Majority of the tweets are in the late afternoons & evenings (peaking after 3PM)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "798326bc",
   "metadata": {},
   "source": [
    "New features will be created by counting number of positive & negative words in each tweet. The list of all positive & negative words are borrowed from this study: https://www.cs.uic.edu/~liub/FBS/sentiment-analysis.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "26830377",
   "metadata": {},
   "outputs": [],
   "source": [
    "positive_words = pd.read_csv('positive-words.txt', skiprows=35, names=['words'])\n",
    "positive_words = positive_words['words'].values.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "7592135a",
   "metadata": {},
   "outputs": [],
   "source": [
    "negative_words = pd.read_csv('negative-words.txt', skiprows=35, names=['words'])\n",
    "negative_words = negative_words['words'].values.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "53bbaef6",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Let's create new features, i.e., number of positive & negative words associated with each text\n",
    "\n",
    "# Defining a function to count number of positive & negative words\n",
    "\n",
    "def count_words(tweet, words): \n",
    "    count = 0\n",
    "    for word in tweet.split(' '):\n",
    "        if word in words:\n",
    "            count += 1\n",
    "    return count\n",
    "\n",
    "data['number_positive_words'] = data['text'].apply(count_words, words=positive_words)\n",
    "data['number_negative_words'] = data['text'].apply(count_words, words=negative_words)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "b6f954a7",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to create stacked barplots\n",
    "\n",
    "def stacked_barplot(data, predictor, target):\n",
    "\n",
    "    count = data[predictor].nunique()\n",
    "    sorter = data[target].value_counts().index[-1]\n",
    "    tab1 = pd.crosstab(data[predictor], data[target], margins=True).sort_values(by=sorter, ascending=False)\n",
    "    print(tab1)\n",
    "    print(\"-\" * 120)\n",
    "    tab = pd.crosstab(data[predictor], data[target], normalize=\"index\").sort_values(by=predictor, ascending=True)\n",
    "    tab.plot(kind=\"bar\", stacked=True, figsize=(count + 5, 5))\n",
    "    plt.legend(loc=\"lower left\", frameon=False,)\n",
    "    plt.legend(loc=\"upper left\", bbox_to_anchor=(1, 1))\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "9063a0ad",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sentiment              negative  neutral  positive   All\n",
      "number_positive_words                                   \n",
      "All                         301      715       793  1809\n",
      "0                           285      654       505  1444\n",
      "1                            15       55       190   260\n",
      "2                             1        6        85    92\n",
      "3                             0        0         9     9\n",
      "4                             0        0         4     4\n",
      "------------------------------------------------------------------------------------------------------------------------\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# number_positive_words\n",
    "\n",
    "stacked_barplot(data, 'number_positive_words', 'sentiment')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b493548e",
   "metadata": {},
   "source": [
    "- Tweets with 0 or 1 count for number_positive_words have all three sentiments 'neutral', 'positive' & 'negative' \n",
    "- Tweets with 1, 2, 3 or 4 number_positive_words are majorly a 'positive' sentiment as expected"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "25cd1588",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sentiment              negative  neutral  positive   All\n",
      "number_negative_words                                   \n",
      "All                         301      715       793  1809\n",
      "0                           189      625       676  1490\n",
      "1                            82       82       105   269\n",
      "2                            28        7        12    47\n",
      "3                             1        1         0     2\n",
      "4                             1        0         0     1\n",
      "------------------------------------------------------------------------------------------------------------------------\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# number_negative_words\n",
    "\n",
    "stacked_barplot(data, 'number_negative_words', 'sentiment')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "82358e28",
   "metadata": {},
   "source": [
    "- Tweets with 0 or 1 count for number_negative_words have all three sentiments 'neutral', 'positive' & 'negative'\n",
    "- Tweets with 2, 3 or 4 number_negative_words are majorly a 'negative' sentiment as expected"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "1a48dcca",
   "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>text</th>\n",
       "      <th>location</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>hour</th>\n",
       "      <th>sentiment</th>\n",
       "      <th>number_positive_words</th>\n",
       "      <th>number_negative_words</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>@CIBC please explain to me why I want to remai...</td>\n",
       "      <td>Canada</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>2.783333</td>\n",
       "      <td>neutral</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>RT @CIBCLiveLabs: We are pleased to announce, ...</td>\n",
       "      <td>Oshawa, Ontario</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>2.650000</td>\n",
       "      <td>positive</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>CIBC World Markets Inc. Decreases Holdings in ...</td>\n",
       "      <td>The Netherlands</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>2.183333</td>\n",
       "      <td>neutral</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Your home is a valuable asset. Use your equity...</td>\n",
       "      <td>Lower Mainland, BC</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>1.750000</td>\n",
       "      <td>positive</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Both #CIBC and #BMO ads at the Chicago Black H...</td>\n",
       "      <td>Northern Virginia</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>1.633333</td>\n",
       "      <td>negative</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                text            location  \\\n",
       "0  @CIBC please explain to me why I want to remai...              Canada   \n",
       "1  RT @CIBCLiveLabs: We are pleased to announce, ...     Oshawa, Ontario   \n",
       "2  CIBC World Markets Inc. Decreases Holdings in ...     The Netherlands   \n",
       "4  Your home is a valuable asset. Use your equity...  Lower Mainland, BC   \n",
       "5  Both #CIBC and #BMO ads at the Chicago Black H...   Northern Virginia   \n",
       "\n",
       "  day_of_week      hour sentiment  number_positive_words  \\\n",
       "0    Thursday  2.783333   neutral                      0   \n",
       "1    Thursday  2.650000  positive                      2   \n",
       "2    Thursday  2.183333   neutral                      0   \n",
       "4    Thursday  1.750000  positive                      1   \n",
       "5    Thursday  1.633333  negative                      0   \n",
       "\n",
       "   number_negative_words  \n",
       "0                      0  \n",
       "1                      1  \n",
       "2                      0  \n",
       "4                      0  \n",
       "5                      0  "
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "5e39294e",
   "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>text</th>\n",
       "      <th>location</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>hour</th>\n",
       "      <th>sentiment</th>\n",
       "      <th>number_positive_words</th>\n",
       "      <th>number_negative_words</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>cibc please explain to me why i want to remai...</td>\n",
       "      <td>Canada</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>2.783333</td>\n",
       "      <td>neutral</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>rt  cibclivelabs  we are pleased to announce  ...</td>\n",
       "      <td>Oshawa, Ontario</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>2.650000</td>\n",
       "      <td>positive</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>cibc world markets inc  decreases holdings in ...</td>\n",
       "      <td>The Netherlands</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>2.183333</td>\n",
       "      <td>neutral</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>your home is a valuable asset  use your equity...</td>\n",
       "      <td>Lower Mainland, BC</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>1.750000</td>\n",
       "      <td>positive</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>both  cibc and  bmo ads at the chicago black h...</td>\n",
       "      <td>Northern Virginia</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>1.633333</td>\n",
       "      <td>negative</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                text            location  \\\n",
       "0   cibc please explain to me why i want to remai...              Canada   \n",
       "1  rt  cibclivelabs  we are pleased to announce  ...     Oshawa, Ontario   \n",
       "2  cibc world markets inc  decreases holdings in ...     The Netherlands   \n",
       "4  your home is a valuable asset  use your equity...  Lower Mainland, BC   \n",
       "5  both  cibc and  bmo ads at the chicago black h...   Northern Virginia   \n",
       "\n",
       "  day_of_week      hour sentiment  number_positive_words  \\\n",
       "0    Thursday  2.783333   neutral                      0   \n",
       "1    Thursday  2.650000  positive                      2   \n",
       "2    Thursday  2.183333   neutral                      0   \n",
       "4    Thursday  1.750000  positive                      1   \n",
       "5    Thursday  1.633333  negative                      0   \n",
       "\n",
       "   number_negative_words  \n",
       "0                      0  \n",
       "1                      1  \n",
       "2                      0  \n",
       "4                      0  \n",
       "5                      0  "
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Convert tweets to lowercase \n",
    "\n",
    "data['text'] = data['text'].str.lower()\n",
    "\n",
    "# Remove non-alphanumeric character from tweets such as '@'\n",
    "\n",
    "def remove_non_alphanumeric(tweet):\n",
    "    pattern = re.compile('\\W')\n",
    "    x = re.sub(pattern, ' ', tweet)\n",
    "    return(x)\n",
    "    \n",
    "data['text'] = data['text'].apply(remove_non_alphanumeric)\n",
    "\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2dcc3f27",
   "metadata": {},
   "source": [
    "- The text has been converted to lowercase & non-alphanumerical characters such as @ & # have been removed "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "39f5da2c",
   "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>text</th>\n",
       "      <th>location</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>hour</th>\n",
       "      <th>sentiment</th>\n",
       "      <th>number_positive_words</th>\n",
       "      <th>number_negative_words</th>\n",
       "      <th>tokens</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>cibc please explain to me why i want to remai...</td>\n",
       "      <td>Canada</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>2.783333</td>\n",
       "      <td>neutral</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>[cibc, please, explain, to, me, why, i, want, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>rt  cibclivelabs  we are pleased to announce  ...</td>\n",
       "      <td>Oshawa, Ontario</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>2.650000</td>\n",
       "      <td>positive</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>[rt, cibclivelabs, we, are, pleased, to, annou...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>cibc world markets inc  decreases holdings in ...</td>\n",
       "      <td>The Netherlands</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>2.183333</td>\n",
       "      <td>neutral</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>[cibc, world, markets, inc, decreases, holding...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>your home is a valuable asset  use your equity...</td>\n",
       "      <td>Lower Mainland, BC</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>1.750000</td>\n",
       "      <td>positive</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>[your, home, is, a, valuable, asset, use, your...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>both  cibc and  bmo ads at the chicago black h...</td>\n",
       "      <td>Northern Virginia</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>1.633333</td>\n",
       "      <td>negative</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>[both, cibc, and, bmo, ads, at, the, chicago, ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                text            location  \\\n",
       "0   cibc please explain to me why i want to remai...              Canada   \n",
       "1  rt  cibclivelabs  we are pleased to announce  ...     Oshawa, Ontario   \n",
       "2  cibc world markets inc  decreases holdings in ...     The Netherlands   \n",
       "4  your home is a valuable asset  use your equity...  Lower Mainland, BC   \n",
       "5  both  cibc and  bmo ads at the chicago black h...   Northern Virginia   \n",
       "\n",
       "  day_of_week      hour sentiment  number_positive_words  \\\n",
       "0    Thursday  2.783333   neutral                      0   \n",
       "1    Thursday  2.650000  positive                      2   \n",
       "2    Thursday  2.183333   neutral                      0   \n",
       "4    Thursday  1.750000  positive                      1   \n",
       "5    Thursday  1.633333  negative                      0   \n",
       "\n",
       "   number_negative_words                                             tokens  \n",
       "0                      0  [cibc, please, explain, to, me, why, i, want, ...  \n",
       "1                      1  [rt, cibclivelabs, we, are, pleased, to, annou...  \n",
       "2                      0  [cibc, world, markets, inc, decreases, holding...  \n",
       "4                      0  [your, home, is, a, valuable, asset, use, your...  \n",
       "5                      0  [both, cibc, and, bmo, ads, at, the, chicago, ...  "
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Tokenize sentences to arrays of words\n",
    "\n",
    "data['tokens'] = data['text'].apply(nltk.word_tokenize)\n",
    "\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "4fd884cb",
   "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>text</th>\n",
       "      <th>location</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>hour</th>\n",
       "      <th>sentiment</th>\n",
       "      <th>number_positive_words</th>\n",
       "      <th>number_negative_words</th>\n",
       "      <th>tokens</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>cibc please explain to me why i want to remai...</td>\n",
       "      <td>Canada</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>2.783333</td>\n",
       "      <td>neutral</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>[cibc, please, explain, to, me, why, i, want, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>rt  cibclivelabs  we are pleased to announce  ...</td>\n",
       "      <td>Oshawa, Ontario</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>2.650000</td>\n",
       "      <td>positive</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>[rt, cibclivelabs, we, are, pleased, to, annou...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>cibc world markets inc  decreases holdings in ...</td>\n",
       "      <td>The Netherlands</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>2.183333</td>\n",
       "      <td>neutral</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>[cibc, world, market, inc, decrease, holding, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>your home is a valuable asset  use your equity...</td>\n",
       "      <td>Lower Mainland, BC</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>1.750000</td>\n",
       "      <td>positive</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>[your, home, is, a, valuable, asset, use, your...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>both  cibc and  bmo ads at the chicago black h...</td>\n",
       "      <td>Northern Virginia</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>1.633333</td>\n",
       "      <td>negative</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>[both, cibc, and, bmo, ad, at, the, chicago, b...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                text            location  \\\n",
       "0   cibc please explain to me why i want to remai...              Canada   \n",
       "1  rt  cibclivelabs  we are pleased to announce  ...     Oshawa, Ontario   \n",
       "2  cibc world markets inc  decreases holdings in ...     The Netherlands   \n",
       "4  your home is a valuable asset  use your equity...  Lower Mainland, BC   \n",
       "5  both  cibc and  bmo ads at the chicago black h...   Northern Virginia   \n",
       "\n",
       "  day_of_week      hour sentiment  number_positive_words  \\\n",
       "0    Thursday  2.783333   neutral                      0   \n",
       "1    Thursday  2.650000  positive                      2   \n",
       "2    Thursday  2.183333   neutral                      0   \n",
       "4    Thursday  1.750000  positive                      1   \n",
       "5    Thursday  1.633333  negative                      0   \n",
       "\n",
       "   number_negative_words                                             tokens  \n",
       "0                      0  [cibc, please, explain, to, me, why, i, want, ...  \n",
       "1                      1  [rt, cibclivelabs, we, are, pleased, to, annou...  \n",
       "2                      0  [cibc, world, market, inc, decrease, holding, ...  \n",
       "4                      0  [your, home, is, a, valuable, asset, use, your...  \n",
       "5                      0  [both, cibc, and, bmo, ad, at, the, chicago, b...  "
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Lemmatize the words\n",
    "\n",
    "lem = WordNetLemmatizer()\n",
    "data['tokens'] = data['tokens'].apply(lambda x: [lem.lemmatize(word) for word in x])\n",
    "\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "39e6c89b",
   "metadata": {},
   "source": [
    "- We can see examples of words being broken down to the root like markets has become market & ads has become ad "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "89462daa",
   "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>text</th>\n",
       "      <th>location</th>\n",
       "      <th>day_of_week</th>\n",
       "      <th>hour</th>\n",
       "      <th>sentiment</th>\n",
       "      <th>number_positive_words</th>\n",
       "      <th>number_negative_words</th>\n",
       "      <th>tokens</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>cibc please explain to me why i want to remai...</td>\n",
       "      <td>Canada</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>2.783333</td>\n",
       "      <td>neutral</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>cibc pleas explain to me whi i want to remain ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>rt  cibclivelabs  we are pleased to announce  ...</td>\n",
       "      <td>Oshawa, Ontario</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>2.650000</td>\n",
       "      <td>positive</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>rt cibclivelab we are pleas to announc cibc in...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>cibc world markets inc  decreases holdings in ...</td>\n",
       "      <td>The Netherlands</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>2.183333</td>\n",
       "      <td>neutral</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>cibc world market inc decreas hold in ing groe...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>your home is a valuable asset  use your equity...</td>\n",
       "      <td>Lower Mainland, BC</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>1.750000</td>\n",
       "      <td>positive</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>your home is a valuabl asset use your equiti t...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>both  cibc and  bmo ads at the chicago black h...</td>\n",
       "      <td>Northern Virginia</td>\n",
       "      <td>Thursday</td>\n",
       "      <td>1.633333</td>\n",
       "      <td>negative</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>both cibc and bmo ad at the chicago black hawk...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                text            location  \\\n",
       "0   cibc please explain to me why i want to remai...              Canada   \n",
       "1  rt  cibclivelabs  we are pleased to announce  ...     Oshawa, Ontario   \n",
       "2  cibc world markets inc  decreases holdings in ...     The Netherlands   \n",
       "4  your home is a valuable asset  use your equity...  Lower Mainland, BC   \n",
       "5  both  cibc and  bmo ads at the chicago black h...   Northern Virginia   \n",
       "\n",
       "  day_of_week      hour sentiment  number_positive_words  \\\n",
       "0    Thursday  2.783333   neutral                      0   \n",
       "1    Thursday  2.650000  positive                      2   \n",
       "2    Thursday  2.183333   neutral                      0   \n",
       "4    Thursday  1.750000  positive                      1   \n",
       "5    Thursday  1.633333  negative                      0   \n",
       "\n",
       "   number_negative_words                                             tokens  \n",
       "0                      0  cibc pleas explain to me whi i want to remain ...  \n",
       "1                      1  rt cibclivelab we are pleas to announc cibc in...  \n",
       "2                      0  cibc world market inc decreas hold in ing groe...  \n",
       "4                      0  your home is a valuabl asset use your equiti t...  \n",
       "5                      0  both cibc and bmo ad at the chicago black hawk...  "
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Stemming the words\n",
    "\n",
    "stem = PorterStemmer()\n",
    "data['tokens'] = data['tokens'].apply(lambda x: [stem.stem(word) for word in x])\n",
    "\n",
    "data['tokens'] = data['tokens'].str.join(' ')\n",
    "\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cac3658a",
   "metadata": {},
   "source": [
    "- We can see examples of words being broken down to root (irrespective of the tense) like please & pleased has become pleas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "5800de0e",
   "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": [
    "# Let's plot a word cloud to see the difference between original tweets & tweets post data cleaning\n",
    "\n",
    "# Original tweets\n",
    "\n",
    "data['text'].to_csv('tweetsonly.csv')  \n",
    "text2 = open('tweetsonly.csv').read()\n",
    "\n",
    "wordcloud = WordCloud().generate(text2)\n",
    "plt.imshow(wordcloud)\n",
    "plt.axis(\"off\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "2ecb3cc6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAAC1CAYAAAD86CzsAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAEAAElEQVR4nOz9d5hm51XmC/92fnOqnDtndVSOtixbthwxtjGYMMyQBpjAHM6kM9c3Hx8T4AwMMISBIQwYBhscwLIlW7JkZbWkVuicqivn8Oaw836+P/bbVV1dVd3VCjbM6NYlqap2TutZz1r3upckhOBdvIt38S7exXcH8vf6BN7Fu3gX7+L/JLxrdN/Fu3gX7+K7iHeN7rt4F+/iXXwX8a7RfRfv4l28i+8i3jW67+JdvIt38V2Eeq2FkiQJAFWTUHUJ2wwQwXfnxN7Fu3gXG4AsQfA2MJAu7+ft2t//4RBCSOstu66nq+oS/bti3PrBHJG48vae2TsASdVQo0kUI4Yky0iyghqJI8kKiiYRy+rIioSsSkiKhBZViKY1uOIWqYaM1FxH0cJ1jIQabqesey/XhSzDpn4VTYNkQmJTv4L8Ns8xersVbjuiY+jh722tMrfdrBOP3fj5/p8A6X+D2yJpKpH+NiT9mr7T9aHKaK0pUGQimzuQI9rbc4I3AK09TWRr13f9uN8LXPdp+Z7ANn3sRoB420dACXj79impGplN+wk8Gy2eoTR8gmiuC1nTkRUNp3yeTbe2MHehTKojSuALJAm0iML06SK1RZu2bUmQJJLtERRVxml4eLaPosm4pk9ppoFZdG7ovGQZujsV5hd9Ng+o3H9vhN//kyqW/bZdOh3tCr/w8yl+4V8XmJkLyGZk/slPpfilXy0xOOS9fQd6K1jrcTf/JkmwijL+9r4eHL5F4/VjLtmcTFu7wtiIi/02PoPrQZIgEpXwXIHrvvX9ybpKfE8fensGa2IBOaqjJqMElotXaSz9bg7N4tesFdvG9/YhBDjTBZR0DK0liVduEN/dh96Wxp7KY08XiO3oRpJl7Kk8Xqn+1k96HaRu3Un8wBbG/+MX37Fj3BBUBTz/ndn19VaQAEWViCZkJDl0D3KtMp/6oTg794Qj4rcebvDENx2iuW4Cz8FpVBCBh6xoCD/84CVFJfAcZEXDdy1kRSPVtZ3K3DAEPmokQeC7+I4JSMiq1vx5GbG4xA/8aJwt21TeeNXhm18zMRvLX6WiGcRaeymNnCDRtQ01kkCLp6nNjpDdcgBFbifVoaNFFNKdMaoLJo2ijR5TiWUNanmb1i1JXMsn1REFAWbFoTheJ9MTw6q4LAxVrnm/NBXef3+UXTtURsd8Hv+OyX13Rdi9U+PsxfBLG+hT+cl/kMR1BV/8aoNcRub9741gGBIvvWLzyuvLRl2S4M7bDA7dpKNp8OIrNq+fcPjcZxKkkhLnL7o884LFa8cdZueWX5KLlzxGx73Vhqy5UzWdQY5EcfOLzSllgKQoiCAASUKsYY0iWoqWxCYSkQ50JYYkyfiBg+VWqVpzlBvT2F511XZ6TCHZqpNqNZgbrhO4AcgSVtWjd0+K4oyJqslE4irFGQvfDYimNTKdEaoLNnpcIT9hEngbt8CJhMTAFpX5WZ+F+YBNWxUymXB6kcnK7NqrkkpLDF7wKJcCFAV6ehUiMYnhSx7eVUZRkmDzVhVVhbFRD9uCllaZnn6F6QmfxYWAVFpiYLNKfjFgetLHiEDfgEohH1DMB3T3KBy5TWdhLuCN1xws8+0ZUQLHJX7TAHpnFmtolsiWTgDsyTxesUb6nr0Uvvna0vpGXytGTwv1sxMEjocSCPS2NA15EgDhekS3dqLlEugdGernJgncvyMD93cBcswgffdeGucmsCcW3vb9X9foyoqEEVNQdXlpShaNSRy6Reeu+yIIIbhw1gXZR1EjRDNdRD0HCQnHqpBoHcCzG7iNMnoih6wozF84iqyG3meqcxueVUeSZWRVI/A81EgcszS7yujeeqfBT/3TJJGIxH0PRDl3yuXc6eWvw7MaBL6LkW5j8dyLyLJK4Ll4Zo3AczHLPsf/ZgwhBLIs4bsBgR/+7DkBCBh6fj4MujS/BxEIfDegNNXAcwM8+9pB7c4OhfvuNvhv/72KaQkapuD1kw5332EQMZorSfDo4yYH9ul89MEoiaTE2fMuFy95VKur99/braDr8OQzFg++L0prTqYlK/OVrze4/94Iu3foHD+1ce87tmsPQcPE6O5CiUbRO7sQQhCYJgiBMz+HMzt9xRYS7antbGm7i5ieRZE1JEkGJAQCIXz8wKVuLzI4+wzFxviK40WTKtnOCJ3bEqQ7DEbeKJHI6kydq9K+JU7bQIzh14pkuiL07k0iKxIzF2t0bovTtT2Basj4Tp7C1Mr3YT0YEfgX/ybFSy/aIGBhPkDXJO69P8J3vh0OJtt2aHgefPCjUf7LL5c5fIvOzj0a83MBYyOrja6uw+59GpmMzMBmlWefsvmJn03w/DM2nT1QLgV88jMxJid84gmJxXmfe94TIR6X2LlH43/8do2WNplNm1Uc20W5KrwkqQpKTCewXSRdQzihkQscb40pwDICy8WaWCAy0I4S0XDmS6jZBFpbCvPCFNbIHJn79q7YRmtN4eYr2JN5AHxDRWp+3ML1sCYXie/tR+/KhetNLG7ovv9dgyYZbIkewhUWrrCZti/RG9lFVE5S8wrM2EP4rJ5yCNdDSUbR2tPfG6Mb+AIJCDyB8Nd/+CLwcc0yaiSGqsdwrRpWaQ7NiOOYFaLpTnzHxGyUEIFH4EnY9SJGPEs000E9P4kWTWIkE7hmFas8v2L/kgQDm1Xi8fBtTWego0vh/Bl36Z2UZAU9kUU1oihGjMKFV4i29dFx8L2ARH1+iuvNV53G2iO6Vd3YfLC7S2Fu3mdqZtnrLJcDbGf5uLNzPtMzPsmkx6c+HmOm+fvM7NrTGc8TTM8GXLjk8Znvk9mzS2dwyGNs3CMIIJO+sQCxhIQcjeAWi8ixOLKu4xYK+GYDNZ1ZFfDMxHrY3vFe4kYO1zeZr1yk4RQRQhDRU6Rj3RhqAiEEpltcdTyr5iErEtW8QzVvk2o1iGc0JAkCLyCW1ki26rRvjiGrEo2SS7LVwLV9EFDN+4BYOwSxBnbt1rh00eOJb1pL6w9e8IhEl69r6JLL0edsenoUdu3VyGRljr3kcO60S7DGuKooEoYhMbBFwXUFji1oNATbd6o89oiF60K9Lti5R+ObXzPJtcjsP6QTiUh0dMn0b1GZnfIYGfY4/rpLvb7yQlL7+zA6MkhyeE+UiIo9X6VyYgy/sfaAKoRASURoefAI9nQB2VARnk9gu5iD0xgDbSQObKL66qUV2zXOT9Lxg/egtWcwL82id2WJbu/GWSwjaSrZ9x3AXahQfe0SLR++Ga09Tf30OPbY+gYocXgb2QcOhLE0AebQNOk792Bemmb69x9Fjupk33+I1B270dIJ7NkChW+8QvW1S6seqpyI0PljDyCrClO/9w2EFxDb1UvLx24nsqkDd75E4fHXqb58HuGt7wTJKOhylOH6cTxhk1RbMOQoI+YJ2vUBkmoLJW921XaSquAVVs/Y3i5c1+jqUYVYSqVW9vCvFeIQAVZlAauy8sEUx08BUF8YW/H3wHepL45TX1z2ihqFyfV3L2Dwgkux4JNIyszO+EyMrZw+J7q2UBo5iVWaJ9m1BcWIUrjwMpIkI75LtIupaZ+2FoXdOzXq9YD5BZ+BfpVsRqG3W0VRwvjrzu0qu3eonDjlkEnLbNusIsuQLwQs5leeq6ZJbNuicuSgzkLe5+Qpl5sP6+zfq6EoUCoH9PcqpJIyvT0qddMlGpHIpGX6elXyhYBiaXmf9XOnVwRRL0fqJF0numkr+D5IMogAWVLIxHpIRFoxnTLHx75Mxbr6RZVIR7vQlCiWu/pltes+F48W1rxfx785txS7nTyzvO1GDexaKJUCMlmZvgGFakVQqQS0tMrEYjLtHTKKAomETHevQiIlUSwEtLYrdHYp1KoBUxM+3lVj7x33GCSSEq+97KBHJFQN/uavG+zao/GpH4zxe79R5dmnLPoHVH76nyb4lV+qMHjeZWbaZ2HeZ2LMJ5OVkWXo6lHIL/gr4rrOYhW/bqNlYshRPfQ4pwv41vqDfdCwWXz4lVWxb2vkCodljbh4YDrM/MmTS0wFc3Ca8rNnmkuHltZTdJnCl57FtQMIAmRFIgjE2n6LIhPd2s3MHz9G5j37ifS1M/0H36T3Fz6B3t2CO1fCnlhk9uzjeKUaqTt2k3ngINb4PO58OdyHCFBzCVo+ehvC85n+n99GuD56d47cQ7dQPz7M7P98nOjWLnIfvBm/ZlI/MbLu/REIPGHjijCebcgxUkobvcYufFycoLF6IwmMgXbkmIFs3VjuZqO4rtHVdIlUi4oRlZG/x+SF1192+M3/XKGnT+HE6w4Toyu/DDM/RaytHyPVilVewKld9rrWNrhv5cNeD7PzPo89aXHrYZ3pWZ9aXbB9q8bIqEtXh8L4pMdzL1rs2qHh+/CtJ0yyGZk7bjW4+aDB6XPOKqMrBMSiEls3qzz5jMW5Cy6ppMRNe3XOX3QZHnO59YjBmXMuPV0K+UJAW6vM4LBLR5tCe5u8wugu7fQqCMehcfHcVfdIRlOiAFhuGdMtr3HVgrI5vcbfN4g1nsG1nku6VaV/Z4yRM3VqpWVPoGdrhI4Bg7MvVRkZctm3X2d0xMOyBAObVE4ed+gdUJmZ8pme9OntUzl/1mV81EcI2H9IZ+9+nYV5C++q+PGFsy65VhlVlRgd8tB1iZ27Q2/96HM2qgabNmuk0hKPPWJRKgZcPO+ybadGe6fM9KRPtWnQu7sVBs9LuO7yMczx/DVvUbpFZc/tSU69UKVWumpEuNY7fK1l10mMZ3tjxDI6vhuG3pCgONmgsVYiWYBXaVB7Y4jYnn7c+RLW8Ax+1URNRXGm8jgzBSID7UQ2daCmYsiGhhKL4FIGBHJEp/Xjd4Ass/DFZwkaYSjI6G9D78jgTOdJ3bYTOaIjxyNE+tuvaXSvRsXLU/EWKHsLeMLBEdbqlQTYY/NE+tpREtEN7/tGcF2ja9Z9inMOuU79bc0kvxlYluBv/mqN0akJz6xRGT8LhNzieFJClmU6+3Vmxx3MeoCmS5j1gGhc5uBdcV57toZuyCBBZ5/G/JRLpfDms5aBpvPsqz7ilWYCUdV4/EIa+/E5kGXkqMHZRR13rgRCIOkq9UWJLz3qIXwfSdeQIoAfINzwPFxX8Npxh688vHztf/voyvjmE09ZPPHU8ks0PAovv/rWR+og8LHcCkIEJCJtdKX3MlU8gS/ehvT7m0Rbj8FdH8lRnHNWGN2tN8W59cEMI6cbfPPrFrLMUqjglaMOrxxdvh8zU/6KQXdizGdy3FzX2E9O+HzlC40VoYfvPG6F4QD/8jFsZJmlGeGFcx4XL3gglo/z6svOmxrsW7p0PvzjnUxetFYb3XcIvhuQaIuQ641h18NZpVX1aJScNW1BYDdDfYEgsD1EIMLErCwT3dpJ7kM348yVcOfLBGbzWTST80gSWkcWJAl7YjHcrgklYiCpStPwSwS2S/nZUzTOTVzz/D3hMG0PLv1uB3Vm7CFUSSfAQyxdxMrpgJKMYk8toqZjN37TNoDrsxckKC24TA9b2Obfj8oIVZM4cGeciSGbTIvKzgNRujdp1CsBI+ctzHpAPKmQSCscvDPO+KBNPKU019M5+ngV135zI0zy5u1IhoZfbqBm4lhj8xg9LeidGbSWFF6pjles4S6UkQ2d6LZuJFlCjumAhJuvYnTnsIZnscYXEELwnWfXGJG/SxAElBpTVMxZUtEutnbcQ0d6F9OlU8xXBnH99QfBdwoTF02+/NvTlPPXNvxrxWavxNWG73qG8Or9CQHCX/n71SG4taJaf1/UVMszJnbVY/w1aYleade8DTpfK1eK7uhF0lRKT53EXawQ29NH/MDmFau7cyVm/+wJ2j55F5n7D5D/xivgBwSmjVeuU37hLPb0FTOC63jqAT5lLwx3hjz9CDWvjCLrBIGPkASqkSCW7aG+OIbvWs11ZSRFxhqd28iF3jA2xNM1Ygqbdsc49ngRq/H3w/BG4zItHRptXSqaLpHOqfieRyqnUlr0sBo+c5MuqaxCrlMj26JiRGRkOXzBbgiKTNfPfxx7dA53No8a0dG7cgS2izNTQGtNISkybr6KO1tEbU2FuSpJQolHkBQZJRHBq5o4s0Xie/uRDI3LL26p/PZ8pZKq0PdLP0rp269TefrEimUdP/VhnJkCxa8fXbVd1ZrjwuyT7Oi8n2SknWy8n0yshy3tdzNXPs9E4bUlb/itQlFB1cJQlhDguwL3iiRkJC4jSRK1kofnrr4vgnDQjTQTrp4jVq0nK6Dpy8cIvDAxdj1IEugRGccKECI8F98TuLZAUSVkhaXBWtEkNF1CAnxf4FhXnYMMmhGeQ+CD6wRLHvPlY2m6hKJKBCIMsb+T6Nx5D937HsCuLnLxuT/FNasEngi92rcBgWmjpmNobWm0XILWj9+BdFWFUGC7WEMzLPzti3T/9Idw50pUXj6POTRDquHQ8uFbKD59EkmS0NrSmMOzOJMhs0KLpthx7z8gkmzj0gt/QXnmwop9J9o3E3guyfYtSJKEUy+hRuLUFieIZTqRFZXy9AVE4OHbDkbMILazF2dmdWL4reL6nq4sEU8pdG6KoOoSXMOxkSSIxyUyOZlIVEJVpZCA7QgatYBCPtgwKVw3QlrPevB9GL7k4q7xTniu4PlHKyDBuatmEJlWle37oxTmPE69VEdcTcx/k4R8WdeQVIXqsYurRuDKiyvjpAzNACA8i8rRq2KoEY3q60PY79AoKxs6krr6C/bKNfzy+uT3Yn2c10b+ks7MXrozNxHVM0TUJJvbbqcvd4ixxVeYKL6BvUYibaOIpxXu+XgLNz+QIZVTcayA489W+PofzmKbAbIC/+S/bqFnSwSz7vM//p8xRs6sfCE1XeahH+9g6/44miHzxlNlvvX5OSqFcEpuRGVu/UCGuz7WQrZdw/MEQyfqfOHXJjFr1x402vsMfvzf9/MXvzJBcd7lP311Dxdfr/K7//codzyUZeuBOF/+b9OomsT7f6idA/ekUDWJuXGHv/39GUbPhueqGRI3P5Dh/s+0kcqpVAoez38tz0vfLGKbQcjU2R3jYz/ZSc/2CIVZlwuvVW/IQ5ZkFUXV8VxzQ661JCsoqoGsGqwoz9wggoaFM5UHIXDzFfxKGP6yJxcJTJvqq4OoLSk6PvdevEqDytFzKPEIgR0aBK9cx5kJk63W4DQLX3iG7INHaFycxF2sMPf5J8h94DCdP/o+hBfQODd+VXhBQlZ1FM1o0hlXwrdNrOoCqhZFEKDHMrhmFUXVaJRmcc0KkiwjAvArDfy6hTW0mtnwduD69YMCFqccIrHGmp4FhDFHw4D9h3XufV+EI7cZdPUoJJMyvi8o5ANGhjyOPmvx/FM2IxuokOruVfjiI+3rLi8XfX7oYwtMjF0j/ipW/1xa9Cgtrjy+WGO97xWE5b5jBvdayP/VM9ddxwscJgtvMFM6Qy4+QGtyC5lYH4lIK5vb7yRmZLkw8501CySuB1WTuO2DWW77YJanvrTIpRN1IvFw5nE5rBX48Gs/c4mb7krx4R/vWHM/PVsjXDpe4w//3Rgd/QYf+rEOZkYsXvhGHiFgz21JPvDD7bzwjQKnX6ygqBLJjHpdgwvgWAGVvEemTSeV1Zgbt+nbEUMzJLIdOqV5F88RPPgj7ey9Lclf/+Y0lYLLnR/O8Y9+qZ9f/7khSvMue25L8omf7uKbn59n8HiNLXtjvPfTbZTzHieeLRNLKrzn+1vwPcEf/j9jKJrEJ3/2xkpkk+1byPXuZer0t3Gt2g1t+2bQODtO42zIRCo8cmzp79O//fWlnxe//DyLX35+ze3Lz52h/NyZpd+rr10K6WRNuHMl5v78O2/+/IpT4XFmzl+1ZA0vyxc0To/xTuG6RjeWUhjYHbvMIFoTji146BMxfvQnE2zaqqKs0CeQ6InJ9PSp3HaXwW132/yP/1bh5OvXdnldB8ZHPeIJiXhcxoiwROBeD8m79+EVqpjnxtG7W0jdvZfKc6fDrOd7DuAtlGicGUNJx0kc2YHR30bQsKmfGMYcnFz2UFWF2K4+Yvs2I2kK9ugc1ZfOIa6oytF7W0nesRc5omENTjXDATcASSK6vYf4oW3IiSiSJCGEwBqcpPLsKaK7+jD62ik/exJhuyBLJG7eAZJE/bVBJFUhedde3HwFo6cVtTWNt1CicvQsfnH5I4ts7SJxyy6QJcxzY6u83PjhbcQPbAVZonFqhNorF64+01XwA4eF6iD52gjJSDv9LTfTldlHS2ILrclRpoonrruPqxFLKey5LcnJ5yscfaSwIqRwJYQA1w5C6tIaKC+6PPPVPAtTDrNjFlv3x9m6P8Zr3ynhOgEH7k0xOWjx9JcWbzhU5lgB5UWXbLtGMqNw9uUKR96XoW97lEyrytCpBr4vuOsjOb79lwuceiGsXlycmuWOh7LsuyPJCw8XOHhPmslLFs8/nMdzBPPjNjsOJzh4X5ozRyskMio9W6M8/pfzXDpRR5Lg6a8s8uEf79zQeUqKRrpzGy0Dh5g5/yzwzhvdv7/47ntZ140UuXZAo+qTzKgo2tpG76aDOj//iym27tAo5gMe/4bJH/9uld//zQrf+GqD+dmQkqNpEnfea/DD/yhBe8e1Dz0/6/MLP5Xn534sz0/84AI/+YN5/ut/WouutAyjt5X4gS1ImkJ0Zy/pB45gDITecub+Q8ixCHIySssn7yZxy07cxQpoKi2fuY/YvuWgfnz/Flo/+14kVcavNEi//wgtn753qWhAySTo+ImHMAba8QpVYvu3ENnUyY08QK0jS9uPvZ/AcrCHZ4gf2IIc0TAHwxHZ6G8nfmQ7ktYcFyWJ6K5+orv7QZWRDI30+w7R+un7UDMJ3IUSyTv30vL990DTsKotKTp/+iNo7Wn8aoPUXftQW1IrzsMeX6D22iDRnX1Ed/Vv+PwBAuFRNqcZnHuaQHhoSpSYnr2hfSzdD10mnlLIzzjrGtyNoFJYjvW6tqBScEnlNFRNQlEksu06hTnnTeUmHCuguOCSaVUZ2B1j7ILJxEWTnTcnUDWZ0oKDHpFJZFTmJ5fLqOsVn0rBo73XQNFCCubClL3kxLiOoLTg0tqlISuhop8WkaiXw1mcaM42NwrNiBNNdfzvoerzvyGu6+k6VoDV8FE0iX13pDj2+OrA8vsfioIET3yzwe/91ypTEz6+H5KoZUWit1/h3/5yhsO36qiqxC23G+w/ovPEo+tn5V0XBs9fxcO9DnvCmc4TP7QNSVNR20Jen5pLIccMtI4M9tgc0W09RLZ0Mf/5b2MNzyAbOtmP3Ebqrr3YI7P4DYvcx27HPD/B4l89gxCCxtkxen7x09SOXcAanCJ5607UbIKpX/sSgWVj9LU3FZI2/pLHD25FkmWKj76M8AP03pCL6M6sXUSwFmRDo3F6lPzfPE9gu3jzJVp+4D1oLWncuSLp94YVQrP/41GE5xPd0Uts36YV+/AWy3iLZfzy7eseR0JqDidrG0PfdwhLggMC8ebodiIQEITJpbcCzZCXkk6SHBpzxwrLvS8n5vTImzuG6wgWp2323Z6ipUtn6pJJe4/BziMJSosu1ZIfHisQRGLLpHZJgkhMwWoEYeLNEUTiygqbqEdkbCs8x8AX+K5AVZdXuL7CnxTOBCUZI54jmg69YllRkZSrPvMg2EDSU2omui6fg0AIsf5098otJblp8K+4QHHtY4bbNIuYRKj/cbnUfPn4wVuifkiSvFQxJ4T/ttBItu5QGbp4YxS+DZQBw/lXqpx/Ze04nSRJKKrglRcc/vP/p8zC3NU3VnDpgsev/n/L/OEXWklnQsGcHbs0nn3SwrkBlafr3SNnKk/mAzcjaSp6Vw7z3DhKMkZkey9Bw8ZdKBM/uBW/ZuLlKyEdpWFhj84R29WHkowiPB+jr53io68shROsoRkC0yG6rQdrcAq9txV7ZI6gYUEgcGcKeMUbU2CSFDm8HlmGQIR8T/fGDFZQt7HH55dI5M5MAVlXkWOhyIOxuQvz4iSiWVljj83hN25MVktTogy03orllCmZU3i+FRpWEb7EqmKwpe0uZEnBcitUrTcXjzZrPhOXLHYcinP6aIXinIusgCxLmDV/xbOX5NDArCWP2dFv0LUpQr3ik27V6NsRZfyCiW2F7IBLJ+vsvztF/64oc2M2SGE8uVH1N6QVXSl4ZNo16hUfqx4wer7Bgz/SzqtPlKgVPXxXcP5YjX13JDn5XBnPFWy5KUYkLnPx9RoigKGTde75eAvZDo1K3iPbodGzNcrplyr4nsBqBNQrPn07owyeqCErEntuTXKtzt3Z3n3k+m4ilusmkmxFkkIjfdNDv7hq3YXhY4y+8pXQ8KyCQNEipDu307rlZuKZbmRVx2mUKU2fY2HoFRqlWdYagLVoimTbJjI9e0i09KPHMkiSjGtVqRcmyY++Tnl2cImadSX6D3+M1i03M3Xq2+RHX6dl4CAtmw6H1yIr2LU8hYlTLAwfw6nfOKNANeL0H/oIrZsOY1UXmTjxKMXJMIacyYUv0mUefyQiUa2EP2t6qApnNgTJVMjnr1YCIhEJTZP4vs/E+a3/t0wyJePYgnpNXNdObUDwBhJplWROZW7MXjOZZjYEX/x8bQ2Du4wLZ13On3G47S4DSZLo6lGIJ2Sc6wjI3Aic2ULo1eaSqJkEladPEj+0lei2bqzR2aZkmozwghUv8GVlLWQZSVVC/uUKWTdB4HpIWvgiS5pKcEWdqAiCDXkAV6L+xiWSd+6h5fvuxl0so/e0Unz89XXXv8wdXHHevo9wroiNL4lQhN6BrKt4zhXn6Qc3HMKSJJlsrJdc+924vknDKeG4NQIRoCoGCaMVXY3jBRaz5XPka6M3doAmzHrAy98s8H0/2833/1wXs2M2qiZRWnB5+it5HCugo99g24E4A7tjtHTpHLk/Q+emCBdeq7E4HQ4s+RmH9322jV03J2jp0okmFE69UF6icr36RImtN8X4vn/cxdRQ+PH7nuBbfzaHWb/+M6wVPSJxhZlhC8cOWJyyicTChHG9EhYQfOOP5/j0P+3mM7/QQ63sse1Agme+mmfoZDgwv/J4ka374/zgL/YyPWzR1qNj1nxee6KE7wmqBY9jj5e495MtZNs0bCugZ1sEx1r//IxEFsWI4jRKBJ5DLNuDBFQXxwj8laEJszR7RWHASiiqQeeOu8j178dzTMzKArKiocczdO68h2T7FoaP/hWN0tUViBLtW2+j56b347s2rlmhUZwBCTQjQbZnD+muncyce4rp00+u6fUqikaydYB4rod0107cRgWzPIeiGRjxLL03fYBU+xaGjn4Bp3HtUOOV0KIpevY9QEv/AazqIpMnH1syuLkWmYe+L8bslMeFsy5dPSqHbtE5d9ol16KgquD5gmMv2hy+1aC9U+HEaw65VhlNl9AjErmczPs+FMMyBd95zKRUvPZ7dF2jq6gSsZRCS5fB4pSzyugKIVhc8Hnp+et7UJcueNx2V+iFJRIy2tuslRzULbzFCrGDW/GKNeyJeRK37cLY0ol5bgICgV9uoMQN5IiODyDLaNkkQcMmsBwCy8Gv1NE7c0uaBEo6jpqMhVVkEMZxb9rcNG4CJRlb8i43CnchrMxRW1P4lTqFb7xM48zo8rW4PrKhLSUP5ZiBkk7gla6YcQiuaUTd+TJad27pdzWbuGHBaz9wmCmfxRceMT1HTM+SjHSE/NPAw/HqLFQHWaheYq58Hj9487zO4TMNvvBrk2w7GCeZVamVAsbOm3hu+BJH4jItXTq1ksdzXwtJ8i2d+hInd+RMg/nfmEIEsGlPjGqxweCJOuMXlqv35idtvvjrU+w8kiDdpuE7gukRa0M8XYD8rMuTX1ygOO+EfN0AvvrfZ1mYcXEdQbY3ysiZkIK2/VACzZB47M/nOfVCeWlMLC96fOHXJtl/d5pUTmX4rMW5Y9WlgcNzBa8+WaRW9ujaHKFW8njh6wV6t0UoF9ZOQM9feonFkVC+MdG2ia23/wAgM/7G17FrK0NWge+t6yTosTSZ7l3kR49TnDyNY1ZQNINE22Y6d9xJPNtD5+57GT56te6toDR9Hj2WppafwKouNFkTEpFElrYtt5Dp2UvnznvJj57Aqs6vOrYkK6S7d+LUS8yef5bK7CU8p4FqxMh076Z92+0k27bQvu12Jk8+tqHnpUWSdO95Ly0Dh7FqBcaPf4Py9DKDYfsujYlRj2eesMhkZTq7YfC8y47dGqViwEvPWxw8YrB5W5iv8jzBwZt1Bs+7vPScxb4DOq0dClMTHu1dCsmUROk6jvh1v0ARhPEkz10/Yzx43qVRv/5L27iClqPqErL89gf6rdFZkrfsovLCabxSDYQgsrmL0iOvhMtHZgicm8h84Gaqz59CbUmTvHMPtdcu4pdqCD+g8syJkBkwX8KrNMi8/wjuQonG2VEA6seHSL/3INkP3Urj7Bjx/ZsxultonBze8HlKukZkcyez/+MRzLOr6SnuXBElFSN55x7MCxPE9gwQ3dFD9ZWrKS/ro3r0DN3//JNkHjgc3pfbdqOm41ediIRkhBxjSVObsoLLH7YfuEwXT5KvDqOpURRZR25OXYUI8AIHx6s3hW7eWoxMBDA1ZC15oFdj7JzJ2Ln15R0nBpeXXXhtnYy9gIUph4WpjcfOr4RZ83n5W0WiaY1sT4zCpMlLj5XQY+GntOW2Vi69sMDclMvY+QWiKY1YRlvFJzcbgloQ4dS3CkSSGpGkQbLNo7Jgo6gSekLn1NEqx5+t0NIfo170mR0v038wS7o3YOp0GVmTiSRUhC+oFx0CLzyI75hN/rnAsxt49o2FvqoLY0ydeRLPXr6HZnkOI5ahY8ddZLt3IytqaLyvQL04xcTxR/GcldxpqzKH7zlE051EUu0k2wbWNLoQzqzyo28we+G5peuhCmZ5nli2m0z3LtJdO5k6/QQiWCM8IpY9EVk16N5zP61bbsGuFRh/42EqsysV18aGPb7/c3F0Iwwp7LlJvxxSJgjAscJddvUodPeqFPM+pWJArlXhgQ/FkOWwnqBvQKVeX2/+sBLXj+kGEE+ptPXqDJ+s466x2+mJjU2tr7TZ71Re1RqeJX3/IayhGfAD3PkSkq5hNXUx3dki+b9+huzH7qD7//o0Qd2i8vxpyk+dWAoplJ48DpJM6w++F1nXMIemmfmdr+FXw4/aGppm4QtP0fKJu8h84Ai1YxdonJ+4oZisCAR+zaT7X3wK4XoEloN5Zoz5v3gCYTmYFycpfft10u87RO6jt1M/OUL9jUsIy71C6zdYFecLu3uEf2ucG2fxK8+R/cjtiEBQee4k1vDM0oPQu1po/dz9RLf1oKRiRLZ2k7xlJ7VXL7L4pWeWiiUC4WOuK3bzvydkVQdJRtH0kHSvGjQKIbMk3RVh880tVBdtCpMNYhmdXF+MyrxFJKHSsy9DLKMx9NIifQeymCWXwuRKQ6QaCp07U4yfKKLqMq0Dcbbc2sKxvx6n/2CWeItOfrzB3GCVZFuEvR/o4oU/GybbE8Usu0gy9O5L0707xdkn5+BtKpzynAal6XMrDC6EqoBmeQ7ftVE0A9VI4DRKKzcWwSqDexl2LY9jVoik2tGiqTXXAbCqi5TnLi0b3CZ816RRmiHduQNVj6LoUbw1+MeB7y2FC/sOfJD2rbdj1RYZf/3rVOaHVq0/N+vz538UxtodR3D2pBtqS4fCatiW4MlvmWEyNCLhOALfB0UJ0zFPfDNM+htGKGC0EWH668d0ZTBiMh3961ek1TcQC/tuofbSWWqvnAuTUxKUvv4ihYdfXLb4QmANzzDzW3+zzIu+ynAFNZPC374Qbsca6wSC6nOnqD5/urmPpgQTYmPOniLT/c8/iXVpmvn/+RjC81DiUdr/0QfJfOAIxYePgudT+tYxSo+9egV/+/I1gG+7TP7Hv1xxXs50ntF/8fvLf/MDSt9s7qN57YW/fXFpuTOTZ/rXv7x6BFzjnrydUHSZWFanUXTwnfDdUbXwxa3X/g4IE0gyrTtux7Nq+J6D8N0Vwgup9giFyQYLQ7WQEeEJJFlC0cIwx/SZEn0HsrT0J6jMWcwNVlfN5htFh3rJoTJnEU3pzA5WMcsukZRKx44kdt0jcAM6tiVo35Yk1RbBswPqBYfynEXgCwJfMHGyRGl6Y+LuG4HvWpjltZOhnmsiAhcwwkFpTUjIioqsaEiy0mQhSKh6bClUJl1DrtBplNeN13pWrfleysjK2rHJwHdBkui96QN0bL8Lu15g/I1vUJkbXHN9IaBaDkvLg8DFsSVAXpFkvGxIr+xSczWutexqbMDTFcwMWyQz6rq5omvq7H4vEAiQIJGSae/WmBpzsK4eLMQGDOT1esKt2MfGb7qSjBHZ0kXp0ZdxphcBiSDl4tesVfXo1zzPtQzjWn9bcR1XLd/IfXibkemJcfCTm3j1C8OUp8MHs3m7zi33RPnLP3jnvWlZUtGVGI5fX6K4KZJOXM9ieVXcwMI1q7j1IpKs4JpV1GhyafuF4Rpbb28l2Wpw4Zl5ommNeItOLKNRmGjQuz+L7wkmT5fYdCTH1ttbOffkLP5VcpGVWYuuXWkaJYeQyCJh1zwuPDtP60CcyoJNokWnMm8ttSqqFx1a+mLkx+qYFffGdUKugyDww9LhtXCdQyl6lERLP6n2rcRyvRjxNKoWQ1JUJFlBvpq6ttbxPWdV4u9GIISgddNhMj17QQKrlg8TeteApkVIJXux7bAUWEKmXJngnfowNqS9kGnXSOZUOjdFljKwf9chSZDJKew8EKHQFLi5GqmkRKW6fGMP3qQzNe2xkH9nPXe/0qB69CzpBw4TP7gNZBklHsEvVqm+dO76O6B5fT1xeg7miLdE8B2fqVNFZk4vzzONhMqm29vJdIcSdfmRKmOvLuKaPpIi0bIpQe/BFiIpDbfhMf56noXBsIqqZVOCviOtRFIa5ekG468uUs+HFKubPtrPzJkii0NhUq9rb5bcQJwzj07SviNFNK2jGspSv7nJEwUWBsuIAPY+1Ev3TTk23drWlAp0mTqeh3p43K5eldZOhUoxYNM2DVmB/ILP+ZMO+w4bJFMylXKAWQ9YmPM5eGuEs8dtIjGJwqJPuXD9Z2cocdriW5mrXcD2w/dZkTW6ErvJm2MsNkYpjryxcqMrOplYVY8z316uy588WWLyZAmA4uRKg3X+qfUpdMe/PrWi0jPfTN1W5m2mzpRBQH6svqJSdep0mekzYVJuYegdqDQTrB0rvQ4ULULvTQ/SMnAQSZZpFKepzo/iOWELLUlWyPXtJ5Jsvfbhr8PnvR5UPUqu/wCN4hR6PEuipZ+OHXcyc/6ZVe2/liERiWSJRlswzXxTQVK6JkXvrWBDqexILKRRNWp/11za9SFL0NKp0tqxkoR+GfG4xCc+EuPzX1geREpXtdW5UbR06bR060wPLWuetnTpdPQbWI0Axwq5f3pUJpU/iWVnCRQdVZEYe6lEbWQx5A9vBJJEoj1KbiBBPW+TaDV43y/s5ZFfeoPieB0tqnDPz+4m0xNn4vVFCCCWM5ZayLdvT3HPz+ymMtOgMF4jktKJpsMpY8vmBHf/9C6qcxbVeZNNt7XTti3Fsf81hFV1OfSpTfhusGR0ew7m2H5vJ2cenaRzT4bDn97MzJkS+ZEqLZuT9BzI8fwfnKc4XqdesDHLDiIQVOfCxqBWxUVToLVD4dZ7o4wMOvRu0rj13ijfeaTO3oMREkmZXTcZHHvepG+zxt5DBt/6ao0774+RSIeE9+e/vb4aU1diN4aawA9cytY0ETVOT2ofAJOVUzh+nZqz+KaLOzb0yDQVkJAUhcB1UaIRfMsOqYgS4PlIuk7QaKypG7L069+BCMzVaN18mPZtt+M5daZPP0Vp6iyuVcX3HBABaiRBPNd7XaP7ViHLCnOXXmL+0kvEcz0MHP4YbVtvw64XWBx5bc0BxfcdCsVBfN/F920UxXjHDC5sUE/XMQNKCy6F2XemfcU7gUDA7KRLe9fqsEhbq8w/+KEED74vQkebwiOPmUSjEj/9D5P83h9WGJ/0+Yc/nCASCQ2Urkn8+u+U+ckfS9LRrvD4UyZPP2txdYPUSFyhe0uUmeHl7HvPtiiJjEq14GLEZNp7jbDqSngkg3l0TcYqB4yNlfCuow97JUQgmDlTYH6wjGf7qLrMptva6dyVoThep/9wK5tvb+dv/+UxihO1sJhBlnCb7eT3fqgXs+Lw/B9ewK66yOqypOX293ThOQGv/MUlzLJD5+4Md/3ETlo2J5g6ef2MjSRLHP+bURYvVWnZkuSun9xJpidOcbzO6MsLBJ6g90COwadnKE03kICWAwa33hNlftpj8IzD/lsijA+7nHnDZssOnT0HDabHPU6/bmNEJe7MRtm8Q+fEMYsde3Wmx12K+fUNZt4cRZE0WqKbSBrtoRfZGENTonQn9zBaenXD9/5NQZYxBvrRu7uxJ6dQYlG8coX41s2hluz8Akghl9q6cPFtOODlPMN3B7ne/UiyQqM4w8Kll/C9lRRSRYugKOvFgd8+CKA8c4FGcQqruogezdC973307vsAZmWe2sLoqm1836ZxRdGQ572z+tUbUhlz3YBYUkFRv3sP8a1CUWDzDoNYQlnRFgVgYTHgm982SSUlfuP3KstK/4MuiiqhquB6AqssuDTs8n0fiXHrEQPTEvz3P67ywz8Q58Qph/mFq9vqCOpNbmW1KToT+IJ4WqUwa9PWGwHAsQN8NyyrbJR9ahUPs35jHpYkQ/uONAc/uYm2bSnUiEKyLYLWpC61bElSng692MvJqqVtFYmWTUlGX1mgUQhfNr9p7xVdJtMdJz9cxSw5S52QHdMj1Rlj5nRp9blc9Xt+tEZtPmylblddfDdAvVzeK5phZLHM7hGESnXPPhaKu+y8yUCWQtGjyyHnuWmPtk6FVEYmkZC5eMZh136Dh/+ywrbdOq6zvmi5Iun0p48gSWDIcarOAk5gYnk1LK9Ge3zbDd37jULf0k/uh74Prbud/J99BUU1QAjkWBQUBSWZwCuV8as1lFQSv1RG7+7EuobmkNbTSfazH0dJJVj43T/Fm1/d5kcEAcL3UKNJFC3yjlzb1VCMsLWN7zmhd7sCEvFsD0Yit3rDtx1iyZsNPJv5S0cxkjlaBw6x9fbPcu7JP8BpvHWqh6zJRLIRAk+gqDKBHxD4AXbp+vUK1+fpirAO3G4EN1p09T2FokphVlniKtWzEJcLzq5sQBgELE3lXDfUeiiVw7K+vh6Vu283SCVlFhYDzDWoIdNDFjPD1orpnxaRGTpRJZFRee5vbqSVtdTUPAiajTUFsqQgSTJ+4BDNGLz3n+1l+MV5nvrNM1g1lx/7s3uXr8ULUDR5bWU2EQ4Gir62BoEQAqXZwghC/QxZlnBNDyEEImAFxzqaWenBeLbfpK4tH29lHb5YFfIx6wHDF11OvGLxng/GGbrgsDgXdjtenPW4dN7h0O0RPvVjKeZnfB79co2PfTbJ3IzHhdM2+bn1B624lkECZqsXaY1vAiQ0OUpcz6HJEWrOAqpsoKsxBAJDiS3FeldAkpCTCQgCgnrj+vN8WQJVAVVFeB7Vl6+qOFyjb489MnrNXeqb+9B6O5FjUSJ7dlCbXy067zsmdqOEHkvTtuVmZi9YV3RFUAh89xrxzTcHszRLLBOWICfbNtMozyCCAEXViWW7ad92G2okfv0dvc1wrSqz55/FiGVItm1h8+2fYuToX+OYKxO2sqojLhtsIZZV49fRe4i3x9n84BZ8N6A6WaUyXkZSpLfH6EqShGMHNK6qf/9uQ9VYmu6HJwZGJPRKr+7cCmH5sqRISDJrNtQsFn1yWYX33G2EjR5TMlsGVGo1HUNfaRGEgFeP27S2yJw47bCwGGCtU8F09T068UyJRFpl5PSNtbWJaCniRg4/cNHVGJ5vNw2uR8WcRlYljITK4lCFwA8YuKWVRPtyI72ZsyUOfnIT2+7tZOL1RUQg0GIqtXmLwA+YPl2k90COjl1pytON0EMWUJ0zmT1fYvu9nbRvS1GeadBzIIckS1RmTEQgqC9atG1PEW8xiKR0eg/kbijP6zkBqq6Q7IhiVV0CN2By1GOy2Wj0S3+6Mq799LfCezc5ujJx9Ce/VQLgkb++dkKp7hapu0USeo6GU8TxTVTZIKIm0FJpCskKMa2TQJaRfYWImlrT6MqxKOmPvA9vbpHacy+vLMF+M3gTH5RfqhDUGhAEeHNrD+J2o0R5+jyxdAft2+8kmurAbpRCnRTNoDR9noXhY29rcHh+6GVSHVuJpjsYOPJxavlxAt9Bj6aJ53oxy3PUFkZJtG2+/s7eZjSK00yf+Q4DR5Kk2rbSve8BJk89tszzlWQirZ0gKyi6gVPKo2daCVybwHOxFmcQVxWCCCB/Po9VsnAqNpIsITYoa3Jdo+u5guKci2MGYSvm7wLuuMdg30GdWEwiGpOIRCUMQ6Klbdl6RmMy//zfpKmUAhxbYJrL/5563eHlF2ymRhzaOtU1I1sLiwFf/GodxxF4fji9/cZjDUxTkC/4PPmMie9DuRzwZ1+ocf6ii6pIxGMhQfoyrtdkMPBZ6lpwI9CUCHGjFUEASBiqwHIryEro7To1j+GjCxz61GZ23t9NvWAze7a0tP3suRLH/nKIfR/uY88He/A9wcJgmTe+NIpZdjj72CTxnM7dP7OLwA3wnIDz356mOmdy6elZkm1Rbv/xHUhSyEO9+NQ0+bGQRH7y4XFu/sEtfOBfH8CqOuRHa+QGEhu+tuqcycJQhTv+4XbMosOZb00y8uLaFUpvB3zhMl09s+ayxIFbIIjiaQHThVHc+fW7BcjJOLFbDtB45Thrqu18F2APjVH84sNIqoJ9aWTNdQLPZmH4FYTwyfbeRKJ1gJSylcBzcawKpZm3I2a8EtWFEcZe+xqtW24h0dJHW+ZWhO9h1xYpjJ9kYeQ10l3biWV73vZjbwTluUEmTz7G5ts+TcvAAexGkbkLzzeLMMJKski6FVnTUYwYkdYurMI8BAFOpYhvrhzYa1NValNvrkvKhtgLtZL3XetACvCeByJ8/DMxFFVCUcL4bChdt7yOrkvce3+kGRcU+EHIF/Z9wVe/0ODlF8NySrdZQXI1AgHPvbg8NOULMDruoxkygS+IJuRmLkLm9dMeqqHw+gkHWYH+3THC8JREe7/B5MUGmi6jGTJ6VGJq8K0H4ut2Hsttenwi5B0HgY8kKfiBh98QvPSnF0m2RkIB8qKNLEt4zYHRdwJOf2OCsWOLYYmqEJgVF6sWemfl6QYv/OEF4i0RZE0mcANqC+F51ws2x/7XJRJtERRdwWl41BYsPCu8kSNH58iPVNGiCq7lY1VcIsmQrD741CyjLy9gVsLjVBdMnvmds1iV5ThfPW/x/O+fJ5YzQILa/Peu8aaSTuOVSqgdrQT2NVwVCfT+HuT4O9MhdqMQpoV15vpC806jzOyF58mPHkfWQpEpEQSIwAs1Ea7yFBZGXqU8e5HA99as9AKozA1y/qk/QpLlVXoOCEFh8jTVxTFUPYokKQgEgefg2bVQBMeqUpkbWnP/M+eeYmH4GL5jrqlCBrA4+jqVuSGE8HHNlbMhz6oy+NznkRUNu7ZGO3shKE6exqzMI8lKk8rmLS2zFmdwSnmECJAkhdrkJQLPDTVGbkQKcQO4MfWTJkQQVmlUK+EH7myQZuU4glo1zJ6Y5uoS1qX9s2xAV/51g+cnQkMd+ILXX2hQKW48SbX/nhR2IyA/49C/K4bvCYpzDnvvSvPwf5+md3uUbfsTGFGZTLtGYc6FQITi05qM6wTAWzcigfBW1baHF7c8pbXKLlZ5/Smu7waUJtfhVQswyy7mOtvbNQ+7tvbH57uC4sTK/ZrNBoZm2eHKcFngiqUCiKVDB6Fhrxfe3pcZYNf7u3ngX+zliz//0hKl7VqovXCM6ME9+OUK7tRVJHoJUJSQMK+qRG/ahSSFlC85YrBKxNR1wb/ObFBVkBRlWWA8CBC+f83tJF1b07MWjrs6eyhd/k9o8GyvEFaFKfJVynnhapKmIYIAz6otG0NZRjL05v6b350kEUg+lt1MQmkyeMrKyighcN0GnrBoegmhPGrzO/edBuY6ZcKXK9FkSUVBC5kcBARi+Rvw7Qa2E2o3y0j4VxCYhQhwqnkkwtmwLCmr6H9CBJjltWcywnPxvbcYLtog3pTRnZ70+YWfCh/m0vy6+VJIYeYKWdPwTXPFS/FHv1Pjj37n+oTuX/n3ZX7l37+1yiRFDTsC7z5o8NJ3alRKGwuNzIxYdG6KEEspNKoe0bhCR79BaS40Ko4V6p0W5h1GzzZo6dJRNIloQmF6yKKtd21aTBiD1XDqHr779ygj+XcQiiajx1XsmrtUqbW0TJXQYyrSBsWU9L4uzBNnUdtbUHJZgvpygkntaCNx720Ym/rQejqRoiETIHHf7STuWy36vvgHf0Hj1ZPrHEmgb+olfscRInt2oOQy4Pu4M/M0Xj9F49UT+IXymrGqlp/8IaIH9qxIigrfZ/43/wj7/LKegJKKoqVjyIaG8HwC2yOwHZREFKMjTe3cFH7dWvJflFyGlh//BPbgOJVHn13Sj47u30H2cx9h8b9/EWd4ElSF6IGdJO69Ga0tB4pCUG/QeP0s1W8fRdjht6H1tJP8wF0Y2/qRNBW/XKX67Zcw3zi7otXVeohrOTZnbiWu55AkCdOtcD7/FLZXQ5YUOhO76ErsQldimF6F6eoZFusjBPhE1CQDmSNkjG4Egqo9x3Dx5bUTot9jvCmju7RxLoeayRKYjbCeWg4NsKSoaO3t2ONjOHNz3xM2t94UIC4uetd1Pq7E9JDF9DoqVwCzoxazo83la/S0mxlde9tUR5T9H+/n9CMTFMb+7r0If5/QuiXJtns7OPnwONW5NzmrUBSUVALheiTfexdBw6R+FbtAjkRQ0ikC08IeGkMf6EVJJfAWC7izC6u8TL+0fmGL3tNF9CPvR8kk8fJFgokp5HgMfaAXvb8bvbeb4pe/QVBe7Z1bp84T1BtIuoaay6L1dYXe8lWIDbShtyaxFyoY7S3IER3h+dizJbRMnEhPjvqFq3Vwrw81lybzifdhD01Q/c7LoZZCewvCcRDNLLacSpD7kY8ROC6VR54hMG2MnZvI/vCHAUHjlVPXPIaETEusH00xOD3/LQLhEdOyuH44CGYjvXQmdjJSOkbdKdAa20RPch8Nt0TNydOb2o8q6Zya/yYgsbPlHjZnb+X84lM3fL3vNN6a0U0mMbp7EE23XDgubmERORqKWyjJFFKhgHCW43kyClnaiBDDpE6JBdroQUGlTgUFFRkFDZ2AgDplZBTipADBAjN4XL9Iw7EFU6MuQSA21On1TWGtsWSd8SXTE6PvUAvnn7jxl/5dLEOSoXVzgu59Wc58c/LN70dTUdtbkXQddzZMmFw9hXcmZyj+1cPh9FyWafkHn0HZsx3r/BCVbz61KgYcNNanYSXuuRWvVKb019/AnZ5DuF6YmDu8j8Q9txE9vA/z1Dkax06sclJqz74Mz72CpGtE9+8h86mHUNLJVcdojMzj1W3cQhWvaoYtxT0f4fm4hRpezV7TUbj+zZKWvFtnZIqgsnq2Gju0G627nfnf+nOcobA1unXmEtF924nffoDG62eXeZprQBBQdwq0xbbQk9pHvjFO0ZpYChFkIj1E1CSdiZ0IEaApUVJGO7oSQ6JAV2I3pldmIH0YAF2JEddbgNDoHv5AC42KR0u3Qfe2GNWCy/En86EDJUJRr713Zth8IInnBlw8VuHcS2U27UuQadM580KRjk1RjjzYwvNfnseq+9z+8TZe+Er4843gLRlde3oadzEfPkhJQngewvWQdA0JCDwvjHNdgSgJUlKOeTGJh0uSHBHiFJglTSsJUiwyQ5ZWiiySoR0ZiRKLGETooJcprq9b6/tQLflEYlfwTVWJvR/sQYuqxFsMuvdmOfG1ccyyw5HPbKYwXufFP76I0/CQpFBT4KaP9NK6JYlVcxl8Zo7zT0zj1MPRXYsq3PYj2yhN1cmP1jj4iX5yAwkqsyZvfGWMyRMFtt7dzp4He+namyHdFeVjv3wYx/QJvIBjXxjmwpMzS+e258Eedryni3iLsTQ9FoHgpT8bZPCZOVKdUXa8t4ve/VnirRGsssOl5+Y4/+Q0dq05fZPCEt/9H+unfXsK1Vj2iEaPLfDKnw9hVTYWu1J0mUOfHGDTrW3EMno4kwGcussLfzLIxGt5tt7dwd4He/jWfz7Bvof62Hp3B4omM/LyPK/8xTCi2Yqo73Ar+x7qIdsXp1FwOP/kDJeencW1fD7xK0e4+PQsZ781Rbonxn3/eBf1osOTv34a1ZC5+bNbsCoOruWz/b5OuvZkieV0PvlfbsGzA1zT4+W/GGL4hWUGRDSlcd/P7aZrTwbX9Lj4zCynH51ENPMEwnZwxiZRO1qJtOUACTkaYcXn43nLBkaWl7w64Tj4lSrC2lhMWiKMy5a+9Aj2pdEVHrJfLKN1dxDZtY3Izq003jjNqlJHACEQtkNgWevOHP2ahVkLPX+vdGMUxWvBL1WoPfUyiftuIbJ7K+bZIRovHsedXr7fxvZ+5HiE3A9/dDl2LIHamiGw7DAGXrv2ORWtKbyCQy7ax+bsLfT4ezgz/wS+cNAUA8urUDAnlhTA5moXqDmLSEioikG5NkPFDrUuCuY4/hXx4B1Hkuy6I8OpZ4pMXayz89Y0Ld3dfO2/jdOoehx8b45D72/hzPMlFE3ioz/Xh6rLyDLsuCXF0Ikq++7JcuiBFibONSjMWGw/kuKlhxdu+H6+JaMrHAffWe11iqsVm6+AQQRbmNQJp2KtdNOgQpUSWakNFQ1T1PHxsTHJkSQixYmTQoiAMhsTn1ZU2LbXINeqcvylBmY9lN9LdUQ59P2bGHx2Fi2m8KF/d4CxY4s4DY/9H+1l7kKZc9+eov/mVh78N/tpFGwmXs+TaI1w78/sIjeQCA1z3UNWJNq3pxi4pRWn4VGdMxl7LU+yNbJUgVXPO4y+PA9CoOoy55+cpjIb8l2vTPTc8tkt3PojW3n1C8PUCw4HPtFPtifGN//TScZfC7OxfYdb2PNgDwuXKiwMLdCxM817/skegkBw5tFJAl+Q60/w/l+8iVre4vhXx+jYlebmH9jMmW9OcvJr48vGeQO472d3s+t9XRz7wjC+G3Dr57YgKzIP/8YZ5i6GMfd4zmDLXe3c+Y920r0vw9TJItGUhmY09QQk2Hl/N+/9Z3soTdaZeK1ApjfG+35hD6muKK99cRjNUBi4uZWz35qibWuSzj0ZIkmNJ3/jNJF0yAO+8NQMlVmLkaMLKKpMy+YkZx+bolGw8d2AwtgV3pcE7/sX+zCLNjNninTuzvD+//smRCA4/UjTOxYCYdnED+3DmZ4LHQX1LX0O60IA1sUR7MGRVQbTr9RwJ2eI7NqG0poNWQbvyFncAC7napoQtkv1qVcwT14kdvM+4nceJHHnQSqPPU/18aNhSFHTCGyXxrFTK5KC5rHTeKXKhjjNgQio2HNUnQWma+c40vX9tMU2M1u/gO3ViWoZ8o0xvOBySKlZPIRMwynhBjbz9SHWK4G26z6P/sEktulTXnS5/3NdRJMqnie49aNtvProIm88WQAJ2voi3POpdh79gykUTSISV9h2KMnJp4v0bA/Fy/OTa7cvux7embfsGnCxSUs5DBElwKdBhSQ5oiTC6jea3kTzvzY2nvBYYAoPd0Ohhcs7UFWJti4VTZdXlNkWpxq8/OdD6HGVH/3Te5g8UeDVvxrmh37/Trr3Zhh9ZYFDn9pEcbzGY79yisqsiaRI7PlAN/f93G7GXllg5KVwhJPkMHTwyC8dZ/SVhVVTt9lzJeYvllEjKi2bk1x8epb5iytjf3pMYfeDPVx6fo6jfxoq29t1lwf/1X6q8+aSoTz3+BRnvzW1VO0Vzxk88Iv76DvUwsWnZ3FqLm3bkkQzOt/5rTPMnCkx+OwsrZuTKLpCfdFeWSl2DURSGjd9pJfXvjTKsb8MZxayJvOen91FabqBay7fT0WXyfTE+OLPHV2V2Ep1RTnwiX4mjxd44tdPY5YcFE3iyKc3c+Szmxl9aYGpM0W23dURhg62JBl/dZHdH+ihbUsq7OgbVShNNZh4I8/M2SKJtgh6XOX8k9OUJtb2nkrTdb72b14j8ARaVOFzf3gX+z/Wv2x0m/CLFdyJGaKH9uJpN1IxeGOwL602uADC8/CbHqCkaawq1bsGpGYgLvxmxNLfJJpJbSAgaHK910AQIFwfydCXQyuShJKKr24/5fl4c3kqjzxD9bEXyPzgh4jfeQjr7DDuxCzO5CzGrs2YJy/iTl6lrLaBkIaETGdiB4qsY3lVDCUBCBpeOLjP1QdpifWzJXcbhcYEiqyhyBoL9WEcv85Y+TUG0ofwhYvlVomoCWy/wXx9WUd3/Fwd2/QJfLBqPooiISth8nXnLSn6d8X5yM/1LZ3y1MUGlbyLqsqkWzXa+iM8+odT3PeZDoJAsDhlrSqx3wjelNFVUNHQsTHXbXB3JXQiJEhRoUiDGnVRpYM+TGrkmSNKgja6qVNpxnJ96pRR0KhRJCAgRwcCQYkFaoQPIkYSmwZ+c1IoIS+9YLIi0agFlNegi9XzFq7lYzc8PNunPNMgcAWu6aNFFfS4SteeDG98ZZTKXBinE75gvik03bY1yfhryx/ozJkSC5cq675Y17tDelxD0eRQOrGJyywHLbr8iCRJItMbJdkZxYipaDGVSFJDNWRkRUKSJaJpHdf0lji1QkCjbKMZKvIq7Yzl9tZXI5LSkGRpSZsBwCo7BEFoBK++wDe+MrrK4ALEsgatW5I8+3vnl2hlviuYORc2YWzdkmTmTIn9H+kn0xMn1xdn+nSR9h1pevZnKU3W8d1gBc/3yuOuCQGnH5lcOh/X9ClN1unYkV61auPUuTBWOTyOXyits8O3Dr/05oj014IiqWT0Pmp+CV84eMIjKifwhEtEiaOgUvNLmMHaCb7AcvDm8xjb+jF2DOAtllBb0kQP7mq2Pw8hZ5IYm3rwixUC20HS1DC04gdLlLLGsdPEbz9A6oN3U33qFYKGhaQqKKk4fqmKO3Xt4heBwPFN2iM9ZCO9eIHNxfzzVOyQ4tVwi5xdeJKuxC66kjvxApeiNUnQpFDO1QfxA4fW2GYyRhe2X6dkr6QAXqu4qzTn8PDvTnDm+WVdBt8Pqyd9L2DboRTTgw2qeQfPCejbFefFv5lfpZG8EbwpoxuOsBvfVEGhS9qMIyxqVJhnpbcxzUhzPZUs7Xi4zDBGnBR1ygQElFlNeFbRcJABHwmJDC1UKeHhIkngeYL8nL/qxgSeQAShhoAIWBqtRLMIQVYktIiC2/BXfNie4+PUPSIpHUlZfintmotrvXlJQKvsUBirsemWVi7uSmPXXAZubsWquEscV9VQ2Hl/F3se7MGuu9hVF0mSSLRHKE012+oEgtJkHT2q0ne4Bafhke6J0b4txeDTszhXagpLEpF0B2okRn1xYikZehn1vE1xos7WO9uZeCNP4Ak23dZGaaJOdY1ihsuFFVdD0aSwSMVaGdZwTR/P8oll9aUBrPdgjmhapzBeZ+L1PH0HczgND6fuLRVbbBRXsxoCTyzJWl4JfaAX8/gZvFIFKXJjzUVvBGKtWvW3CAmJqJxAlULdgJI3R1JtoeItEFfSyCjU/fWpl8KyaLx6BrWnncwn3kdg2gSWjV9t4BWWt1PSSdIfvi/U3XC9ZgGBTO2ZY0uCO97sIqUvPUbi3iNkP/PBkHvcXL/29LHrGl0Q5M0x8ubqfoGX0XCLDBVXa00ACOGz0BhmobHxPoWX4doBg69V2XlLmvGzdRpVj1SLhucIZkdMCrMOu25Pc+b5IlY9rMxNtWhU8s6bImZd13JGiNFOLwoqi8xQp0IHfQgCTOrESdBKFxIStmSxIKZopQudKDEpwawYp8g8lqiv65ho6HQT1mQbRAGJdnoxiGBSIyCghy1oGESkGDNiFBB0SAOMiwsE+OTopEvaRF2UyTOLLRdIJBWCQOBdZzS6emngCZyGh5FQV5T5qoaCkVAxy04oKKMqyzt4C7Q4zwl48Y8v8tFfPsxHfukQVtXFrLg8/4cXqC2GxiPbG+Pmz4Y6ta9/eYR6wUGLKNz7j3dhJNWl85g5V2bwuVlu+aEt3PThPhzTY+xYnpNfn1gVWlCMCNFcN1ZpDu8qo+uaPt/5zbN85JcO8bFfPoxVc2kUHZ74jTNLXvSKe7hO2MJ3Ajw3CKvirphmajEVLarSKNjYNZfSdIPuvRlc26dRtJk8nuf+f76X4lSDyry14eTf0nG960/7tL5u4jfvR23JAmBfWN1D6+8yfOEy60wgEEhIuMKi4E7jCYcFZwIJCSe4hrCNAHt4guLnv46STSEpMn61TlCpo7SkcWfCwdCbWaDwF19HjkWgWWThV2r4C8UV/FvzxAWcyTnUXCosuvB9gmoDb3F9VS85EiFwriERB6AoxPbswZmcxCu+dYWwq+FaAU/8+TT3faaDH/p3m5FkiWrR5bkvzzE7YjI/ZnLfD3Twtd+uYjd8aiUP3xU0Km/O0dqQ0ZUlmWkxgotDgE+VImlakJFR0YmSYJgztIou0rSQIsc4g3QxsBSjvRaSZPHxKbFIlDCWU2aRLjYhNY+RJMc45+mgH78ZC+4Q/SioS9SymigxzyQmdRQ/9HS3740wM+Fimxu/QXbNZfJ4gb5DLZx9bIrSdANZkejcmSaa0VkYqqw5lV4XQmBX3TB0cdn4wAprH2uJYCQ1vvZvX6M83cBzQlnEyxq3WkwlmtaZOVeiMFYnCASJthRt25JLIRAAVZdJtkU5+9g0p74xjmuF3rm76volFC1yzcRNsj1C4Ac8+h9OUJ238Gwfu3Zjxq+et5k7X2HLHe0MvTBPvWCj6jI9+zLIqsTCcBXfDZOKPfuzzJ4rYZZdGsUisaxO1+4Mw0fnlwx94AvsuoeR0NAia9/LjcKdmaf69FHcqdnQi1sjKbwS3/MU1woEBNjByrCF15xuuxtVX/EDvPn8VRKRElHLQCNOHRvhuPgTi/hC4ItrPH8h8BeL+JeN7HViuZKmEdu7F/PSJfxy07O+HNO+7MQ0KwDVXA53YWFlzPtqR+caQigP/84EglALBWDsTI0/+lcXaVRDIa/pwQZf/c1xIjFlaZZsVkPbderZIkPHq1QLHoEveOLzMzz9xVkalTc3e7mu0a1RJi5SbGYPk1yiRhkfb0VwvkEVBxu/+U+AoJeteMKhzvUqy6SQ0UADi/qSkb7yGGHyTNDDVmxh0qBKQEBwBcHn8rG95sAgXLAaAXNTLtYNNs60qi6vf3mUB//1fj76y4eZOlkgnjMYuLmVU49MML2Gpuy1IAJYGK6iKBL3/swupk4WEAIuPT/H9KnwBe3cmSLwBJ7jL4U3FFXCqrp4dkB13mT2XIlbP7eVVEcUWZXp2pNGiyhLNCiASFIj1RllYagSVrFqMpGkhqLK2HV3SZ5TkiQCzyXw/XWlrntuyuKaPr4boKhhw0FFk7EqDv4Gs7bVeYvjXx3l/l/Yyyf+8xGmThfJdMXpOZDl2F8OUxirEfiCxaEKhz+9iaEX5rGqYcfj8rRJbiDO8a8uDyqBJ1gYrBDL6Lznn+xm7lyJIIDBZ2aZu3CDVYyeFzIKNoigSRGTE/GwTHaDlLG/q5CQUWR9qdRWlsKZmx+EhjUb68d0yggCImoSL7ARftBcT8IP3HWTdGrCoOu925h89BzJLS2oMY3CiWWOuqRpxG66icQtN2Ns3kTj1CncYonUXXciKQqNs2exR0bJfvghhOsix+OYFy+S+cAHUBIJgkYDe2wMJZmk9sYbqKkUscOHKD/+7TXPp1Fd6XR4rqBW9Mj1x6kXbOyah1XzsdbojuPagvLCFeX3dR8u1zdJ0LopQXnWXMOxWRvXl3ZEotLs75wii0mNGAkMoqEXjLwimSYjIyMxxyQeHgoaKjqaZBATCVxsHK58WQUmDdK0NNdXkZCJkUAnQpQ4AT4yEtNMEjTX0TGW9ulgN9eRiZMmQOBhEwTQ1a8xctHGbIQar8WJOp4b4LuhPvDoy8tC3jNnSjSKDgiYPlXkkV96g70f6qV1cwK77vH0754LWQJNnm7gC2bPlXEa3nUbBOZHqjz2q6fY/UA37TtSNEpXeLFRhWozJvrx/3gk1LsNBL4TcO7bU7z850NU5yye/8OL3PSRPrr2ZjDLLm98ZQwhBLn+OIEbhPFoTaaetzj8qU0c+Hg/AJ4bsDhU4eifXlpiTqhGDCOZwy7Pr6nxYCRUilN1dj7Qzff96s1LGryu7fPGV0Y5+fAEwhdU501Gjs7jmB6ypBLXc1heDddfZhUMvTiPWXHZ+2APbVuTmCWXJ379NMMvzC8J9EydKTF8dJ6pUwV8J0BSJM4+HvJ282MryfhTp0ImxI73dNKxK0N13iJohjeqCxYjLy0sPaPLmLtYxnsTmeYlCIEzNkXsloMYWweI7t2BPTgaxmqb2gx+qYK4lmjOW4EiIykKkq4veXuyYSDpGuKy2hOQaI+E73OTNbX0Xl7tFCKRjfURj7TgeA1kSUFVwph2xZzFcqtLhQmaHCEb76PUmCQRaSemZwCJmjVPsTGx5ul6NRvhh7kTt2qhxlZ27xWuizU4iJJIUD9xAr9WI3HkMObZc1ijo+Q++pFQmGYxT/XYsdAYSxL1E8eRNJ3Ils0Eto3e04NsGOj9/ViXLt3wbe0/mKUyZ1GcblBdsIjENYykil31cEyfVGcEu+oR+GGITFYlqvMWqqEQy+jYdZett7cye7FCadqkumBddxZ8XaOrohEniY9HicUlSopFAw0dC7PplQpqlIkSx6SO0TSbAQFVCtREaSkcsdLoQpUiOgYKKgXmEQRo6NSooKKRJkedKlFi6ERwsDGpURUlJGQ0dBpUKbOIQRQNHaE6pLMKjiVo7VApLvoEnuDs4ysrwr72b19b+vm531+p3jR/scL8xbMAKNE4mZtuJXXTLuoTlzCnx3FNnxf+6CJIEsmtezBaOhG+T+nsMfzGylLfwBOMvLRMNbsS2+7u4PCnNvHCH10gPxrKJyqaxJa7Orj9R7dx4TszFMbqLA5Xeeq/nV33WelxlTt/fDuO6fPofziBUw+LPFJdUW75wS1sv7eTwlgtLCiwapTGTiOEIFiDV73/Y/3s/VAvT/yXU1QXLEQQhi52f6Cbu39iByNHF6jMmiuuKapl2Nx6B9Pl0yzWroiPNgexy179WlgYrPDlX3hleRNf8NKfrf0ReXbAxadnufj0avGS8VfzjL9a4Gor8/Ln32K8VgjMU+dDozvQQ+b7H8KdnCWwbCRNRdJUyl97PCx+eBsRPbQPras95BFrGmpnW9h5QpaJ330rxs6tCNdFuC5BsczWrjyF0Sp6TMU1fep5m9JUDf+qzL0kKSQibUyXT9OR2omhJsjXR3G9BtlYH5a7HLZwA5uGW0KWNVRZp2otYLkl2pI71jW6SkzDyEYxclH0TBQ9G0PWFAJ32RsUlzVbLocMZDlMwAUBkqwgNcXfQ1GgALW1BaOvD2dyCiWZDEVuZmbQOzvR2lppnFxP9+LayPTEiKQ0Ml1R0l0xXMtn/lIFSZboPxh2uxCBIAjAd3w6tqcI/DAxuzhaQ9Zkcn0h82b45UXKM9cWiL+u0TWpY1IP1ZZ0DWG6q9gHl1GjBAiSZBAIFBQqFKlSokpp3WP4eKv2ucCycUySRUVDI4KMgk2DCgUqVxVKFFk2aCpQWPQp5k2yrWGc5q1E5ITv4ZYLZPbcHErBzU6GHgaAALdWQYkm6LjnIWpj51cZ3Wuh90AO3xWcf3JmhYeW7o4hK/KKqrJrQY+pDNzcylO/fZbRl5fvRXXeYv9HPVTjik4SQqzfPUCCbfd0UJ23OP/kStpN194Miq6s23XiuwE1EkeNxLGrRWQ5LMIIxf5lJEnGSOZwGhUkScKzwucgyTK+s75OgxyPhkpa16ia8uYXKf3114nfdTPGzq0YO8Lkb1A3cecXQ4PxNiN+xxGiB3aHz+0qDm/s4J7lX4TAmZgiOP114i0RIimdykyD7n1ZiuOr6WpCBFhula707pAfH7j4gYMQAYqsk4p2kIp2UrcXcX2LbKwXy60BgoZTRAhxzQ5swhfMvzSKb/uY8zXcymqOuLBD8e/krbfSOHcOZ3KK2L59RDZvwhoawh4bI33/e0ncegtqJo07P4eSTqNUqkuJN3NwkMwDD+BMTb1phsjshTJO3WPTLa34XsDEiQLVeZvNt7Rg11ziLQZ6VGHklTy1vM3Nn+pn+lyF0VfzmGWHgcM5Rl/L0707jZG4PqtrQ7wvSdfQtw2AL66b4a1TYZLldTZczHAN1Chhc9lACAIZNDmGHzgoskEQhC2eJaTwZ0lF+B7FOR3HbbA44/ImukqvQODYVC+dIdo1sMZSgTU7gVct0X73h25438XJOrs/0MPdP7mTqZMFZFWibVuKPQ/2MHu+TP6qjgnrwXd8agsmBz7ej6JKNEoO8ZzBtns6yPXHef1Lo7j2Bm6ECD3PfR/p466f3MHiUBXVkOnak2HPg72MHVtcJdf43YSiR0l2bkPRp0l0bAoNqyRhlxexawXUSBxZ1ZEUFUWPIqsaBAHFsVMEbjMuG4sS2bl1iSamZFI4oxNY564xRfUD7EujOFOzyNEIUpO9IvwAXA+/vnKgdcemWPjt/xmGHsrriOF4HrWnj9J47STCcVcl84p/+TeUvvLohnpMCtejYFWbpdcSmd44VsVZM/QlCFisDaEqRhjTFYRls0IwXT4NQlAxZ8M4rhBMFN4IW6NLgiBwCYKA6dLawvAAge1RHVpOzq1lBYTrUn3pZSRVJbCssFCkXAZZImiYCNeldEWMNrAs7PGJ0PMVAmHbCM9DSSUxnzh3/Ru0Drbf04Fn+0ydKpHqjCCCy0VPUXL9cexaGD70nCDsnFJwMOIqhz/Zz8z5MkKwFK7cCKRrtRqWJEkAaL2dGDu3IEyL+tE31s0QqloMSVZw7XBEXA+yrKLqcVynjghufHRKxDppy+5ACIGmRilXJ5deStut0pbdRcNcwLLLFCorEyWSoiJr+lKmM3CdZisOCUnTkBU11JEIgvADvYLKIikq7fc8hG/Wyb/6zKoWHmo8yZYf/b8Y+/LvYy+s34Hgaii6zJ4PdHPTR/tJd8WQFKjMmAy/OM/rXx7dOF1KCtun3/pDW+nZnyOS1LBrDtNnypx6eJyJ4/kNvxhaVOHg9w2w+wPdJFoiCCEoTTUYfHqW4387hmcHSEh0pffRnzuCImkUzSkMNcZE8Q0Wa0MkjDa2tt5FwmhHSAGF2iiD888gEGxquQ1ZkhlZfAlfuMiS2vybwsji0WtmybVYimiuB0UzMFItmIUZJEUFBJ5VJ5LpoDZziVhrH5ISDsBuo4pZnF7yfLXuDpRMGr9SDUu021rwK1Wc4fGN3aA3ibDoQEKIAFlWCS6//1IYZxXrPiCpOUuRlrQHrn+w5v9vYIonSRBN60hymIQNPIGsSXhWgKxKtG9Lkh+toRoKjbKDpisgQ23BJggEsbQWFuooMoEX4DQTsTfE9tkgtM4Okrfdhj06Sv3EmwstSHKYVJZklnoqXn4E4XU0u2SLZjhEsNQ+TVHlJWriZUO91GxViHWHyQ15ukGtgTsxG07B1oARzQICI5pFiyQxq3PYZhkhfDQjie/ZeE4DI5oGJCRZIZbsoF6ewvddPOfGpA5dr0GlPgMIPM8mEB6aGkWWFAwthWUVl7vNXsFbUWIJsjfdRrx/O7KmEbgOheMvUh08hRKJkjt0F9Gezciajm+blE69QnXozLU5hG8DfCfg7JNFBl8O8Bo1vPqyVyQbUdR4BK++gYomAfnhGt/6T6dREym8enXVwLARaMksvt3g2F8OL5UAr4V0tIfezAHGC69RMifpTO2hJbFpabnrN5irXmRw4Rk0JcrerodoS04xWzlH2ZymL3sIQ0vScAroapxUpIPp8ulr05IAt1HBbawvo1ibDc/ZKq8vRuLOzOPOzC85EH6pfEMluG8GsqySSHWHxtZ3icXbKZfGCAIPTYshyyrVShhm040UiABVixL4LgLQtCiSJFOtTK8wvOnOCNnuKEhQnrUoTjVnhW/CzkXTOgM356jnbbbc3oZr+lTmwjL4VEe0WcFp0rs/S+ALWjeHAk8XnpnDrnkM3NyC7wratyWpzJrYdY/B59+ZVkzu7ByFrz285rKW/rABQWl6ZQitc0cSs+pSnglDTSIIq+EuE6Gu9CcDX8Bas4SmObg6MbtRh2ZjRtd2QuV5y17l5erRNLmuvZi1+TCeFkmhyCrJ3CZss0g83YXv2lSL46Rym7Hqi7hOHUXVSbVsxqwtUruG0ZVVPdTqDS8L33WwnQq2s/5HJyGTTHTheXazo244PKW27yfev53Fl5/Ezs+iROL4dvhQAtehPnGJ8vnjBI5NatdBktv2Yc6M4dXWP9bbBS2RJnfgTsz5KYonl6tu1FgCNRrfmNFtQonFye67jeKpl3CrpRs+l8zem6mPDdKYGb3meqlIB6ZbYaF2Cdc3ma9eJBvvX1ru+ja2VyMT7UWWVALhE9PDxETVmsMXHuloF6ZTImG0IksqVWtuvcO9rZBUFbU1uyRyo2RS+KUKznWUsFbtxzBI3HIYVBW/UMA8d3GVst5lqGqURLIHWVaQZRUhIJXuxTSLqOqyQ6OoBslUL75nk8r049gVyqUxkuk+GvXV+hCtm+Lsvr+dve/r5Pg3pnnsN67fzmfN8zNkFF2iNGWiqBJTp0M2j2P6JNsMKnMWsiLhWj71gk3gCaoLFvW8jWf7eG7AwlCNVHuE2QsV7FrYxulKL1fRJPSouqLCUJE0DDWB6Za57NELIUJ5WElFIPCCG2OF3P0PNlPL2zz5uyvDRQ/9q12MHCvw1O9/7wphNmR0heMiRQzUjlas80MrDG8s2YFVz1Oav0g83QNIlPPDbN7zEIpmYNfDoHs00YZtFinOX0AzkqRaNhOJ5cjPrp+NV2MpWnffQTTXDTQb7p16jsbi2hnTpfMloFKbWvE3SVGI92+lMTlEffwSIFYYMhEOeaR2HkCNJ9FSoeKTrK7dCWKjiPVuIdY1QOB51McH8RpVkpt3IxsRZFWnMTNOY/IS1sI0jZmx5jQ5RKSth8TmnVjzYVJRMaIkNu1CiYUxS2thhtro2rEsWdNJ7zoc0mymR2hMDhPt3kSsawBZ1ahPj2DNTRLv246WyiBrBk5pkcrFE0v7UBNpklv3Uj73GsEafaIUWVtKvkDYYsjzl9fryx4hHemgaE7i+TaB8Jdq+h2/QbE+QVtiGwvVIdoT26ja88t94d5hKLk0WldHKMQtwi4KwZvp7isEgeMS274Nacc2rOHRdY2u55mUCqEKVhC4KIqBED6+52IRJgWFEPieQ606jSTJuAt1PM/CtsqAwPfsVSGIoZfyjLxaoHPbao3dG8GW21qIJFROPjqzatnCVa2PKuuIxy+O1FgcCXMQ8Zy+yhvsP5ildSDOsS8vf8OCgKTeRlRNIwjwAhsvsMkYPTS8Iq5v3bDRff1rU2+pNP+dxMYSaUpI5RC+v6rKpFGZoWfbfWh6DN938T0LEfgI4dOozJJq2UKjMk29PEXHwG2oegy7XsSxKtiNAm09+5kbO7bmcfVEluzWQ0Qy7QB4doPi0PE3fbFh2MBirXlXcstusgfvonLhDeoj54l09BLfvHNDCYz1oCUzpHcepHD8KGo8QXLLHmrjF0nvPMjcC98KjWj/dqz5SYI1MutOOU9gmRitXdRGzyNrOqlt+8gffwER+KR3HqIxM0Zgr/bOFCNKdfgcvmOS23879uIsTnEBv1FD0SO03HwfM4tfIzGwg9rYRZzyIq03v5f6xBAEAVo6S2r7TRTPvkqwnhEJHFQlsmRIZUlFUyJLy3uzBxnNv8RM+TRCBPRkD67YfrF2iZ7sAdLRLjKxPs7PfntDAkor7rFs0B8/wFD1lRV/bzH6kCWVBWtlTF+XoziBiV+qYlbrS9q17nTsTYWRhONQf+VV5GiEyLYt11w3CDwa9etPtYXwsczVEqaN+vrhksATS1zlN4t9D3SSH3/7uprUC6vTZ7vua1/1N02OENPSeIELCBy/gSJpxPUsVXchfKducDycOFFad9n3oJHNCmzM0/VD4RfRsFbZK8eqMnb2W83kkx8uFgEjZx5FBD6V/AhCBIjAZ+zcY2El1BKVQCzJ0K2GhJ7IYqRa3uy1XXURArdaxsi2IanaUvKM5vlGuwZwSnkqF08RuA5atg1Ze2sCKFoqi2/WcUrzBI5JtLMfNRrHrRZxivMo0QTxvi3Iqram0Q0cC8+so+nL5+HWytjFZixS+CiaxlpOgGfWcMqL+GaDwHVRk2kibT1EO/qQpHBAkCQJr1HDKc7jlAsIz0XRDVAUWg7fS230AnZhft1gVakxRWdqDz2ZAxTr47SndhDVM8vn6jdIGG1EtQxtyW3EtPQK4qDlVSnUR9nWfi+ub1I0N94JQiKkhymSRlQNPTyZsOGjED6qbKDJBrKkhu8fgriapjW6mYnqSQLbCR+/HFKxJFVd/TXKcuhwNFtQQfNbWIuadC2DLUlImhpyUgOxxD1dcRxVWc7Q+P7bKpCT6Yrwnp/aSt/+DIomM3epxuO/eZ78eBhaO/zxHm77bD+dO5L4nuCOz4UMnf/1z15n4mRT0S+rccun+jjwUBeKpnDuqTme/eNhGqXQGkoy7Ht/J7f94ACpdgOr4nL623M89yfDCAHb727lPT+xle49KZDgwIe7AHjkV89x6luzjJZCvvxSOLD5W6iXu+z5xHM67/2ZrWy5tQXNkCnNWDz8H86wMBwOFjd/fy/3/qMtxNIar39tikf/3/Mrb4aAbHeUT//Kfvr2Z8iP1fnGr54nP7o82ESSKvf8+GZ23deOEVepzNs8/lsXmD5hIkkaQghkSUIIUBUj7B58A2GxjXm6qhpmdsen1qynDtZgIIjAQ1LDLqbCcZeIzlcPMoK1pwCyqhFr67sinvvWIHyP8tnXaL/nIdpufx/m/DSKEcGrlqmNXsAp5UntOBAm2VSV1O7DK158LZVFS6TR4ilkVSfS0YtXr+KW88i6gZbMoqUySIpCpLUbhIRvNVAiMfRsB2osAYBv1kOFsytPTpJQ4ynUWAJJUVFjSTyzFsZzY8mQlxpPhfeyGQZp3jzWc8WVSAw900oQs8IQiRBEsm1UL50i8Fw6ch3NfQRXnQsgBLPPPExiYAeZXYcpX3hjzYRc1Z5jJH+U3uwhWuObWawPM1l8YynEcH7222xuvZPdXR+kUB9htPDKUonpZcxWztObOcSlhWc3nJWXkOiIbiVtdBIIL4zha610xXYgSxpFewpZUmgx+kloLdTcPEV7hv7ETcTVLH7gkpdm8JIK+qa+sOggm8IaGsNbXPYwIzu2kbz9FrxiCWOgHzkWxbo0TPnJp/GLpQ2dK7JMdNd2kvfdg9rst1Z75VXqr70RCnvLMtGd20ncdTtqNgsiwJmZpfzYk3iLa7QSfxNwLZ/CZIPj35hGUiQe+NltPPQvd/PnPx/2hDvzxCyDLy7yI799mEsv5Tn6v0Klr0ax2XAyonDrp/vZfHOOx3/zIo7p856f3Mp7f3ob3/qv5/FdQbojykf+9R6e+N1BJs+UyHZFaSb7ARh9tcAXz1f4/v+4n9K0yZO/F8ZarWqz1VeznHgtJ/Ty12IkVD77awcRfsBj//UC9aJD25YE5bllr+PVr0xy4pFpPvtrB9fkkqu6zMChLC98foRjX5rgjs8N8P3/v338yU8ew7MD9KjCx//9XrLdUZ78vUuUZy1a+mPkJxp0ZPbjuFUadoFMYiCczdsFgvX0itfBxvQZhUDNZdD7u3BGp3Anro75SKiJZNj3/LKKvGUiR6Jo6SyN0SEi3f2Yk6MbnsLJmkG8fdMNXcz1YM6MM//coyQ27yIxsAPfMrEXwmupDp4KpRIHduA1ahSPv4is6QROSOBODGwn2jVA4LtIqkJm3y1Y89MUj78QhhF2H0KNp6heOk28fxuR9h5KZ45RvnCcxKYdCN+jOnwGr14NE1SBT+DamHNTCN8j1rMZSQlbT0fae6hPDhFp70WJRBGBR7SzF2t+GnN2AuG7IASN2fEl3umVCFyX2sh59Gwbih6hdO417MI8jdkJIu09BLZJ6fzrBJ6LtTBNYJsIEdCYHsG3baz5KbxahcVXniSz91aUaGxlMvGKgXexNsRivZmUuOqLqViznJj86jWfia5EcQOTherFDT9HTY6gK1Ema2eQkNiSupm03oEb2NTdWVoi/ZScGcrOLJP1swwkD4SKXI1B0kYHk/XTofdpK6gdbThnL6JkM2uW8Op9vXjlCvmvfA0lESd1/33EDx+k8vRzK9uPr3d9vT1kP/Zhqkdfwbo4iN7fR/KOW3HnF7GHhiEI8MoVai8dw11YRE0myX7sQ8RvOUz5m2vrCNwo6kWXZ/94Oczyes807/2ZrUu/23Ufu+7juQK75lFdWHkf4jmdLbfleP1vpzj/TBjiUHSZ9/3sdnJ9MRaG66i6hOuE9LB6wWH2QnXFBMm1Alwr1KJ1Gv6qY2wEW27N0bY5zh/9+MssjoYhtclTq/U2XCtYt+RbBIKJkyWOfWUSBNTyDj/1+dvo2ZNi7I0SPfvS9O7N8Ff/8vjSvqfOhP+XopN4gYMsydStRRyvjuPViWipG7qODRpdEEHoYWld7auMrqQoGN19BLaF3tKOPTeN0daJORmOmJIEaip9Q61I1EiCaEvXjVzL9S8j8GlMjdCYWi1y4lsNiidfWvE3zZDYeiDJ3KhK3D3N7DPHMKIKkZhCcd5BBIJMh048VaVx6UlKJQ89KlMrrvQKrz5e+VzoYfhmnerQaQCqQ2dCetoVqI2cozayMlF2JRvhyqTXlQhsk8ql1d1XLx9r6foiCvbsBRQJhAKN0ePgBbiLF4klNBJ9ceYHX8Zr+EhyKG2pahL9BzIsjjXQIgrlZkIl0xkh3+zisEK3dx0osoahJOjJHmC+chHLuxGR7yvfo/AnSZKRJRUfj5nGBSJK4orlzX5zV4azmlqvfqmM3t8LsoS7RvlmYJrUXz+BOzWNq8ho7W0YWzcjx2MEleufc3TvLvx6g9qLLyFcD69QJLZnN9HtW7FHRiEIcKdncKfDb8pbzGMNjYRe79sEVZfZensLnTuSRNMa7VsT6FEl5JVuwAfSIjKtA3FuerCLvv0ZAGIZjUhSJZ7TWRiuU5g0ef5PR9jzQAc772tj/HiJE49MU8u/9eKoy2gdiGOW3SWDe12sYWwCX1CeXQ6T5sfrBL4g0xNj7I0S2e6QFjc/tLogqWJeafeW4+uOd2Nx8I3FdF0X++JIGHdaKwotgazrzXppBbeYx+joQs/m0HOtOIVFtGwLWq4NZ4NFA7H2fhQ9cv0VL5+CpiAbofiHHIvg15vdUF0PJRlDjhk4kzfWjkWPKiRbNKYGG7T1RYgnVbKdOsU5l0reRUjQPhAh06ajGTKnny/Rsz3G2RdKN3SctwpJkUnt6qAxWQoHx0CEteGqTGB7KLEwvKBENZSIRn28SCSusPX2FmRVpjTdINUeFkCouoJV83Atn66dSeI5nUtH8xhxhV3vaWf6XIV4zqBRCXvEtcXizA/XyHRHMaseWlRhYY0X9krIkkZPej+d6T2YbpnxwjFkw0BrbcdZmEOSFYTvIYIgjKUKsaLvnhtYuL5Fb3wPTmARCJ+SPUtnbBtZvZuqu4CEREJrYWvqVmpeETcIE6gprZ3e+F7mzWGcwERta0FrbyWwbNy51YmqoN5YrhLzA/xKFTkaRda1DU0q1UwavauTjp//6ea3I6G1t4b7bPKC1dYW4ocOoPV0oyQSaK05rEs3Lsa9Hu760U1suTXHyW/NMnWmjF3z6NmT3vgORNjJZGGktsIYnX58lvxYaAADX3DsyxNcOpqnb3+awx/voWN7gr/596c3zF+97mkEAkV7i+XnkrSig4okhx1XLjcyCAKBrErvKGV7w+0fhGWDoqwZ4BeeR+3cqeW8lOdROfkqIhBYM+H0ufjSsxsn6ksSqd6dGz01APTOHIkj2zHPT6C0plBTcfyaiV8zEU5Y3nijRte1AwozDsmcRudAFLcroDTvYMRkFFXCsQLseoDSKVErusSSCl1bopw7WnrbXrSNIL2vi8TmVjI3daMmIlhzFYQfELg+vuniVi3cikWkPUng+jQmikiKRCSpAYLitEnLQJxYWkfRJGpny0SSKvWSg1330AyZbE+UdGeE0deKFKdNGkWHaCr0mlzLJ9sTo1Fy0QyZheHaNYn5gXCZKp1kpnKWQHj4gYtiJIj2b8Lo7AqZMp5HYFlImoY1MYp/hdEVBMyZl5rMhNCQ+cKjUSmFy5s3f94a4XIFV4CP77ucKT4ZyoI248d+qYKazYTaIms1plTVFa3ZJVUFz1vW3bgOhOthT0xRefKpcEBswq/WIPBRshlyn/gIXqVC9Znn8MsVUu+5BzkW29D+r8SadkKCmz7YxalvzXD869OIQNB3U2bNlT3LR4soqyRt7YbP4liDhZE6Jx+dCTUUmp2hfTe8pstec368TnGqge8J3v9PtyPJ0grNBdcK0CJymDMUa/tw62H6fBUjodJ3IMPkqVKzOkwKWRtr+YJrXKOsSrRvSaDqYTVZ3/4MCFholtovjtaRZImBIzkGn2/2PLx8DW8T62FDRleOR9F6u9B6O6l95+iad+pqg7rUrrrJVLhWh+BVx1M0kt3bNrw+gG86ONMFAi9Aj0dwZgoIz0drT2MOTqG3Z25ofwCOGTB8PJxCzgxdwfu9IqY5dqbG2NllIzMzdG2FoY1AUjWUVApJVXEXF9aOg8syWmtbmJwMZIQXUJ8oIiER6Ujimi7IEuZMmfhALmQq1G3UmN5kmgismkt+vEG96HD869Mrk6RXJUynzlSYOhPGdQtXNIOcPB3Gu2bO31gPMF+4+P5yUk14HvbCXMhIiMdBBGi5FqzJcfzaas9ZEOBfNbKtqmRbIzHniZXvoZJMhB0Oag0CczWDRM1m0Dva8BbzyBEDY/MA7mKeoHHVc76sliWt9MTskTGMgX68QjEsOQYkXW+2Ugc5FkNOJGg8fxRnYgo5HkNtayWob2wKrWgSqY4I8ZxOJKWR7ozQsy9NbdGmumAT+IL5oRoDh7IsjtZJdUTYfFtuTYM08mqB3fe3U5xq4Jg+Qy/naRRdanmbU4/PcOQTvcSzOuVZk1R7hMAXvPzFcTwnYMutLey4u425oVD7Yfd7Oxh7o7RK5Gb01QJ3fK6fI5/sxap6TP3/2fvvKMuy67wT/J3rn3/hfaT3mWWyfKGqUIUqEARAoEgQoGjFkVptl7pHa3WP1Bqpu9Wi7KwR1ZKmpaFaFEVSJEEAJAzhC+VtVlZWmkrvwvvn7fVn/rgvw2S8iIw0ACmt+f6oynjmvnPvPXefffb+9rfPllcq6G6BseN5Lr+1xOf/zkFOfXuWRsmjazTOh9+aoTDZQNUE6V6LRJdBotNAURWGj2So5R0qi1ExBxKyAzFe+Ot7KEw3ePDFIS69ucjC5WiOzZwrc/7lBT79P+6nf3eSypJDdjDG5TeXmL2wcR/E28EWwws+QbH8EyO4xXqGUa3EbX3Hz5Wp5iIDYF+dWfe+O3n7/ek3xHoKxj2FMThE75f+EkZvHxP/+O/jl0rrPqOl04z+j38Ld2Gexa9+mfJHUQxXS5joKZPm/Eriq3ZtvYdvV0POvXQTzUVu8O+fAELHpnl168m0e4XGh+tj36sROg6Jh4+i9/WhJOPovb2UX3plOelmDA+hD/Rjbh9F6+wgcfQBglIJ+9p1wlqdxtnzmNtH6fjcZ3Dn50GCmkpRffNtvPkFwkYDv1gi8eD9GAN9KMkUQt06YyfRaXL0xSF6dibJTzVQNMHTf2UH02dKnPjGDM2yx0v/6jKPfHGEA5/oJT/Z4Nv/6AJP/PK2dff43T+aIAwl2x/qxK75TJ0u0cAj9CXnXlqgWfbZ+3Q3fXuS1HIuV9/NLWsPFKYaeHbAjoc6kcDUmVLLs177Gx9+cxrNEIze3xGxKiYbWza6MoRv/YPzPPj5IYYOplF1QWnOxm1EDl4so3P/zwzSvy+1HEt+6v+2g9nzFU58fZp6weXym0vMXazQtyfFyP1Zrry1xHt/NLnmN176V1dYuFJj29Es/XtTVJZsvKZ/z56JLQneCEPHOrwPva+bxpkL+DM/vlJNRTcYeOin6Tny8RUZwhZ8p8HEq39IZXLjKrafJDrTOyjXZwmC28/EbgZzdPutjW42y7a//b8uG11n6t4KtQhVw0hkMVKdaLEUmhVH0c0obq+oESc78AkDn8Bt4jdrETe4XsJvVn+iDHSh6ZjpbsxUF1o8hWrGUVQtium3hIt8u45bL+OUFvEaW3MgrP17ST/zFLUPPkSNx0BVcadmcCYml5kL1r49mCPDa74Xeh6N02cJWvdNWCbW3j3onZ2Rh16qYF++EnnLioLe34e1cztC1/AWc3hLOfSebprn7lw56z813NDPVVNRrFlLprDnplE0LeLVyxDpe4TNyEALrSWKLgSKriNUFb9699WMmgamKXBdiWkK0ukovhuPK1y/7iMlbNsWhWAKhZBmU6JpAsOAWEywsBAS9eS8S8EboBVnC5b1RtVsCnPnMM0LY8g227KtQCgKerITq6MPK9uLle1FT3RsyFpQdZOBh3+ankNP3dHvAdRmr7J49vV7on3adEp3pJL2FxZCweroIz1ygGT/dvR4OjJgmomi6QhVi6rPWlJMUdFLiAw8As8l9FwCr4lXK9HMz1KbH6OxNNk+li8UVM1AhgFC0ZCBF/GJhUBRdRRVx65uvjuxsr2ktx0iNbgn4jOb0cKgqHrE774xziAg8F1C18aza9jFBSoTZ6nOXCb0b1HqJATe7ByN+faOhn3pCvalK5seQtoOzTNnaevP3cReuAF/8R7uzP4TgN7ZRWxoO145amWlpzO4uQWM7l78ahVF01Ati+bMJIqmk9i9P5pXYUjgRLKQ98LoHjmiMzCgEgRQrYaUyyGuCwMDKuPj0Tzu7VVxXUl3l8LQkErTlmiqIJcLyS25t2QSbrEiLUTIqOQxrESxD2HoaD0diGtTSHtVQ7ibo/CroMfTxLqHiXUNEe8ewuroRzViURsSRUNRo6qcmz3cGxCKSrx7uO17W4XvNFCEhqpZhDLA0OIIoeIHNmEY4AVNtrKP6Os8RF/XYS6Nfw/Hq9DXeYhMcgQhBJPzxzD0OL2dh9AUg5mlDzH1BEHoU6nPsG3gScZm3yIM76DWvw1u6FurGvQNaMxM3u5CIIj3DNN733MkB3ai6FZkZEX0XvuvqAhUUAHdRFtFNJE9IZlthwh9D7deonj1JIXL7+M3V2KzsVQPqe5tSBlG3rLv4tkVBALViCHDYEOjqyey9B55hsyOI2hWqjXWTcapqJGcZyyJke4i0TNCdscRmvkZ5k/+iNpMpMWxwaW5ZxgZUnn4fpMfvtbk8AGDIwcN/t3v314s/Hbxl34uwRc/n+Sv/80cC0sr1mBkWOWXv5Dkn/7L2+wr92NCbHg7imnhz964HiFqLIHZOxB5waqGls7izM8hVBW9o4vQcQCJO3Edo7P7lr/R3/sAPV0H+OjCHy2/JoTKfQd/mStj36fRWKKzU2HbqEq1JkmnBbouUAQMD6ucOAG2Dfl8QCKusOeAjudJKlVJb49Crba15OoWK9JUUBX8m+K6Wl8X2Z/9BErcovqjY6AI4o8cQsumKH/nLdyxtWWdI8/8AumR/ayeydFz/eOV1LsZyVgPlpqiVJ8kHR/A9etk4kMkY92ML76H691aNHyhcJ5ManT5YU/EesiVLlNtzOP5TVKJfurNRfKlq3h+k3jHvhuMUnQ1di+fZQaGNPbcHyORVNB1wVd+b2ui57S6LPQe+Tid+x5dLnXd0IBtEUJREIqBUHViZhy3N0/xyombPqOiaCa+20BRdZx6ATOWRSJxGiXMeHb9cVWd9Oh+Bh/5DGam947GGnVgUNHMOMmB3ezq20H+wrssnH51XdhBej5htXZHuyJFgVRSQdfBcSS1uqRWk7z+jk2zKbFMQU+XQlengu9DpRpimoJkQhCGUK+HOKtyfqoK8Vh0ruGqrL9lRkah3pA0bUl3p4IfRO/X6iFf/Wadjz1mobbye4mEwDIFnVmVTHoT+tUmnXW3fA1SCcLq1jis5VPHabfwubn1OwwFk8b1y9hzLfsi5ZaoqJpqohtrc0VRAaZcJizPzwecOOFSLkfaujdy2Ksvx9WrARBw5iPvxs9v5muuH8etPwIEIUrMjFT2V/162LCpvnwMNZ0k/tBBKj96j9qrH2DsGMI6sGOd0b0hwny3D/XdwnarGFYcRTEIwwDHrQIKDSeP72+VfdBSK25hZulD+joP0pnZycziCQqVMXqy+9g28CTz+Y8iYpMQKIp2zxeZ2Wmf5qSLpgksa2vHFqpGamgv/Q++QLx3dFm05l5CCEHgOdQXJyKDtgqeXSM/dRqv2X5L2CiuTYZqsRRd+x+j/8FPRp0g7tH4hKrRffBJzEwPcyd+QGNpavkBdK5dj6rG7gA7t+v8wosJpJScveDxo9eb/MqXkvR0q/zzfx1di/sOGfy1X00ThpKv/VmdZ56wGB3SKFVDXnmjybmLKzuhgT6VL72YxHYkigDXk1SrIYcPGjRtyfikzze/W+ff/L+7eem1JvWm5KXXmiyu8m5NQ/CrX0rS1aFyq1ydsXOEoFQlyBdv/+QVBWN0kMyLn2TpX/3uFqtQt27gQ9fBnr1JaXAL1q6dmFIoA86c/4/Lf587t7JLXH3Idoe/1fsbYWtGV0BYraNn15a7BdU60g+QroeSjBM7srtVx57aMo/xzwNe0GCpHGXKm040qerO1uvchVDJJkeImR1kU6OUqpNYZgbbKWPoSSwjjeLrBIFLELqYRhrXr9OR2oGmWstdV+8KNxnuke0aji2pVbdw3RWFzr2P0Hffs5iZnrsfyybwaiWqM+tjnp699fibnsjSd/9zdO1/HKUdl/YmBJ6L9J1loRq1lQDcCEJRSQ3vRdFN5o5/NxJBv+kpGuiLYtlz81uotlNhzw6NqRmf3//Kyq7jh682eeHjK7q518Z9/o//b5kvfC7Bk49YPHDY5H/5xwVq9fVPsO9Hx200Qvbs1EklFU6fdXjnmM0rb9r8b3+rg1ffbGJaCn/4J7W2xxjoV7FMwW//QZXRYY2f+8zGXOD4Q0dQ4jFKX/8BYXnrIRCha1iH95L+zHMYo0MgomrBRLyPTHoETbPwfZtieYxGYwlV0UmnR0nEe1AUDdspUyqN4Xo1hFDo67mPpl0gEe9FU00adoF84TJSBiiKRjo1TDLeh6IauG6FYmlsWWtbCI2ujt3E4924bn2NZjFAV+deUokBQhmwsHhmjUa3rsXpyO7ENDOApF5fpFi+jpQhsVgXMasTpCSR6EXKkEp1hkp1mq0sHlszuqqKnyuutH++MSHXTEyJYhqoHZkNawsrUxfx6reOISWH9mIkO9Z5xGHgU5u7hle7g9W3hfri5BqS+h1BSlyvztT8MTy/SRB6eF6DMAywCxXqzUU0LYbjVXBKVWrNSM4vDH3C0KfezLUVCbodKKuUx3bs1tm1TWdpIdhSUUbnnkcYeOhT6PHNa8ZvdAq2C7PY5SW8epnAaxIG0YRXDBM9lsJMdWFmejCS2TWLgQwDGrkp7MJ6fdatQjVj9Bz6GJ17H9nQw5UyxK0Wqc5cprE0hdeoIgM3MrqKiqpbGJkukr3bSQ7uQjXWd0ARQiHRt52hxz/PxKt/iF1au601DMFnPhXj/RMuM7M+8wubX+hQcktvUtejzLgiwPMlm60nthNxTIWARlPS3SVwPdB0gapGlVQScJywrcGFyOFUlEizS73FxkboGoknHgRdo/QH31zPS97oO08+ROpTz6D1dEbjRZBODbNt+Gmadp6mXUTXYugtAygUjUxqhFD6SBnS03UA00gzMxdJdW4beZpGsxAZNCnZPvIsihAs5s4jhEYqOYQiFILQpyO7m1isi8nptwkCh+7OvQwPPU6xNIZlpunu3LfG23XdKk3VZO+uz1EuT64xuqpmkk4N47hlVMVk++izBOMu5cokiVg3O7Y9T7U2S72+gGVm6Mzu4vrEy9TqWwhz3PITEHmymRSsKgP28yXqb50ktB3Cpk3pG6+AHyBMI/Jy2xi2/MVjW1IN2/6JX8FIrq89D32PwqX3qUxdbPOtrUGGwXLBxh0fg5C6vUTdXkn0eP5aMnvgeuu6W5RrW5UuXAnaCaO9iLq+KnEwN+3DksPSfEAitfnTlBzay+DDn97Q4N6gEDqlRfKX36cyeYHAaRD6XksnOVwOYgmhLCc5FM3ATHeSGtpLevQQVraH0PcojX10x9dbKAqZ0UN07X88kpxsM9bQc1g4/QrFqycJnHpUudZm5RGqRkF/HzPTTe99z5IePRiNfdUiIYQg1j3MyNNf5Nr3f5vQW2Hl5IshFy55RAqNm4dwggAuXfV44mGL3/yNLk6edXj59Sb/1a+n2LsrWjimZ326O1V+4//ZQS4f8Lt/XEMI+Id/t5NiKeQ7P6xz/ORKULfRDCmUAjRNMDvvY5qCN9+z+e//Wprnnorxzvs2pXK4xg9KJgS/+gtJHrrf5L/4tRTf+n6Dej3kf/rrGUqVcNNdv3Q8EILEQ0cQoaTwh9/cnKWkKKQ/+wlSn3gCEY9BEFD5zmsoQqOv5whNp8D45Ov4gYMi1OWKQN+3mZ471nJCJL5vk0mPomkmvh9RMRuNRaZn30VKiWlm6Orcx2LuPEHgMLdwIpKKlZLOzj30dh/G0BM0A5fB/ocplceZnH4LIRTi8V5iVnZ5yNXaHLX6Int2fXbd6ThOmYmpNwhCF4FCItFHJr2NciWiZgohWFg6Q7k8iWGm2L39UyTivffO6BKGOJdasa0bdzUICRutmxAGhJVWwLyx8Y0J/a1VpW3siUoCz9m4dfiW0Uaf8i8QpOcRtur9jb4BvMWbhK9VlcThI8t/mjGFZz4RI7cY0KiHTI6196L1RIbBRz6NFm/fYUBKid+okrvwDrnz7+A79U2DVTe/41bzVGevMnfiB8S7R8hs20d16gKqFmmPylCiGQqes7WdhpHqpu+BT6CacW6mEUgZ4hQXmHjjj1tx2M3vpwx8/KCGb9eYeO2P6Dn8DL1HnkE14+sMb6JvOwNHP8nM+99ZNuDNpmRs3Mf3IZe/9SIyPunzG/+siCIirzcI4H/++4Voo4iCFArfe81FtJJevh/yJ99y+dNvN6PF5KZL5Lrwu19uPWMSEJIggL/9G9GuL/JzJL/yX63MlVpd8n/9XpV//x+rhGH0mY/Ou1FV8y2qWqsvv4U+2Id1eA+xh4+QdRzKX/8BYX39s6ekEnT84ueIP3IfKAphtUb5my9Rf/tDNHRiVpal3IVlxyRYVSmoqSb9fQ/Q1bEX00ihaTFq9bllYaIgcKnV5wkCt3UdKqTTIwCoqkFvzxG6O/djmRlU1cS2C8v5iXi8m5n548ssoVp9DsvamuaEYaQYGXqSdHIITbOi8KG9ssNu2kVsuxSVsfsOQeCgqlvrMrNlnu6fu9z6PYBqxAjcJkYyi5ThhqGOG5/TE1kE4NajVikIgWYlCD237QJipXuwK+1pToqqo8fTyMDHbVY3DMEABLUqfiGPHB4h+/QzeIsL+NXoO4ppYe3cReLwkeXFqVwM+PbX6szPBRu2mheKGrU+6hhom8iUUmIXF5g/8QNK4x+BDDHjKkZMxffC5XVK1SPdCc+5oesraVYi0riiCmLpqPeXU5rAHCyCCNj+QJZ6yaW86PDgp/t5549v7fELRaXn8FNYHf1tx9oszDH91p/QWLz9opDQc1g49TIg6XvgedSbxOqFopLdfZTK9GWqM1G/saFBlb/8ywkqFcl7HzgcO35rByIIWKMW7fuAqmINjqAmkpFY+fg8hp7AFXUysQHqTp5AeqiqjhdGPeZC6RPTM2iKSRD6QIgvPUzDRFMMvMDG0BQECnU3z2pzGgRrFSil3FpeKyhWKPzen9D56z+PdWgPiSeOEjZtqj94Y5mrj6KgD/eT/cKnsA7uaXGaF6l85xUaJz4CP0BqGqEMN9zhDg0+Smd2N1fHvketvkhv90G6uw6sGq9cE4pbHR7o67mPvp4jXB9/mUp1mo7MDoYGHln5bBiFwW7gdpLF+3Z/Hs9rcO7SV3GcKof2/zyrr2sY+mxWWLYZtm502yDTswcZ+lTyY6iaRbZ3N75nU8mPgwzJ9u7FtSvUy7OY8U4SmQF0M4kQCq5dppIfx3dvsxGgUDBSXRjJDkLXxq4sEevoJwx83FoRI9WJomg41TyqbkY3SUY3oHPXg5SnLmCmulCNGF69RLM4j2+vJDu0WJLuvY9RnjyPnsigWUm8Rhm7uICUAbGOAZqFubZGd/ShF5k99zKaESfwbBrFWQK3iRAKXTseIjt0gNB3WLzyHtXFG2It6xHUatTPn8PatiOqTvvlX8OZmkIGPnpHB+bwCPXzZ4ntWtGnmJ3e3PuyOgdIjxyI+Kpt4NWKzB77MypTK1VQo0cypPtMajkn0vlN6lhJDd1SaFa8SM/BCzn/Wg63GRDP6hx5oRffDbn0dp7ubXHGTpZwWmWazYq/rhZ/w/F29NO556G27/nNKksfvUFj6S6q8GTI4pnXMNNddO59dN1CpMeSdO59iEZuksBp4rlwbcynu0ul2bxzB0QxTIyeXkLHJWzpk2iKgWaYWFoKQ40vS6CWm3M4fhUQdMZHURUjchaCJopQ0bU4AkHZnsdU4wih4PhVvNvsJ7YRglKFwu//KR2/8DPEHjxE6tnHIQipfP91pOdjHdxD5rPPYuzcBorAuXSdyvdewz53ZYXhFPrU6gtk09solSdw3AqqGp2H61YxjRS2U6RplzCMJOn0KKqyNXaKYSSiuKxdQNfjZDPbUFcRxiu1aTozOylXJlGESiY9uqYLBayiHN50/+NWBzOla/i+TSo5QDzWg23fG07zXRndbO8edDNJtTCBEcswvPd57EaBRmWBIHAY3PUUi1MnqZdn6R19mERmAN+to6g6VqKb4vwFpq+8dltVXapuEusYQI+nlktQU4N7KY2fId41hFD1qC/YroeQMojkARWF8uQ59EQ2ii9KGdW3C0Fm22Hyl1bp6EoZGfTlz0WXKLv9CMXx0+ixFI5eAHs9F9ZIdDB69HN4zSpCqFQXr7F4JYpFde04SmHiNGaig+6dD1PPTxEGGxRHSEn9wjkQkHrwIcyhEYyjDyFdF3dhnuIrP6Jx8QJqJhOJx98CQlFJDuxs6zVCtPWeO/GDNQYXwIirkRpTIEl3m8QyOjKUkcK+pYIAtxksr/iGpaCqAi2hYiU1OocssgMWmT6LwAtpVn2yAxapLoPqLXRWuw8+gaKvl/aUYUBt7hrliXN3nRCVgc/CyZdJDe7BSHWueU8oKom+7cR7RqlOX6JcCTl3wcN1XSYm7yInEEQVeEZ3htBxaIxNUnejbhW2X0HAcq8wL7Bbsc/IILhBg7qTxw2aqIreUlOTeIFDXUTx6UDe2wrJoFCm9CffQ/o+8UfuJ/X8k1HJbb5I6vmPoQ30gJQ0jp2i8r3X8KbXxjTD0GNh8QzDg4+xc/sLUSPOwGVx6SPybpVc4TIjg4+zb/fn8H0bhMB1t8YzL5bGSCUH2bvzM3i+jSTAcUqtdyXTM8fYvu1Z9u36GRy3im2X0ZNRAi9mddHfdz8xqxNNtdg++nGazTxzCyep1maZmf+A7q79pBKDeH6dpl3YcmeTW+GujG6ztkSqcxSEQDfiIASqZkblnTJAt9LYrZbRixPvg1AIfAchBP07nqBr+D7mrr+NfxtGNww8FE1Hi6UoXj8FgG/XsUsLxDoHaRbncGtFuvY8hFMrRYZTUQmcJn6zilPJYSSyOOU8jdwkyYFda47vO008u4ZTWUK3krjVAvWFcVJD+6IyV9/ddJuydO19KnOX0a0UPXueQLdSeHYNzYhRnrmAoptsf/QL0XZrI6MLSMehdvoUzWvXUGJW9HkZEtoOfq0KQcDS176C0DT8YoHtuzR27dU5/5HL3E1erxZLkRrc0zb7L6WkNP4RpbH1wi/pHpP8dAMzrnLpnfxyiEG2KMpCiaT9bsRoy4sOH35nHqGAawe897UZGiWPsQ9LSCnxnJB3vzJNs7p5JZ4WT5MePdj2vcC1KV4/TdCmGeedwK0VyV8+Tv/Rn1rn7RrJDhJ926nNXUfVfO4/bBAEUK/bTN5iZ7ERQs+leuFMNIdkSBA0ILjVuUgWq1e4k1bk9wJ+rkD5Wz9CaBqxo4dIPv8k0vNQkwmk61F9+R2qL79FUNyAc93MMTb5CoaeRAiVUPq4TkRDK5au02zmUFWTMPTxfRshBK5XR0rJhct/guuuFFjMzX/I4lI0VyvVKa5c+y6aZhGGAb7fRAiB50Vx50pthstXv42mWQShh+83l2ljrltlYfEMqmYyNfN2xLqXYavrMszOfUC+eCVq+6QOklAlKTQOJJ5E9S30pRpaEM0X37cZm3xljWreZrhro2taWVTNJJbsoVldQNUsNCOObiaRMsSuR/xXp1lGUVbKfO16Hk2PI5TbHUJUJqrH08Q6+mnkp5c9xtrCdTp3P4wQgvLkRTQzhpQhmhWNxXeadO15hMBz8OrlKF60ThdA4jUqdO97DL9Zx7ejmy9lSKxriNTQHjQzgVsvRW1zViEMPMqzl3HrBdxGmV4hUA0L320S6bqG+HYVRTXYUn1pGBJUygSV9tsav7DCLd69P0YsJti1T2duJlgTudDjKRJ929v/hOeSu/Demkz9DZz4zhypLoPSnE3g3XpLHXiSRnnlmpSbkYFYnTgrzd/aaKSH9kbl4eLm5JnEq5fvir1yM2QYUJ44R8+hp9BuUrYTikqybweFeIp6vcD7Jxx+6hMxdmzX7tjoqvEERncvzsJc27b263CjMcDqxqU3aJvL/79JnPam8ig1nkKoStRy6U7ikBL8xTzFr34HYRpYh/YgDJ2waVP5/htUX34L2dzsXCSe18Dz1i8uUgY0VyWobkajuZY/73q15c7AHfFRgtCjXJ+lfahOYjslaA1Ni2uQAFM3EapAKnXS+2IsfRTlYRRDRbMEYUUh8Fwajeh1RxRZEJfZFjvMojOOJ1069IHlWLMkxHa2Hnq4O6NbXSKUAfF0P1aym2pxini6n0RmEKGouHYFz6mhKBqZ3j10D92PEUujqga6mUQ3k7f+kZugJzJ4tTILp1+hY+cDlCbOUrgStXD36mUWTr/MRuyE3IW315U3Lpz60doPSUnuwjvc3Mtk7oPvAlBfGNtwbPXcJH37PkZ+7EOsdDdGIkvPnsfxmtXI84+lkWFA6Ec148uiCWGI0FSUmElQt1EMLSpIsSNNXEK5/LmNMDnm89QnLBbm1xpchEKsc2idQVke88IYzs3Jv9a43EZIwfEQugWBE9HEFAWh6YTNBkI3gFZ329Z1FboOYXhX3WwTA7tQ1HZTU1KduYzcIgtmq/CbVRqLk6RHD6x7L9Y1iB5Pk42VOXzQ4A++Uidf2DyskVAz1IO1D6EuTBJqlrJTQEtnUVNpQrtJ4+qlTY+Vuv9B0o9/jJl/8y9QEylU3UTPdtKYGcfqGcQt5Yj1j0Sc6KnrGB3dqPEkQtOxF6YhCOh44AlC16F2/QJu4dYt4DdCkC+R/90/ofNXXiR2ZB/S8whK5YhetgkUoWJoCQQqbhAxYjTVwgsaaIqFlEGrYEgQSh/Hr6OrMTTFQBLieDU0NersLISCHziE0qPplvBbHUG2gqEnRqjOVBl8fBgpJYWLOfSUzsjHt6EarU7SoWT23Wnq8yshDk86eNJBRaUZ1lodiiWKuLOmuXdldH23jtsskUgPYMY7KC9dRVE0YsleFE2nUY5I8enuHey870Vy06eYvfYWnl2lo28fowd/+rZ/02uUcetFYl2DlMbPbMAC2OQmbHWlv4PWDwuX3mTovp9i9OEXkYFPYeIMiqphpXsoTp2ld++TKKpGozSPlAHGYBeh6+EvlUFTiR3YRvPSJNauwcgzKdZQU3H8QgW/2iAobFwZNLpdo5gP0fWbEgWKSrx3tP0pypD64gR+c219vJbJoHV2Edo2imGgZbLYU5MohkFQq2KOjNK4cB6jtxclkcBdWGglh2zUWJzQdXBmppHe7Qv6KLqJle3lZjHw1oCpzl1b85JqaghVoMV0nJKNDEKMtIka03GKK/Sm0A1QTRUJhM5aLzVwmjTzs22NrhZLYmZ6qM9Mkc8H7NqhoSg+S7mN58ce6xGu26epBNFiFlNSjJgHUFAp2Ys0xq5EC9ZtzjE9lSXWP4LVO0jQbBAf3oHfqKLGk6R2H8LJLxIb3IZqxQkcG6TEnptEhgF+rUxgr/I0FQVjZOs9CKWUrdbwAeU/eyU6ryP7STz5EEG5uiyEtfZL4E7NkrJ66UrsIGwlAUuNKQYyh6k5S2iKiRs06EntodKcR1U08vVxOuPbCKWPppgUG1OkY/0oQsULbKr2ArZXZaTjKOXmLIvVy8vdhDdDZbJMebyEoglkENkBr+5G5fmGGo236uA328/borfAgLmLUAZI5B0nLO/K6ALUS3PE0/0YZpJmbQmhaPR1bkPTY8yPvQtAPN2PECozV9/Ed+sgFDQziaptjde2GqHv3V3W+scIu7LE+Ptfx0xkCX0Pp16MkoRCoOoWHSP3oZkxSjMXCMMArTcLfoC/VEaoCmomjpqKYwz3IACtOwuKQImZeKc2lw+8UZmk3GSrhKIQ62yfQAtcG6e0uC5MosSTmINDyCCI9Al0Hb1eR00ksF03ojsBQtcx+gZQY3GMoSG8pcUo1pdK4S4u3pHRNVKdbUMLAEhoLq2lm2X3dpPe3YVbsXGKTfKn58ju70FPmKiWRu7kDPH+NMWLi/Q/sY3CuXmai2sXmdD3cKr5lszkeu/F6hygC4NDB9UoUajBUm5jbzvnTbLNPMSkcw4QbLMO48omk/Y5hKJi9PRhdHYjw5DKyfdvcUVWCfD4HloyTWNmjNTuw9iL0xgd3QR2Y9mgysCnOR/pEghNj/IQtSpepUjQWDGMwjTo/LUv3OK3V49CRhQw30f6AcKIWj2Zu0bp+IWfQTrrr4eUkqV/+m/JWEPoapyGWyRl9jJXPkfVWaQvtY+x/Htoionj1ZgunqI/vZ/u5C4SRielZqS/ETc6UIVOqTlNsbGiuVBpzt1W4rB4JUpY5s7emWxmxV8irqYxlBiObNwx3f+ujW6tPMPo/k/SrOdwm2WEUDFjWXQzSb08C4DTKAKSvm2PUM2Pk+wYobN/rdrYfy4I3AaNm2lwUhK4TXLXj7dUjaJV2b40tVyPKR2P+odXCOo21XfOgZTIIIyq+6QkbGy+qk6N+9z3kIHnrp0FQlE21FfwmzW85nrv2VtaIKiUopCBogAC6bkIVY0y7pejmKozN4tXLEYe1cR4q6BDIoQStaO5A+jx9Ia0Nt9trBuv1RUndAMaMxXSOzspX8mjxXSSoxkSQxlmX7vGwFM7qM9V6HlokLm3xtscWeLbdQLXbhuGsTLdLFyCUjmkM6uysLS5VzXrXiMkZHfsITShM++OM+tewZM2aiKJouuo8fhKeGn17ktVyTz+MeK79+FXygTVyvJjEjQbyDDAyS+SPfIYpbPvk9i+FyPbjWpYUShXRvNrdbLXq5XIHDgKQHN2AgChKhg77k4m9Qb0wd62r99gl/ihQ9MrUaiPkatdBcDSUvihg67GQEpU1YgWeCEIwkizpNycIQh9vKDJUPY+/PDehpVuF6OxIyy6E8iWob9Tpsg98HRnUHWLenkuyv41ClF1RmjRbCXRCvMXMeOd9I4+TP+Ox6kWJpm79jZDez7OnVaGWT0DxEd3Ufjw7Q1DBmZ3P11PPE/5zPvUJ1Y0UxXdoPvx58m99zLhbfRuuyVu1ncTSmvyyxZVLVxztqsNqfQD/HyU/fXt2x9Td5/CD77VYO+BtQZL0YxWRdd6+E4D314vvSc9j2ATLzVoVctJxyFwtrbFih8ape+Xnmb8N76yaQxQsxIbxHMj49FuSx76IYEbICV0Hu7D6ooz/dJVdn3xCL4dYBcbDD27k+LFJUKvfQIscG0Ct9nW6OrxDM2m4JXXHZJJhXx+/RgU1nrIi94kmjDIan0seuME0kOgEDQb2LMzyFCiZ7Lr5m7qvgdIP/wYxVdeAqDj489By/v2G1Xy77+GDAKmvv47hJ5D+dyJFq1REroOlUunW0Y3mosyCKhPXsVemF4710OJXyi1vRb3DGFU2FBqzjCcvZ9tnY9Sas7Q9Mqois5E/n16UntwvBpxPcu+vuex/QozxdP0pvYwkn0QP/SYr1y4qRBBkLEG6MvsjyieWozZUpu2Szd8Otn6Q7mhcggylAhFLHP4twZJ0ZtdLtDYSkijHbaoMraKQBzKGyKUIATNZp7j3/8Hy2I4EsnZt/9ti6qlIFqlcfPT7zN7/a01CaHc3EdblH1rA0VFteJoqQyEIX6jjqLrKLoJSPxGDSc3T3Py+spFVRS0eBJFN1CsWBR/9KOa79YJtv69crdUxWhVnwQrnxEgUNbx9nr3fIzAa5IfO4FqxBk89Ak6Rg5HYYfjf4pbv3Ohnlthbjrgpz4XZ+zqWoOmx1MbSmmGnrO1DPo9gNBU1IR1yx1ZJGrfflreHHuODizoOTpIYjBN8cICtakyvQ8NM/SCiVuNkixLH8zw6N97gff/3o/Wf78FGXgbdpHQYkl27tT5S5/X+ehcxKEtlVfmbYfWz/7YE2uPh4xE8hWLBxOfRCKpBEucd49hDgyiGAaNifWykemHH6N25iS1j04BYPT3E993IGr7nkzgF4oIU0eGAbFDB2ievwhhJEEmdA3pRuegDQ5Ez0WhADKSMERRUOIxZCiRnsfc//rPo8/fgy4qm6HhFri8+Oqa14qNKERYzx+jIz5KsTHJeP7Y8vvTpdPA6eW/K/Zq0SRJ2Z7l9PTX0ZM66YEEcT2GogpUQ8Vr+BgJHdVUCf2Q4vUyscFROh9+Bul5GF191McvYfWPoCgqU9/4nS09BxLJ/uQTNIMoTDPnXMUOt6hdvQq3NLqKaaF3dS+3xRaqglAUQtcjsJvomSwyDNGzHTTGrqHG42iZLH6piBqPtzidMlJ9n51Cy3YgfW+5nYq7OH/HmW6js4f0vvtRYwkaE1cxsl0opoVQVRoz4zQmr658WIDV3U/m4EP4jRqqFSe57xCNyTGk56HE4suFE4RhFCOTkpQ1QODZeG6dMPRRVQPfb2KYKWqV2TXj6Rg+SH78Q0CQGdhLvHOI2bM/Ij2wj759TzH14Z9tej5C19fEQfWOLoJmg9BuIgwTs28Av1LGLxdBUTD7BqOy4qVFEkmFqXGfenWtSWtXYHADUY+ze9O94l5huSVQGwT++gdDBiEzr1xj4fh05BAAZ//1uy1hatDiOqltWWbfGserbfxghYHfvq0Q0W7BdeHSZY8rVz0KN7EXakGJy81bxWajLLiUIdJ1MYZGMPuGyL/yvTWfUdMZvEJh+W+/XIqen55uzNEhmpeuYrQMqppOY24bAQRadxdhs0nzo3PRmC0LLZNG64qKPqTvo8TjUfdjJxKpEjELd3Iab+7WIi0/Trh+47akVVej92AXnTvS2BWX6lydwaO9lKdrdO/tYOKtWaS/sqDE+kaYe+lrxId2EB/eSe7dl+h77vNYvUM0pm+tm7zgjJHSOrHDOr70fnzhBTWRJLZzN6Ft07x+ldi2nUjfJ3QdvEIBs38QVGVZXV9LZbCGt2FLid7RhVeIiiPUWBzFtIiNbF/2PEPbxq+U2rbY3gq8apnih28TH91Fcud+3FKe0kfHEJpO9+MvrDG6QlEwuvpozI7TnJ1Az3SAUNC7ulFjCfR0lsBuRIUUjTrOwhwIUEOLeLynRbgOIvnGRoFkeoB6dX45PgugGhZOrYBqxEj27qA8d5n8xCk8u07/gWc2PReh6ST2HKB2/szya4ppLa/AQlWJ79qLuzhPrVoGITD7B1DjSXqSFXbuFUxe99eVqAp145JKGYYrVV2qQtenj1J64zxCEWSfOUT9/BTNsQUS+4eQITQuTBHbM0DyyHaEqeNMLlE9NUZYj+K35mgP5mAHQaVJ/OAIQlGofniN5tWbpB0FWNt6ie8bonL8Cn5hVYLnRivztuNd75GVx4pILwC54kOvLjUWmoKiq8y+fn05Y93+4HLDWnqhaBSKkumZgGxWoVRZa3Q9aZP313egFigIBOEqBQahG2jpTDT3a+vj6aHdjFrQt6CY0aIZFIuEPd0oiTioClpH5Owo8Xi0o9Q1hLPquoUBzsws5vaoeCnIF1Di8agiTlVRs1mkbUdb7D9n1N0cdXd9x+qtIvAlhSslKnN1nLKDois0ck1K42WMxMr8D12HxtR1FMNCS2Vpzk3h1Sqo8a1RV1NaBwKBLsyI/y8MPHn7uYtbGl0Z+DizM7i5RYJqlfql863XI4nEoNmAMEAKgfQ83MU5/HKR0HPxy+VlTqVfqRDaDeqXzi97lKHvt/oc3RlueKOEYSucIaI4qqqtKy2OVK7CaOvaKtCoXb1AaDdRDAujqwe/UcPs7qUxORYZOwllWURVdcIwyuSHLWlIz2use0jdRplE9yiKZmIluyiMn0QGfmRIbkL66GNoyRR+pUxzcoz47n0kDxxBTaZoXLuMUFVSRx6kcvoEQb1G2GzgLi2s7AqCAGduBmtkO7VKyHuveyCg2bjJ6G7WpUKwHEkRikL6oV3YYwsITSX7zCHUdBxntkDivu14CyVkEND/a89RPztBUG6QfmIfem+G/HdPIF0Pc6iTvl98msrxq7iLJRRDQ+jqut80B7vo+bnHaVyZi7jIN9+o2+hXVp+rEdu+E0Op486uN3xexWHpxPrXbw8h9UbI28e2NlcVVAaN3fToo6hCx5VNpp1LFPxZpO/RGLuK0LS2Qv/1C+dJHr4Pe3IcgPjufQhVJWg2sa+PIT2foFTBUaIwnXSciButqkh7ZXzu/CIEAbbjLIcCvaXcyoIWSlAEYX1r7XTuGgLUhIl0fBRLJ/Si58noTRHUbNxCHUVvXRMZ7S6iEn6VMPQjGVEhCNy1Ri53qUBpokIj10QGkuLY2vlkl1auyXL4KAxbMqVRnmUd5WcDJNUuxu0zdOvD6IqFGv6YeLp+pRx12byRibypOupGS+Tlvx1n2ZCufa++/P69gVzeTiIlTnEJGQT0PPlJEArFU+9EHvDug8TsHaixBM7iLF2PPovZM4AwLIJaldB1CBp1vEoJZIi7tIBhhJimQhhKQl+iJ1Tsmk88o2OldMrzLs3q+uqaxctvs+OJXwIpyY+foFFoUV46h3Cqa7dPqYNHWPzeNwjq9UgofGYSo7Ob6pkPCT0XQRRzVTbQ012NSjmknGsfl9ssfCCWE32ADLGn8hgDnQhDo35hitjOPhRDQ0vGqJ+bpOtTR7EnFsh98xjSD3EXS/T8/BOU3zqPl2v1i/JDqsev0Lw2R1SFt4ry5AVoHUkG/srz1M6MU3jp1Lqk2nKIpw2Um6sXhSC2fSdaJntHvOo1h1KUtosjRKGH28n3Dpp7GTL2sOCN4YZN4mqGfbHHudJ8n5w/TdDGw72B8rtvoXd00PcLv4JXyNO4fpXE/oPgBwTFUvsvtdG5lbYNikJQvqk093aaed1DWANZYqOdmN1pzL40zmKFoOGimBqVj6YRZZvsQ9uw58soQRbNTKGaMaTvR1KkQuA3KpSvf0TgrNgVp+zilO8wGX6bl8CRdUatA8SVLNWgQNG7M3H+WyfSpPyJ36CtwF6YwV6IjFp98iq0Qgnlcx9EH2iNuTG1Vgd47odfW8mVrT6vG1vX0OPw8/3UCh47H85SmLaxaz7d2+IgJZMfVajl2y8clfmrnP/+v0A14tjVxeWFqrZwHeemHmH5135I58eew56fpXrqOKHrRjFWO5pQEpDe7T3s7bAZOyMSH4+mgAwl9tQS5lAXCKidHid1dBdqOo6asggqTYy+DPmXxpe9U2e2gNBU9M4UXi4yJF6ujLtURvptjKCUDP23n46qgX50ui2LIRJLb29AVfOmjg9S4s7PEd+7H2/pziutoHUtNkjghV5UQWhZYrk/2WZpiGFjD5ebxyn4Kw9l06gwaO4h528uaxk2Gyz+6VdWXpCS0huv3M6pAKDEDJJPHKZxdgzF0AhbZbp6fyderoy/WIraat2ojAwl0nYJPR9WxUERAiURQ82k0Pp7UFNJhBbRCG8JKam+8i5IiV9zUE0Dv2Zjz5cQqoIMQsze9PJzqMQMzL4M3nSAokVJ8SAMsYsLIMPIGdlCE4QfF8Yap0lr3ZhKjIqfpxneWSfnu6aM/YXDzQvERn9vZMwkLE00KEzbeE5ALe/SvyfJ5OkytZLbmmobTzivWVnXbLGycHXthxQF6QdUznyINTiMms4SOjaKFcPoHcCvlBCajppMoXd24VWKSN9Dz2QJfR8tnSWo19A7u9FSafRsB25+qS0TxG9WkVK2DTOourmSaAslznSezJP7sSdyePkqzlSOxMERhKriFarIIESNrXjeih5V8YRNd/mCRl1D2l9cLZsg/70TJA5vo/NTD1L43ol14QXfqW/onWux9YpqoesiHZtNe91sAYpmtJgv6+E3qqST8PRjFo4juXzV5/r4xlZXEwZOuHYnZId1NGEsl1Jv2mH4VnNYUdAHevFmNkmACQU/XwYpsfaPRsyGpoOIW/ilGkLXSD59hCBfRe1IElTqhI5H/fiKroXQdaxDe0g9/yTmnu1RGOM2IMOQ6mvvQSDxK01yr2+umZF7ZbXK3Rjxvm00FiZu6zc3QmA3ac6Oc0Oo31mai/jl89ORJsUW0G2MkHOn6daH6NAH8F3vxxPT/c8FimERuhtfIKOjBxn4eJUik2eim1ArRF7i4thdKloJgZnsxqlGlTBCUTC6e0AI7LkZ/FIBGQQ0x66hd3YROk0UM4aXj7w3NZ4gdGyCRhSiUROJKOasKPjlEmo8gSjk2/IGA9cm9Oy2fcFUM4a2isPrl+oololQBEG1Sf3iNOlH9mBP5QgqTernp0g9vIv6+SmCWpPkAzvx8lX8ytauT9BwKPzgJI2L03S/+Bh+vkr5nYtrMsx+s7YhdctIZKJ4/SqWgVBVWk3Cll9TTBM1k8HL56OQwY2CAU2L6FFSrjN6qmG1vUYQidj39wr27tY4d9G7ZTPnsp9j0NjDvHcdX3oYIsaAsYeSvxDd+95+lFgcv1yKEs23uZNULJP088+Q/72vbPiZ0HFxppfAD/Dm8gSVKP+hJGME1WaUU6k08HIl/FIV6fmoiVjUXgJAQPyx+8n+/KdRUok/lw7e98rgAriFRRZe/RYQFYjcKBJZeueHWz5Gtz5KLSiS0ruRMsBU4njBfxJGV5DReyPeond3W8KtQtFN0nvvo3R2Y1pP6Dp33TttI6i6xeDh5xl798tARN+pnj217nPVsydX/VXCXVwbM6qc+mDN37U2x7gZMgyxyzkSPSPr3tNiKbRVvdJC2yNs2ISuT1BrRsbx849RfifyQAo/PEW3ZTDw689FGfG6TeEHH+JXb6N9UihpXp2j9MY5Op47gpur0LgwtbzzcKvF1nZ+PRTNwEx1rWkaKYOAoFZDMc01rxn9A5jDI4S2HXmWocQvF9G7uvFyOdy52TVFLHostWElnFNZIqz5dHZY7NsNuVzItY11jxh3zrDdPMKB2JMtTreg7C8x41xCyoDQsYmN7sAaHMHNLdAcu0bo3Prh1Xq6iB2OKjmFaZD82KNIKVGMyAOtvXdiJZkWhAStYptmuT2/uf7hZYLWe0pylcEFtL5uks8+jpJKQChx5+axz17Gz5e2TvGUbLjr+U8RIR4j1kEW3XHiyuZNXTfDT9zoCgSWmiC8y8THVqHFU3Te9ziJ7fvR4ykq185idQ+ip7IopkV9+hpeKU/X0aepjV+iOn7prpMyN0PVY8TS7Uslf9yQMsAuzrc1uqoRw8p0L3uPfrnO9P/5XaQXENoezavzXP87v49fih5Mv1hj8atvo6VjoCgIJJmdHTRNlaARUjs5RuPSDIIQoSp03jdI/mQUw2xcnGbiH3+N0PWjGN/UAsVv1rDHltaEerxGBb9Z2yAkIoj1jNzUqTeKSSrWCh9ZSSTQOzsJanXUWAwt20FQq2KPX8fo7QNFIFR12XgouoGR6WnLD5ZSYhfmyS+4/Ob/xyUIom68m6EaFLjUPIYhYuiKiRs2cWQTX7oI3UDv6qE5MYZfLUeCQqk07q2MrqZhjAwRlKsEtRrmrm1ofT14cwugqGgdmYjnbW8xUS3lssEFCGtrF05z5za0rqg5rHP5OoXf/zpBqRLtSm7HM9/ks/pIL5lPP0nt9ZPYlyfuOn9xA4lHD+LOLOHNbE1jIfXCoxjb+ih++Udte8DdwPXGSRSh4oRNGkr5jsuS78jo6orFUOwAcS2DoVhMN87R8CuMJA6hC4t5+xold46dyYfQFYuSO8+CfY1dqUdQUFGEyqK9iatwD+E3qlSunUPoBkvHXwUZkt59BHtpltrEpShpIyXN+dsX0VF1i6H7PkV5/jLlmQuMPvR59DZxR0XbuEDhZnT3KKhqtAtOpQSptMKZUx6qCh0dAolA11kunpMSHAdiMaKW3Gr0RqMuqVajbXRjcZLOPQ+vM2JRE8YdaLHjUYltKNdwZqUfEEvpWDuHsXM1rJ4kbrmJUBS0pEHl8lLEOmtxPUPbRbNUhj61n+LZOTIHetFTJo2FKmZXgtD2MA73IgOJDEP8mkNqNEvlytIKt1aG1BbGSA7uWs8xFoL08D6KV1Y8fum6lN95c83HgnKZ0huvL1Olog9GF6t26uQ6Q6CZceLd7XUIvEYFp5Jn3y6VZFLhxCmHDWoo1o5B+vjCIwxDfOksd7+Vnktz4npLlF7i1bfGUReaiohZeLPz+PlCpMsRhgSVKjIIUBPtS703PqBA6GqUrEWsvya9XSgxC+n5lP70B/gLEY/WsgSeJ9G0VjcLVyJlxLoyTYHnShRV4LY0QKL5EVXgtmREluFNL+JOzK2peBVWtNuQro8wNKTjIgw9EtmBNaEoAGFoy73upeNGVbLnxwhbpfRC10CLumpI14tEnFohtCgf4VB/5wx638dv2ZfelfbywnBzzP52cOeerpDMNC7ghk12JR8m707TDGoseNcZiO1FaXkNU41z9Fjb6TG3EcqAK7VjDMb2bXrozRq+bUTr2fR4N+J9LQ9WhiFetbQmrndnTeZkVCXVOn56YB+N/NS6yilVSnSzvZ7tms+p8OIXYjSbUafX6SkfpTUh+/pV9h3QSKYULCuaxD19KrWqZGLc5/6jBqViyPh1n0IhpFQMqVYjHmJjaSrSFWijwRDv3YaV7YuM7s1QBKmdXSweGyc52kFjrkL2QGQ0K1dzxAbSqJaGZun4tVa31nKT+nQJe7FG0PAoXZin68FhvJpLckcXQhHMvXyZgU/sQRnKsnhsfN21r05fpvfIxxGKtm6hSPRtQ4+n8Rqrkh/t7t3yPb9FUopI2SzWNbj+GEAzP4vXrDK3EPCrz1qkUoIrVz3GJjYORcWVNDutB0ipXRG3W4YsehNMOecJTIXkofujIqEgYPHPvrrhcdYM2w8IG030vh4U00TcZCBuZ/bqPWlUy8Da0Uvj4ixCU/DyVcJVbBJhGlHRU72BO7ai7PXcUxZzCwGlUsCOUZ2pWZ8gBFUVHN6vMzUT9ZF77wOHciVkdETF0BUyaYXxSY/c6mo+eaOQJVocjZE+kk/djwwl7tgs8Qf3UvrWm2Q+8yT2xXHChkPj1JU1ZcuZn34CkYghGzbVt06hxGN0fOFZKi+9j3N1iuQzD2KM9kchksl5Gicv0fHzn4iMsSIo/MH3IwP9E4yC3LHR9UOXoFUKpwptWaTZD13yzhQKGpaaIql1UvPyhDKMyuakJLxF+dxG2WshxIYZ5s0Q2HXUeJLEyG7s3HpunZbMYHb1IT0Pr1bGXpptc5Q2x/UcZk5/f2XcnsPEB98k8NZuUfR4hj1P//qWjlmtSibGfIJAMjsT0NkVreK1asjCfEA+F2JZgnI5pLcvJJ0R1GuSUydclhZDKuWQWFzg2CuzyGtWqC9MkGmnF2vG6N7/OPWF8fWx1FBSvrpErC+NU2yixXTKV3LLhShOoYFqqpHo+g1IsJdqCFVQvryE3/BoLtYiQZ+6gwwkge1RuZLDKdZRbnghq4xhMzeNXZgn0b9jzXCEEGhWksyO+8ide2tL1/NWEIpGx66j6zoCQ8TPrS+M4zerVNyQP/lW5N2Uy5uHn7ZbRwikz4Xm23ihi6nEGTEPMGTsY5LL2FPjSMdBibdP3LWF7+NOTGHt2YlimdjnL+FX6/j5Ikq1jqw3l3UXboX0gztQTB0ZhOg9KdS4SdBw1hhdfD+ikXn+moUqnVLYMarzw1cbDA9pHDpgICWcv+SSSiocOWSClPT1qtTqIY89ZHHitEMyITCMjZNxQlUxdw3RPH0F+8oUXX/5M1FyryNNWLMxdwxRffPUOp2I0HYJSzW8xQJBqUaQr2Bfnlr2iKUf0DxzFefqFB0//wma58eQQYAzPkvYcCLRqZ9wVd4dG901almElLx5dMVEEzpOWKfi50jpnWhCx5cOVT9HtznMcPwgppqg5G5Md9mo/5VQVIxkZ9v3NoNfr1E+f2KZA1odO49fXcWblWFUey3ZMHO+FUyf/h5Bm7Y3gWtTntu8QwBEztnrr9gszK881LMz0b8rFcn5s2sXq4kxH8MUVMqbL9N+o0p19gqpod0obcqC06MH6Nz9UNQx4yZUr25entmYLq17rXI5iqU15yMeY+HU+oqwwumNq8RkGJC78O46owtRUjS7/TCVqQu4lTur11+NWM8Q2R33tX3PrRapL4wjA59EWvDCsxaWJTh23OHD0xvPkw6tn9O1l6mFJQDqYQlFKIyaBxmvncHzcgTNBoqxsQNhDA1hjo5QP32GsBE9D0GhRP3Yh7c+KSGIHTxA4vBhUASlH72Mv7RyH2vnozi7dH0C20NNmMvb8RvwSxWk66LELJREbDnWeeyEQzIuKJZDTn3kggDHkThuyMSUjwR8PxIECkN457jN3HxArS6p12+RKwllFIto7W7c6SVih3bgTM4TO7wT2UZ9r/bWaYxt/Vh7R8ELcK7dxIP2fMKmExlhVSFsRs0vvdkc3nwe6XrLIY2fFO7I6Pqhw6J9nSD0kIRcqR7DCeq4QdRuPJAebmgzWT+LIlQC6eOFTcZqJ7kh+rlZENqtFtq+LlSNWPcQQlFvk2kgqU+tcGUdZ60n6ter1OpbJzrrhsD35bp8W2W+vdB46DvMX3zj1qOUrDG4t4Jtg23fel8kw4D63DXs4kLb2KWiGfQf/SmcWoHqPew/djeoTJ6jsTRFvGdkTYhBCEG8dxsdux5k8czr6wTYbweKbjL46M+sL7ogCkHVFyeWBfNHhzSqNUlPt0ostrlnFMoATaw1qLowCWRLdCaexC9GuiUbUca0jizWrp00L15aNrpbhpS4k1NIx6Xnl3+R6nvH1hhdZ3rtYhW0ofy549ME1TpaV5bY/QeovxMZ+/HJlYU/X7h1Iml6JnpOFxZvel4VBWvvKPH79xBsH0CJW9jXZkg/e5T4g3tpnr6Kly/R8XPP0vjgIuG2vnU0P6FrJJ95ELUzjWIZhLaLdXgXsSO7MIZ7aGTXayqoiRhqKkH8wb0E1Qb1986SeOII5t4RUk8/QP29sxG/uQ0SSpZ6WCapZtGEQcXPrdHV2CruyOhGrSpWPDo7iAzWzTJnTriWqrJVGbRGbpqVsrHVEMQ6+kn076A2e7XNN3/8sOIKz/98J6Wczzs/KG2Z6BC467OiWjZB8v4dVN6/HBUYCIHek8FbLN3bQRNd08rUBaxsb6sr8Nprq8VTjD79Jabf/SaViXP3hj4nxHLC6HaPF7g2CydfYvvzfxlxUxdjRTPoOfQ0TnmJ8vjZOxqropsMP/Eiyb7t7RtgNsoUrhxfDrlcvurTdCLlsnp984Vu3h1jb/wRZpwruLJBQskyaO7lavMDjIE+kgcOE9pOtMO68uNZ5IJqdVkZ8E7gjk1hn7lI8tnHSH3647jT81F79btse7+MMMS+PInzL2cACa3EYOGPX4oSZkGkj7z0r75C6Pq4E3PLRldRbtTC+LhvHcd1W3lS38eby+FcngTkiraFjCQD8r/zbVLPP0Llh+/hTi+S+dTjAFS++w6V778bfb6NHsYNbIsd4XrzQwbMPSgo+NKlFty+XOtfyOKIxtIUgbNeUFoIgZnppvvAE7i1Im6lwN1GwHW9VfUoVzj2ihL9X7QEYaI2OAK7KfHckOKSRyK9cTmiMDQUS0doKtJtNW0UoMRMhKLgl+tRWaaiUD8/Reh4CF3F2t5H5mMHyX3rGEG1Ealn3StISe7CeyQHdpHo27GO4C+EwEh2MPLUF8l3DVG8dhK3tjFndsNzV/Wo0MCMEe8aIjW8j/r8dfKXbi19eDNqc9cpXH6fzn2Po6gr11sIgR5PMfKxn0fVTcqT5/GbW2QBKApGqpve+5+lY/dDbctKZRhQvPohtZmVnUsiLnj0qMHsfMjM7Ob3ZcI5S0jAkLEHTei40uFa8wSL3gRiRsPNLUbl3rc0YAI1nUaJxZBSElSrhKsU+ZRkAjWVjoRgGg2CSmVLRlHrbvW/i8VQDCNSPCuV18ZLQ0n52y+jdncQu+8A3f/1L1N7+R3sq+OEjeZ6KsIGCEqbVHuFIdKNxtvfr+B5gmw2ClfEYiqNRkgQBMzPSzLJkFRKxfMk2WxU8GJago6swnvHHJYblciVY94M6fnYF8dJfux+YvftxptZIijXWgyOW0MSktK6aQTlZSHzO8FfSKMb+i6lsTN0H3hi3XtCKGS2HwagcOUEjdw0frO6wQQQUQ13q8RT0U0Cp4FXLwGwbYeKokYdQEXLMGpalEPo6FDQdLCbMpJlEHD2tIduKLiuJGu2FOjb/KrR30F8/zB6NolfqSMUFWcuj9GTQe/JYE8tUXn3Esn7d2Bt72Xpa2+DIkjevwNjoIP4viHqF6YISvdWAcqrFVk4+RLDH/siZrqr7Wf0WJK+B58ns/0wlckLNJam8OqlqJ2N5yADH4lEKCpCUVE0HdWIKtu0WBIz3Y3VNUi8awgj1Unou1Ht/B0gcJvkzr+L1dEfLRQ3MVe0WJLhp75I8vppyuNnaRbn8WpFwjbdglUzjpHuItE9QsfuoyT6tm1Qxy+pTl9i8cxra151vWiOPPW4iedJlt7ZRJuXgAnnLJPOeVSh4ctI+PzG8QVgdEUsBHt6Y6qims2QeuJx0FQUw8AvlSm/8gpBqYyaSZN94QWUWBQakWFI9dgxnGu31oXtevHz+NUq+AFqOo1QFcqvv4l97dryc2RsH8bYMUyQLyHDAL2/h45f/jxBpYafK0Sx0U28wmhQkqX/8/e2VCBxYL9Gb59KV6fCwmLIju0a0zM+r74aXecjR3T27tU5e9ZlcEDj4qWIRmnFBKraXho/PpjGztWRoUTRVcxsDC+Xp/HDt3CLjUjM/TaKN4reHAk1w6x9haTW0bqvt4+/kEYXKSlcPk5m22H0eBveq6qT3fkAid5tNPIzuLUigV0nDHwE0ZZW0VuGtmVwVcNCUXWK10+SO/c2AAcP61Srkp5eBUURnDzhsnuvxrkzHvsP6SChVAqZHPPp7Ioe+FRWJZVR0TfJxEovQLEM/HIdY6ADoWtUT11DS8fxa03M4W7gEvWz42jZyJsPyg1qZ8ZACCrv/vjiqpWpS8wd/x5DT3wefVU12moIoRDrHMDq6Cd0bdx6Eb9Zi0qKA7/FXlARqoqiGWhmHNVKoFkJVN1gS2IoW0SzOM/8yR8x9PjnsTr614UCFFWjY/dR0iP7sYvzONUCfqNK4DaRMkRRNVQzjh5PY2Z6WuGVjRMn1dlrzL7/7XVtjBJxhZ4elakZv227nnaQhPgyWgAUVHr0EZa0RayhEfSuHhTdwJ6d3tA7VRNJau+9T+PiBfSuLrKf/CTm8AiNUpnUE08gDJ3id76L9H1SH3uS7HPPsnB9bEseqN7dQ+4rX0W6Lulnnib1xGM4ExPLIvrxxx8k9fzH1urtCoGaSaFm1j+Tbc9/ebt46/GUKxLXC3j/fRdNE4xd9ymVQxYWomtz/XpAIR8yNx9Sq0nK5ZBGXdBoSDyvzfEF9D46SmOugpYwqE0UifWnKJ6bp+NgL17VoT5ToblY3fJmueDNkZbdZLSokOZWLKyNcNtGVyAwlchQhDLAlbdRAroB4kqKRkuxJ6V2Uw8KNPOz5C8do+/+T7Tl5gohMFKd6MkOovhNsCwoLm7EElsanDcQ+h61uZVY8HtvOQgFVE0ggEI+JL8UUK1Kvv9nTaKu0xLHliRmojHUKgGNWoBhKRvO7bDpoMYMnFId6QeocZPYzn5QFfx8FTVhtbdLgVyvP3sTNCuJZiZwawWEqi97dYmuIexKLupM0IpxyvDGvyWKZoIMCQOP0ljUR2vwsZ/BTG3MBhFCoJoxYm0STT8xyJDa7FVm3v0mI8/8Akayo22Rh2YlSA7sIjmwizAIIo1nIu806kaxSjy4/Q9Rm73G7LE/wy6uL0/PFQK+/NU6oYyKTm4XqtAZNPew0JjEbzTwSpdQ44lNwwFePoc9Pk5Yq+MhonLnFs0sfugQSsyi51d/OTp+MtnyWtUtxXGd8XH8XJRcs69fp+Ozn1mrKysl+P5dBe9uh/t+5YqP60o2Un6dmQmYaRFebu7c0f7HIbB91JiOFtdRDJVYb5L6TIz4YJrA9vGqDs2l2pYr7HbGH6Dul0moGXw8an4hKpi4TWzJ6GrCpEPvxw0buKFNUu3AkzZJtYt6WMQNHZJqlpK/sFyp0WtsRyCwwzqGYuGFDlJKTDUelUWGDUwljhPWSWs9eJ5LRusjqWVp2GUC3yV/8T2sbB+ZbYeipEwb0Y0bD1MkN3d7KBZXtnw3UC5F/y7c5M04TvR3sxZy4rVqq7ld++MGDYfQdpGej5evIr0Av9qk85MPEJZq4Hlo6TjdLz6OOdoDAoovn8aZK9Az2kPPl56i9NpHeEvrs6i6lUK3ksQ7hnAbRVQ9hqJq6FaSZM9OhACnXkQ1YjjVPI3iLDIMSfXuoFGYwW1E28XS2Bl8u87Qo58l1j284fX9iwAZBlRnr3D9+7/N9k/8yorHu8F4lRsiOFs5tpQQBlRmLjN77NutUMj6h9DzYHa+fSxXFyZJtWPT39GFiYYBYYgzO9Wqltt8zkrbWYmz3hB3b52zErOoHXufxtmza7+z1cTZamMfBOscm/I3fkjlO7cvKbkOtwpBtHAnC9mtMPPylaicXBFRQc/1PNIPGZ+LHLzQ31gRrx0UVHLeJJJhQoIN20rdClsyugk1gx3UqAdlVKGiKTpe4FLxl0hrXSQVlTn36pqeQZaSiGTQjCEqfp6Kn2PA3E3Bm6VTH4paXgA1WURTDEwlTiBd3LC5POfdaoHZ97+NlJL08F4U3fwLYRj8VdsZQauFOhJB1JhTugG5bx5b973g0jgvfirBkYMm/6BWZe7frVc4mvwnX9v0t71mGTPRge9UsStLZAb3RwkJGWJXl5CBj5HIEvouVqqLRnGWMPAIfXfttlqG1GavcP1Hv0vf/c+S2XYYLZbaUFP29hGVh4ae0zbGevuHk9jFOa5+97cYeOhTZLYfQbMSd1ShuHzIMMBrVChc+ZDFM68ROHcWQ89qvRyOfxxPbhznveGARD+8UvJ8ixFu+I4zMYHW1YlfKhF6re4kN18L0fpPm2dGHxxEWBYEAXp/P36xtMYQS8eNymr/AsBKqmR7DYSARsWnvNQ+lqqbCvGMRr3k4btyufPzsmvVWr+CNnXciayGZiiUlzauTrPDGqEMiakp9BZl7E6wpSfMDmoktQ5CAkIZtrYNIYZiUQ8rhDIgrXVT9QvL+pIKKpaaoBoUccIGISG1oIilJCn588SUZMuI68t16QKBL701EoVOeYnpt75K5/7HyG47jJntjTrG3qbxlVISuM01qvN3i5iRRQgVKQNURUdVLaT0CAKPurOeWRGGkjPnXdLpOzcUvtOgNHN++e/89Q82+fQKNuIQe7Ui029/ndL102R3Pki8ZwQz1Ylqxe9oJb9xnb1GBae8RG3mCuXJ87f+4hbhN6vMvPctKlMX6dj9ILGuYcx055bFraWUhL6LW8lTX5qkeOUEtblrdzUmCcy4lxizz2z4GUPEOBB/8q5+ZzXKr7xG5898hswLLxCUy61EW4naBycQpok5MoLW3YUwTWJ79qDG47jz8/i5iKOrpVNknn4qYgHs3EH9w5Oba/z+OWJoT5xP/Nogu4+mOftGkd//X9vTRYf3Jfj4L/Xzw38/w+yVrXObVV3wzC/007cjxh/8vWt4TvvFcNI+T1LtYMmNKGnNO+gEDFs0uo5s4HgrJ9F0by36W/TnKftrY2Or/64HpeV/zzrtDcIN+HadxVOvUBk/R6JvO7GuQcxsL0Yii2olUHUrqkVv6aaGvkfoNvGdJn6ziteotChmeeqLty9ssxFiZie6auIFTWJGB4pQkUiaTgHbqxCEa1dkAWQzCoViSLhZk8Q/B9TmrlNfmMDM9hLrHMTq6MNMd2EkO9FiCVQjhqIZy8ZNhj5hECADF99uXedmFbdawKnkcMo5moXZTTWM7xSh71KeOEtt/jrxriFivSPEOgYw013o8TSqGY8q7xSBDAJCz8F36rj1Mk55CbswT2NpimZhdsMOwLeDRlDGl+6mnq5E3han051fQJ48RdBqeSVdl/rpMxG1C3BnZsh/81tY27cjLAu/VsOdjVokCU1FzaRR43Eqb7wBErSOjsibJVokGhcu4uVyqOk01XfepXnlyr3j4N5jXDtZZfyjy/zVf7p30/BrYd7hg+/lqCz9eDz0QXM3AgVferhhE0PEaN4Bg+HHxl64U9d7M9ilBezSAopuRZlyw0LRjUgYRVG40dFVyhDpe9G22nMJPRvftVda8twjVOozaKpJKAMUoWEZGQrVcVy/Rri6MaZQSHSNIJqTWKagu1Nteep3aXg1NZJlvIu+c8IyI55iq9GoXZjDLsxFfFszhmbEomus6hFjobXD6OjV+PjPZfnw1SITZyqEnkPgORFrYANDpmqCv/y3h/juf1hkYWrlwUhkVP7bf7wNw1I4/lKJN75RwGmG6777K39zkJf+KMfcWHS+gdOgOnuF6tw1NCuOaiZQDROlNdaoR1sYLRC+R+DaBHa91dzw7q79ngfifPwLXfz2/zZFI6zQCDd3RHzpMe58tOXj+7nccqILQHoezQtrWS3e/ALBUnE5r6EoKumOUWrlWZonPyIMNzYI0vOonzy15fH8eSPwJWG4ccdmgPKiS3nxxxcS+Yk1prxTtOticK8QejZuG42DHweEUBkYeIi+/gfQtRjV2ixTk29Tq83hhw5+GBmAfHUMRSgEoc/ND7RmxBi+71OMv/VvOXvRZSkf4K/ydJPJAXbv+RlisbVMgiuXv00ud37d8QBQBMboEErMwj5/mWV5PrFKpm91CObm16RE6DqxI/txxiYJcmu9MBl4+A0Pv9HemCR0k+7sAIqTp5nfWgm1osKu+xLo1tqwRaMa8Hv/aJrnvtjF4C6LdpECocCuIwne+EabEnEZ4jdrWy6QuBdId+rseyi5LLF5a8h1FZnxtIJnSzz3zhYAVTXJdu8iDAM0I06lMIYZ6yTduYNGbZHC4sUtxI1vDSUWw9g5gtbXjWIZGyYw10FC5Xuv3ZIdoFsKL/z6IPd/opNYUqM47/CtfzXJ2OnqlogFiir4n37vMB39JosTNl/5J9eZubw2vKCZgqd+vo9HPttDqlOnsuTx0u/M8NEb7XcfRz/VxfO/OshX/ukYE2ej+/aTa0x5t1AVlJgFoYwqWW6CkohFYhv3Op6kqpEOZ5tOqevQKoxol8kcHn6C7TteQFWjJFQ80Uc83sfZj34fx6mgGhZh4EcaALoR9cG6CUaiA1W38HwYGdK476DJ5evu8ikLRcM0U1hW9qZT2LgnlTE6ROqTzyAUgZpKYu7bSfkbPyD7C5+jceIMQhFovb2o6SSh61J/+zjG8CDmrm1g6tTf/gCtI0PiYw9j7d9N4+RZ7LO3FuW5G3iO5H/5S5fW2QEZQn7epVb2sZLtvQfflfy9X7l8r/Xl/1zxa//zMK9/vcDF43e2WEgkrlMjDDwMK4WqWehGnGZtiUZtcUODG9ZqhPatnwuhaySefIjMiy+gpJK02mBsHWFI5QevwyahNM0Q/Nrf383wvjg/+Hcz5Gcd+nfGKMw5W9ZKDwPJP/v1sxx6uoNP/OoA6k1MJlUXvPjfb+O+5zp56XdmmLveoHvYYmnaJgwk6g0usox2VIee7uJn/8Y2vvpPVgwuRCLmGa0XXTGp+oWVxOht4sdudNVshtRzT4CUlL75Eje3Uc383Keovfk+3sTW5BS3CmO4n9iDhyh/49Y9kLSeToSur2v0pygGnV37UFVjTeIunR4ikejDly4jD36W8uxFilNn2fbwz6IZsXVboBtylEKA40pOnLY37Sa7FbhTczSOn0L6AfaZC2gDfagdGcJKDXPHKN78ImomSf29D9H7e4k/fD8oCs1zlxCaRuzIfqqvvI2SSdM8c4Eg115kaDU6enUGtpuomkDV1vJeY0mFge0WibSKlFBYcJkdc5a9wOE9Ft0DBmEouXamQb2y9UW2f5tJ3zYTAVw9U6dWuvV3hQJd/TrJjEYsqVJqtYjv7NOZG3cozHvopmBgu0W6U0MoUMn7zF63lz3PrgEd3VSQIfQMGchQMnPdodQme949oJPt1Zm81MS1JZoh6Bs16ezVCQLJ0rTL0ky09e0bNegdMTnyZIq5MQcrrlAp+ExdbuI5W/d6w8ClWopyFPVq5HU1qps0q2wh98db0PBVBImnH6XjS5+JnJeItI4MQkSrPVDUbw5QlWUKn5SSsNYgKJYJq7Vb7gJ23Jdi99E0v/03L3Htw2jHdPXE1hpFrkbgS5o1v22uZGBXnH2PZvjmv5jgw5fyIFn+rdUQKhx+OsuzvzzAn/6zCT56PfKCBcoyS6niR4nIDr0f2LqezGrc0uiqHRmUZCR+rWZSOFfGMXeO4s0tYuwcRTENvLlF3PFptL5u1M5MJLJs6DROfESQL9I8cyHysFpQ0knM3VEZprKB4r3Qdcw92yNPuOmgZpLYF64iPR9zzw4Uy8QvlnEuXkVJJjG2DUX6BskEzQ9WYmfCMrH278K5Mo70PIydo5FGZ6WGffE6ajZN4smjKLEYztVxGqfOQ6sWW9MM1DZC2gCmmUKWA6oL13BqkcGKdw6Tu358nbiNZiXpGDpEEMDFKy6mIe5NzkKuiLr7cwvEHjiEc3Uc6/A+hGGgJOJovd2EjkMwMY11cC9aTxdhrU7zzIUW7XNr/Nxsj8bP/jd9ZLp1SksepqXQM7Ti1XcPGDz3pS7CQGJYCh3dOv/hH00zPx6FX/pGTPY+mOCTv9TNP/2vr3HpxNa9hO4BgwMPJ/mpX+nmN//6GGffvXU4Q9MET3y6g71Hk7h2iJVQmB936OjTWZh0+NN/PU8irfLcF7tQ1MjDyXbr/PAPl/jonej4D348zQMfzzBxoUk8pSIUkD8srRjdVjSnd9Tk03+5h3o1YH7CwfcCDj6S5GOf68Cxo7LfWELhW/9ukclLTXqGovNJd2psPxgj1aEye91hfty+LaP744TW3UniqYdB1whtB/vcZdzJWaTjkv7cCwhVofH+KbzFPIqho3ZmMXdvR+vtxptdoPLtl3GuTd4yOdczYuE7IVMX7m3J+2pkew2ESvQbm1zekX0JRvYluHiszNnXV5yQLn2ImJpa02Iso/UwZZ/H5sdhdLNp9ME+lLiFsW0Ib24J6779qJ1ZEOAvFUg8eZSgXEUf6MXYNYp97gqy1lgfyxGAqmDu2Y7W1YFzfQo1vV5+DUCYOsbubRCGCF0nbDSxDu2l8cFHBOUKYUMn9fyTOJfHUDNJYvfvp/nRJYJ8cblThLBM4g8fiTiHYYg+2Ie5exvO5TFiDx0mKFUIKjVCxwVVxS+U14QYgsAlCP22/bpct0YYuOTHTy6/5jt1Fi+/S3hT5wg9lqZj6BCmITi838AwBO+ftO8uohKGhI0m8aNHoiKM2QUSTz5E4/1TGDtG8KbnELqOmkkjHQfn2jjuxDRaZxahaXhL+eV+WoknHqJ55gLu+NSGP3fw0RT920y+/JuzLEy53PexFLvvW1kwF2dcvvs7i5TzPoYl+B9+cweHHksuG90Tr5Q5/VaF577YXvNhM5x9r8rFEzWefvH2tJQNS6G46PLKV/L8jX+5gze+XqBRDXj257tIZjWqBZ8ffTlHOeeBgC/8d/0cfDTFhfdr+H40D7LdGt97v8rYuSaqypr4q+9JOnp0PvfXesnNurz8lTy1UkAirfLU5zu5crrOO98pommCX/1bQ3ziS138h38wzbn3alw+2eCFX+zm5T/OceF4LVLJ+gsUOjF2jKBl0xBKGsdOUf6zHxGUIo2T5CeeRLFMGifPY5+JkntKPIY+Okj6p5/B2DGKsXMbztWtMYXusMbgtnBDW2UzBL7kyokKO+9PMXIgydiZaPHVFAM7rEc1BC0YinXHeatbGt2waYOqosRjUZfVkQGCchUlEcO+cBVvag7rwG70wajxor9UwL0+uaGKvdB11FQSb2YB9+o4fqG04W9Lx8XPFVHTSbyFpagZX0ea2P0HQYAxOrhMCA/KVdxrk4T1VgBdCKz9uxC6SuXbryIdF310kNiRfWg9XajpBGo2jTezgL+QR0nEcK+tbfkcBC7FwhUymVEUZSW+WqlMUautD6Jff/sP1hlcaAm3jB1HSslgv0ZHVuGD03dpdKXEvT6Bv5BDui6h45L/D18lKFeofP81QtvBnZ5DMYyIRte08ZcKUX8oBKFtI32f+rsnEIYe3edN0Ddqkpt1mZ9waNZCpq/a5OZW7rFpCZ74bAf7jiZJpBUGdlicffemmLS8c87AnXwv8CWlnM/8pEPgSyYuNoglVaSU6IZAKHDwsSQPPJMm3anR1a9z5u3qGiMwN+YwddmmUV1/s4JA8ut/Z5jAl/zhH+ZotEImhqVw9BNpDj+R4jO/Hj0XqQ6NiYuN5TznjW1wGNxzUs09gdbdGTFbgoDKj94iKK5s+aXrQcxC6Cv3N2w0cS5dpxKGdPzqz5H+1NPYF6/iXt28jfriRBPNUNh+OMXl4+11bFdjSzm8mz5TnHcIfcn2w0lyU/aGseL5602+91vT/NRfHeIX/84O/vVfv0B5yWPJmSAkWKMs1gyqPz7Bm7BaR03FCR0X/9IY8Ufvp/ryO5jbh1EzaYJCGa0rS/29Enp/D9IP1sY0VRVhGghdQ7FMZKs8Vs0kEfEYanaTVsZSRtuTVvNIoWsYO0fxl/LY564QO7LSa+1Go75VX6b50UX8XJHY0UPU3/oAfzFP8+xlaq8fi+TwaisZTmEYiJi1LvE2Pf0OQejR3/cAqhajXBpjauotHGd93MlttJ80YeCxePk9LFOQTins32OgqqK9UMcWkUwrPPVCjFPv1zEMwfxMSFgpoauC0K6jKxC6Pkpg092nMV/1I2/KcTEtwQsvJjl9XLIw3UR6TTQhEKbA92TbHaEQy7ehdU6RUbvx3n/5G9twmiH/4R9MkZ/z+L//H9vv+NzuFaSUURVSa8z+TcSSv/Q3Btl9f5zf/YczzI3ZfPav9tLRs3ahcOyQYINEUM+gztVTdXYcinPo0SQfvFJGhhFLo1YM+ON/PsvVMyvbZs+RK8VoN168iwJLISCbFVhWFCLyPYmqCebm2lnxVgfkYNUF2QRK3EJoGkGthj+7lm8vHRcUgWLd1PlCStzxaezzl0m98BSJxx/EvTa5KXvh+ukq598u8Sv/+y5e+b1ZSgsuvdtjXHi3xPSlehSaSakkMjqxpIZmCHpHLeoVn2bFJwyjSrRkVqNr0MJKanQPmVRzLtViVJk2d7XB2TeLfP5/2Ea6x2BpwqZz0GT2SoOrH648x0EgsesB3/wXE/yX/3wfv/h3d/F7f/cKzWoUbsyoPQzH9qOik/OmWHDH78gb2IKn24zaPldqkVd7cA/uxDSyaRN/5AixI/tonL6AP7+EkoyjtHorIUQUStg+TOy+/SjJOMlnH6P+xnHcsUmSH3sIc/sQ7rXJtm04ZBDiF8uEtfryVtpfyOHNLBB/7AHUTAr70vWoBNZxCQqltWWMrosR1BETF+HIUdIHRqhcmkbr6ST9mWcBKH/rZWSjiTe7gLFjmPRnnqXy7VfWlD+Goc/M9LvMTL/b5uqIdZq/G0GGIUHY4MJlF1W9PTGQdhgY1sh2K4QBjO4yKOQC9t9noaqQWwgY2qYzM+EhBDz+bJxv/VGFSqmlXTqk0d2vIYCHn4qjKJDKKFRKIQuzPtcurr8fuVmXPQ8k6Oo3yM259I0YdPavGKhtB2J87V/OkZtx6ejT6R0yuPbRTVVBrR3eZt7KRm/deP1eVoHvPBzno3eqTF1uksyo9I2YhLdRi18pBvzB/2uGR38qy9M/10lxyePaRw2cZsjsmE3XoMGZt6v4nsSMKcuLFESerudKMl0ahikIZau8/DamhWnCx5+x2L9PI5cPMAxBLhfyjW82cW+6hWayk86RI+QnT+PWb12kIVQ16h3WRms2rDcQirKc61kN6Xr4C3mk42Lt27mWwtgGYQBf/ofXeOpL/Rx6qgMjplCYdaKYqoR4VuNTf3WIXQ9GzlkMlV//h3uYudLg+//XFIU5l0Mfy/LCXxlqiZ9Lnv+1QeSvSr7zW9NceLtEGMJ3f2uKpSmb+57r5L6Pd1BadJm6UGtpbwgKc9EOVYYS35P84f9+jS/+P3aw/7EMJ38UxXe3xY5wpXEcX3oMW/uIKymqwa0T0Dfj1uyFUFJ/a6XUNP/bXybRHcMwmwTH36FaiAZrZQ3CuRnsenST9O4U5kgPXr5C4823IzpGwiSxfwC/0sA5cwahKtQvzbQVxZBNm+bxtWWVNxgO7vjaPkj+UgF/ae3J606VAXOe5Oe3c+Yrx9jzwggnTznU3zy+/BmBQkxN4y4Uqf/pKyhCg9BHEVrLOCjowsIOam3jN6oRY/ShF1ePevnIN11EGoU5itde58BeA0URW6bDbITpCY++IY1SISCZFnR0qyRSCmfet0HA8A6dZFrhyjmXQi5Y06hyZtJjeszFdSSGKejq09A0ePfVBvuOmFxroyx5/liVQ4+n+Nn/po/FaZd4Ul3WoJDAh6+WefiFDF39OrGUSq0cLJ9jtkfj0GMpOvsNzJjCE5/uYGiXxbn3asxPOBx4JMmOQzEOPpYimdb4qV/uYX7C4dgPSqSyKoceT9EzZBBLKjz2qSx920wunagxdfnuuNonXi1z/1MpPvfX+tANQSKjkp+7vS1j4Es+fK1M95DBJ3+ph2pxjtysx6tfzfPkZzvo6NVx7RBFEXz0ToWz70aJFxnC+y+VePrFToZ2WUxdtjn9VgWncRtxQgm5XMA1DWr1kO5ulcnJoC0zxkp2ke7bRXFmayXZoetBECJi1orI9I1zLlcQuobW3dGWpyxdD+kHKJn0LT35dGYUoai8+h/HefU/rg/Z1Us+f/qbE6QzowSBQ722uO4HT71S4NQrmxs/35W8+41F3v3GegU5ieTYny2tea285PHbf/Pymtc86RJX04QyRBU6CbUDRWjrKm9vhduXdlQEqf44qqkiQ4lmafQd6qQ0VaORt3Hr0aSK7x/G2tGHfT0SsA4dF707g9BVzOEu7IklUJSWAtDtjuLWMFMRzaW20EBRBdnR9RqgHcYAMT2LlAG+9PBCGzeokzJ6KTvzCBHpRzhhY1k2cjVk6FNbiESjJRIjkaF7x8MUpz6KGA1CYCY7yQzupzx7CQQUigG7d+qt1up3bnldR3L5nIuUkstnXRq1kEtnHGqVAKEILp52UFRwnZALp9fGsYIAzp9yqNdCxi67jF91kRIadcn1S+0renJzHl//N3NsOxBHVQULUw7vfq/I4lREC/vmb82z54EEuqGwOO3y3vdK+F50zcIAmvWASt7j3//9KFnnOXLZaLtOSK0U8P4PS9G1lGC3jE8YRv+uFHx+9x9Gi63vSvxbFBRExrCCEOC5IV/+Z7NUiz52PeBHX85RKfi8+tUcs9dtEmmVcs7n/R+WEErkccoQzr1XY+Ki3dYQjl9o8Me/OUsYRspzr34tz+774jhNSeBLTr1RITfr0t+i2NVKAVNX1i4SN66ZYSkUFj1C//bmg+3AW2+7ZDIC14WOrGBuPlxzr1O9O8kM7CPVsx0r3cfIA58hcG1k6JMbP0l1cUV3YrVTGlaqSM9D6Dp6X/caOqU3uwiqit7fE4UYb+oOIWJm1OJ89WtCiTRow7UrQlfPfhTVoFKaRG5oCARd3ftpNgs06kt3vUu8U9T8Akk12+ps7qELA1XN/viNLlJil12sjEFmJEnH9jSpvjj56xXM1AqFqPbROPVzUze1wphBiZuYg500Ls2Aqiy3Sr7XCNyQRHeM7EiSzFCSRn69V2RqSQzFwg2beIGNLkziRgZTS+L4dTRFx1DjVNwlQtpss3yXxavvLf+97ZGfZf7im+SuHSPwPQRRPy+nViTeMUgm/IiRIZ147O7TtYEPi7PRmNYbSsnC7Mp4xy6v9d5kCHNT0fu1ylqDklvY+H7MT7jMT7Q3ysVFn/d/2D6mXSn4fPjqxtzLa2caXDvTXqCkXgk49cbt8zbDkDWe8PEfRWNzGnDxg5U468nXNj723LgDtC+xzs955OdWzreSX3uOnisZO99k7PzGAkulJZ/jL906eXQrlFvdoJvNNsZISjy7JUDvuzjVHJ5dQ8oQ340Se6N7LTp6dTr7dF75WuQxegt5woaNmklhHdi9xug6l8cA0EcGsA7vjZpWtkJ7SjqJuWMEYegE+eKyX5HObsM0M+SWzhHedjNRycz0u4Rh0Nb5+Ulhzr3CvRDov22jK0MoTVUR04LFC4XlIJsM5JrxBOX2ky2o2Xi5ym3pWN4JqvN1zn39Gr0HO3GrHvNn17frnm9cWZZjZF0jzJVQwebUkFXlvF3bKE2fI2j1FZNA4Nk4tTwdI4eZvRwwO++TSun3NDZ5WxCg6grBBn2k/v/4zwvVpXHqhWkEAtWIs3T9A5qVlmcmQ4SAUs5nccZdw7t2J6YJKlXUzgyxBw9SffXd5apRd3oOb3oefbif7Bd+Gq23G/vCFZR4nPgjR4g/eAgBOJci46yqBqn0MIaRpFS8TkDk8cpldcGoGwxSi6RRQ3+VcRWoqkEQuFHXki1DoKo6QbC6VdLKsW4874qqIVCQRIL/q39XUbRlpT0pA/xVWhaKoiFliBAKqmIika3F5NZ2bctGNxYXBIHEdWiped1oQX4HxnMDg6soGqpqoqoGiqKjtERLorjRDTEbnyD0CQOPIHDWbVduwMqaJHpiXHs52pLufn6Yqy+vjQVLwptG325cWz+/RmmOnj2P4zYry4ZX1U16dj2CXV6gp1tloE9D17YW0223qguhousxVNVEUfRVkyIkDD1838H3m223anpCJ9YVo/dwN/OnFqnN3bpiaKtQVRNNM1v3rTVZW03kJOGK8EzoR/znwF13fqpqoBsJfN9BNxIoqk4YeLh2hSBY8ThVzcQw0yiqjgx9nGaZIHDQjSSKouLYFRCCeKKHwLdx7DJC0YgnumnUFjf1loRQiCd7cewK/iplPVU1sOKdNOt5wtBDCAXDTKPpUScHz2vgOdXlYyuKhm6m0DQTIRSCwMN1KgSrKIWxRDe+FzknhpkCIfCcOm4bZsydoyUAJcPlf99MCO7s08l0afQOG0xeinYHYaWGc/EaWkcG6floHWn8G/ocQUDle6/R+Ws/h5pJkfnsc2Q++9yaYwbVGrU3j6OpJnsP/BzpzAiKotPZvZcgcJkae53c0vnWuafZs/9niCf6CAOP+dkPWJw/Qxj6xGKd7D/8JWKJbqbG32J64s1NwhAriMW7uO/oX+Hc6f9IrVWtl86McODIL3Lqg3+L61Tp6tlP38BRTCuD61RYWviIpYWzhKFHLN7N9l3PY8U6EAjq9UWuX/k+nltDCJVdez9Ls5HDtLJksttx3SoTY69SLW/Mdb+BLRldRYHd+zRiMYXj7269JnqrMM0MyWQ/iWQ/iUQf8Vg3hplE0+IoSjREKQPCwMPzG7hODdsp0WjkaDZyNO0CzUYe329N4JTO0NFeunZnWuMX7HhmaJ3RBYjHe0hnRm9rvIFvk89fXqfiNHfhNYbv+xTbH/35SG0L0Iw4bqPE/MU30QKf46ds9u3St5QlD1eJ5yiKTiLZRzazjXR6lHiiD9NMoaoRbScIXBynTL2+QKU8Sbk8Qa02v2aCJvsTdO7tJN4dp/dwN/XFOvI244iroaoG8XgP8UQvqeQg8UQPlplFNxKtRUFtLQY+QeDgunVcp0rTztOoL9FsFmg289h2ESlDkplhRnY+g90soRtxVM1CCJX84nkWpj4gCBw0PcHQ9idJZUeAyEhWihPMjL9Nd/9hkukBxi59H0XROPLof0F+4RzXL36XeLKPgw/+Cife+hcE/sYJOEXV2Xvfl8gvnGfy6svLr3f3H2Zk13OcPf7vse0yHd176R95NJqfQuB7DeYmj1EujoMMiSV6GN7xNJoeQygqmh4jP3+OmYm3l7fXew9/kUo5imXG4z2ouklx6QrTY2+sDEgIhKZFVK+Wkp7QVJRkEtlsRoVDzYhzfSdljpIoXt+oBuu+Xnv7BO7MPPbZK1FJ76ovNc9cpPrKOySffAglm14uHpJhiJ8rUHv5nYjl5HlcPPsVRrY/g6bHmbz+Cr7vrFn4Uplhrl/+LmNXX6Kr5yDdvYcpFcexW/Pj5PHfYs+Bz9/WeTUbeUrFMXr67qNWnQckfYMPUipex7FLpNLD9PU/wMLch1TKU6QzIwwMPYJtlygXx/C8GgtzJ6nXFlBVk70Hf5aunv3Mz6yQCrr7DjM5/jpTE2+iqjqeuzUN3y0Z3TCEelWy/6DGB8cc+vpVDh7WGb/uk0gKLp/3MC1BT5+KaQr6BlXOn3FZmNt8EhhGip7eI3R17iGZGsAwUpsIZ6vLnlA83gNE3nYQuDSbeUqlMaan3sG2CyiKwEzrxDpMsqMphICrr7RbgQQdnbvYu/fFNu9tjHp9kXJlCtdZa3Tt8gLjx/+URMcQmpkAJL7ToF6YxnfqxCxBsRTw8ps+3i3DWnLZi9f1BINDj9HTc4hEond5IVoNRYmh6zGSyX56eg5Rry+wtHSemZn3lj226mwNoQhKY2UCx49CQncAIRRSqWF6+46QzWwnnuhted3t2ikpKIqGplmYZgZa+cwwDHDdGo3GEosLp5mbiyazFe/E85pMX38Dz2vS3XeI7r7DlHJXadQW6Rk4Qkf3XiavvUyjtogV62T3oZ/Fdao4zRLZrp0oik4s3kUYeAihYpgpkqkBGvXcpgYXIPAdFqZP0Dd0lLmp9/GcKkIo9A49RCl/DbtZJJ7sY2TXcxRzV8jNnQahMLjtCYa2f4xGdQHPixaXhdmTuHaZMAzo6j1A/8ijFHKXqVciFo5QFDKd25kZe4uFqQ8QirZu52YMDYCUqF2d0W7PdfHzBfTBfpyxSfS+HoJKjaBWi+iVtwsZ8VN7hg1ubhrizy3iz7VPEsmmTfUHb+COz2DuGkXNpJBBSJAv4lydwLk2sVwgFcVio5K76N9rPdVqeYb80kXC0Kdamaaza896safbnqqShdkPGd35HFYsC1KS7djJ5QvfACCZGiCW6KGrZz8dnbtQVZNYvBsr1km5OEYY+CiKRv/gUVTVQNNiWNbalkyV8hSl/LVNJTTbYcvhBceRhFISiwmOPmJw7bLPkQcM6rWQjg6FUjFk734DoUqaDfjCLyX4N7+5cY18NruD0W0fJ50eRdOsO2rDI4RA00xSqUFctwKt2Ktddrn+6gyL54uUJqNun55992LVW4Fv1yjPtVfr8gPJyJBGOqnww9cbm4rehGFIKAM0Lc7+/V8g27ELTTM3/sIqKIpGKjVELNZNKjXIlcvfwnEq+E2fwpVbczQ3gxAKIyNPMzD4MJaVbbsAbG2MKpaVwbIyLCycWnlDwuLMSarlaFdSKU3Q1XcITbNACLr7j1CrzFBcuhy1KGoUqVfn6Ojew8TVH6HpMRTVINUxSqU4SRj6GEaKRHqAWnn9TqcdlubO0D/yKB3de1ic+ZBkZph4spexi98FIJUdQVUNFmdOYDej61nMXWHn/k+jGTE8r47n1ijlVsT5l+bPMLjtCQwjyWrT6DTL5ObPbhjyUNIp3KkZjO2jyCBAHehD+gFaZwd+roCazSBMM9pRbGB0pZQErXZNShvlOs8JiSc373DdDmG9SfPDs9jnr0QiOKGMSu5v7VGsgetUlhkJq2Oqd4tabR7XrZHJbkdRNFy3tjyvVM2M7lFxbDnkk1s6T621II7ueJZEso+FuZO4To1Esm+djfLc2h0l9rb8xEQtmASaLkimFGanfR542OD4ux6/9teSXLviMTXus32nRrHgMzG2QRmwUOjq2s+u3Z8mFuts26V1LSVk9b9F6xhrPx8EHuXyFLa9kgn2mj7xLovRx/oQikLgB5z8/fXGUEpJ2KrDXDnuyvFvdzHQrRTxzkFUPb7mtHyngVu4zI4RnQP7DF55u7lc398OYRittAcOfomurr2rYrdR0m/1NVp9DVePV1UNurr2o+zXuHjhT3DdrenebgRVNdm370W6ew6jqhrtHoz2927j6+nYFXK5FWKw79trYqlSyuWKCgHE4p2ksiN09R1cM65KaQLPqRGGAZaVIZ0dpbB4gXiyD8OKPN3psTe3dJ6+16CwdInOnn3k5s/S3X+YRm1h+YHU9DhWopMHnvjvlktDhVBRNXN5YTStLP0jj5Dp3IFupiJpUCnX7uSkxLHLmz649qWrEATUjp2I0hsi6obhTk4jgwB/MReNYbPQggxxqkvoVpLBQ5+guhQluKqL12gU55AhGJags3djKdHNIG1nWcdjg0+0knZK2+oWea8SCzch8G2K+atkO3ZimCnySxeWdxKeWycMfcrFcWy7tDxDb8zf3oH7mbj+KvmlCy26Wyu/dPN53QG2ZHRTacH9Rw127NbYvlPjykWPX/z1BGNXfYr5kCsXPRxHcuZDl55ehW07NM6daWd0BZ2de9i778VWKGG9sQ0CZ/nBs50omREEHoqqYxgpLCsTufuqiaqZgMBxShSL19ZchPRgguGHevHdgHquSWaonbCOpJC/zKXgT9D1BLoej/5vJNC1OJoewzTTaNrWerKlB/ay47EvtZJGax+CRnGWydwl3j7eJJSbFuksj23btmdIpUYQQkFKie/buE6FcmWSRn0Rz6sDAsNMkf3/tfffQZKl53kn+vuOT5/lfXdXezumx2Jm4A0BkgBFUKIRteJKK1HS3pVdcaW7N7Qbe+9ViFJsUNJKWvGSosSlIJEEAZIABo4zGGAADAbTM9Mz7X15X1npM48/3/3jZGV1dlV1V/d0k4iNfiKA6crM47/zfu/3mufJ7SGVHsQwMq0E5LqBU+jq2sfu3R9kYuJPWpnbu4eupzhw8Cfp6zvRMhwb92NdrSMI7PgcvRqe12gbz/g+5tH1ZMzc1koCAiyvvHvLkv/2r6BrV6gUJ1mYeZ2bn3fgu3huHdcuk0j3kcwMMHHpRVQ9QTo7jGFlqVXnd3StUkasLp5h/9Gfoqf/MLmuPR3HCwMHp7HG1LU/wfc6WaYa9VU0zWLf0c8ghML0tZdpNlbRtAQnnvqrWx3s9iezTtDRWhat/1q2GopuHWfboV6cZeb0V+jb+yR9e5/Gtys01uKQW+BLXnux/ECJZxynTL5nHz29h/DcOKy0k4ShYWTQjSS6kUTKiGSqj8C3cd07l9pJGVGvzjMw+AiJZC83rn6V9TtYLk7Q3XuYsT0fpFi4ghAKmp6gXLyBYxdp1JfJd+3BtUtk87swrBxUd7ZSuhN2ZHRrVcmLf2Tz4h9tlIG9+5aHQEWg8ZXPx8XCUgpe/CMbRdl64k2nB9h/4CcxzU6+hXWDUqvNsbZ2hUp5alMSaB3ryZtMdpRsZoxkqo9mc5V6rZOPNwolpakqURhRX7bpP9S1aV8AjlNiaWm7Jbdg774fY3T0udsSiq9j8PD7Kc2dZ/HCt9uJtJuuEssQZDMqjUZ0x5dN0yzy+b3xljKiXl9kceEtVlbOtYxtJ2Z4lUxmhLFdL9DdfQi9lVWPqRs1evuOUSpNUChc4m5naFU1GRl9Hz09R9oGvX1VUmLba1TKUxTWLlOrzrV4KW49Rjw5ZNJDZLJjZDIjmEaGwuqFHWWj149VWL5Ad1/s+TvNUvtlWS/XadZXyPfsI/BtHLsUx3m79xL4NoG385ina5dp1lcYHHsaSUS5sCGGWK/MEYQuhpmhWVsmioI46aeoyChE6DqGlaW4colmfQVF1ejuP/RnWmMqw4DizBmKM2c2fadqgv4RHSlEmxVuE4RASVoo2UyblwEhIAyJXC/m0C1Xt/W4i4WraJpJV88BgsDBn2/iuVXq9aVWiCoeL4FvUylPEQROvCruO0K+e5z18bNr/IPUqgvMTe9s1eI6FVyngu838W6aIF23wuS1b9A3eIK+geOtePJsO9QwcfXrDI8+w8DwSWqVOWYmvo1sqZ1LKalV5/H9Ovfi7d4zibmKQc4aIpIhoqXPFEQ+VXd5yzIuVTUY3/sJEolOaj8pJZ5XY272NVZWzuE4ZW53IWHoUavNU6vNsyjeJJHoRQixKZhtl1yWLxZRdYX+o92sXLqXWKbcsUEAUPUE5flL+PbWM7imCDJphXRm5y6FlJJabYHr175KpTLN7e5NrTbPtWtfZffuBkNDT3bEgE0zR//AI1QqM63BQoslSqAmEwhVxy9upWsnyOV2Mzj4+KaYchSFFItXmZ/7IaXSjTvcK4nnVllzq6ytXUHXk1hWF83m6m222byPlYV30M0Uo+Pvb4eFoihgZeFdXKdMs77MyJ7nWVu5BIBrl9CNNM363XUyBYFNpThBd/+Pszz/FsFN8lD12hILUz+gb+gR8j0HWH8m9eoii80SYeCxtnyRfM9erGRPXMrnNXGam2vFfxQwMGZw/Nk0xRV/S6OrduVIPHokpnvs60bNpuMYrlCQQUDUtAlLFfy5pRaz2ExcTXETgsBmYe4UzJ3q+Hx1qbPV37GLzE5tVG8szp9icb5zm7uBqsZx7GLhWke5HsQO19ry93jqeII/fLGz8sBuFlqe8VaIWFqIE7+5rELlLiv87tnoWlqGvDXMmj2DqaVRhIqmhDS8IpHcbHT7+k7Q1bVv0+dh6HL16lcorl256yyglBHN5jYteFIiI8nqlRK1pSZe88En0iqLV+gae4T6yuSWXTe6IYgiydUbPjttxIuigIkb36BSmWp9cvv2Yd+rMzf7GqlUP11d+9thESEE3d0HSaX6KJdjo6v39KKYFnpfP1o6Q+nVb23ywDUtweDg45syt1JGFAoXmZx4uWU4727G9/0mvt850Bu1RW5cehG7sWGIm/UVJi6+iN0obFzfjVexUj1xcg2J79ttg1Yrz3H13Bdw7XK8z/oKNy59mcB37moCRUo8t4bn1igXrneOTRmxtnyBRnURw8q22lv9Vr1wXHi/OPM65bXr7WJ8u7HaqgfdiKvfuPQiwR2qKe4EYyCH2Z+lfnkB6d9bd6eiCkqFgERK3TS8zEN7yX3m4+hjgyjJ7cNscvcIiROHST55gsYP36X+ndeJ6jsroXoQEELFMFJ09x1GUTRqldktVxqJhOBTH01uMro7QTol+POfSfFbn7u7XMk9G92GX2K68jZhFFATK+1IXLRNSGBo+IlNZUVSSqanv8Na4eJ9X3pZOZO+Q10UJ6rYpXtXy70b+E6dvv3PkO7bhV1eJgrc9vh162vUp74DwBOPmJy96G5LGXgzCoVLlMutzh6hM5DYR9GdQwiBH7qAQBEKmmKgCp2aX8BxSqwsnyWTGUHXN5igNM2it/colVZtqNB10scfoXHpAmoiuSUjVDLZQ2/f0U3PrdFYYW72te0nvR0ijlbE+w5Dm1q5s7QvDBxqldk4kasKolASBPa2lQhh6FIpTrb/jiVt7lywfiuEopHKDOI01mg2Nq8ApIziGvHm1qrXgW+3pXRu/uxm1NZjhOuGTqz/Qfwcbv57fSStPx5FQCTxi3WCSnOjnX6dxk2yk8QBELdaN6shCxNOp8Hdv4eev/qzqD359jnJSBI1mkSNJjKKUBIWajoFmgqqgjbYF7P4CUHtm99F3kp3RlwLPXTgA/QMHwckqzOnWbrxOlKGdA0dId01Bih0DR8hCjyWp05RmDmNjEIQCv27n6R/z1MoikZp6RIL175LeItQbSo9wJETP4fvNZid/i7NW57TC89a/NSnkriuRNNg95jG3/ilDOm0wu9+sU4qqfDpH0uiaXD2osd//UKdWn3j5vT3KvzVX8zwqY8lGR5U+dLXm5y9sLN8yT0bXUlEEMUHCe9gMPP5cSxrM+t/uTzJ/PwbDyTWFYURekIjvysde7kRNIsPVkHYyvTSWI1JmxWhIHSr/V2omTiOxPMlp047eDtQgJVSsrT49kaXEyp+5CClpNscJZQhftTEUBJ4kQs3tSuvrl5gdOy5jiSgEIKe3kNMTr5MGIa4C/MEpRKRbeOXS1vE4wQDg492ELgDRJHP8vKZVrjj7iAEdI1YRIFE0xV8L2LXiSyLV+tk+0zKSw5hEKGoClEYEYVQK3iYaY39T3Ux9W4ZRVVAgtBAVQXF+e2Jqe8WimrS038IK9lDz8BR5ie/h+fcOWlzL9j9Nz6KUBXsuSLFVy/S/f7DWCPdyDBk+Wvvknt0N3pvBkVTaE4WUJMG9UvzOItlxv/uJ5n6N99k+C88S+j6LP3xW0Suz8BPniQ53oezWGLtlYt4hTt7YfVKQFdfipG9JvM3YgdFpJLkfvrjscGNIvz5ZRqvvU3znQuEpUqHQReWibl/N+n3P4V19ACKZZL95AdwLl3Huz616XjDBz9Erm8vU2e/AorC7mOfQigaC1e/g6Ia9O95mrW5d7n+5u+Ryo3QP/40br1AtTBJ9/AxBva+j+lzLxIGHmNHPsbo4Y8xfe7FzmuqLfDmD/5V66/OwZHLKhzcp/P5P24QhJJ/8LdyvPCMyfdedzh12uWf/EoX33/DoVyN+Fe/XuGXfj7D4IBGrb6x2lkpRHzpG03SKYV/+mvlu+pLefBqwAiy2V3oerLDW4qikOWld4nukE1XEgmEYRBWqwhVRevuJqzXUbM5hKrir65sOZvKCHJjaXoP5onCiNANee3/OLvFEe4fZk9/5bbfJyxBGEClFu3ISHhevUOhIpQ+zaASVzMQ0QxK2EGNvDGIHVSw1A0mtSCwqVZmSaUGuLnSIJHoxTRzsYcahoT1+KUMipvjjYqi0dNzeJOX63l1VlfObfo9wOiIiqLC3NzmDicARRMcfqGH0ItoVgPmLtYwTIXAi9AthZ6xBP17kkSRRDcVFq81sKs+vhthJBQe+UQ/djUgmdVJ5DTKSw6NyipOS9mhr09h9y6Vt0/7OzbEmYxgbFTl+o2ACJWuvkMoikbk/pBHjkwgD5ucPetTWNt6ebp/v8q1awE7ENjtvBeGRuHbF3CXymi5JNZwF2uvXsQcypN7dDdqxqJ+YY7K6SkQkH9iHGusB3MoT/XtScKGS/H7l0nuj0US9a4UqYODTPzv28Uit8a6Q51IbeQarMN70Qb7QYJzdYLS575EsLR1/F3aDs65K7iXb5D+4DNkP/1R1HSK9PufonhjutNAC5X+XU8wde5FqoWYoW/R/D5jRz/O4vU4OeY01lia+CFOvUDgNsgPHEC34uT70P7naZTnUVQdRdWx6wV6Rk4wff6rW3j2Ww8AXY9LYG07olKLf5NKKkxOB1TrEelU3KY/MxfgtjTr1C3SMFEo2+T+d4MHbnQ1zSKZ7N1URG/bRaq1reMsN0NJpUgdO07t1Bvo/f0kDhzAX1pGSacRisAYHqL+5pubtqsvN3ntX2/O1D5ImJnN17mOKPQhKNKVV3j2SYvf+i9V3Dt4u436UkdSMiKkEcQJwcXmhgdTDqcJI7DDzoh+qTTB4NDJTV1+2ezYjsICiWTPplguQKUyjW1vZaThE58wMXTBb/9OY0sjFIWSibfKQMz4j4TysgsCGkWPKII5u4aiCJxGQLMcICPQLYXykos/baNqguqqh+eENMudlIhPP2Xw3/+tFH/us2s76PqLsW+vxl/5pST/9FdrrKw0uXr2DxACjh3TePbZBJ/96Rz/+H+u8NJLm8NUgwMKf//vZPgn/2uF+YW7e/sWPv862cf2kNzdi7NUQU1bmMNdRG5A88Yc+af24pdbWoMSvLU6qd4MyfF+lr64ObmkGBqRd/e5C6EIwkB2SBLpQwMoyUTMs/Dit7c1uDdD+gGNU2cw9u4i9fSjWAf3bApZaUYCRTfwWjF3ALdZRtUsFDUm3PHsSjsnsl6OuC7AaqV6iMKQ/MCGakxx4fxdXW+tLqlUIt7/vgTVekQQwqWrPk8/YTK+R+ftdz18X95x0i6WIrq7VD71sQTvnPVYWtlZTP2BG13DzGCYm2tyG43lHfUqB8UiimWhptMkDx/BvnGD5IEDoChEjoOavbeC7geB0Uc/hZneQjxRSmqrkxQuvsiNKZ+YH/rObphtr90x+WOZ8OEPWZQrEW+c8jpm3UZjecuMfTo9cMdjA2QyI2xVEB7XRG9GFMG3XnFRBJuUC9pbR7B0vbN0q7QYW+fy4vax97AecKNlrO83JqcCfuu3G5TLNymPSDh/PmBmps4zTxvbbru8EvFv/32dteLdGVzF1Oh+4RBCi3mp3YUSzRvLGF0pQsfHni5sctScxTK5J/ciwwi/aqN3p+h67iDmQJ6w6VI5PYm/1mDo557FL9Qovz1JUL7zO6YoMdPYjfMbMWclmUBoKlG9ibtFiGA7RLUG/sJyHO/NZjYNn9B3iQIfI5GDUhzTNpN5At9u6wvezhHz7Cr14gxzl7/V+cVdxJdcV/LqDxx2jWo4ruTSVY8r13zqjQhdF0xM+2iqwDQFrid58ZsNllc3v4eFYsR//C81gkDi3IWK8wM3unHDwa2yHnFt544yt1GEMz1FYv8B1FwOb3ERc2gIb2UZb3FxU2nKnyUWzr+EaNXzxly6JsnuEfr2PkV5/iJHDho8ctRkftHfUbLf82p3XAkEISQTgq68xhunOi1dbLQ3H8hKxN6rkoyTZ1GjgdE/iLey1PG7VKp/y2PWblMkPj39Z6uyeC+h3UpFcvbs1uNIrrN+boNmU3L69N0LFEZeSOHbMcuW9EKCusPqS+dQTB2kJGy4rPzJWSJnY9+R7bH6jXj1FrW0Bpe/+g5CVYlsj7DhsfTHb6JYBjIICes7i3cEgSST00ikFKprAU4zQgZ+LIUVBOy41AbiZJsftFuCb713UeRTmH2HwX3P47sNhBAM7nuO5clTO8rtLE38gJHDH6VRWaBZWcJI5FA1ndLipe03EgLVTBAFAYqhg4RCNaBwOS55U3QDGQrOXIHI2zzxX5/cfmy8+c7dJ+kffHhBNVFVq+MzKeN+6yjaWbbPvn6d7F/6y9TfepOoUceZmSZ14hFSR4/TvHqF5oW7W148KNjlpU2f1ZZvIAOX3vEnuHJ2glI5omlHWykUbYLv23esLTVNwdXrwZYx1CjyCUMPTev01Aw9phHUe3pRNB13cYHU0WN4a6sbHVCAZeY3HS8u09ucse/rVfjiH3TT06Pynz/X4N/82waN5jpDGnzkwyZ/5ZeS7N2rUa1GfPlFh9/5z01qNUk6LfiVf5jm+edMursVshnBzEzIL//NEtdv3MHTt+Af/09ZPvoRk+WVkCtXg3b8zTLhU5+y+IWfTzI6qlJci/jd32vye5+325f54pd7GB1ROXfe5//5P1eYm9+5x/p7/7WLgwd0ZmZD/v7/WGZy8u6Mk7/W2c0WNlyI4lilUIBAoFoWkeKimhaR7yF9EKqGaioITUNEJqpu4jcX2sY6bNydIUikFFJZBcPSePR5hTdeqhCulYlcDyVhxarAt23zvQmqippOgqbiz8XsXrdi7vIrDB/4AHsf/2xcejh3hqUbrwFxGC70nZsMsCTwnTaXbmHmHaSMGNr/AoaVxXMqLF3fSr9wA0ZXH2bvIKqVRMoQRTdRrSSKrhPUa0SejWImcJbnac5P3hSkjQU/73ei/+7legwNocVtoNIPkGEUS+5EEqEq8ZLE8dvuvqLoqLfEOdfp/naMIGD5t/9j+093ehp3+vaZc6EKNEPFt+PjmBkdt3ZvksnvDRLfaWCmu2k0JZMzOz+HMFonW94enifZNabyxEmD3/lcY5PhDUMH6GyBVjULYRiYQ8Po3b2Yo7vwlpc6DC6Apm8W3dyKBxdgtRDxoY8W+Nv/Q5pMWnQsKwcGFH7pv0nyzZdcXn65wsCAiqbHXiLAZ386wZHDOn/h59YYHlL51/8yz9/+e+U7GlyAX/j5JM8/b/AP/1EF35P8g7+fJp2OrW4QwsxMyP/xb+pMToU89qjG3/nbGc6e8zl3Ph4XP/mZNZ5+Wucf/L3NS+E74ef/YokDBzT+9a/l7hspfe7w40jfjZfnuknoNIkCH9VMoCaSLQrHmLXLXpkjMbgbr7x12dpOUSuHLEx5WAmB3YifrXttirBQQh8eIHnyGI3X39nREl7r68bYMwpBQOPUu1tyZ0ehx9zll5m7/PKm70qLFyktbui4+W6dqTNfav8tZUSpPknh+2c3jdftEDRqGF29hHYdhIJfKaHoBpHnougGiq7jVUoE9UpsZFvb6XqSdGaYem2x3VB0P3BXRtcc6Y65K2SElkkSeT5qwkTLJwkqNjKKDbA9sdxeFq1rI92MqEVm/SCR7LboHs8yeyrWaDv4iV2c++LWscj7hUz/XhS9s2tLVQ16xk/SLC1ss9X2iMmnbz/QFQV8H7pyypYv/lYNJ4qiIl2XxoXz6P0DsfLrFse5dbIEbhsSkttwSji2ZHIq5OgRjVLJ4OxZn5nZDeHK3h6FUimi2ZQUixGeH3u/O8GPf8riq191eOONeNX0yisuv/DzcatyGML8QsiJ4zof/IARZ6UjSX+/CjfJL70XR+YeKGxvvz/fw6+WAEFoN1AMA79Wwezuw69X4mW/IpC+h7u2jKIZhG4TuV2OQFFIP/ck9sWrsWL2FpARXDndGWf3F1eof/cUuT/3cTKfeD9hrYl7bXJ7j1cI9OF+0h98BmN8DPvMJewzt1ny3yuEQB/sI2raLTn5OyNybeo3dibI2bGdDFqE+RmCwCaTHUHKCNetton6pYzw3NpdNXbdldG1dvcR1mzCpouWS6JYOloujtdq6QTuchmhq4hb6yv+lLVpUn0JDn5yF/ldGVJ9FoqmMHyyow59OgAAPONJREFU74Eb3Z7xk5ipjWy/RMQMT/Uiy1d/8ECOKYSgVou4dmO7Eqmt7n2rDCaTQUtnCGqVra3HfXpupbLkt/5TgxeeN3nhBYPPfNriq19z+OrXHHwfvvEnDs88neVf/GqOMIS33vK4dm1nk3I+r7B6U5KjVIra7G1Dgwp/82+k8Hy4di2gXo8z0qq63d7+7FGfvkJQK2/63K9urXbbXJjc8vN1CE0l/1OfJCiVO42uqmIdO7D1RjJmLQuLZex3LpJ630m6fuHTeBMz+AvLBMUK0nWRUVwxoWTS6IO9GLtG0HeP4E3OUv3adzZ1pFknDqGYBsI08Wbmka6HeWAcxTJpvnOBqNHEOnoAtTtPsLyKc2WC5GPHUDJJ/Pll3BvTWAfGMXYN403PoQ72oWbTuDdmUJIW1qF9NN+6f2WhMoraVRS5rnEUoRJFPgODjxFGPqLV1bJ4E7H5TnBXRrd+Zio+mVDiLVc29NEiiQCiIIiD+s5GrFauS4TcBEVREcqDG/lu1aM8XSORN7HLHkSSU//hwgM73joWzn9r03VJGRF6DuEOWeVvhiLUVkxpe29316jK0cM6ldrWLpeqbs68R1GIYllo+S6EqhI2GlsKhG7NobEzTt9bMT0dMj/f5KWXFV543uDTP2lx7rzPjRshMzMBKPD1bzhMTIYUCmFbbPFOWFmJGB3buOc9vQqaFo/LkVGVgwd1fvWf1zh33qerS+Gv/XebQyY/StjK4D4IKKZB91/+7PY/WO9oUwRoKvpAL1pfdyyv7sfy7JK43ExoGsI0YnVvIdBHBsn/wqfxJucof36jftY8tBfv+jTu5HUII8yD44TVGv7yKpmPPEfjzTNog3003zqLdDz0oX60/h7sdy9gHT0Y8zssLJM4eRxhGERNm9SzJ/HnlzH37gb9waWoksle1gqXQUr6Bo7HpE6SLXlf7oS7OsudBOjlLaq5W4UShFB3TH4tFHjmU90M7ra4dKrKlbfuHFsJ3JC5t1ZYu16httRE0QRR8OAZnnynjtxW6VS0l/Ey2tmyKDZwt/c2l1ZCrl73OX5U35LdbSujG4QOWlc3elcPKAqJfQdBRtTePtVRDbIuf3TrOW01EZhm3PyRSICVEGQyIpYntyW7dqk8+YTB0nJIpRwxMqLG0uqtSMXQoEp/n8LVawGrqxFRFHujOylM+YMvNPmVf5jhwoWARlPysY9aJBLxPfM8MHTBrl0qrif55b+eahtkiIvkLUuQySpoOmQyCqmUpNGIi94tS5DLifi7lCCdFniexPNA0za+VzXR2jYOkWw7RyoCocbSO0LTWq28UVthAYh5g3W93eYrfb8zZqMoG9tKGS+xb50wNTX+jZRx/Gmbc9Hy2a2/2wpSxox1pgHm9iV0SImaSqDs34O5Z4zyF74G6+3ukcSfXyZcK6Nk0xi7R1G7c0SNJmG5hppNE1WqhKuxV584dhDr6H7UrmxMjh5GhPXmhiJFtQ4yQu3OYR07QOXFb213VjtHm3tCkMoM0t1zgGazQKU8zdDwU4ShS2ntOrqRIgw9DOMu7mELD7x6IQzcDlFBiGcHw0ijKPodZ4nB3RZI+PpvL+E2d244VV3BSOskey32PD9EZb7BXCu++6Cw97mfZ/b0izHFXxjERjgKQAiyAwfo3XsS36mzeu0NnNqdi811PdEycNv/pl6XlMqSV151NxkpTbO2NLqeU8VbjLPd5sgY7twsem8/ufe9QOPyJfzV+D5txVmqKAqW1bWpOeKX/nKKT/+k1XY2/sNvdnHhgs8/+sdVPFeyZ4/Kz/1sAtOEa1dD/u2/azA/HxuLTEah0ZD8h/9fF5EE15F89esO//7XG3fM3Xz5Kw79/Sq/9JeTLK9EfPEPbZ59xkBKuHo14Pf/oMlf/Pm4w+0P/9jh4qWAajXe6Y99wuKv/LdJUq1OrP/9X+RYXAr5u3+vgmkK/sHfS/P44zqKgF/+6yl+8S+l+OM/tvnPn2vyoQ+a/LW/miKfj+PE/+yfZllejviVf1ShUNh6nJrju8l+/APYl66SOHoIrbcbb3aB4u9/CWk7oCokHztO+gPPoKZShI0m9Vdfp/nuhXg21VSSjx0n9cxJtHwWGQS4EzNUX/ouYTl+VkomRe6TH8Y6vJ+oadM8fX7LeVuGEe7k3XNSwJYUHVuj1dTR+VH8gXQ8/NkF3GuT+PNLRLaD1tONMTaE1tdD5Lr4C0u4Vydp/OBtpBcQNpqo+SxKwkTNZYjqTexzV0g+9SgEYWyE3wOsvEnv4R5qi3UqM1VqldkOoclGfZE7EU7tBDs2ukLVMJI5vGYVZAgoIGKeztvBD5r4/maPKZGIWaI87/ZG13cjdFNh16EkhQWX0vLOXPlEl8nQo70A2GWXwz++u8PoCkVlvYdPSolQVIxEBt9tEAX3RvSdyA2y64nPxFpXgUtp9jyluQsIoTB07MM41RXMdA99B97H3Ltfiw3ybWBaudtoxsWIIjh/Yet7kkj0bMkK1WwZTMUwUXSd1LETaLkuSq98E3N0V9voblUaBpDNjm4yur/xmw1+4ze35qtdWo74tX+59QuRzwv++a/m+NV/UeOdd3yklDxx0uBf/GqO/+t3mjQatx/gUbT52H/whXi8BYHk9z9v8/uft2MPUddQ0hZRzUfN5/jGmzpffbmKUDXQVKJmE8IIYRo4JPgn/8xBaBFRs4lQlLhSxzRQUklefsXm5W+5rf3qcZOD54NUQGHbrKI+MkhQLFN96VXCegM1lYwNLmDu3UP+pz9J9evfxp2cwTq0n+yPfQhvYZlgaQXCiLBUpvbt12KdtIF+cp/6MIljB6m/FndlJk8+grl/nNIffp2o1iD9gWe2jM0L16H6L/9PfF8iEG3yJc0QsXaeEAgBXqvoX9NjwiFVhe5ejdXlgDCIVwTBDlfX3tRcOxEnPQ/3+jTmwXGsYwfxZhfx55dQe3JYxw8SrBZxLt9A7evBOnaQsFzFnZjB2DNKWK6hDfQSVmp403NYR/bTOP3ey0aHTg7w/D96hvO/e4mzn7vQ0em4gfdO8rEjo6snsxjpbhLdQ7j1IrTYfhAKzZVpQm+zUV2H59W3lIlJpfrRtMQdJWSKSx6zV5vopoJQdp7YiUJJ154s9ZUm11+eZdezgx3fJ/MjWOlu3EaRKAwQQpDIDRK4DSrL13YcArgVXqNEZfEqeiJLfuQIjbUZAreJZiZZuvRdNCvN2GM/jqKqhHcwujFX8L3HvlPpQWIL0IlGIzaqfrmEkkgiqhWCchFrz178tQ0PfOsmCEE+v5fl5fvTYm0agu4uhWZD4rmSVFqwe5dKuRyT3dwvmPv2oFgGaj6HNzWLsEyM4QH81SL+8ip6X0/bhQtrdfThQaTnIQwDf24BfXSYqNFEaPHzcG9MEzWaGLtG4sSxEESNJvrIIGGpSlirExSKm+I9MghonrnQ7vK6eQSknnqUYGmV+g/eis+jVCX19GNYh/ZRX1oBKXFvbJRKhtU6iUePxp1fAEKQfPQIzuXrOBdiaar6t0OSjx3bdD+OPGqR61GxG/F9tpsRXX0quS6VuUkfXReYCcG5ltLJngMGa8sBfUMaB49ZzE/71Kshi3M+lR104wndQK1EpIb3Ur9xCT3fg1dYju/RTc6jfboz92K/fW7T37d+Vv3qK3c8/p2gGgq53VmsnHXnH79H7DCwKrBy/UgpUVQdPdODohkIIbDXbi+B4nsNbHutrfm1jmSyj3R66I5crFEEUxfuPglVX25y7gvX8Ro+TsXl/Beud3yvaga6mW6L9a0TjwuhxGU4t5lItoOMQpavvIZbX0MoGpn+cRTNBM9GiFhVwKsXW2VldyYyT6cHUVVjy9jqTtDVtW8LT1dSq8ZLprBaxXbi0htreJSgWiGsb3ikjcYKnlfHNDMde8h37cUwsi0x0PeGwlrEv/0/6/wP/480iUTscRUKEf/v/28V27lP1GGAkk7Gqxo/QBg6ai6LSCTQB/uJ6g304UGEqRM17bbnKrQE0g/Qx4YxRoeQjkvYiGOKwjKh2UQYOlp3nqBcwdg9ij7Uj12poQ/2E6xsXilEtQZRY+vxrPV2Y4wOM/A//s34AxFzIGj9G8T/+sggySdOYAwOoGTSaP091L/3Rvt7NZ8jvKlUKyiWtny9Bsc0snkVKaG4GjI4qqEbgiCQ6K3OetMSJFIKnisZHtOREYwfNEllFfYcNFhbCahVoh0Z3eTwHtRECmtghObMDdJ7j1BcW8bIGPQc7KI8USYM1mv+IQoiFFUQ+jFbYOiHBE54zwrWd4KRMek+0I1QH3yl1Y6Mrt+oUpx4h1aRLgiBmetHhj6Bfac4iqRWncP37Y6XV1E0BgYfZW3t8qaY7/2AUAT9R7roPZDnnf9yhcxQqkM9orY6Qa0wFYv90UpKQPzfuyG7vglObZX86FFWrv6QRK4PK9NL794ncaorIGIeUakorZDMnQePqpotr/Kduz4XXU9vyZ1Qry/jeQ20fBd6vgutpxdFNzBHx1j5wu92/DaKfIrFawwNnWx/FsfjM/T1HWN+/vadQDtBGMLn/kuTP/xDO+bWleAHG40TG8eNF1fr3q+itpzIHb6D9tmWIRLEBfvT8+2VkwxC/KUWAZCUrVI/2f4bRWC/e5H2+JfECSwJ7o0pvMk4QeopAoTA2DWCv7h1/kBuEedsf+d5uBPTVF/5fuc9KsWTmz46RNfP/ATu1CyVb3ybyPXIffLDnftwvFjVoQVhmVvHdCMoFkLeeb3ZaohYp/+UrEcMhUL736+93CCKJFPXvPhZtNi1dsIJDbFKiV9Zw+obRqgqimGQ250l2ZugazxLuj/JwullBh/rw15zcCouez+ym/J0BSOpE7ghS2dXqc7F6t5CFW36z7YhFnEIJA6NxO6zjOLqqq3qmIUi4s4/IUgPJOk9FPOmKJqCaqgIZfM2offel187jOnKTVl5p7jzYv9SaQLHKWEY6Q7Pq6fnMAMDj7Kw8Cb3I1ZyMzJDKVK9CZpFB81U2fXMIDde2Vguy/hpxP++T8dcuvgdxp/9OQYOvQBSUpo7j2ok6T/4HI3CNH0HngUkXrO8I0FBIQRDw0+yunruLptJBP0DJzaJf0opWStcjPdVr8XcC6U1glIpzobfAikjlpfeYWDgkY5VyjopfaUy1UE9ea/wfaj4t38KI4fSDB9OceqPl0nlNZ756UGuvVFm9uIOkye3FtJHUcdzl95tnke4MUbWFQlQIQyDmEBeT8QrEt9BIhFzVYSMCFvF8zuFc/k6qWeeIFgpEDWd2CEwjXa9q9qViyeAd87jLyyj9nah9XYTrG3U8LqzCxjju1BSSWQQYh3av2VM95UXb71vW9z/m26Z1yZ0ube3JahXSe89jNHTS8/TH6I5Nwl+hGYo1JeaOGWXRN5EM1RUXSW/O4tbc2PDGsXq1zKMQIKe1nnqbz3O0b9wiLd//QxnP3cBzVIZemKQ/Z/aS9/RHhJ5C6fqUpmuMvHSFDdemsKrbeRqkn0Jxj+8m4FH++k50EVmNI2ixavPk7/8KCd/+dHN1+AE/M5Hf5/If2+VUDuL6ZoZzGQeRag0azFzVTLTj1BU7PoqqmYRBg6+WyeRGcBtljp4cn2/wcrKWTKZYYTYOKQQgvHxj+O6VUql6/e1Sy0KIqIgwswZ9B7IE7gPnhinsTbL1W//JsmuYQLPoVmaJwq8WDwxkWHo2EcwrAzLV3+wpZzPVsjnd9Pf/wjLy+/u+AVOJnvo6zvekrPZgO83WFu7EnvyATEXcb4bJZXCL2xdTVGvL1IsXqe393D7MyEEqdQgY2PvZ2rqZWx768L9e4GZVBncn8SwFAJPMnW2SjKrYSZV5i9vGApNEwztT5HIaJRXXFYmbQxLYWBfEiulEoUw8U5ly24z3VQYGE+SyKog4frbFYyEysCeeNta0WfxegMrrTK4N4mRUKmseixPNDGNDKn0IEqr7DEMXRRFxfcd0qnBtnKzomisFi7eVQim8cY7GLvH6PqZn8BfWo2JWnIZKl99mbBcJSxVkLZL6tmTBGsltHyusyRMShqvnaLrL/wkXT/9KYJyFa0rh3Tv70oypXVhhzUiGWAoCdJaN7VgDT9yUIVORuvBiRo44Ua+xi0sEjbrNOcmUVSVxuwEMvCpLXQa/8KVjdWoUASDj/Xj1T1Kk5VNxk4IQaLHItWf5MQvHuXQZ/bHNJVuiIwkyd4Eqb4kdslh6tXOSo1Uf5LR9w2TGU7HSigrTdKDKYQqcEoOTsXblAgN3PC+eGg7Mrr5vv2YiTx2o4BorIGM0M00VqoXM9mF79Swkt2sLpwl33eAtcVzeHZnBcDy0jv09h4lnx/v8L50I8WBg59mbu4HrCyfu+s4oaLopDNDANSq8+3QgF10qMzX6c92kx1Nc/1b90c++U7wmhW8ZmeplZQRfrPC7NtfiYupd2hwYyjs2fMRwtCnULh4x9CHYWQZG3uBbHZ0k5dbKFyieZM4otbVjTk8SmTbbJe1CgKbpcW3yGSGO1ScFUWlr/84iqoxP/8G5dIEdzciBaaZJZMdpVKebOulHftAN11DJpVVD68ZIgRohsLek1msjMb85bgDy0pr5CQg4PFP9vHlX5tg/LEcI4dSlJddAk+2+ug3n9PuRzLsO5mjuOAQhXF54fijWUYOpaisejz+yT6++RszjB5OM3QgSXnR5cSHe3jpt2ZpFG1su4gglpZ3nBK6lowVboMmilDjVmmhbAqbBWsl6t8/RbiNkmFkO5S+8CLW4f2xVxtGeLMLRM1WTH+lgv3yKcw9owgEzsWrOG9fJFI3rtGbW6T0R9/A3D2KDAIab76Dc22CYHkjvpw0uml69z5R6sLEpUFErGYylDhI0LyAHzkIFLrNURpBCTdsIFtqJlb/CKFroxomWq6bjJmgevnd2x5HRpLF07cv88yOZjj51x9hz4d3Mf/GIivnV7FLDkIRJLos8nvzLL69hFPubF+vTFd5+zfOoFlxYjS3K8czf/cJzIzB9KuzXP+TqbbM/cb5xIo07xU7MrrN+gpWqhsr1UN55SqGlY2FAY0kmpGgvHyVTNduUtkhnGaRwN2cKPD9JhM3vsGJE/8Nxk2xXSEEltXFnj0fobv7IGuFy5TLE+0M+60QQsE0s6TTQ6QzI2Szo1hWF8XiVRr1JcLWMlIogvqyzdK5a0SBxK3efRlYzBuxuXpACLHj5o6bIaPgridKIQRWoot9+z9JOj3A4uJbLcXkzchkx9i9+4N0de3bVJ9r20VWV853JOUUTUMGPn5hZcuONGgRjJRusLJ8luGRZzqk6FVVp6/vKOn0UEve/WIrfr916ZiqmiQSPWSzo2SyI6RSA5hmnrNnfhvfb6KZgqEDKS5+r8jkuxuGqbLiMXO+xr4n8+3PPDvi2psVps9VOfBUnr49SQb2Jli42uDCd7c3KIom6N+TYHXG5vQ3VkGCmVLZ0zK6hRmbriGTgfEkvaMW02dqXD1VZmBvkrGjaS68WtxEAB9LzgNuedvjAoTlCs03393yO00xySaGcPwa2oVlFFHECxsk9CyqyCMsQdroxV2oIxYnqDdnSRndZKx+VmpX2/ux1AzJZbBnz2NqGaxIQbwzR0ZN4xkCQ0uRsfqZLZbbBnE7jCSOsOzcYDz9OCvOJIrQ0IRBrznGRP00gfRwojrNsML6hBtIl0ZQIqJzPGnJDIpuktpzkNKZH5I9uHn5fi8YeKQPRVM4/R/OcONPpmiuNIlaMV6hxoY39KNNCTiv7lO4tOGABHbQbqCqLdRZPrNKdI9Cn3fCzoxuZRG3UaJn+Dh9o48RRSFOs4holkhlhwlDl0Z1nmz3LiqFyW0bHqq1Oa5e+wqHD3+23dkEsWHR9STd3fvJ5XYTRT6eV8d1q/h+EykjVNVA0xJYVj6WVVa09v+EEFQqnSKAVs6kezzL6uXSVqdCd/cBEoluNC2BqllomoWmJdA0C12zUFuNBZqW2GRgLaubkyf/BlEUEoYOge8QBA5BaBMEbvzv1v8q5cm7lBmPYdtFXLfa8lhVLKuLsV0fYGj4SaqVWWq1BXy/jgRMI0u+aw+p1GBLFmljySmlJIp8VlbOUi53eqOh42AmUiQPHkEGAdU3t06MBYHD7Oz3SCR76ek5yDrlXfzsVBKJWGGiv/84QeDieVU8r0EQxFUbmmZiGll0I9V6ZnqbMESIjX2FrbiulVZRNNFKgBDHNpW4bnR9NW0mVYyEgqoppLp0GkUfGcXbqnosYLlVaEGGkjCQWCkVTY8VE6JAUlvzuPR9j/OvrqEo4DmSvjGLRFZDUSE/YFJbu7f67Z0gaw1ie2W8oEE6PU7VXiRt9WNpGZJ6NxEBNWcFVWjoatw044Z1ElG+rVUI0J3azXL1ErpioSkGGasfL2hSdRbR1SRB5O44jJfQsuiKSULN0WOOUvfLlLwFes3dqOLuxAMi36Xr5PPUrpzFr5XbShB3wjqD4XYw0gYTL09x/vcuE7qdRlKGkmbh3ip/EscPY1+bbnVoSoRpIG0HYehxPXYkEQmLqN7YYafIBu5odIVQ6Bk+Qa5nL0JRmbv+HcxEnr6Rx/Ddesx1GYU4jSLp/BiuvbWRA0BKCqsXuaYajI9/DNPMdhgIIRQ0zURKA11PtfS9tjqnO5d13EmYctfuD5HP7+FObbZbHUtRVAwjs8WvN+P69a/GzQh3RWUlqVSmmZp8hf0Hfpzu7oMoioqq6ihKlt6+Y/T2HWv/9uZruDWkIGVIoXCJ2ZnvbnrZpOMAEi3fhTt7e6pM161y5fIfcvjwz5Dv2tvh8caGU0VRku2JcTvc7tnJCE59aZkP/uIIT//UALU1nxf/1SSH3tfFk58eIN2l86n/fg9vfXWFtXmbxz7exws/O8zFV4uszTucebnAB35hmOMf7sFrRnzxV6/j2bcsESVcfq3ECz87zF/6p4cJfMnn/z9Xufxaief+/BA//St7CX3J7/1v17h6qswLPzvEyU/1M/lOhZkLWyftNB0URdyUbNoamhbn87Z6R92gQVdylLpbIIx8/MjF8atoiknTK5Iw8qTNPryggaVnSRk9NL0SqqKjKSZBFI9tz6/Tm9pHID00xSKSIZEM8EMHIVQyZt8dG27W4YQ1+qxx1twZusxhqn4BXzqbvNidoDFzHXt5HkXTEIpK4c3vdHxvZg2SPRZei35V0eIJsf9INwunl/Ea2xGJSy5+8epNBvfm8SVpKyND6987ew/1gR6UXA4lk0K6Hu7kDEQRiUeP4k3PI1s8M/bZS3ctpHBHoytlRGH+DIX5jWJ4p16gsrpR92pYORLpPuz6Kp5z+5islCHLS+8Q+DYjo+8jl90VkzaLrQ3HTiC3oPe/kzBlzBAk7vpY93qOd4MoimjUl7DtIpcvfZEDBz9Dd/eBdlvwLWey7X6CwKFQuMSN61/bkpJR6+7BXytQO/0WXR/6KFw4e9tZ2/NqXLjwu+ze8xH6+o5hWXmUWwh+7uXZ3Rx3XZ2x+cI/66ypPvftNc59u7MDbnlycwirtOjypV+7A+uWALvi863fmsZ1Yp6EREpQXnb42r+bjPt+WiVRa3M2X/s3k7hO3HllGAJFiakqb97f2F6Drh6VM6dswgCSKQXXjQiDuNY1Ju2SHDxuMTvh0WxE3NrI2fAKNLzOut66u0rdjVdJVWeDIH+tsXGNi5XOTqxCY2Lbaw88d9tYrkAhqWQICXGieHKp+QX2ZZ7mavV1sno/EklCzWAoFkktRyBdBAJTSZJUc7hhg1CGWGoaSYSlptu6fXqum8yBE+i5LiLXoT5xCXt+qn383c8NY+WN9hK/s9TrNmO8GVCZjnMoQtUxknEXp1CUNvE5QOg76Iks9g4pVoNCiSiUeAtL7e5CJZWMpYh8n6hpoySsu+ZghvvEvaCbaRCC8ur1O/+YliEvXKTeWKKv9xjd3ftJZ0Y2KQbfeT8S161Sq81RXLva4ck112yufH2aVG+ChXdXSQ/cKhn0owvPq7VCEhG+3+Da1S8xMHiS/v7jpNPDHV7mVghDn1ptntWVcywuvr1tHXRYr6ENj5A8eBh/rbCjZVIYukxO/Aml4jV6++LEaDLZ1w4V7ATrIY9mc5VqdRbfu38E0XdCOqtw6IRFIiW4cNoh26XSP6TRqEY4TkSpEDI4qrM879M3qNEzoHHtvIuiwv6jJrMTPtPXb2LRk+A0I0SviqLA2AGDfI9KGEiW5gL6RzSyeYULbzv0D8cNCIuzPiuLwf2uknxPUIRKTutHoDDvXQEBrtlgLZwhMG2qqSV8aWPKFBVvGVWo6MJECJV6UIpDhEoCIT38KB5vhpJoG12rbwh3dYHiW6+ipXN0Pfpsh9GVUuI1AqpztXaTROhFKJpCeJvYqtf02/FaVY/lsTQzftdD38UuL6EZCZyqR6p7BKe8tKMqoOa757dpA37vuC9Gt1GZp1G5fWfaVnDsIrOz36dQuECi1aGWTg+RTPZgGBk0zWqVmMkWW5mH59VxnAq2vdbyBtew7eImPbFUX4LRpwYw0zrFySqHfmwXP/z1Da9gYvIljE3abfcf9frSHZc0drPA1atfQWslv4LApV7fmJF9v8n83OuUitfIZsfI5feQTg+14tsx1WIYujhOmXp9kUp5kkp1FrtZuO0Ai1w3NrYQdxu0kE6LOKuvge9JDEPQ1a3g+5LFxQgpQ0ql69RqcyQSPSSTfWSyY6RSfVhWHl1Po6o6Qqit5+YTBA6uW8N1yzTqyzSaKzh2EdsudkwKAoGq6EQyQgCGliKMAiIZEEY+qqIjZYSiaOQSQ1TshdiIyyj2uqVsxTi3fmGSaYV8j4rvS/LdKs98KIXrRJTXQiaveRw8brHviMFLfxTw2PuSIGF1MUBRwUooFJZuv5R87H0JTn2nycETJvleldJqSCKpMDiqk+9R6R3QmJ/arB32Zw0VDV86GCIBQGY0R25/N75XJh9047lV/NDGrjUoe52yVPWgcxWyYF/etH8ZBJi9g8goQs/mibzOldfiu6tIKWmuOXfXdXZTvDd0m9RXp1FUFRBEgUfgNVEUjSgKqK1M3lEU4E8DD5xl7E5IPvEIieOHWfu/fp9y6QaKaqAqGkKordjTuvck2zHKdbrIMNz+5VJbUj2hFyIUSHR31qyKjx7DLhSpv/pgyMXvBkFgUypeu+1vpAxpNJZpNldZXb2AqhoIRUWgIBAcTT5HIyhzvXGKMHTbxnaP+Qh5rZ8Lze/hy06PV8vlEIqCu7RI/rn34y0vMjok2DWmslaM+LGPW0xMBqytRezZo3L2vM/SUtR2iIPAoVabp1ZboFC4FBtaRUMRCqDQq4+yyzyKJZKIyGHKPsWaN0cY+q1kq+Bw4hkC6XPdeRuAjDVI0sgTRSFeaKMIBV1NoCoaqtAx9DRVexHbL6MInb70gZZBjg1z2uyl1JztWI7fCrsZUa9GIGDyqsvB4xYLMz6zEx5PvT/JykJAvRpSXIm5BurVkHROpVgIadQ7JzFVg137DI6dTLC8EDA/5fPUB2JWswunHQ4cMxnepXP2lE2lGNd5DozolAohYQgppYu91iNMuxephhsJ15SS50DiCea9a6z6s2w1zjVhsN96Ak/aTDnnibj3WvSIkGqwhqbEE7+eNjDzFkJCGITUpivoSZ17S0tBY26CKAwxewcJahVK1zuVHOrLd9/qfyvi0szyps/DVpjBrb03WaP7hT9zo6skE2jdeWCDe/euho4QqNkMYdOOW5taCOyAdF+Cnv05+g53UbzRWTur5rNtdqcHAWEaCF2P++zv4+wqZUQQ2B2lXwKBJ6oEsrGJp8FULBJKGnEL14NiWSQPHMbo6ydsNPBWl1pdWgqZrGBoWEdVodGUzMyFHDig3SZ8JQnDTgrPhJJlv3Gcml1k3j+HKnTK3gJ+tPFyCUAVekdMVxCXT4X4pPVUnMU3exFCwQ0bOHYt9ri8CorQSBk9aIqJ69dwgga6mkDZhiRIQWV1IWJlsdG2YZNXPN76fpMojNtav/CfynEIL4KXv1xDWW+Fnd16VIYBnPpuk7e+1ySMYH7KR9NiDokogpnrPpK4tfY7X6uv3672kFAQaMJoTVQbUIWKpaTRxPbctQKBpSQhuqfQYud1SJ8ubQAhVBphOc6DOAHLp+ZxyvZtKwjuCCGQgU9z7gb2/ARC1UiO7qU+cX/kfBTdAOHe9J61GHTaSbSbzn0nvJQPmH7hz9zovlcoCYvcpz9B7ZXv4y9seDeNNYcLX5ogO5zCbwbUFreuHX1QsA7tRx/sp/a9Hz5Q4w4gkZxtfvuutokch+qp10HE/17H/HzEwkKLfu+msfk7n7s7T6RXHwEkl+3XqYdbV7RIJBea3+v4rOIsUnE6W4uDyMUPHfywc0KpOkubPFrb3756Ztx6lHKwzFrQGQq7mZrw5h6RKNy2Z6TzOqKNjlkJ+De1NAc3xQW3ivTUoiLvNF6680EeMAQKIQF2WAagfHWN8tW122+0k/1qGqk9h2Iy9vXPdJ3kyJ77ZnTTB4/jhzeQnodiJeJVXhTFzG9Oi+azXkXPd8dy65GMBXQ9D+l7BPUqMpLtjjcrn2i1Tt/OOCvoRgIhVALfvr8aacIw0EcGURIJtK4c7sw8aiqJmk1jX75OVK2hpFOYe3ej5rIgJf7SCu6NKdbJcazD+wlW19B6utEG+oiaTZyrE0TVW2gdhYg1k5JJ7LMXIYpQUsl43915ZNPBnZ5tszdZRw9hjA1hHdoXs89XqvgLSzhXbpDstujalWkLU574mX2bNdI0lcSjx1BzWcJSGefaZKuMKvZUzT270Pp7kUGANzOHP7/Uuic6xp4x9L5eUFXCUhn3xlTcOaRppB4/QeLEEZRsnGCUrotzdaJjUvhRQORuPRncD8fcFAkkEjd678vGpnebMsQdQkVj0NhLIyq/53396OG9P7CIkGLw3nk0boWiG2T2H4u5FtY/U7X7KtelaBpW3xBCNzC6egjtZixDZTewF2dRNJ2gXiU5Oo6UEi2diRU8Ah97fpqgXiNwQpqFJqn+JMNPDpAdSVOe2r4SS4iYOlXTE9jNNRq1nd+7OxpdJWGSft+Tsbdu6CSffAxvZg6ttxsUhcab76AkE1gH9xLWGiimSeqpxyl98St4M/OgKqSefhyEIFgtIn0/JpM2TSJaRlfGV5E4epDUc0/RPB2XdgnLJP2+p9D6egjWiihdeczD+6m+9GpM6hyFcZGypsdy8H6ADMNYmPLHxu4oTJk4cgDCeEZMHD+Mkk7ROPUOCEg+cgzryAH8lQJCU0k8cpTqN78dc7HqOolDB4h8P9720D7UbIb6G6eJmY3CWJqlJVMf+cFd1unG0IVJnz5GRu1FFRqB9KgGBVb8aSJCdGExYhwgrcZimOVghTlvcxIDBGm1i27tCIaSoBnWWPRu4MpO7z+p5OjXd5FQ4xpkL3JY8+cphztT3NCFSb++m6zaS5c2gCYMjiTeR0SEHdWZcS/iSwdNGAwb+8mqMcl8NVxjxr1Vw04wbOxHIikHywzoe0iqWYLIY9mfphKutH6l0K/vIqf1owmdQPo0wjIr/jS+dEkoafr03eTUXhJKhjHjCL3aKAClYIkF71o7vKEKnV5tpLUvg1D61MIiK/4MgXRRUMhrg+S1fkwlBUjqQYkVfxpXbkwuQ/o+VKFRCObo1/eQVvOE0mfFn6EUxBNvXhtgxDiAQMGXLvPe1S1XBAKFEeMAWbWXiJC1YIGiv9BRKyta8fO8NoCCSiVcZdWf6Yjha8KgT99FTu1DQaEelVj2pjrOO6P20K0NkVAyKELBiRose1PtiSqtdjOojzPjXqRf30VG7SYiohyssOrPbFm/G7oOhddfxq/cVKqmqFj9Q3cYTTtHY+oazUIT1UoS2g1C10HPdeEszRE2G+2qmsb0DZARQtMRioJQVfxKCZDYRZulM6v0Hu6h+0AXz/1Pz7D07gpuNSbdMTIGiqrw5r87HV+CopFI9uLYRTz39pzgt2JH4QVhGNjnL+MvLNH/d/4a5T/+Gua+cbSeboSmEayVqHzzOzErvKrQ+1f+Iua+8djoru9D06i/doqo0QBFQfo3SWCHIeb+cVLPPUXjh29jX7wCUYQ+2I8+Nkz9u6/jzsyhZtJ0/blPYR3aT31pBefKDYLCGqlnn6D5zrnYk5SS0FR3JEwZNW1q33kNGQSknn2S1MlHsM9ciI3qI0donD6Hc+EyqCpdP/MTpJ5+HG9qlqhpU331NaQTJ/Lk809jjA2jvHueqNGk+c55tO4u1O48jR++RWQ7d+0+miLJwcTTdOuDlIJlvMgmoWRAg4I/S0Rc9F4JC4T47LVOIlC2NLoJJc0e8wSubCKR7LKOktf6udh8DU/Gy6+c2sex5PtBxMZbEtKlDeBE9R0bXYnElx7NqEpS5kiQoR5ViAjxoiay9VJGMqQarhERMW4+iibMTUZXAF3aICk1T7c2iIKGJ20sNYMVJqm03u9x8xGGzQOUgiUC6ZEQaXqt0VbyKZ7PA+nhybgRxJEN6i0j4kTNto+oC5O91mMMGONUgwJ2VMNUEiiin2KwQCBddGGxxzoOEppRFVXojFuPkda6udp8g7CVjchrA+T1fjJBT8xTEDWxlDSWkm5fnxPWKQZLZNUe+vU9rPnz1NlsdMfMI/iRTTOqkVLz9Ou7ucZbLPsbnmOPPkJCydCMqujCZL/1BAklw5RzjhAfFY391hN0aYPUwgIhIWPmEbq1YS43X8dpTb67zeMklQz1sISUMGTsp0/fxen6N/Gli6UkGTL2kdG6UdFphCVSSp7+xB5UVBb8LUpGowi/UkTv6kX6PkG9ErfRm4kdjamdwK+UCGoOQa2CV1pFRhJ3eX7TSu52op9ew+faV2/Qe7ibwUf7GX5ykIETfYR+hFBAURXcqts2ulEUImWEbqTvuuRxR0ZXen57+U4UEayVMEaHEelkO5GV+fALmOO7UBIWWnceb7YzbuZOzxHWarFn2bFzULMZ8j/1SeyzF7AvXm3/Rs2kST5ylMSRA20qRKHr+OvkHVJulIDcJI+yLkxZnKrSLMQ3fqtEgDe32CaU9heXUJ9/GlQVYRpYx49gHT4Q86YSTxrOlesxV6dpkHriMRInjqBmMyipBO7EzAY9ooxo865uI9tye8ReXq8+wrnGqxSDxZgysJUgiFo98yEBpWCRUrDELnOzOsA6IiKW/UkWvetIoFsb5JHURxg0Yq9FoHAk+T5A8m79W9it4njB1oQx2yGQHiv+NILYiGXULmbdiwTSa+1Fts4npBwsUw5WGDEO3uYuCFJKlgXvKovuDSKi1jltjKF+YzfVcI3Lzdfb3ysoBMQxNieqs+hdp6H2MmwcoODPsuRN3HQ28f/3aCMMGfu4Zr/FkjfR3tf6+QK40uZs41UiGbSfx5h5hGHjAEk1Ry3ciIEmRIaF8Dpz7uUtz9uRDRa9G9hqlX59z7b3wJcu55vfxZcehjA5mHyaXdaxjrh0KANuOKephmsoqIyZhxk1D7Hmz1EOV+jRR+jTx7hin6LgzyCJqyOezHySIXMfk068srxi/5BQhu3z7PFHOJH6IF3aICv+dPuZ6Jicrv8JIQGa0DmR/BAj5qGtjW4LZs8Aod2Ija6qkRwZpzl9+4qdbSHjd9yre3gNv+P1Wu8Oi9y77JqTsHa1yCv/r++x/5Pj7P7AKJmRDIom8Koe1YU6S+9scG4oikYYOPfEBb5jPl3WXz9JpwSJEPT84mcJKjUK//G/EpYr9P7yX2JTh5jvbx16EnEFg3PlGub+vZg3pnEnptvxYHdqlvKXv0FQ2BjQHeqp29gEzVQZf/8IuZEUQhEEbsj3/+W7nYfWbqKZVNQNjlshCAtFSl/6Ot70BiWcDOIezuzHPoh1YC/Fz38Jf2mF9LMnMfePb3fz7hqGMMloPZSCJcrhcvul39783d4wetKmFCy3jXU5WMGJGqTVbtZDDwkly6Rzhma0Ece6t0ihvHm0bOo42/zb2x+lHlVY8xfaXuStvy748wyZexmzjrLmz+NE9U2lcfFRopv+3bkXFZ2s1ks9LFMMtj9WPItGmEoSlZg7IpR+zIdwS5WBHdUo+HO32dfOrn/Nn8WXHpIIV9qUgxV6rFE0YRDIuEmjFq7RjKrIeP1DKVhmzDyCqaQhXKFf30MoAwLptZ45qELDj1xSSh4FlYiQsEXVqLJe9hefWxxK2cC8d5WA+Ni+dGlGFfr13bSrBrZCJNFSWfRcN6qV7Ai3CUND6BqR48UOl6qgmAYyCON3XQiEZcROluPhN3ze/PUzvPnrZ+PvFQV9pC/OJ62W4NZmCkWJNeyiCKFrSNvdOIbvI/0Q1DjcYFd9zn7uAmf/84X4mIB0Y0VmoWsoKQsZRLFenqKiqzujA7gZ96V6QevrpXn2IlGziTE6hN7Xh7+wc+XdoLBG+cvfJP3+Z8l88DnCeoNgeTXmD/V99P5egtW1uBUvkYjJTFoPnTD2KtWuXCxN0pK0TnSZqJrg1G9eiH+/hbdp7hmNlUcdF3PvboK1IjIIiGwHf6WAMTKEPztP5AeoyQSR8JG2g9bThbe4RFAoomYzGGOjCOOml07Gk4yaTqKmkoRSxgZ7hz3aitDQ0GlGtftSzC1lRCg3Jqo4DOCiCQ0FpV0Qv97++aOEUPpt47IVptwz+NKhRx9hQNtNPSqx4s9sW9u6FRShoAsTT9qEt6HOTCpZRoxDJNVsa98CQ6zXf3fWGYUE+PK9V634srMWPWg9xzh+Hd+XUPpENxmx2PSGaEJrl5VZapoDiSc67oiPhx3VEa2ytRHjEF3aQGuCEqhCQxXqpgoqJ+rMBUjkHfkcnJU5MvuPkz30KEJVqV1vhZOEwBwfQc2nkJ5P893rmPtG0fJpwmoD5/I0+mAPxu7BWP330hRqOol5cAwA9/ocRJL0848QVeo03r1GsNzZ6qwP9aD3d6GkLIgkzpUZ0FSMsQGUhEHjhxcw94+gZlOgqTjnJwmbDomj4yimjju9RFAok3zySLsqwjkzGUuw30R3ulO8d6Mrof7Dt0g+ehxjdJjIdWMm+x0oI3TsxvNp/PBt1GyG3E98jNLnv4y/vELz7TNYRw9iHTmIjCKiap3GqdOxcitxuZNz5QaZDz5H8tFj2BeuYJ+9SBRKEl0WI0/049sBURgx+8POicBbXCHz4RdQ00mEZVH//htI10O6HvXX3iD11GPkP/sTyDBCuh7Nt8/gVms0z14k88IzdP35n4z1sgx9k+6VOz2HeWAv+c98krBWo/HGO3gzO+P0lTIiJGx5T++9aFAIBVVo7Xd3/SWzoxqSqP3y6sqDF+W7a8itJ8x1BNJn2j3PkjdJVuthQN/DwcTTeFGTSrhRDH878yulJJR+K4arbPPjOJTQow1zw3mXarhGKH16tBH2JR7fYqfc0YvdCbT2GFhP9sX10qHcmMAVEXum641cCgoKKqGMqUR96WKHNa4032gb7XUE0iMkZEQfZ6/1KNed0xT9hTg+rqTJpT+x6ZzuhfAmqFcpnX0D1bSIfLeDJEbJJNC6c5j7RmieuYG5exAlaRKslkGA1pvD3DNI82wc3ks8so/Ekd3Q4liwz08QNR385SJhZXNpqD7YE/PrnthH852rmOND+Csl1FyKzPMnaJ6+ijHaT9R0UNJJtP4uopkl1FwKY/dgq/TMI3F0D41TFzH3jWC/eRVd7/TYd4o7Gt2w3mzLRUvXY/U3fofIcWmeu4jQNKTnUvvOD2L1UVWNqc6iaGOpHoRUvv5KzF6/hSFunrkQl5cBUSM+ltbThXRdpB/QPH0Wd3IGJRVni6OmTVgqt7eXrkflG6/EDRaKQliqgJS4NY/iZAXNUFBUnfAW1vnKiy8RNW20bAYtncWvlPBvUm91rtxAbUikFdfh6UoCy7Eg1U1wbZbS0gpW9wCGkaU8dR4UsUE2DYi1Js1v/oBAC8l07+4Ij9wM3coghIpnb1yTJx2aYYUhYz9JNdsRK7wXGMIiq/a2QwcpNUdCSbHsTSKR1MMSvnQZ0Pew5E1sWp6jKKgZK16GKQqR7aImTcLavfYn3S+IdqzUlQ1W/QZO1OSkPkxW6+swuqH043ikMDftJcCnHpZalQZdmzy5+Ei0vyv4s+2wQZe+NRPe/UK3NsS8e4WQKFZlULtxombLA46RUbswRRK/RUCzXs3iShuQrPqz7Lf60RSTst/JBRw3zUhSah6AJe9G+/n3a7s3NW3cFW5tRIhCQrvz3hpjA6jZFPb5GxijcYigceoC+mAP6Q88iju5gHN9jqjpkHr6KN70EsLQsc9N4M4sE9VtItshatiEtWZb4v3W0whKVcJynbBYxdg9BIaON7NMcHQ8rlV3ffzFNbS+EKGpWEfHkX6AO7GADCOiWhPCCCWVoHHqIkJCFAb3pKd4Z083DOOlfQvedOytRdU6QlNiT8R2OioVbsVWqqjriKq1jnrdqN7AqzdokYAh/aC1/W32Uavj1TqXxnbJ5crXN+gK9390rP1vM9VNShvA08oIzyAV9rK6NtNa/gu6R08goxBVmiiehl11aFanyfXvw0jkSGQGsKvLNK9eRR84CHWXTO8ewq48gdskkR0gCj3Cpo2IIkxDI9+1H8cooJtZZOQDCkYiQxj4hH4T3UrRLC8iZYQkYtG/Qbc+xOOpj7PkT+BGTUyRRFU0rttv40sXXZgklSyaMFCFjqWm6NGG2xUE6wksRWiMmYfJaj2EMmDY2I8dNdoZ8BCfq/abHE+9n8dTH6cQzBHJkJSSw5Z1Jvyz6H15EgeGkX5IaHs41+fv2ehqGKTUHJowYtIURdCjjcQeWVTbbPS3gS4MHk99nHpUphlVQMaZfIHaLs1qj4eohh3V2WUeRcMgJKARVloJKUkhmKc3HONY8v0se9PYUQVDJNAVkynnHM2oRilYYsw8zLj1KHZUp1sbJKFk79qjFQiSSg5DiSdDVWhktV5CfDzp0mg1KMTGMMeJ1IcoB8uk1W769FFu2O/E1I0tTltDSXIw+TSlYAldWAwZe1n1ZqmH8TJ71Z+hVx/lcOIZCtoIdlTDEAnSap557zor/hTlYIVR4zCHEs9QDBbJqN1t473j6xICrNgzF4aOuXcMb3KOyPHiuK2mIQwdYej4MzG3iPQDrL0jaD25WObIMkg8sh99IPZOURWs/aNYh/egJExkFOFcmiL3yWcxxoeoffddwmqdoFAh+7GnqKkqztVObm0ZSYhkS0w0DvVp+QzW3hGUZGsSjqL274gk+CHJRw8S2Q72hQmQoA/3xqKaukZ9+Ryry+e4F4jbLd2EENt+qSYN+t5/gPLZOVRLJ7A9ZChjNU4AVSFseCh6HMT2ig0iL8Dsz8bExEFE6PqY3Wn8uoNQFYhk+zNUAZEkaHoENYewuXMC6d6DeWqLDfqPdrep4o7+1F5e+l9iqep01xhRFNKsLKKbKdLdY5QWL7PeOjiw5xkqq9fpHj7O2vw5sn3jlBcvk+3fj2dXCQOHdNcoxYWL5AcOUlubIpkdwEz1IKOAwuwZrHQP+YFDrM2dif87f5Zc/wECr4mV7gEEupXBtyuoukVx4QJ2tTP8YYkUI+YherRhVKHhSptVb5p57xqSiAF9fMulrURy1T5FwZ9jt3kcU0lS8GfZZR4loWSohUUmnDOxoboJObWfUfMwGbULkDTDOnPeJUpqgdSx3SgJE79YRU1Z+GtV3Mnt4/a7zKMM6OOtLHfnkrZXH+NQ4uktz3vCeZclbwKBYJ/1BIYwuWKf2rQPiL20XeZRevURDJEkIqAelphzr1C5icdgHSk1x7j5GGk1T0TIvHu1o07XEBbDxgF69TF0xcSPHAr+PPPelVYM3GC3eYwePa7zLfrzzHvX2Gs9xpx7uX3MvdbjpJQsl5qvtxNON8MQFseSHyC5RRLGlx6nal8hpeTZZR5lwbtGjz5Knz5GJEMWvessehOE+GjCYNx8hEZUQSAYMvajCY2CP8ese7mjBlcXZrsEzFBMvMim6C+x6F1vl4yNGAdb+9CphgWmnHPstR5jLZhn0btBlzbIPutxrtpvUr1pFTFuPkKvPsrb4Suk3v8kIFCzKaTjEdYbRLUGQakSUyG2wkXupZtq5lvvftsrFiJOfK2XlQqBUNW20QTiSiEhOquhbt3P7bBeaXS79mZVbQuapp8/QVhrEqyWsQ7uonnuBmFx++YJKeW2ccF7N7opg8yhQZyFCvkndqNoCt5aA2soh1AE3loDGUXUJwqYvWnqN1YIag6jnz1JaPs0ptew50p0PTWOaqrIIMJeKNOcLZI/uRtnsRJHM1WBs1TFXd65dtq+j4xSmqry6M8dpHC9DMDu5wZ58e/H0tZWqocw9PGdKqpukeoao7423RLTFGR69+DUC6TyI9SLsyRzg3h2lWR+CM+uoGomMgrw7Bqp/DBOYw1VT6AoasxqpBmEvkMUBqiagaqZ1EuzZHr2EEUhQlFRVQ2nUcRMduM2ighFpVmevyv12Id4iB8lCEPHOn4Q6bhIx0XJpEBKgmIFYRoouo5fKKLlc7jXp/6sT/euoPXmSRwbR4YRQbGKc3nqtgb7gRhdoQr0XBIZRiRGuwiqcaZWyyUImx5h00MxNULbR03quKs1Iieg/6NHaEwVcBYryDAitacnZgIT4CxXkX5Eck8P9nwZPWuBEPiVZnv/O4We1LByBrXFeLYfe3aQ2R/+aLXhPsRDPMT/PfFAjO69QssmCKp/+gkYoYq74+l8iId4iIe4R/xIGd2HeIiHeIj/u+Oeje5DPMRDPMRD3F+8hyK8h3iIh3iIh7hbPDS6D/EQD/EQf4p4aHQf4iEe4iH+FPHQ6D7EQzzEQ/wp4qHRfYiHeIiH+FPEQ6P7EA/xEA/xp4j/P5RRmvYFIGMZAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# tweets post data cleaning\n",
    "\n",
    "data['tokens'].to_csv('cleanedtweetsonly.csv')  \n",
    "text3 = open('cleanedtweetsonly.csv').read()\n",
    "\n",
    "wordcloud = WordCloud().generate(text3)\n",
    "plt.imshow(wordcloud)\n",
    "plt.axis(\"off\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0911201f",
   "metadata": {},
   "source": [
    "- 'cibc', 'co', 'bank', 'thank' are some of the common word occurrences in cleaned up tweets\n",
    "- 'HTTPS' was a common occurrence in the original tweets but is no longer a common occurrence in cleaned up tweets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "c362daad",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Dropping text column \n",
    "\n",
    "data = data.drop('text', axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "74d64459",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Unknown                        412\n",
       "Canada                         342\n",
       "Toronto, Ontario                70\n",
       "The Caribbean                   64\n",
       "Toronto                         57\n",
       "Jamaica                         35\n",
       "United States                   30\n",
       "Barbados                        23\n",
       "Bahamas                         20\n",
       "Edmonton, Alberta               17\n",
       "Vancouver, British Columbia     16\n",
       "Calgary, Alberta                14\n",
       "Sioux Falls, SD                 14\n",
       "Ontario, Canada                 12\n",
       "Bay Street and Howe Street      10\n",
       "Chicago, IL                     10\n",
       "London, Ontario                  9\n",
       "Ottawa, Ontario                  9\n",
       "Winnipeg, Manitoba               9\n",
       "Toronto, Canada                  8\n",
       "Corner Brook, NL                 8\n",
       "Toronto, Ontario, Canada         7\n",
       "Chicago                          7\n",
       "Worldwide                        7\n",
       "Oshawa, Ontario                  7\n",
       "Name: location, dtype: int64"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Clean up locations column\n",
    "\n",
    "data['location'].value_counts().head(25)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a81ecf5e",
   "metadata": {},
   "source": [
    "- This column is too unclean and will be dropped from analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "82ddecf9",
   "metadata": {},
   "outputs": [],
   "source": [
    "data = data.drop('location', axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "982ad746",
   "metadata": {},
   "source": [
    "### Machine Learning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "aca62c5d",
   "metadata": {},
   "outputs": [],
   "source": [
    "data['sentiment'] = data['sentiment'].replace({'neutral':0, 'positive':1, 'negative':-1})\n",
    "data['sentiment'] = data['sentiment'].astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "a09a4fc2",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Assign features & target as X & y respectively\n",
    "\n",
    "X = data.drop('sentiment', axis=1)\n",
    "y = data['sentiment']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "8d8d3df3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# One hot encoding of day_of_week column\n",
    "\n",
    "X = pd.get_dummies(X, columns=['day_of_week'], drop_first=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "6bea8dda",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split data into train & test set using stratify to maintain the split of sentiment across train & test sets\n",
    "\n",
    "X_train_full, X_test, y_train_full, y_test = train_test_split(X, y, test_size=0.20, random_state=1, stratify=y,\\\n",
    "                                                   shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "c2f41636",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1447, 10)"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Shape of splits \n",
    "\n",
    "X_train_full.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "d471bca1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(362, 10)"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1b41dcca",
   "metadata": {},
   "source": [
    "- There are now 10 columns & 1560 and 391 rows respectively in full training & test sets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "66856662",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       " 1    0.438148\n",
       " 0    0.395301\n",
       "-1    0.166551\n",
       "Name: sentiment, dtype: float64"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train_full.value_counts(normalize=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "2feb3063",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       " 1    0.439227\n",
       " 0    0.395028\n",
       "-1    0.165746\n",
       "Name: sentiment, dtype: float64"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test.value_counts(normalize=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dbb83119",
   "metadata": {},
   "source": [
    "- Stratify has maintained the split of neutral ('0'), positive ('1') & negative ('-1') tweets in the dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fce2f2d1",
   "metadata": {},
   "source": [
    "### 5-fold cross validation scores (default model parameters)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "af77e8da",
   "metadata": {},
   "source": [
    "We would like each class to be predicted correctly. Tuning the 'F1' metric will ensure the maximum possibility of correct predictions across each of the target classes\n",
    "\n",
    "Further, the following additional points need consideration\n",
    "\n",
    "- Vectorizer 'TFIDF' is chosen to convert text data to numerical feature matrix \n",
    "- RandomOverSampler would be needed to handle class imbalance across target classes\n",
    "- Further, choice of ML model(s) that work well with textual data- Multinomial Naive Baiyes, Linear Support Vector Classifier, Random Forest Classifier & XGBoost \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "4e0520c7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cross Validation Model Performance on Training Set - TfidfVectorizer\n",
      "\n",
      "SupportVectorMachine : 0.38172497048147835\n",
      "Time : 0.4990891933441162\n",
      "\n",
      "\n",
      "MultinomialNaiveBayes : 0.4179577924814408\n",
      "Time : 0.15718913078308105\n",
      "\n",
      "\n",
      "RandomForest : 0.37329228832046796\n",
      "Time : 1.3154309272766114\n",
      "\n",
      "\n",
      "XGBoost : 0.38960427193273184\n",
      "Time : 33.96854658126831\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "\n",
    "# Initialize vectorizer TfidfVectorizer with default parameters, & stop_words as english\n",
    "\n",
    "vectorizer  = TfidfVectorizer(stop_words='english')\n",
    "\n",
    "# Initialize oversampler \n",
    "\n",
    "oversampler = RandomOverSampler(random_state=1)\n",
    "                                \n",
    "# Initialize ML models with default parameters\n",
    "                                \n",
    "models      = [(\"SupportVectorMachine\",  LinearSVC(random_state=1)),\n",
    "              (\"MultinomialNaiveBayes\",  MultinomialNB()),\n",
    "              (\"RandomForest\",           RandomForestClassifier(random_state=1)),\n",
    "              (\"XGBoost\",                XGBClassifier(random_state=1, eval_metric='mlogloss'))]\n",
    "\n",
    "results  = []  \n",
    "names    = []  \n",
    "\n",
    "print(\"Cross Validation Model Performance on Training Set - TfidfVectorizer\\n\")\n",
    "\n",
    "# Loop through models & calculate mean cross validated scores              \n",
    "for name, model in models:\n",
    "    \n",
    "    scores   = []\n",
    "    times    = []\n",
    "    \n",
    "    kf = StratifiedKFold(n_splits=5, random_state=1, shuffle=True)\n",
    "    \n",
    "    for train_index, val_index in kf.split(X_train_full, y_train_full):\n",
    "           \n",
    "        X_train, X_val = X_train_full.iloc[train_index], X_train_full.iloc[val_index]\n",
    "        y_train, y_val = y_train_full.iloc[train_index], y_train_full.iloc[val_index]\n",
    "\n",
    "        # vectorization\n",
    "        \n",
    "        num_tweets_train = vectorizer.fit_transform(X_train['tokens'])\n",
    "        num_tweets_val   = vectorizer.transform(X_val['tokens'])\n",
    "        \n",
    "        feature_names= vectorizer.get_feature_names_out()\n",
    "        \n",
    "        bow_df_train = pd.DataFrame(num_tweets_train.toarray(), columns=feature_names)\n",
    "        bow_df_val   = pd.DataFrame(num_tweets_val.toarray(),   columns=feature_names)\n",
    "        \n",
    "        X_train   = X_train.drop('tokens', axis=1)\n",
    "        X_val     = X_val.drop('tokens', axis=1)\n",
    "    \n",
    "        X_train_f = X_train.reset_index(drop=True).merge(bow_df_train.reset_index(drop=True), left_index=True, right_index=True)\n",
    "        X_val_f   = X_val.reset_index(drop=True).merge(bow_df_train.reset_index(drop=True), left_index=True, right_index=True)\n",
    "        \n",
    "        \n",
    "        # oversampling\n",
    "        \n",
    "        X_train_f_over, y_train_over = oversampler.fit_resample(X_train_f, y_train)\n",
    "        \n",
    "        # model fit & prediction\n",
    "        \n",
    "        start = time.time()\n",
    "        model.fit(X_train_f_over, y_train_over) \n",
    "\n",
    "        y_pred = model.predict(X_val_f)\n",
    "        end = time.time()\n",
    "        \n",
    "        score = f1_score(y_val,y_pred, average='macro') # get avg f1_score across classes\n",
    "        time_ = end - start\n",
    "        \n",
    "        scores.append(score)\n",
    "        times.append(time_)\n",
    "    \n",
    "    names.append(name)\n",
    "    results.append(sum(scores)/len(scores))\n",
    "           \n",
    "    print(f'{name} : {sum(scores)/len(scores)}')\n",
    "    print(f'Time : {sum(times)/len(times)}')\n",
    "    print('\\n')\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c725465d",
   "metadata": {},
   "source": [
    "- Best scores were with Multinomial Naive Bayes & XGBoost. Multinomial Naive Bayes takes the least amount of time to fit & predict, while XGBoost takes the maximum amount of time\n",
    "- Random Forest is easier to interpret with class_feature_importance built in the module with not much compromise in performance score & time \n",
    "\n",
    "The next step will involve tuning hyperparameters for the vectorizer & the chosen model in an effort to improve performance further"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5520730c",
   "metadata": {},
   "source": [
    "### Hyperparameter Tuning - GridSearchCV - RandomForest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "7e556cd3",
   "metadata": {},
   "outputs": [],
   "source": [
    "pipe       = Pipeline(\n",
    "                    [('TfidfVectorizer',TfidfVectorizer(stop_words='english')),\n",
    "                     ('oversampler',    RandomOverSampler(random_state=1)),\n",
    "                     ('RandomForest',   RandomForestClassifier(random_state=1))])\n",
    "\n",
    "parameters = {\n",
    "              'TfidfVectorizer__ngram_range' : [(1,1),(1,2)],\n",
    "              'TfidfVectorizer__max_features': range(1000,10000,2500),\n",
    "              'RandomForest__n_estimators'   : range(10,100,25),\n",
    "              'RandomForest__criterion'      : ['gini', 'entropy'],\n",
    "              'RandomForest__max_depth'      : range(3,15,3), \n",
    "              'RandomForest__max_features'   : ['auto', 'sqrt'],\n",
    "            }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "aeafa545",
   "metadata": {},
   "outputs": [],
   "source": [
    "def GridSearchCV_(estimator, estimator_params, X, y):\n",
    "    \n",
    "    kf = StratifiedKFold(n_splits=3, random_state=1, shuffle=True)\n",
    "    \n",
    "    best_score   = 0  \n",
    "    best_params  = None\n",
    "    \n",
    "    for params_permut in ParameterGrid(estimator_params):\n",
    "        \n",
    "        scores   = np.empty(kf.n_splits)\n",
    "        \n",
    "        estimator.set_params(**params_permut)\n",
    "        \n",
    "        for i, (train_index, val_index) in enumerate(kf.split(X_train_full, y_train_full)):\n",
    "            \n",
    "            X_train, X_val = X_train_full.iloc[train_index], X_train_full.iloc[val_index]\n",
    "            y_train, y_val = y_train_full.iloc[train_index], y_train_full.iloc[val_index]\n",
    "            \n",
    "            # vectorization\n",
    "\n",
    "            num_tweets_train = estimator.steps[0][1].fit_transform(X_train['tokens'])\n",
    "            num_tweets_val   = estimator.steps[0][1].transform(X_val['tokens'])\n",
    "\n",
    "            feature_names= estimator.steps[0][1].get_feature_names_out()\n",
    "\n",
    "            bow_df_train = pd.DataFrame(num_tweets_train.toarray(), columns=feature_names)\n",
    "            bow_df_val   = pd.DataFrame(num_tweets_val.toarray(),   columns=feature_names)\n",
    "\n",
    "            X_train   = X_train.drop('tokens', axis=1)\n",
    "            X_val     = X_val.drop('tokens', axis=1)\n",
    "\n",
    "            X_train_f = X_train.reset_index(drop=True).merge(bow_df_train.reset_index(drop=True), left_index=True, right_index=True)\n",
    "            X_val_f   = X_val.reset_index(drop=True).merge(bow_df_val.reset_index(drop=True), left_index=True, right_index=True)\n",
    "\n",
    "            # oversampling\n",
    "        \n",
    "            X_train_f_over, y_train_over = estimator.steps[1][1].fit_resample(X_train_f, y_train)\n",
    "            \n",
    "            # fit & prediction\n",
    "            \n",
    "            estimator.steps[2][1].fit(X_train_f_over, y_train_over)\n",
    "\n",
    "            y_pred = estimator.steps[2][1].predict(X_val_f)\n",
    "\n",
    "            scores[i] = f1_score(y_val,y_pred, average='macro') # get avg weighted f1_score across classes\n",
    "\n",
    "        scores_mean   = scores.mean()\n",
    "        \n",
    "        if scores_mean>best_score:\n",
    "            best_params= params_permut\n",
    "            best_score = scores_mean\n",
    "            \n",
    "    \n",
    "    print(f'Best Hyperparameters are:\\n {best_params}')\n",
    "    print(f'Best Score is:\\n {best_score}')\n",
    "              "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "3e2ddf58",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Hyperparameters are:\n",
      " {'RandomForest__criterion': 'gini', 'RandomForest__max_depth': 12, 'RandomForest__max_features': 'auto', 'RandomForest__n_estimators': 85, 'TfidfVectorizer__max_features': 1000, 'TfidfVectorizer__ngram_range': (1, 1)}\n",
      "Best Score is:\n",
      " 0.7199566069352127\n"
     ]
    }
   ],
   "source": [
    "GridSearchCV_(pipe, parameters, X_train_full, y_train_full)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b74c696b",
   "metadata": {},
   "source": [
    "#### Building the final classifier & making predictions using the test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "8f99d223",
   "metadata": {},
   "outputs": [],
   "source": [
    "# The steps can be built into a function \n",
    "\n",
    "def sentiment(X_sample, X, y):   # X & y are the features & predictions for known dataset \n",
    "                                 # & X_sample is feature set for data where sentiment(s) are unknown\n",
    "   \n",
    "    pipe_final   = Pipeline(\n",
    "                [('TfidfVectorizer',TfidfVectorizer(stop_words='english', ngram_range=(1, 1),max_features=1000)),\n",
    "                 ('oversampler',    RandomOverSampler(random_state=1)),\n",
    "                 ('RandomForest'   , RandomForestClassifier(random_state=1, criterion='gini',\\\n",
    "                                                             max_depth=12, max_features='auto', \\\n",
    "                                                             n_estimators=85))])\n",
    "    \n",
    "    # vectorization\n",
    "\n",
    "    num_tweets            = pipe_final.steps[0][1].fit_transform(X['tokens'])\n",
    "    num_tweets_sample     = pipe_final.steps[0][1].transform(X_sample['tokens'])\n",
    "\n",
    "    feature_names         = pipe_final.steps[0][1].get_feature_names_out()\n",
    "\n",
    "    bow_df_full           = pd.DataFrame(num_tweets.toarray(), columns=feature_names)\n",
    "    bow_df_sample         = pd.DataFrame(num_tweets_sample.toarray(), columns=feature_names)\n",
    "\n",
    "    X                     = X.drop('tokens', axis=1)\n",
    "    X_sample              = X_sample.drop('tokens', axis=1)\n",
    "    \n",
    "    feature_names_in_     = list(X.columns)\n",
    "\n",
    "    X_f                   = X.reset_index(drop=True).merge(bow_df_full.reset_index(drop=True), left_index=True, right_index=True)\n",
    "    X_sample_f            = X_sample.reset_index(drop=True).merge(bow_df_sample.reset_index(drop=True), left_index=True, right_index=True)\n",
    "    \n",
    "    # oversampling\n",
    "    \n",
    "    X_f_over, y_over      = pipe_final.steps[1][1].fit_resample(X_f, y)\n",
    "    \n",
    "    # model fit & prediction\n",
    "    \n",
    "    pipe_final.steps[2][1].fit(X_f_over, y_over)\n",
    "    \n",
    "    pred                  = pipe_final.steps[2][1].predict(X_sample_f)\n",
    "    \n",
    "    feature_names         = feature_names_in_ + list(pipe_final.steps[0][1].get_feature_names_out())\n",
    "    \n",
    "    return pred, pipe_final.steps[2][1], feature_names, X_f_over, y_over\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "6a6cc320",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Final Classifier Unbiased Testing Performance:\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "          -1       0.84      0.52      0.64        60\n",
      "           0       0.70      0.91      0.79       143\n",
      "           1       0.88      0.76      0.81       159\n",
      "\n",
      "    accuracy                           0.78       362\n",
      "   macro avg       0.80      0.73      0.75       362\n",
      "weighted avg       0.80      0.78      0.78       362\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Testing \n",
    "\n",
    "pred_test, tuned_model, feature_names, X_, y_  = sentiment(X_test, X_train_full, y_train_full)\n",
    "\n",
    "print('Final Classifier Unbiased Testing Performance:')\n",
    "print(classification_report(y_test, pred_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "824aa7ae",
   "metadata": {},
   "source": [
    "- The final classifier is able to achieve a F1 score of ~65% for negative sentiments, ~>=75% for positive & neutral sentiments"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "db2c0af0",
   "metadata": {},
   "source": [
    "### Feature Importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "8ed438f2",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import scale\n",
    "from collections import OrderedDict\n",
    "def class_feature_importance(X, y, class_labels, feature_names, feature_importances, max_top = 20):\n",
    "    \"\"\"\n",
    "    Saves feature importances by class\n",
    "    Reworked version of https://stackoverflow.com/a/36201964\n",
    "    \n",
    "    Arguments:\n",
    "    X, y - array of features and labels\n",
    "    class_labels - labels corresponding to values in y. For example, if y has [0, 1, 2], class_labels can be ['neg', 'pos', 'neutral']\n",
    "    feature_names - names of columns in X\n",
    "    feature_importances - obtained from a classifier such as decision tree\n",
    "    max_top - the maximum number of most important features for each class to keep\n",
    "    \n",
    "    Returns:\n",
    "    result - dictionary of feature importances by class\n",
    "    \"\"\"\n",
    "    X = scale(X)\n",
    "    \n",
    "    result = {}\n",
    "    # extract feature importances by class\n",
    "    for c in set(y):\n",
    "        result[class_labels[c]] = dict(zip(feature_names, np.mean(X[y == c, :], axis = 0) * feature_importances))\n",
    "    \n",
    "    # sort by importance and keep the most important\n",
    "    for cname in result.keys():\n",
    "        result[cname] = sorted(result[cname].items(), key = lambda x: x[1], reverse = True)[:max_top]\n",
    "    \n",
    "    return result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "6910a630",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{0: [('http', 0.005254903628625936),\n",
       "  ('cibc', 0.0035686977763002403),\n",
       "  ('outperform', 0.0012783284104765282),\n",
       "  ('upgrad', 0.001105422848748587),\n",
       "  ('canopi', 0.0008237733183353656),\n",
       "  ('million', 0.0006391237213643123),\n",
       "  ('affili', 0.0005750117591376953),\n",
       "  ('pot', 0.0005617869930377333),\n",
       "  ('pharmhous', 0.0005171659119547022),\n",
       "  ('stock', 0.00048245308190496906),\n",
       "  ('compani', 0.0003972520080258514),\n",
       "  ('80', 0.0002840897183251984),\n",
       "  ('loan', 0.0002778209808019329),\n",
       "  ('recruit', 0.0002509295685181296),\n",
       "  ('reaffirm', 0.00024354833156878063),\n",
       "  ('provid', 0.00023361912402952383),\n",
       "  ('specincanada', 0.00021524804917062436),\n",
       "  ('gold', 0.00018683605095486946),\n",
       "  ('price', 0.0001651093407695493),\n",
       "  ('market', 0.00015872398701660342)],\n",
       " 1: [('number_positive_words', 0.024674967297793084),\n",
       "  ('thank', 0.0075800359501043595),\n",
       "  ('new', 0.006824029545674138),\n",
       "  ('latest', 0.004987000695066227),\n",
       "  ('read', 0.004323775844940422),\n",
       "  ('beat', 0.004250830164816125),\n",
       "  ('daili', 0.0033137842100420677),\n",
       "  ('click', 0.0030861135368116816),\n",
       "  ('gundi', 0.002906447759835412),\n",
       "  ('isg', 0.0021729156451157695),\n",
       "  ('look', 0.0019328118519697814),\n",
       "  ('love', 0.001404756888657215),\n",
       "  ('market', 0.0013694959613604462),\n",
       "  ('good', 0.0010149851813869692),\n",
       "  ('great', 0.0010121532625143286),\n",
       "  ('firstcaribbean', 0.0009225602339887848),\n",
       "  ('wood', 0.0008998155189470076),\n",
       "  ('home', 0.0007702234465535215),\n",
       "  ('game', 0.0007399270199097473),\n",
       "  ('cool', 0.0006142506427804738)],\n",
       " -1: [('growth', 0.020508279448409432),\n",
       "  ('mortgag', 0.019366596158985994),\n",
       "  ('flat', 0.01593803535685603),\n",
       "  ('low', 0.013638174657405067),\n",
       "  ('ceo', 0.00991227325635339),\n",
       "  ('number_negative_words', 0.009865706883072735),\n",
       "  ('singl', 0.008695714600743245),\n",
       "  ('year', 0.007091239388495461),\n",
       "  ('thi', 0.0036762564809373643),\n",
       "  ('sorri', 0.0034910600221196805),\n",
       "  ('digit', 0.0034875670207551697),\n",
       "  ('forese', 0.0022939260500948098),\n",
       "  ('nint', 0.001975114963768389),\n",
       "  ('expect', 0.0019119478668486187),\n",
       "  ('head', 0.0019059296524093635),\n",
       "  ('close', 0.0015397788130216727),\n",
       "  ('prioriti', 0.001397608266141923),\n",
       "  ('poll', 0.0013932492910597326),\n",
       "  ('fli', 0.0013557985039408462),\n",
       "  ('wait', 0.0013532427413267963)]}"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "class_feature_importance(X_, y_, [0,1,-1], \\\n",
    "                         feature_names, tuned_model.feature_importances_)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "69e2c97c",
   "metadata": {},
   "source": [
    "### Summary & Conclusion"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "64a290f4",
   "metadata": {},
   "source": [
    "- TextBlob was used to assign sentiment labels to tweets. There was inherently some imbalance in classes of sentiments for the dataset; negative:neutral:positive ratio was 0.15:0.43:0.40\n",
    "\n",
    "\n",
    "- Data cleaning, visualization was followed by text analytics (removing non alphanumeric characters, tokenizing sentences, lemmatization, removing stop words & performing vectorization using TF-IDF). This was followed by randomoversampling to handle target imbalance\n",
    "\n",
    "\n",
    "- Four different base models were fitted to the training set. The best cross validated scores though were achieved with MultinomialNB & XGBoost; however Random Forest was chosen for hyperparameter tuning as it's easier to interpret with the class feature improtances, & with not much compromise with score & computational time\n",
    "\n",
    "\n",
    "- A pipeline was further built for TF-IDF & Random Forest, thereby enabling tuning for the best hyper-parameters. Post tuning, performance was enhanced to achieve a F1 score of ~0.65 for negative sentiment & >0.75 for neutral & positive sentiments on unseen test data \n",
    "\n",
    "\n",
    "- For the final model, feature importance was identified for all 3 classes (negative, neutral & positive tweets). As expected \n",
    "  - number_negative_words had high feature importance for predicting negative sentiments & number_positive_words had high feature importance for predicting positive sentiments\n",
    "  - The words of importances associated with negative tweets were found to be 'flat', 'mortgage', 'single', 'low', 'sorry', 'close' etc\n",
    "  - The words of importances associated with positive tweets were found to be 'new', 'thank', 'wood', 'latest', 'game', 'love', 'home', 'great', 'good'\n",
    "  - The words of importances associated with neutral tweets were found to be 'http', 'cibc', 'company', 'recruit', 'reaffirm', 'provide'\n",
    "\n",
    " The feature importance words & tweet sentiments make somewhat intuitive sense lending confidence in the explainability of the final model!"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.8.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}