RNA_EHF.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "RNA_EHF.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [
        "GVAmioM8h2W9"
      ],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eJcEWIjtK6pM",
        "colab_type": "text"
      },
      "source": [
        "# Rede Neural Artificial - Edificações de Escritórios\n",
        "\n",
        "Código para desenvolvimento de uma RNA para predição de fração de horas em conforto em Edificações de Escritório com ventilação natural.\n",
        "\n",
        "## _Neural Networks - Office Buildings_\n",
        "_Development of a ANN to estimate thermal comfort Exceedance Hour Fraction (EHF) in naturally ventilated office buildings._"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hG_SxQ83LR_H",
        "colab_type": "text"
      },
      "source": [
        "## 1. Importação das bibiliotecas\n",
        "### _1. Import libraries_"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KM_7HUToKylL",
        "colab_type": "code",
        "cellView": "both",
        "colab": {}
      },
      "source": [
        "from __future__ import print_function\n",
        "\n",
        "from google.colab import files\n",
        "\n",
        "import math\n",
        "\n",
        "from IPython import display\n",
        "from matplotlib import cm\n",
        "from matplotlib import gridspec\n",
        "from matplotlib import pyplot as plt\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from sklearn import metrics\n",
        "import tensorflow as tf\n",
        "from tensorflow.python.data import Dataset\n",
        "\n",
        "tf.logging.set_verbosity(tf.logging.ERROR)\n",
        "pd.options.display.max_rows = 18\n",
        "pd.options.display.float_format = '{:.2f}'.format\n",
        "\n",
        "from datetime import datetime\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6BbkyoG1MDgW",
        "colab_type": "text"
      },
      "source": [
        "## 2. Importação dos dados\n",
        "### _3. Import data_"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "52l3I_81LHwC",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "uploaded = files.upload()\n",
        "\n",
        "for fn in uploaded.keys():\n",
        "  print('User uploaded file \"{name}\" with length {length} bytes'.format(\n",
        "      name=fn, length=len(uploaded[fn])))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "d4XtezVCMVTB",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "dataset = pd.read_csv('dataset_08-14.csv')\n",
        "dataset = dataset.reindex(\n",
        "    np.random.permutation(dataset.index))\n",
        "\n",
        "dataset_validation = pd.read_csv('dataset_validation08-14.csv')\n",
        "dataset_validation = dataset_validation.reindex(\n",
        "    np.random.permutation(dataset_validation.index))\n",
        "\n",
        "dataset_test = pd.read_csv('dataset_test_08-14b.csv')\n",
        "dataset_test = dataset_test.reindex(\n",
        "    np.random.permutation(dataset_test.index))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gmIJxHUNPlzY",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def preprocess_features(dataset, features, lhs_to_sobol=False):\n",
        "  \"\"\"Prepares input features from dataset.\n",
        "\n",
        "  Args:\n",
        "    dataset: A Pandas DataFrame with the dataset.\n",
        "    features: A list of strings with the features names.\n",
        "  Returns:\n",
        "    A DataFrame that contains the features to be used for the model.\n",
        "  \"\"\"\n",
        "  selected_features = dataset[features]\n",
        "  processed_features = selected_features.copy()\n",
        "  \n",
        "  if lhs_to_sobol:\n",
        "    processed_features = processed_features*2-1\n",
        "  \n",
        "  return processed_features\n",
        "\n",
        "def preprocess_targets(dataset, target):\n",
        "  \"\"\"Prepares target features (i.e., labels) from data set.\n",
        "\n",
        "  Args:\n",
        "    dataset: A Pandas DataFrame with the dataset.\n",
        "    targets: A string with the target name.\n",
        "  Returns:\n",
        "    A DataFrame that contains the target feature.\n",
        "  \"\"\"\n",
        "  output_targets = pd.DataFrame()\n",
        "  output_targets[target] = dataset[target]\n",
        "  \n",
        "  return output_targets"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gm-NYTpmR3R4",
        "colab_type": "code",
        "outputId": "e2011bde-88f1-437f-9e76-b76929f44337",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 89
        }
      },
      "source": [
        "print(list(dataset))\n",
        "print(list(dataset_validation))\n",
        "print(list(dataset_test))"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "['open_fac', 'v_ar', 'room_type', 'area', 'people', 'floor_height', 'roof', 'shading', 'ground', 'wall_u', 'absorptance', 'glass', 'azimuth', 'wwr', 'folder', 'file', 'zone', 'temp', 'temp_max', 'ach', 'ehf']\n",
            "['open_fac', 'v_ar', 'room_type', 'area', 'people', 'floor_height', 'roof', 'shading', 'ground', 'wall_u', 'absorptance', 'glass', 'azimuth', 'wwr', 'folder', 'file', 'zone', 'temp', 'temp_max', 'ach', 'ehf']\n",
            "['open_fac', 'v_ar', 'room_type', 'area', 'people', 'floor_height', 'roof', 'shading', 'ground', 'wall_u', 'absorptance', 'glass', 'azimuth', 'wwr', 'n_floor', 'ratio', 'zone_height', 'wall_ct', 'bldg_ratio', 'folder', 'file', 'zone', 'temp', 'temp_max', 'ach', 'ehf']\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sY2pyLW6MkaD",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 638
        },
        "outputId": "bef254e7-23e7-4444-fdf7-a0ee6793a2a1"
      },
      "source": [
        "# Define the feature names and the target\n",
        "features = ['area', 'azimuth', 'floor_height', 'absorptance', 'wall_u', 'wwr', \n",
        "            'shading', 'people', 'open_fac', 'roof', 'ground', 'v_ar',\n",
        "            'glass', 'room_type']\n",
        "\n",
        "target = 'ehf'\n",
        "\n",
        "dataset_features = preprocess_features(dataset,features,\n",
        "                                       lhs_to_sobol=True)  # False)  # \n",
        "dataset_targets = preprocess_targets(dataset,target)\n",
        "\n",
        "dataset_validation_features = preprocess_features(dataset_validation,features, \n",
        "                                                  lhs_to_sobol=True)\n",
        "dataset_validation_targets = preprocess_targets(dataset_validation,target)\n",
        "\n",
        "dataset_test_features = preprocess_features(dataset_test,features, \n",
        "                                                  lhs_to_sobol=True)  # False)  # \n",
        "dataset_test_targets = preprocess_targets(dataset_test,target)\n",
        "\n",
        "\n",
        "# Double-check that we've done the right thing.\n",
        "print(\"Features summary:\")\n",
        "display.display(dataset_features.describe())\n",
        "print(\"Target summary:\")\n",
        "display.display(dataset_targets.describe())"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Features summary:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>area</th>\n",
              "      <th>azimuth</th>\n",
              "      <th>floor_height</th>\n",
              "      <th>absorptance</th>\n",
              "      <th>wall_u</th>\n",
              "      <th>wwr</th>\n",
              "      <th>shading</th>\n",
              "      <th>people</th>\n",
              "      <th>open_fac</th>\n",
              "      <th>roof</th>\n",
              "      <th>ground</th>\n",
              "      <th>v_ar</th>\n",
              "      <th>glass</th>\n",
              "      <th>room_type</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "           area   azimuth  floor_height  ...      v_ar     glass  room_type\n",
              "count 100000.00 100000.00     100000.00  ... 100000.00 100000.00  100000.00\n",
              "mean      -0.00     -0.00          0.00  ...      0.00      0.00      -0.00\n",
              "std        0.58      0.58          0.58  ...      0.58      0.58       0.58\n",
              "min       -1.00     -1.00         -1.00  ...     -1.00     -1.00      -1.00\n",
              "25%       -0.50     -0.50         -0.50  ...     -0.50     -0.50      -0.50\n",
              "50%       -0.00     -0.00         -0.00  ...      0.00      0.00      -0.00\n",
              "75%        0.50      0.50          0.50  ...      0.50      0.50       0.50\n",
              "max        1.00      1.00          1.00  ...      1.00      1.00       1.00\n",
              "\n",
              "[8 rows x 14 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Target summary:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>ehf</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>100000.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>0.77</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>0.18</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>0.15</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>0.64</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>0.79</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>0.92</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "            ehf\n",
              "count 100000.00\n",
              "mean       0.77\n",
              "std        0.18\n",
              "min        0.15\n",
              "25%        0.64\n",
              "50%        0.79\n",
              "75%        0.92\n",
              "max        1.00"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Y3i9ev2CTVGj",
        "colab_type": "text"
      },
      "source": [
        "## 3. Análise de correlação\n",
        "### _3. Correlation analysis_"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hxZYAaLSTMUA",
        "colab_type": "code",
        "outputId": "cfe76b33-a8c8-467f-deb7-4cf942246183",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 540
        }
      },
      "source": [
        "correlation_dataframe = dataset_features.copy()\n",
        "correlation_dataframe[\"target\"] = dataset_targets[target]\n",
        "\n",
        "correlation_dataframe.corr()"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>area</th>\n",
              "      <th>azimuth</th>\n",
              "      <th>floor_height</th>\n",
              "      <th>absorptance</th>\n",
              "      <th>wall_u</th>\n",
              "      <th>wwr</th>\n",
              "      <th>shading</th>\n",
              "      <th>people</th>\n",
              "      <th>open_fac</th>\n",
              "      <th>roof</th>\n",
              "      <th>ground</th>\n",
              "      <th>v_ar</th>\n",
              "      <th>glass</th>\n",
              "      <th>room_type</th>\n",
              "      <th>target</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>area</th>\n",
              "      <td>1.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.27</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>azimuth</th>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.01</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.02</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>floor_height</th>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>-0.01</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.01</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.01</td>\n",
              "      <td>-0.21</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>absorptance</th>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.01</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.01</td>\n",
              "      <td>-0.01</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.15</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>wall_u</th>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.01</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.01</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.15</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>wwr</th>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.01</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.15</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>shading</th>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>-0.01</td>\n",
              "      <td>0.01</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.16</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>people</th>\n",
              "      <td>0.00</td>\n",
              "      <td>0.01</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.01</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.01</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.23</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>open_fac</th>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.01</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.01</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.53</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>roof</th>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.01</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.03</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>ground</th>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.01</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.01</td>\n",
              "      <td>-0.10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>v_ar</th>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.01</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.01</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.40</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>glass</th>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.14</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>room_type</th>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.01</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.01</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.22</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>target</th>\n",
              "      <td>0.27</td>\n",
              "      <td>-0.02</td>\n",
              "      <td>-0.21</td>\n",
              "      <td>0.15</td>\n",
              "      <td>-0.15</td>\n",
              "      <td>-0.15</td>\n",
              "      <td>-0.16</td>\n",
              "      <td>0.23</td>\n",
              "      <td>-0.53</td>\n",
              "      <td>0.03</td>\n",
              "      <td>-0.10</td>\n",
              "      <td>-0.40</td>\n",
              "      <td>0.14</td>\n",
              "      <td>0.22</td>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "              area  azimuth  floor_height  ...  glass  room_type  target\n",
              "area          1.00     0.00          0.00  ...   0.00      -0.00    0.27\n",
              "azimuth       0.00     1.00          0.00  ...  -0.00       0.00   -0.02\n",
              "floor_height  0.00     0.00          1.00  ...   0.00       0.01   -0.21\n",
              "absorptance  -0.00     0.00         -0.01  ...  -0.00       0.00    0.15\n",
              "wall_u        0.00    -0.00          0.00  ...  -0.00      -0.00   -0.15\n",
              "wwr          -0.00    -0.00         -0.00  ...  -0.00      -0.00   -0.15\n",
              "shading      -0.00     0.00         -0.00  ...   0.00       0.00   -0.16\n",
              "people        0.00     0.01          0.00  ...   0.00      -0.00    0.23\n",
              "open_fac     -0.00     0.00         -0.01  ...   0.00       0.00   -0.53\n",
              "roof         -0.00     0.00         -0.00  ...  -0.00       0.00    0.03\n",
              "ground       -0.00    -0.00          0.00  ...  -0.00       0.01   -0.10\n",
              "v_ar         -0.00     0.00         -0.00  ...  -0.00       0.00   -0.40\n",
              "glass         0.00    -0.00          0.00  ...   1.00      -0.00    0.14\n",
              "room_type    -0.00     0.00          0.01  ...  -0.00       1.00    0.22\n",
              "target        0.27    -0.02         -0.21  ...   0.14       0.22    1.00\n",
              "\n",
              "[15 rows x 15 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4E6ESpl0bv-8",
        "colab_type": "text"
      },
      "source": [
        "## 4. Normalização dos dados\n",
        "### _4. Data normalization_"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nyBlqjlLcVjm",
        "colab_type": "code",
        "outputId": "4e9f7100-ff52-4af0-9458-8b609942e6db",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 736
        }
      },
      "source": [
        "# Check features' histograms.\n",
        "\n",
        "_ = dataset_features.hist(bins=20, figsize=(18, 12), xlabelsize=10)"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABB4AAAK7CAYAAABVp0xUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\neJAkSZIkSQvjxIMkSZIkSVqYvwFbSq/HcyfirgAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 1296x864 with 16 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OK-BZtrZcX2K",
        "colab_type": "code",
        "outputId": "af4168e4-9b38-45c9-bb49-9830f17a787a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 736
        }
      },
      "source": [
        "# Check targets' histogram.\n",
        "\n",
        "_ = dataset_targets.hist(bins=20, figsize=(18, 12), xlabelsize=10)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABCQAAAK7CAYAAAAqdl6lAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nOZCQJEmSJEmd+xtL6DjTq0cXmAAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 1296x864 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7JjTWr-Pb0c9",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Functions to normalize the data.\n",
        "def linear_scale(series):\n",
        "  min_val = series.min()\n",
        "  max_val = series.max()\n",
        "  scale = (max_val - min_val) / 2.0\n",
        "  return series.apply(lambda x:((x - min_val) / scale) - 1.0)\n",
        "\n",
        "def log_normalize(series, base = 10):\n",
        "  return series.apply(lambda x:math.log(x+1.0, base))\n",
        "\n",
        "def inv_log_normalize(series, base = 10):\n",
        "  return series.apply(lambda x:math.log(1/(x+1.0), base))\n",
        "\n",
        "def pow_normalize(series, base=1/2):\n",
        "  return series.apply(lambda x:math.pow(x, base))\n",
        "\n",
        "def sigmoid_normalize(series):\n",
        "  return series.apply(lambda x:1/(1+math.exp(-x)))\n",
        "\n",
        "def tan_normalize(series):\n",
        "  return series.apply(lambda x:math.tan(x))\n",
        "\n",
        "def asinh_normalize(series):\n",
        "  return series.apply(lambda x:math.asinh(x))\n",
        "\n",
        "def atan_normalize(series):\n",
        "  return series.apply(lambda x:math.atan(x))\n",
        "\n",
        "def clip(series, clip_to_min, clip_to_max):\n",
        "  return series.apply(lambda x:(\n",
        "    min(max(x, clip_to_min), clip_to_max)))\n",
        "\n",
        "def z_score_normalize(series):\n",
        "  mean = series.mean()\n",
        "  std_dv = series.std()\n",
        "  return series.apply(lambda x:(x - mean) / std_dv)\n",
        "\n",
        "def binary_threshold(series, threshold):\n",
        "  return series.apply(lambda x:(1 if x > threshold else 0))\n",
        "\n",
        "def get_quantile_based_buckets(feature_values, num_buckets):\n",
        "  quantiles = feature_values.quantile(\n",
        "    [(i+1.)/(num_buckets + 1.) for i in range(num_buckets)])\n",
        "  return [quantiles[q] for q in quantiles.keys()]\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "up1zSxIoRnZo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "## Training dataset\n",
        "\n",
        "dataset_features['roof'] = binary_threshold(dataset_features['roof'], .6)  # 0)\n",
        "dataset_features['ground'] = binary_threshold(dataset_features['ground'], .7)  # 0)\n",
        "\n",
        "# room_type from num to factor\n",
        "string_roomtype = dataset_features['room_type'].apply(lambda x: '0_window')\n",
        "string_roomtype[dataset_features['room_type'] < .6] = '3_wall'\n",
        "string_roomtype[dataset_features['room_type'] < .2] = '1_wall'\n",
        "string_roomtype[dataset_features['room_type'] < -.2] = '3_window'\n",
        "string_roomtype[dataset_features['room_type'] < -.6] = '1_window'\n",
        "dataset_features['room_type'] = string_roomtype\n",
        "\n",
        "# v_ar to discrete\n",
        "string_v_ar = dataset_features['v_ar'].apply(lambda x: .5)  \n",
        "string_v_ar[dataset_features['v_ar'] < .5] = 0\n",
        "string_v_ar[dataset_features['v_ar'] < 0] = -.5\n",
        "string_v_ar[dataset_features['v_ar'] < -.5] = -1\n",
        "dataset_features['v_ar'] = string_v_ar\n",
        "\n",
        "# Validation dataset\n",
        "\n",
        "dataset_validation_features['roof'] = binary_threshold(dataset_validation_features['roof'], .6)  # 0)\n",
        "dataset_validation_features['ground'] = binary_threshold(dataset_validation_features['ground'], .7)  # 0)\n",
        "\n",
        "# room_type from num to factor\n",
        "string_roomtype = dataset_validation_features['room_type'].apply(lambda x: '0_window')  \n",
        "string_roomtype[dataset_validation_features['room_type'] < .6] = '3_wall'\n",
        "string_roomtype[dataset_validation_features['room_type'] < .2] = '1_wall'\n",
        "string_roomtype[dataset_validation_features['room_type'] < -.2] = '3_window'\n",
        "string_roomtype[dataset_validation_features['room_type'] < -.6] = '1_window'\n",
        "dataset_validation_features['room_type'] = string_roomtype\n",
        "\n",
        "# v_ar to discrete\n",
        "string_v_ar = dataset_validation_features['v_ar'].apply(lambda x: .5)  \n",
        "string_v_ar[dataset_validation_features['v_ar'] < .5] = 0\n",
        "string_v_ar[dataset_validation_features['v_ar'] < 0] = -.5\n",
        "string_v_ar[dataset_validation_features['v_ar'] < -.5] = -1\n",
        "dataset_validation_features['v_ar'] = string_v_ar\n",
        "\n",
        "# Test dataset\n",
        "\n",
        "dataset_test_features['roof'] = binary_threshold(dataset_test_features['roof'], .6)  # 0)\n",
        "dataset_test_features['ground'] = binary_threshold(dataset_test_features['ground'], .7)  # 0)\n",
        "\n",
        "# room_type from num to factor\n",
        "string_roomtype = dataset_test_features['room_type'].apply(lambda x: '0_window')  \n",
        "string_roomtype[dataset_test_features['room_type'] < .6] = '3_wall'\n",
        "string_roomtype[dataset_test_features['room_type'] < .2] = '1_wall'\n",
        "string_roomtype[dataset_test_features['room_type'] < -.2] = '3_window'\n",
        "string_roomtype[dataset_test_features['room_type'] < -.6] = '1_window'\n",
        "dataset_test_features['room_type'] = string_roomtype\n",
        "\n",
        "# v_ar to discrete\n",
        "string_v_ar = dataset_test_features['v_ar'].apply(lambda x: .5)  \n",
        "string_v_ar[dataset_test_features['v_ar'] < .5] = 0\n",
        "string_v_ar[dataset_test_features['v_ar'] < 0] = -.5\n",
        "string_v_ar[dataset_test_features['v_ar'] < -.5] = -1\n",
        "dataset_test_features['v_ar'] = string_v_ar"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dOdKR6BYVB9O",
        "colab_type": "code",
        "outputId": "4567b71f-dab2-40c7-bcb6-a0e54e809e84",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 736
        }
      },
      "source": [
        "# Check features' histograms.\n",
        "\n",
        "_ = dataset_features.hist(bins=20, figsize=(18, 12), xlabelsize=10)"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABCQAAAK7CAYAAAAqdl6lAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\np5AAAAAARqeQAAAAAEankAAAAABG9/8UE1GcYM4hzQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 1296x864 with 16 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EhBU7F9XwYC6",
        "colab_type": "code",
        "outputId": "fa4f375d-b993-41e4-d33b-90fcc61c0897",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "# Double-check that we've done the right thing.\n",
        "print(\"Training examples summary:\")\n",
        "display.display(dataset_features.describe())\n",
        "print(\"Validation examples summary:\")\n",
        "display.display(dataset_validation_features.describe())\n",
        "\n",
        "print(\"Training targets summary:\")\n",
        "display.display(dataset_targets.describe())\n",
        "print(\"Validation targets summary:\")\n",
        "display.display(dataset_validation_targets.describe())\n",
        "\n",
        "\n",
        "print(\"Test examples summary:\")\n",
        "display.display(dataset_test_features.describe())\n",
        "print(\"Validation targets summary:\")\n",
        "display.display(dataset_test_targets.describe())\n"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training examples summary:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>area</th>\n",
              "      <th>azimuth</th>\n",
              "      <th>floor_height</th>\n",
              "      <th>absorptance</th>\n",
              "      <th>wall_u</th>\n",
              "      <th>wwr</th>\n",
              "      <th>shading</th>\n",
              "      <th>people</th>\n",
              "      <th>open_fac</th>\n",
              "      <th>roof</th>\n",
              "      <th>ground</th>\n",
              "      <th>v_ar</th>\n",
              "      <th>glass</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "      <td>100000.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.20</td>\n",
              "      <td>0.15</td>\n",
              "      <td>-0.25</td>\n",
              "      <td>0.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.40</td>\n",
              "      <td>0.36</td>\n",
              "      <td>0.56</td>\n",
              "      <td>0.58</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.62</td>\n",
              "      <td>-0.50</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.25</td>\n",
              "      <td>0.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.12</td>\n",
              "      <td>0.50</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.50</td>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "           area   azimuth  floor_height  ...    ground      v_ar     glass\n",
              "count 100000.00 100000.00     100000.00  ... 100000.00 100000.00 100000.00\n",
              "mean      -0.00     -0.00          0.00  ...      0.15     -0.25      0.00\n",
              "std        0.58      0.58          0.58  ...      0.36      0.56      0.58\n",
              "min       -1.00     -1.00         -1.00  ...      0.00     -1.00     -1.00\n",
              "25%       -0.50     -0.50         -0.50  ...      0.00     -0.62     -0.50\n",
              "50%       -0.00     -0.00         -0.00  ...      0.00     -0.25      0.00\n",
              "75%        0.50      0.50          0.50  ...      0.00      0.12      0.50\n",
              "max        1.00      1.00          1.00  ...      1.00      0.50      1.00\n",
              "\n",
              "[8 rows x 13 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Validation examples summary:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>area</th>\n",
              "      <th>azimuth</th>\n",
              "      <th>floor_height</th>\n",
              "      <th>absorptance</th>\n",
              "      <th>wall_u</th>\n",
              "      <th>wwr</th>\n",
              "      <th>shading</th>\n",
              "      <th>people</th>\n",
              "      <th>open_fac</th>\n",
              "      <th>roof</th>\n",
              "      <th>ground</th>\n",
              "      <th>v_ar</th>\n",
              "      <th>glass</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.20</td>\n",
              "      <td>0.15</td>\n",
              "      <td>-0.25</td>\n",
              "      <td>0.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.40</td>\n",
              "      <td>0.36</td>\n",
              "      <td>0.56</td>\n",
              "      <td>0.58</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.62</td>\n",
              "      <td>-0.50</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.25</td>\n",
              "      <td>0.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.12</td>\n",
              "      <td>0.50</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.50</td>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "          area  azimuth  floor_height  ...   ground     v_ar    glass\n",
              "count 20000.00 20000.00      20000.00  ... 20000.00 20000.00 20000.00\n",
              "mean     -0.00     0.00         -0.00  ...     0.15    -0.25     0.00\n",
              "std       0.58     0.58          0.58  ...     0.36     0.56     0.58\n",
              "min      -1.00    -1.00         -1.00  ...     0.00    -1.00    -1.00\n",
              "25%      -0.50    -0.50         -0.50  ...     0.00    -0.62    -0.50\n",
              "50%      -0.00    -0.00          0.00  ...     0.00    -0.25     0.00\n",
              "75%       0.50     0.50          0.50  ...     0.00     0.12     0.50\n",
              "max       1.00     1.00          1.00  ...     1.00     0.50     1.00\n",
              "\n",
              "[8 rows x 13 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Training targets summary:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>ehf</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>100000.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>0.77</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>0.18</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>0.15</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>0.64</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>0.79</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>0.92</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "            ehf\n",
              "count 100000.00\n",
              "mean       0.77\n",
              "std        0.18\n",
              "min        0.15\n",
              "25%        0.64\n",
              "50%        0.79\n",
              "75%        0.92\n",
              "max        1.00"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Validation targets summary:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>ehf</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>20000.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>0.77</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>0.18</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>0.16</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>0.64</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>0.79</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>0.92</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "           ehf\n",
              "count 20000.00\n",
              "mean      0.77\n",
              "std       0.18\n",
              "min       0.16\n",
              "25%       0.64\n",
              "50%       0.79\n",
              "75%       0.92\n",
              "max       1.00"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Test examples summary:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>area</th>\n",
              "      <th>azimuth</th>\n",
              "      <th>floor_height</th>\n",
              "      <th>absorptance</th>\n",
              "      <th>wall_u</th>\n",
              "      <th>wwr</th>\n",
              "      <th>shading</th>\n",
              "      <th>people</th>\n",
              "      <th>open_fac</th>\n",
              "      <th>roof</th>\n",
              "      <th>ground</th>\n",
              "      <th>v_ar</th>\n",
              "      <th>glass</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "      <td>20000.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.20</td>\n",
              "      <td>0.15</td>\n",
              "      <td>-0.25</td>\n",
              "      <td>-0.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.58</td>\n",
              "      <td>0.40</td>\n",
              "      <td>0.36</td>\n",
              "      <td>0.56</td>\n",
              "      <td>0.58</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-1.00</td>\n",
              "      <td>-1.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>-0.50</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.62</td>\n",
              "      <td>-0.50</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>-0.25</td>\n",
              "      <td>-0.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.12</td>\n",
              "      <td>0.50</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.50</td>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "          area  azimuth  floor_height  ...   ground     v_ar    glass\n",
              "count 20000.00 20000.00      20000.00  ... 20000.00 20000.00 20000.00\n",
              "mean     -0.00    -0.00         -0.00  ...     0.15    -0.25    -0.00\n",
              "std       0.58     0.58          0.58  ...     0.36     0.56     0.58\n",
              "min      -1.00    -1.00         -1.00  ...     0.00    -1.00    -1.00\n",
              "25%      -0.50    -0.50         -0.50  ...     0.00    -0.62    -0.50\n",
              "50%       0.00     0.00         -0.00  ...     0.00    -0.25    -0.00\n",
              "75%       0.50     0.50          0.50  ...     0.00     0.12     0.50\n",
              "max       1.00     1.00          1.00  ...     1.00     0.50     1.00\n",
              "\n",
              "[8 rows x 13 columns]"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Validation targets summary:\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>ehf</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>20000.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>0.76</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>0.19</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>0.00</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>0.63</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>0.78</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>0.91</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>1.00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "           ehf\n",
              "count 20000.00\n",
              "mean      0.76\n",
              "std       0.19\n",
              "min       0.00\n",
              "25%       0.63\n",
              "50%       0.78\n",
              "75%       0.91\n",
              "max       1.00"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "guj6j3wLVPQl",
        "colab_type": "text"
      },
      "source": [
        "## 5. Treinando a RNA\n",
        "### _5. Training ANN_"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "h3LbVih7USdR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def construct_feature_columns():\n",
        "  \"\"\"Construct the TensorFlow Feature Columns.\n",
        "\n",
        "  Returns:\n",
        "    A set of feature columns\n",
        "  \"\"\" \n",
        "  \n",
        "  area = tf.feature_column.numeric_column(\"area\")\n",
        "  azimuth = tf.feature_column.numeric_column(\"azimuth\")\n",
        "  floor_height = tf.feature_column.numeric_column(\"floor_height\")\n",
        "  absorptance = tf.feature_column.numeric_column(\"absorptance\")\n",
        "  shading = tf.feature_column.numeric_column(\"shading\")\n",
        "  wall_u = tf.feature_column.numeric_column(\"wall_u\")\n",
        "#   wall_ct = tf.feature_column.numeric_column(\"wall_ct\")\n",
        "  wwr = tf.feature_column.numeric_column(\"wwr\")\n",
        "  glass = tf.feature_column.numeric_column(\"glass\")\n",
        "  open_fac = tf.feature_column.numeric_column(\"open_fac\")\n",
        "  people = tf.feature_column.numeric_column(\"people\")\n",
        "  ground = tf.feature_column.numeric_column(\"ground\")\n",
        "  roof = tf.feature_column.numeric_column(\"roof\")\n",
        "  v_ar = tf.feature_column.numeric_column(\"v_ar\")\n",
        "  \n",
        "  terms = ('0_window', '3_wall', '1_wall', '3_window', '1_window')\n",
        "  room_type = tf.feature_column.indicator_column(tf.feature_column.categorical_column_with_vocabulary_list(key=\"room_type\", vocabulary_list=terms))\n",
        "\n",
        "  return set([area,\n",
        "              azimuth,\n",
        "              floor_height,\n",
        "              absorptance,\n",
        "              shading,\n",
        "              wall_u,\n",
        "#               wall_ct,\n",
        "              wwr,\n",
        "              glass,\n",
        "              open_fac,\n",
        "              people,\n",
        "              ground,\n",
        "              roof,\n",
        "              v_ar,\n",
        "              room_type\n",
        "             ])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LwssxZ4tVd4x",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def my_input_fn(features, targets, batch_size=1, shuffle=True, num_epochs=None):\n",
        "    \"\"\"Trains a neural network model.\n",
        "  \n",
        "    Args:\n",
        "      features: pandas DataFrame of features\n",
        "      targets: pandas DataFrame of targets\n",
        "      batch_size: Size of batches to be passed to the model\n",
        "      shuffle: True or False. Whether to shuffle the data.\n",
        "      num_epochs: Number of epochs for which data should be repeated. None = repeat indefinitely\n",
        "    Returns:\n",
        "      Tuple of (features, labels) for next data batch\n",
        "    \"\"\"\n",
        "    \n",
        "    # Convert pandas data into a dict of np arrays.\n",
        "    features = {key:np.array(value) for key,value in dict(features).items()}                                           \n",
        " \n",
        "    # Construct a dataset, and configure batching/repeating.\n",
        "    ds = Dataset.from_tensor_slices((features,targets)) # warning: 2GB limit\n",
        "    ds = ds.batch(batch_size).repeat(num_epochs)\n",
        "    \n",
        "    # Shuffle the data, if specified.\n",
        "    if shuffle:\n",
        "      ds = ds.shuffle(10000)\n",
        "    \n",
        "    # Return the next batch of data.\n",
        "    features, labels = ds.make_one_shot_iterator().get_next()\n",
        "    return features, labels"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Kuxpwb51V6g0",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def train_nn_regression_model(\n",
        "    my_optimizer,\n",
        "    steps,\n",
        "    batch_size,\n",
        "    hidden_units,\n",
        "    target,\n",
        "    training_examples,\n",
        "    training_targets,\n",
        "    validation_examples,\n",
        "    validation_targets,\n",
        "    periods = 10):\n",
        "  \"\"\"Trains a neural network regression model.\n",
        "  \n",
        "  In addition to training, this function also prints training progress information,\n",
        "  as well as a plot of the training and validation loss over time.\n",
        "  \n",
        "  Args:\n",
        "    my_optimizer: An instance of `tf.train.Optimizer`, the optimizer to use.\n",
        "    steps: A non-zero `int`, the total number of training steps. A training step\n",
        "      consists of a forward and backward pass using a single batch.\n",
        "    batch_size: A non-zero `int`, the batch size.\n",
        "    hidden_units: A `list` of int values, specifying the number of neurons in each layer.\n",
        "    training_examples: A `DataFrame` containing one or more columns from\n",
        "      the dataset to use as input features for training.\n",
        "    training_targets: A `DataFrame` containing exactly one column from\n",
        "      the dataset to use as target for training.\n",
        "    validation_examples: A `DataFrame` containing one or more columns from\n",
        "      the dataset to use as input features for validation.\n",
        "    validation_targets: A `DataFrame` containing exactly one column from\n",
        "      the dataset to use as target for validation.\n",
        "      \n",
        "  Returns:\n",
        "    A tuple `(estimator, training_losses, validation_losses)`:\n",
        "      estimator: the trained `DNNRegressor` object.\n",
        "      training_losses: a `list` containing the training loss values taken during training.\n",
        "      validation_losses: a `list` containing the validation loss values taken during training.\n",
        "  \"\"\"\n",
        "\n",
        "  steps_per_period = steps / periods\n",
        "  \n",
        "  # Create a DNNRegressor object.\n",
        "  my_optimizer = tf.contrib.estimator.clip_gradients_by_norm(my_optimizer, 5.0)\n",
        "  dnn_regressor = tf.estimator.DNNRegressor(\n",
        "      feature_columns=construct_feature_columns(),\n",
        "      hidden_units=hidden_units,\n",
        "      optimizer=my_optimizer,\n",
        "      model_dir='ann'\n",
        "  )\n",
        "  \n",
        "  # Create input functions.\n",
        "  training_input_fn = lambda: my_input_fn(training_examples, \n",
        "                                          training_targets[target], \n",
        "                                          batch_size=batch_size)\n",
        "  predict_training_input_fn = lambda: my_input_fn(training_examples, \n",
        "                                                  training_targets[target], \n",
        "                                                  num_epochs=1, \n",
        "                                                  shuffle=False)\n",
        "  predict_validation_input_fn = lambda: my_input_fn(validation_examples, \n",
        "                                                    validation_targets[target], \n",
        "                                                    num_epochs=1, \n",
        "                                                    shuffle=False)\n",
        "\n",
        "  # Train the model, but do so inside a loop so that we can periodically assess\n",
        "  # loss metrics.\n",
        "  print(\"Training model...\")\n",
        "  print(\"RMSE (on training data):\")\n",
        "  training_rmse = []\n",
        "  validation_rmse = []\n",
        "  for period in range (0, periods):\n",
        "    # Train the model, starting from the prior state.\n",
        "    dnn_regressor.train(\n",
        "        input_fn=training_input_fn,\n",
        "        steps=steps_per_period\n",
        "    )\n",
        "    # Take a break and compute predictions.\n",
        "    training_predictions = dnn_regressor.predict(input_fn=predict_training_input_fn)\n",
        "    training_predictions = np.array([item['predictions'][0] for item in training_predictions])\n",
        "    \n",
        "    validation_predictions = dnn_regressor.predict(input_fn=predict_validation_input_fn)\n",
        "    validation_predictions = np.array([item['predictions'][0] for item in validation_predictions])\n",
        "    \n",
        "    # Compute training and validation loss.\n",
        "    training_root_mean_squared_error = math.sqrt(\n",
        "        metrics.mean_squared_error(training_predictions, training_targets))\n",
        "    validation_root_mean_squared_error = math.sqrt(\n",
        "        metrics.mean_squared_error(validation_predictions, validation_targets))\n",
        "    # Occasionally print the current loss.\n",
        "    print(\"  period %02d : %0.4f\" % (period, training_root_mean_squared_error))\n",
        "    # Add the loss metrics from this period to our list.\n",
        "    training_rmse.append(training_root_mean_squared_error)\n",
        "    validation_rmse.append(validation_root_mean_squared_error)\n",
        "  print(\"Model training finished.\")\n",
        "\n",
        "  # Output a graph of loss metrics over periods.\n",
        "  plt.ylabel(\"RMSE\")\n",
        "  plt.xlabel(\"Periods\")\n",
        "  plt.title(\"Root Mean Squared Error vs. Periods\")\n",
        "  plt.tight_layout()\n",
        "  plt.plot(training_rmse, label=\"training\")\n",
        "  plt.plot(validation_rmse, label=\"validation\")\n",
        "  plt.legend()\n",
        "\n",
        "  print(\"Final RMSE (on training data):   %0.4f\" % training_root_mean_squared_error)\n",
        "  print(\"Final RMSE (on validation data): %0.4f\" % validation_root_mean_squared_error)\n",
        "\n",
        "  return dnn_regressor"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "3cT2HGeOWYMJ",
        "outputId": "48b7c808-f7fd-4d90-920c-98a3f79a4f78",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 558
        }
      },
      "source": [
        "EHF = train_nn_regression_model(\n",
        "    my_optimizer=tf.train.AdagradOptimizer(learning_rate=0.05), # AdagradOptimizer\n",
        "    steps=150000,\n",
        "    batch_size=50,\n",
        "    hidden_units=[50, 20], # duas camadas\n",
        "    target=target,\n",
        "    training_examples=dataset_features,\n",
        "    training_targets=dataset_targets,\n",
        "    validation_examples=dataset_validation_features,\n",
        "    validation_targets=dataset_validation_targets)\n"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training model...\n",
            "RMSE (on training data):\n",
            "  period 00 : 0.0180\n",
            "  period 01 : 0.0159\n",
            "  period 02 : 0.0149\n",
            "  period 03 : 0.0144\n",
            "  period 04 : 0.0140\n",
            "  period 05 : 0.0137\n",
            "  period 06 : 0.0136\n",
            "  period 07 : 0.0135\n",
            "  period 08 : 0.0133\n",
            "  period 09 : 0.0132\n",
            "Model training finished.\n",
            "Final RMSE (on training data):   0.0132\n",
            "Final RMSE (on validation data): 0.0136\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbAAAAEYCAYAAAA9AaOpAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAJMkdZIBJknqpJ/S2t8ekifR/QAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fqaBY2D6IlY7",
        "colab_type": "text"
      },
      "source": [
        "## 6. Avaliar desempenho\n",
        "### _6. Evaluate performance_"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "66c59sovWXzs",
        "outputId": "9ca499d1-47c7-4445-c8a0-07d1cd1696dd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 269
        }
      },
      "source": [
        "predict_validation_input_fn = lambda: my_input_fn(dataset_validation_features, \n",
        "                                                  dataset_validation_targets[target], \n",
        "                                                  num_epochs=1, \n",
        "                                                  shuffle=False)\n",
        "\n",
        "validation_predictions = EHF.predict(input_fn=predict_validation_input_fn)\n",
        "validation_predictions = np.array([item['predictions'][0] for item in validation_predictions])\n",
        "\n",
        "d = {'pred': validation_predictions, 'simulado': dataset_validation_targets[target]}\n",
        "plot_df = pd.DataFrame(data = d)\n",
        "plt.plot([0,max(plot_df['pred'].max(),plot_df['simulado'].max())],[0,max(plot_df['pred'].max(),plot_df['simulado'].max())])\n",
        "plt.scatter(plot_df['pred'], plot_df['simulado'], alpha = .03)\n",
        "\n",
        "filename_validation = 'ann_validation_'+datetime.today().strftime('%m-%d')+'.csv'\n",
        "plot_df.to_csv(filename_validation)\n"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nHpoAKAJwAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qaXrY610Ejuv",
        "colab_type": "code",
        "outputId": "026a1422-02cd-4504-9228-0e8c28cb6968",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 269
        }
      },
      "source": [
        "predict_test_input_fn = lambda: my_input_fn(dataset_test_features, \n",
        "                                                  dataset_test_targets[target], \n",
        "                                                  num_epochs=1, \n",
        "                                                  shuffle=False)\n",
        "\n",
        "test_predictions = EHF.predict(input_fn=predict_test_input_fn)\n",
        "test_predictions = np.array([item['predictions'][0] for item in test_predictions])\n",
        "\n",
        "d = {'pred': test_predictions, 'simulado': dataset_test_targets[target]}\n",
        "plot_df = pd.DataFrame(data = d)\n",
        "\n",
        "plt.plot([0,max(plot_df['pred'].max(),plot_df['simulado'].max())],[0,max(plot_df['pred'].max(),plot_df['simulado'].max())])\n",
        "plt.scatter(plot_df['pred'], plot_df['simulado'], alpha = .03)\n",
        "\n",
        "filename_test = 'ann_test_'+datetime.today().strftime('%m-%d')+'.csv'\n",
        "plot_df.to_csv(filename_test)"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzsvVusZml63/V7T+v0HfaxqqunD9Mz\ntgePZXAEk4zAvohEggyKZAkhRBAXBoIlRLgI3CSSAcMNxhIXlrCMTBSFIIsICRRNJCcgwQUSIsR2\nIivJ2Ehmxp6ePlXVPnyHdXqPXLzf3r27uqururqq9q6q9ZNKVXvX2t96v7X1/de7nsP/ESklJiYm\nJiZeLuR1L2BiYmJi4ukzifvExMTES8gk7hMTExMvIZO4T0xMTLyETOI+MTEx8RIyifvExMTES8gk\n7hMTExMvIZO4T0xMTLyETOI+MTEx8RKir+vEx8fH6Z133rmu009MTEy8kPzu7/7u/ZTSrUcdd23i\n/s477/A7v/M713X6iYmJiRcSIcQfP85xU1hmYmJi4iVkEveJiYmJl5BJ3CcmJiZeQiZxn5iYmHgJ\nmcR9YmJi4iVkEveJiYmJl5BJ3CcmJiZeQiZxn5iYmHiGDC7wK3/3D/jjk/a5nvfampgmJiYmXnb+\n3vdO+Cv/yz/i+/dbjucl/87PfO25nXsS94mJiYmnzGZw/PLf+QN+8//5AW8fNvzmX/g2P/2jx891\nDZO4T0xMTDxF/vff/4hf/Fv/mI/WA3/hZ77Gf/QvfYOmeP5SO4n7xMTExFPgZDvyn//t7/Kd33uf\nf+q1Bb/+b/1z/Im39q9tPZO4T0xMTHwJUkp85/fe55e+80/Yjp6/9Ge+wb//p3+EQl9vvcok7hMT\nExNPyPvnPb/4t/4x/8cf3OVPvLXPr/xr/wzfeG1x3csCJnGfmJiY+MLEmPjNv/8D/qu/8weEmPhP\n/txP8PP/wjsoKa57aZc8UtyFEH8N+HPA3ZTST37G/wvgV4F/BeiAn08p/YOnvdCJiYmJm8D37m35\ny//zP+Lv/9EpP/Ojx/yX/+o/zVuHzXUv61M8zs79rwP/DfA3HvL//zLwY7s/3wZ+fff3xMTExLUS\nYsL6QIwJHxPOR1rrONlaVu1I7wPWRu5tW/7hH33Eb/2/3Rd6/e9/7z6/8rf/AT/15iHffPOIg3lB\nbTRNqZiVhkIrrA+MNpBEojKautDPZYf/SHFPKf2fQoh3PueQnwP+RkopAX9PCLEvhHg9pfTBU1rj\nxMTEK4T1kXZwuBAwSjGrDIWWl0LtfCTEdHm8kgKjJSA4a0c+OO+4u+r5cNNycj5iU0AKQYwJKSX3\nVh3fv3fGh3fhwwTxS6z1vQjv/f6K77634sffO+NPvnPE8WLG3kyzqEqMVFRGUBmJlAIfPSHCvHr2\nAv80Yu5vAO9e+fqHu+9N4j4xMfGFsD5yth3RSlAajQ+R++uButDY4AHBundsBsuq9RRFopAarQXn\nG89HbctH5z3fu7vm/bMz2h4GD+sNJGAF+Gew7rtrmH10Tm0U3/56idaSlCwhRA7nJU1ZIwAfAlaA\n9YL6Gde+P9eEqhDiF4BfAHj77bef56knJia+ICEmeusfGVK4GvqQUlBo9cS70nZwaCXQKu/URx9Z\nDSOr3rKoDe+dbfloNeBTRCRoVwElEj5G3j/r+PBsy/vnHR+eOk572DyNC/EYeKBLMLjIxloO5hWD\nC2gliSRciFRGoZHYEIjx2Uvv0zjDe8BbV75+c/e9T5FS+g3gNwC+9a1vpc86ZmJi4voJMbEdHKOL\nGC1wIfHBeYcA9puSZVNchkp665FCoJQk7r5+krhyiInNYBFS4PqEDRHrAyEm7q8GhITRR5SCVeto\nh8gYHKUSvL8a+PB0zUfnHT+4B2fP5rI8lB4gQKEkg41oJbBOoIVEIAkpAgoEpADyJsTcH4PvAH9R\nCPE3yYnU1RRvn5h4sbE+EELCaEGIiVXnkFKgpGD0gbPtyMG8JMSIFOJSrKQUEKG3OfgxOI9IgrJQ\nl2GIz9rlX5zjrHOcdyMhQoqR9eh3N5GIEZL3zrcMPvLuScvQO+pSIgT8/vun3DuBswDDNV2zWQl1\nKSAltBCI3XsTgEAQU2J0kUorCq2e+XoepxTyfwT+NHAshPgh8J8BBiCl9N8Cv0Uug/xDcinkv/2s\nFjsxMfF8iDGRSCghWY8WrUAriY8RIUErQTs4Ci1RSn7i53rrOesspZbUhUYqQTcGrI8oKTHqk7v8\nQitONiPvnbese4v3MAbPqrMkGVFW0Q4O6zz/38mW1dbiQ2C0jvu9o92C1rC6RmH/egU/9uYRR7Oa\n1/dqlBLszwxKCgqpkAicz6GZg1l5Y6pl/vwj/j8B/8FTW9HExMS1I6W43G26ECmNIqaEQKCERCvJ\n6DxVoS534RdiPbiAkux2+ZHaKIwW9GOgKhKlMZfn2Q6Os23L3W2fQztCYvFsxpw03YwBA5y2HR+c\nD3y02uIizAvBvW3ifAsWqF3eWT5vXlPwz/7IjB+5fciPvbHHnUXDrWWJlpLSSGpjkBLE7pp+mXzE\nF2XqUJ2YmPgUSkpscIyDx4VETCHvQpXCKIkPEaNyeKG3HmIOt0COnRuZbwAXN4fKKHwKhCAYbMCF\nQO9C3tGnyOgjZ61F5qgG1gXurno655FacbIaefd8y2YFQoO1iY+2H+/U++d8fd5p4BtvLrm1V/HN\nr+zxtcN9vvH6klJLlk3xnFfz2UziPjEx8YmKl7T7elZqjJIERs42jsN5wazUpJTwIXEwL7Lg6xw2\nOe9HamOojCakj1+rsw4XA90Q6cRIXSrawbPuHWftiPWBwTvOW8vgAptdotT5yNlmQEhNH/IOPgLB\nQjdcXwjmp27Doio5nld89XDGW4dLvvH6klmpETfHfWAS94mJV52rFS9CCFbdiPOReWWoC82smnE0\nC3Sjx+8aiw7mn6yWSQkKLXM3pk+M3uNDRAhoSo11ic2YRV51ktPNwN1tSzdEUAmJ5LSzdGOg0JLB\nerrOsu0tSlrubS3na1iH6wm/XHAAHMwabs1L3rm14Jtf2eO1ZU21C1vV5uZI6s1ZycTExLVgfe7g\nhFzlElOiLBTWRyCXNZZGoaVgVplP/GxvPaOLSAkpCu5temyI7DclAjgfLAJBShEXPGet46zvOds4\nEImUEquV5WTrqIv8FNANjrGPJAJ9cKw2iWGE4ZqF3QBvHMLrBw1/8ut3uL2sOJxVHM5KtHq+8fTH\nYRL3iYlXnBgTSuXdshQCoxSJRBIJKQTWBwqtPrM2e7QBKXNnqU+RplRID+t+pDKaZWXYdJ7Trsd5\nCDFytnE45xk9bL3H+4AUiY/OBxSKuoayVrz34cjJKnHaw/YarsuDfP0Qvv2N1/ipt2/zo3cWLMuC\n23v1tfu2P4xJ3CcmXnEuKl0uRN4gaa1Hy1y/7nzutPyskEMSCedT3rmH7N3SGIF1DhciKgk+2HSc\ntwNaSgYbuLsd0Cqfo0iw6UY23chZPyIirD6EpoR1Cxt3M4T9R+bw5kHN12/v862vHbOszXMzAHtS\nJnGfmHjFuax4EXkXLwQUKodhrA9oKS+PebApqTKa7dhTCY0Skhgd2yEQQt7VRxJdHxhtZMARgoAY\nOd0MeCEQlaEbHPfGkc0WKgNdD2c764Bn3+rzaN6awe0Dw8Gs5rAuXwhhh0ncJyZeeZT82MSqHR1G\nKuaVRgAxXQh7wHqPlhIktINnO+TdfTcmVt0A5F18N3iEBKnzcauxR0mFQnF33RFiZDuODNZx3irW\n7YgfQSf4cAPtlbWF67ggVzgGvnpoeOP4gB9/Y0lTKWIkh7Cec936F2US94mJV4wHjb6UzO39ApiV\nu4RpSggpqI2mt552tKSUcCISE7TWIVJurN8Mjvtth0bhQiCEyNZ5bi8aGqM5bhrOh5ExRgqjcTES\ngqQyBRbYDGAt9PaTwn7dlMBXjuGbbx5zZ6/hVlMzrwu2o+VwXn0pH53nwSTuExOvEA8afTkfWXcj\ns1KjdbYEiClR7XbyvfXc3fTEmEsdt72lGyOzSjE4z+nGMqsVtTLc3XTcPe+RCqpCc9YOCCTrwdL2\nliQkSkCKicJAEoJ21dNt4ZzcxXkTEMBtAVUDx4uKkMAoQak1Rzs/HfjYR8f68Mzte5+Em7eiiYmJ\np8aDu/QQ0yeMvkJMaCXwMaLJAyWiT5xtR0JMhBQRKde+n7QDy6pAKYn1kXvrkc04cNIJNr1ldA6U\nYNuPrHvPH51saYykUYbNGFh3HQdNwelm4Ky1+OTZjImBbCFwE2iAAw2HezBvCg7nNbNacXs543Be\nMrrAvPy4HFRKQQhfZtzHs2MS94mJl5TPsuNtB8+8NleOidk7fSdQMSZGH+hdDjecrC3nbW4+Iglc\nTAzO08nEqh8ZfbYQGL1l00cKJVgPASEDxkjOe8cPNxv2FzXd6Ll7vmUzWHoLZy2MHtbXdYE+gzmw\nv4Rby4bjg4a50cyNQUpFkpFNDwez8vL4i5vmTWQS94mJl5SL5qSLUkfrs6fLqstNRhfxdh/iZczY\n+kBIid4FBhfYjoHCSJJLjCKw6S1SSlZtj4uR7RA43fb0PrHuB6JPLJuS3ibafgSjcSHx4XnHqu35\n4Cxybq+3GelhvCagWUJdSpaziq8eLtA7hey9w1qJ0QJ/5UZ407pSr3IzVzUx8RLzNCcXfR4Xdevx\nyg6+KjQn2xHrI02hSbvk6KIyFLtdeztGrI+74RiBFAXWRwqlsN4y2shm8GgjOOtGTrcjymiCh9N1\nhxIKROBkY0kxsrGOboi0A9y3NycEc5U94PAAlrVBS8nxokQpQaElKQpuL2r2ZxqJoB0D8zJgdK79\nv4nJVJjEfWLiufI0Jxc9+jyB3g9YnyiVoipyEnRWaBBw2o4YrZBCcNpaTtoRhcyt9EaSIpRGcra1\nhBhJaJyFKCOjD8QkuL2sUBK2rUc2BZvB8u7pGYXWdIPDRbARPljfrEqYq9TAXp3f67wqch5CKoIP\nOCnYqzRVIYgB5rWm0DIL+w1Mol7lZq9uYuIl42qoBJ5NxUUekefxMZKiYLCOLnlSC/PSsD8zO4/2\nbPRllORgVmB94L3znkWl8TERQsK5yP1tz7I2VEJw1rdENPNKs1qPbGJERLApMriID5F7q0RdOCoD\n52vY+psr7BLY13DnoKAyGhIcNiW3liWLuuCsHdivS2SSLGtDU2iUzDflm84k7hMTz5GLUMlVnnbF\nRZ47mkV7cAEfIz7k6UYX491czJORPumLImhHx2ADWkhs9GipuLNo2DrHH5+uGDzMSkFK8O66px8t\nm85hgyeRIEXmM9hsYNjt2q/LmvdRLIAELJcwrw2zUtNUFU2Rw2W1NpR7kjcOG76yP0MICDE3fd3U\nJOpVJnGfmHiOXCQ3r4rD0664iDFByvHyEBPLusgxdAu999m50eVY+EFTUhtFiImTzUgMcNYNGCPY\ndB6loClyrBkks1Lw7v0t769b2q5nsJ4+JAqp2PaeszUUBQwOhIMx3swY+yF55ulyAceLkkVTYrTg\nsNG8dTDn63eW3F7UGC3zzcsHKqMptcj2DM9hBuqXZRL3iYnnyNXJRZeGXY+ouPiiCVgpBYjcgKS1\nhARxN2BjOzp6rXjzsGEMgfNuhKbivLe8e9aSIpy1I7NaE0QgRsF6cBBgtJa7W8vpume9Gbm7tvQu\nN/146+l83qXrIe+Ib1Ioprryd6nhcAlLo9jbK7mznPHWrQWFkviYmNUFR/OK/aZAAJWW6N3s15tu\nOXCVSdwnJp4jFz4u1uc2fblr8X+YWDxJArbQCusTLgLh4sYAlVakBN0YWPUWIzXboecHJy02OGLK\nu3YbHMPao4ykKTSzQvHRdstq8Kzajrvbjh+eW1ab7N6YHLQx+8DcxBBMDcyAxQKUgNcPCu7sLzie\nl8Qo2G8My7qkUBIpBIdNQakvZsZmS4aU0qe87G86k7hPTDxnrhp1PYovmoC92OULErVRnG4tRkNl\nNO3gctOSTpxuRlyKBJe433WQFKWG87XFxjxBKY3w/tgijaRrHSfna/oEq9Yy9rlWfRizc+NNq1uX\n5KlJWsPxEm4tKpaLmlllCC5wNKv5ylHDojAIKXltUYNKiCiYVQolEj5ElnWeh/oixNgfZBL3iYkb\nzBdJwF7d5RdGsVcXrHuLknLXFZq4v8m+6klAbRT3h4HSGAqp6KxlTJ714EhxN1ijc6y6Duc9H90H\nUYB3sPHgntdF+IIsgaMZSAWvLQSzuuFrt5cMQaBkRBaKo72K/aYk+ESpBbNaoySQJPtNQV1I0q4i\n5iY3Kn0eL96KJyZuCM+qGcn6SDu47LAYc0KzLD5O4D0sAXt1l+98pB0dUuQB1UJIYoxooQgEJILe\nRQabUCqw8Y4gEsRE2zlaGwg20AfP1kacy2K57mD1pd/hs6WS2Rd+VgqOD5bEINmvCqKUtKNnplV2\nrhw8y9JQasW6c9xaVNzaqyi0vPSiF4Ib3aj0eUziPjHxBDyrZiTrI2fbEa0EpdGMPnCyHTmal5SF\nemgCNsTso44AgWA7OrQSNKXJNespURpBUyt6CyFFtoNFS4ELkc4GTrYdZ2225nXe8dF22CVjoXdw\nMt688MuDCKAqIUmQUkKE/caQFBzUJUbC4aJkVpRIAaONNKXhzl7N4azE7JwxCyOpi+KFFPULJnGf\nmHgCnkUzUi5HHHAxEJJi9JEQ82zSD9c9x7PycgLSVdGxPrLqLK212SogRnrrKLTmbD3SR0+MWfS1\nFCghOO8dJ6uBzjp6lxiDZ9UObNoRFyOLsqIdBkaXyxoN0D+NC/cMMWTjLySQYH9eURjNsjGoKCm1\nwiwqZkZzUBuqUmNDoNGa1/cbQoyPleR+UZjEfWLiCXicWPgXCdtcPAnY4HPrvvX4mFBCYLRgtJHS\nqE+9ro+J884RUx5t51OgHwNnneP+eo3WgkWVXQxHF3Ey2xKMg8PHSOs8NiZSDGhpkMJhtOa027Du\ncjmjJM8xvak9mYos6pVkVyEEe5XiYN4wK/PNcGsdom1553iP1/cblJT01ucnIfJN9EUpcXxcJnGf\nmHgCHtWM9EXDNhdPAuVO2LUSjD4wBlASlMhf5/iwpSk1CMG6s/TOUWkFMuJGONlaNsOIJ1Cqknbw\njMGz6QcWdUlpJGUl0Rb26oLRRu7ZSDd0tOPI6CIfnmbR9NxMUVfkdWlyVUwzBykheKgqOJjP8HhG\nC6GKvHE4Y68smZWaEBOj84w+sKg0x/OKlLjRU5WehEncJyaegEc1I33RsM3Fk8CsNJx1IyZJrMvD\nMpRQVKXifjtSG5lDLCJbBaxHm8feDXnHPziHkokxBozUeJeQIjE6Rwjww5OWt48XGFGgVeDd8y0y\nJtph5N52YDNEbIBturnCXpN36h4oFBQGZhWkBHdu1RSyYF5DCJr9pWFelnxlf8ayKVl3Fh8DhdYU\nRqDI1/ymT1V6Eh7rXQghfhb4VfIN86+mlH75gf9/G/jvgf3dMX85pfRbT3mtExM3hkc1I31RD5mL\nG4TRkuNZzf22Z/Q+n6vUbPrcZLTaWlDQ2kCWXoH1gburHk+uhvEx0g+ezjlqo2mMRghJWeRE6/vn\nHVE47p6MnG1HNtuWMxtZr3OJ4+ZZXrinwFLD8R452Qssa01ZFMxLg0hgjGIxKzmelRit2GsK9mcl\n80pTmRwe8yFSGfUJIb/JU5WehEeKuxBCAb8G/Fngh8BvCyG+k1L67pXDfhH4n1JKvy6E+Angt4B3\nnsF6JyZuDJ/XjPQ4HjJXY+eJbEplgKbSFINiUYFREhsi7RhIu8aag6qgtZZhjJdCNdqATYHBj0jE\nLvSQ8CHw4VmP33nNFFKysQPDMLLqcjjndIycnEOXbnY1zJK8a68KqEvF6/tzYoJFmQeClEYjpOAr\n84okE1JKhISjWYUQ+ZoYLTmaVUAipfx7ysM31I2eqvQkPM7O/U8Bf5hS+h6AEOJvAj8HXBX3RL72\nkH3v33+ai5yYeNH4vLDNRTy+HR1GKspCIYFAdm/snaMdPQlyiMUGCi2RMuGlQAqB8wmfAipm58ez\nsWe/rhh8oBsjQ3DcXXe8d29NWWiaUqOS5MQHtqMleI8NgZMtOJ/LHW+ysJfAosx2B7eXiv15w96s\nZF4VKAFaapYzRSU1Q/DMC0NIkVlZMq81RsHoAsuqwOhsg9ZbTwwJJfOg8NGFnUumfymSq48j7m8A\n7175+ofAtx845peA/00I8R+SbRz+zFNZ3cTEC8rDwjaQRcX6QKHUzuArUBd5aEY7eqSQ7M0MzicS\nCaUEYpdsDSohBRgl6EdJUqClZFFUnHcjg090nSOSCFHk4RNC8NGqo1SazeA46wKQP/zrLseqt9d4\nrT6PirzOPQNvHipeO5jRVCWF1pQmV8WUquC1/ZqEQBJprWBRl7iQOGgKutFRKIna1f0L8g231Ire\nBWRKDC5SG4XWz26AyvPmaWUO/jzw11NK/7UQ4p8H/gchxE+mlD4RwBJC/ALwCwBvv/32Uzr1xMTN\n5LPCNtvB0VvPeWdRUlCZPNnH+fxR8SEwqwwJjZaRdvSc95ZN71jWBXt1AeSfs0WkKSTblJBKUBjN\npus5H8ZcOuITrx8ueP+s5WzrQXpShPNttuF13FwLAUUOUR0aIMLREu4cznnjYM5ekxuQYoSy0Nxe\n1BwtSraDQ0nF4axCavBeUGhB1RQczUu0kCgp6EZPTBGjcjw+xHgZooFnM0DlOniclb8HvHXl6zd3\n37vKvwv8LEBK6f8WQlTAMXD36kEppd8AfgPgW9/61k1MxE+8gDyvmaSPe34lJSHGT60nxMR6cMQY\nMToPvOidx0dFZbIDoZAghcAoSWcDm8ETff7eeTcyusCiKmhKhUDQDp5Vb/E+YoQiipxk3WwHTtue\n88Fw2o2ECNZC1+fa9Zsq6pBLGyU5WVooMDW8fXvB24dzSqPYnxc7ES/Zrwta6zjdjjRVTkZXhWTd\nOZpaMC8NCbAhoo3Eek+hFPO6vBwanjtSP+nP/jIkV+WjD+G3gR8TQnxNCFEA/wbwnQeO+QHwLwII\nIb5Jfpq69zQXOjHxWVzEr1MCpeRlvXJ4TmPQHjy/D4mz7UgI6VPruQjR2J1nzOgiCfAxMNrI6CI+\nJDqbq2R8iKxbyxAcs9JwPK+RQrLqR862A0RBVSpA0DvLGC0yCbaDY9N7fEjcX285WTtON1nYT7m5\nwi6BrxjYn4GWuSLmx7+65Jtf3efWYo5LOzOw+Zy9ymBkHpyhpOBoXnJ70TAGj5GSO/sNjdF0Lk+k\nWpQGHwM+JHzMoi13+YsQ06fG5r0MydVH7txTSl4I8ReB/5X8tPTXUkr/RAjxXwC/k1L6DvAfA/+d\nEOIvkZOrP59SmnbmE8+c5zGT9EGu7tStj2glkerj5iWtBD5GNPIT63E+Xs411VogRd6dhxiZl3A0\nrxhdYHB5BN69Tc/GWppCUxtNJOFCYNUNVEahtCO2kiE4Sq25v7WcdSP94NnYDmKeCbrewtkzuRJP\nl5ocTbIO6hpuzQ3zQjNrSu4cVIw2cNjMUSphQ+J+O5BS9ls/mhf4mChNgVGSWaEgJeYqx9YB2N0c\nrup4ftISxJS+0ACVF4HHWv2uZv23Hvjef3rl398FfvrpLm1i4tE8j5mkV3mw8zRYT0gfi0SIWeyv\nnv9iPSGmnETV+XUQaTfiTrKoDCmBEFl8TruB03ZECglIXMgDN866gVU7cDCbIUPgg7MNUgsqLRns\nyOADSkdikviQJyXd1J36VUpgaWBvDsu6YL+piSmCFtyeV7lCKAiOZiXLukAKCCEhhMhdpykRYhbk\nWalZNtmH3YWPw2MICAH0lZkbF70FecDJ4w1QeVF4sW9NE688z2Mm6VUefFLQShJiwvmIKhRKSnyI\nnxCGj2PxAq0UNlyU3OUwznbwhAg+OkLMo/Aao7mzN+Nk27PqLb0VfLjqWHUW6zyr1lOXudrDdpHB\nBs6HjtONpdCCZVGwJXHWhhtt+KWAI8BoOFzA7eWC/b2CAslmGJFJUBaadrTMK4FUgsEFQFAUitEH\nWuuYC8OyNozes9q94fwUJUkpEUKkUAohIlrkzcDVHfoXGaDyovByvZuJV44nmUn6ZXjwSeHi/C4E\nKnLidHTpMhRwsZ5CKYYYSEQkAusjSghKpQhFwgaPEpLOekYXsDKXKKYoGJxl1SXO25FV75EyEUJg\nYwNDjJyve/pxJCZBjI7TFtoh0bWwjblN/6YhyY0x8xLevFWghaSpCuaF4Z2DGXdbx1IKjIIYI5UR\nfO1oHy0Egw0sG0NTak7bAR8TCPAxMrrEQSOJKd94u9EzKzW6yCWOOiSU5KXaoT+MSdwnXmi+6EzS\nL8uDTwoX9dIuZLtYrQQH8/IT9rGFyo/8RkliShgJPspczpdygtW5hE+OFHOooR0cZxuLwyOlIBEI\nISBkbnTSCja954f3zgkEZkXFxg70Q2LTQudh/UyuwJdDAa8pKKucON1vSg725gQXCHFX0y8lb+3V\nrPoRXWjeOpozKwwues6HgeNFjfeJIXgqadBK0PYeLQQHTUFpdvYChaFB40IeGyilYF69vGL+IJO4\nT7zwPM9H6s96UkCQZ3Puyh8h2wJArpQ52XaQoC70ZWx3sA6jJYuq4Kx1uBjQUnI2eFZDz2gTZ93I\nXmVwNtINgUgu7TvdDqy3jvvtyHr0KAHO95ytwXrYxJvpva6BWyXcXsDR/gKRPFVRUxcSrzXOBQ7n\nBUqCVII7Bw13FnP2FwWFlpy1keATIcCQPLUxDN7T9p6DWcmyNhS7wdZK5qcro/NN9EUbbv00mMR9\nYuIxuVol42KOqxstL3fmV+19t4MjRAgxd4MKKThtLVpDKTVVoemt593Tls1oaftAWShcCHgPPzjd\nQoqM1oEQOO8ZQ2K1Htj0I6MdGVwgBZAaztdw3+ZStfF6L9NnUgJHBuY1HM8bfuLOHqe9I4rA7XmN\nkpKTbY+PCa00bx3NuL1oLv3pfQhUWrPfVCiVcD4hisRhUzK47F8pRfaJSQjq4uO4+ote0vikTOI+\nMfEYPDh8+jKWrtVnlmMGmxi8R0vJunesBocECqUI0eZa6ySY1Yr9uuB0u+XeactMl5xsB9a947Ax\nJCFYd457m4HOZhdHRcTGRHSvUg95AAAgAElEQVSJkGDdw/sWwvVeos9Ek4W9kbBcwO15za3jBW/f\nXmA/XNE6wegie43mqClRRnK8KPmR20sgse5g2SgqLXcVMbknYFEbCi13QzY0TSHxMaKEpDS5E/XB\n/Mt1N7s9byZxn3hl+DIf7qsCflEd40KuXVdSfKrDMZHoXR5v53bt7SHBST9iDIw2cTgv8CEnDJWA\nwSU+Oj9HCDicl2gJpxvL909WrLcdnY+IANJI+iHiAmw2uTHpJgo7ZGHfMzBvoC7AaIEAzlvLvNYo\nGXExseocbx41LOoSH7OpWlMYjpdmVwYZ6cbArDQoGZFSsup8bnaaV3n+qZJUD8m/PKuZtzeZSdwn\nXgm+7If7okomv05AyRx/dz4wehh9RApQUu5cB/OO1Og8dEMh6LwDASkISpPLICFbEow24GJg9IlC\nC1SM3N86/uC9+4zBk5LCushgQftItwUb4OTZX7onYrb7W5GdHPebXGN+vD/n9WVJU2pG7ylLTRkD\nUmkqo2iMYr9uOJwVxATLqtw1eUmUzDYK65OewiQOZrk3YDsGbi3FZbnpZ+VfrqPZ7bp5Od/VxMQD\nfJkPd9h1onqb69G1yt2lcVfZIkgMzjMvDT5GNhsHImFdwvlASBGtJcIJGqPQWhAjfLhukUnROsuq\nG/noZIuLcLRXUpYF9z44Z2MdWiaczROSBgu+y97rNzFpuiCbkkmyuNSK7N7YNJRKctRojFLMqgIX\nE521RKm5vWxIJEJI1KWi0OrSglfK/FS0GfL0qUWt8CGxshYpBU2hsD4y/5yk6fNudrsJTOI+8Urw\npAOtIVe8XJQx9sERUv6+AASCwgikzN2Q550lkjBSEolYl+0Gts4RYqQLEdfF7ME+DEhUdnIcI04o\ngnC8fz6gteX+uqdQMFroHHRb6OPNKnE0D/xp6tyQlEQOw8xLgyQyrytu71eU2uRu25R4bd6wHSSt\n9ygJtTEsG4P1efC3FJJN7zA6u2ACdNZRF5qqUPig6EYHSaB3u/aH8byb3W4Ck7hPvBI86UBryI6M\nUuVdvws51u5DbpPvrUekbEA1+EBZSPoxsLWemGAMWeCL3cDrzeBQUrBfl6y6kbvrlro0LIpE33k2\nFoLPlTD3BkuRyBYCITcktddy9T4bAxxIMAUczGB04CIcLwv2ZhWlUAQCTW14+2hJaSRCJPbrks46\nOhzLquD1gznrwTGvFQezktF77q4GXtur0Co/5WwGm33ry4JE9rQvlKSeV2j1yKU+92a3m8DL+84m\nJq7wpAOt+9FTlZphZ/BllEIIiRL5xiAQl1UaIUU2fS7bczGyHS2rraf3HkHumFwUBfe6nqYoUFqz\nPy8otOasjdzfDCQZ6IZETJ5awqaHYczx6psk7IIcV5/P8iQnLeH4uGIYPZWWHC0qeucxUbOsCiQR\noyTH8zmVlrTjiJSSo3nFwbxgURfUhcKniAuJvdowK83HZadC4pKnLGRu4pIX4bH8O3hU3uR5N7vd\nBCZxn3gleNKB1j5F2sFjtEDvQjMpZIevECKFkYSYfcJ9SGz6kSRgO3h66/jDu2sWtWZWlKQU2AbY\nq0o2o+Vk23HWOqQGkyS6kpydjfTOUWhDrRVnKnDubs4IvJK8Y58BpoRlBaXR+BhQEY4WDSklnEtU\nxrBXKIwS7NWGrxwsd7NdPbf3ZswKQ4pQKsmyUkjE7oYrqU2OnyfIVg0KSq3yUA0Shc6eMQnBvDIY\n/Wj38pfRP+bzeHXe6cQrz+d9uBPs4rd8It6uhCSRchBZ7A7cDcS4+tPWRc63DhezO+Gmt6yGMSdU\no6KzubVICEka8jFaCkiRk7OBwQXW/cB57xECUnSc99B3N0fYa2ApYa+B2QyM0sQUUTI3ci0WFcRs\nx7CcaeZlkc3SpKIsNPPCUFcKIRLL2ly6OkohICWi2Bl98fFN14WI1vlGaoykMprzbiTGxLwyaClB\ncPn7mviYSdwnXnnCTpB94HJCUjs6Cp1H4Gkl6a2nHzydD/RDQMrEflOCACMVi6bgfttTJs33728I\nMdAOkb15wTBEqlLRD4GmTnSj42BWEWT+WaUSp6uO1SbS50IbBg+rkCtPrpuSLBQLA6/t55r6WVFy\n+3CJHUaGkFjUGh9AiEhZFMSQk5xfu7VkWRd5piuBUhm0FlgXc+nj7mabb2gXXj0CvwubhRQhQSLf\ncKUUHM5z+Mco+Uo0Iz0pk7hPvPJcmHpplWeZhpRjvErmuvUQcieoS5HTzYiUUEnNZnT0NlIaSTlI\nhJAM1qKlYFFWnKx6WhsvY8JSwbZ3rDpPUwRGH/BEBp9IPuITOAf3w81pSmp2fzywnMPebA7JkqJg\nHD2zpsLEAB6U8gQvUVIyqxWLpsTFhA8JYwRaCRJQSEmS0BR6d80DpdGXT1W99ZeNRyHkyRr7TXmZ\nDxFAU+lXKsTyJExXZ+JG8LRbwx9W1vhZ57jaoHTBRdK1LjT3u4ExeAYbUEpcCs+md/iY2A6JWaGx\nIRJTZFkZfEwcLEo+XA+oCMPoUQpGL3DB8966Y7VxyJi4u265v4VNl3fqN0XYYWcf0MBCwdduLxEC\ntGgoTb7x1QVIJzkdB2Yobi8rjhYFKYFR+TohI3t1Q6UlNgZmZYXWuSJp2HXxFrvfV6HVZW5ECjiY\nlYTIJ3IjL3uVy9NiukIT187Tbg2/+noIQT8GzjqLEoKm0Bj9yXMkYNM7OucplKTUihAjLu06LUXC\nSMXKOwqt0Cq/5nvnORoeY+L1/ZqUEpvBMfiA94mtdeAj2xDZ9h4ZYDs4Vv2AEQKExHnH/VVk08GG\nmzU1qSDvkucFHDaGQkiqUnK0rHBBMFqLjJKu69BB8MZrM/bqkkprmrIgEZmVmkJptBSUhSRZ6L2j\nJM9AJWUb3gd/J1d35Vfnz74KVS5Pi0ncJ66dx+0efdzd/cXrJbi0CogusrWBVW+ptKIymhgTJ9uB\nEBODy6V6PgbaIQ94WFQG6wNaSGQhWNYmd1WOgffOW07WA2UhMUay6Swn3UBvI0WhaKTa1bkHXAh0\nG4dNAZ/AGFhtLc4Ftg5WLZw/r4v9CGbkp4c5cPsACg2zwnC0bDhaFjRFgU2wXwu8KbEh8na9j9nN\nKk3kp53Decm2txwvSppCozRIIdmfZWtfUqSz6ROVLg/7vb9qVS5Pi+mKTVw7j9s9+ri7+4vXG3bC\nToJ2DNgQmVcKFyOb7bDrMM0VGifdSK0Uy9qgdcKFhNiFZspC0Y1ZcPqtpXeO7WARKqG1YK4LzoeR\n99cDOsFSFnw0jJxses6GEZUkDsdm67jfW1IC76AbYAiweg7X+HHQwP5uu74/g+Ws4mheUheSpi5o\nCsOiNgyj47Bu6H1AqMSyLFBakIIEIoicTL2z31AbzegCoBBmFy8vs+Xxuhs/VcL4slsCPE8mcZ+4\ndh6nNfyLeMNcvN7FsOrBe2KKGCnQUtK6XO3iffYFn9WSZWmyt0mChTHYEFi1lsLk1netcgihKiTt\nmEM35UIjEfTe8cFqYBgdtdbcP+/5wUnHYHuUMnTjwHk3shnBDdAG2JKrKm8KRwIOFlCZnKw8ns85\nmBe4FHFDwBo4nimOZxVDaVjUBWrITUWl3l0fEqUqWY0981Jze1kzOI+WglmlgcRm8MxKhe9tnhvr\ns+/OBS+7JcDzZBL3iWvncVrDH7W7vxqySUCI2ffFhch28IwuYRSMLpBSBEQeVG0kMUWqQtENubv0\nZBtpR09K8NpeuRu6kdBCIETizl4NCM5ay8ZazrYj563FBse6Dcwrxeg8J+1IiiMuwXab1/ThTcqW\nksscX5/BrIajRQke9hd13lGLRHKCW/sVi8rQlIoU4SuLGWtrL6ccaSU5aAogd5O+3Sx5fa8BkdjT\nBTElRhcQEqTMdsm5OkbQjp4Z+jLBOiVLnx7TVZy4dh6nNfzzdvefFbIJRBCwbnPh+KLW2fjLBgol\nCbt2+HlpGFzYmVlFztvA1lr2qpLlTONC4mQ70hQKXRhkUpx2I2dby0fbNrfF+0BTKNIQGJPnfIhs\nx4FtD22bHRwHbl6ydClgMc+WvLf3muyJXhiWRd6J+xi5szQYrRFJEL1gy4hSkr26YF5rvE3IIt9I\nlYKDWc1ryxolBZveUVf6SnWRhZhvxsVOzJPISehZmZOqTyNZ+qoN5XgYk7hP3AgelTT7vN39Z4Vs\nDJIYA7eWFc5HVqMlhcR+U2STLxeIu4HWVVK8vx7wDpQGLRRKQ200PiaEyM00J+2Qh2HHhJApe7L7\nhJCSZQPB54TrD+5tuLvxjD0MN0zYF0AlYV6CquC1uaKp5hzWijHAvhHcWTSMMbIdLMtZgUwCKSV1\nqenH7JMzrwsWpaJaSEqlL5uNtJIMzmOUQkoYbaDZCXyhJKVRxLizFJD5GjsfnpoIv4pDOR7GJO4T\nLwSft7u/Wqc+2JBnagqwLnB7r6GpskVsbz2D94QQOV5Ul7XqY/AcNBXO5/mlqXQED+s+OzhqJRld\nwIZEO3jubnvawZFSoiwVR6nk/mbg7qZl9BFrLcHDNuQPmL/ui0cOv5TAm8ewrAqCDyxnNUeLisKo\nXL8eIzHmEk2SZ9kYBhuojabSCiUkxgheO6iYFZKm1OzVJUrmxqPcJpBwIeZqpSQYgqeKubMUAc4n\nlBR5CLbIvzutJFKIpzI441UcyvEwXq13O/FC87DdvZQC5yOdzeGVbBObPcE3vWNRG5QUzCpDHTVC\ncPk6e03BB+fZbzFETe887Sqwcpb1AFpKOpcTss5FQsqv63YJVyUFtxcN99Y9q23PenRshkTfZxfH\nnEa8PgTZ6KsGSgXWAz43X1VGUJeKw6KkS4FSa5QApUAm8DF3kxZGUxa587TWNZr89KJEFmmjJD0C\nRERLubv2+d9S5BtHCHG3M0+78lL1iaevp1Ul8yoO5XgYk7hPvPAUWtEOI34n7BdqutcUjC4wWklT\n6Ycm7EQSJJHFatN7Ru8ZBseH2wHvIj56toNjMwaMlDSFpigEtg+c2JF/+P27fHDao0y25237PCUp\ncr3eMA1QCxhTnoy0t4B5KZFFSRFzs5WMkr29isMEnbeQFEoL5rqi0BLnPSSISXJYl9xeVjSlRKQ8\nKjCHYAqkhBCyq2O+2Qa0lBRafqpXYd3lCVVayUthf1pVMq/iUI6HMYn7xAuPkoJCyzxrNH68Uw4h\n4lNkM1p8DLvxbtlKtrce5yO9y97row+7hKsnxMgQAjLlLtN75wORSKkU97cdzkWaukKKwP3zLe+e\n5nBOXcBmzLXr1+m9Lsgf7Aog5eTp4R7sNZqIwEg4mtUs65JFo7g1r0hANShQ4ELChsB+UVDVJcvG\nUGqFEIIkIt2QmJU6T0dKOQxWKIUU8TKZqWUuF1Xqk6KqpGDZFJdx8ac9OONVHMrxMB7rHQshfhb4\nVfK827+aUvrlzzjmXwd+ifzZ+r2U0r/5FNc5MfG5GC0pfI7Jd85jZHYMtC7RFJJ5VQDspisJjMqT\nk07bkdEHQoSYAuveonUWv270GCFYNIbOBlrn6R2kFLm3XnOy8WxbQEFZwr02WwhcNxrYF7uh1CJP\nSTK6QCuV4+elZK+u+Mm3DkgklBLURrEsDSElPlq37FUFd5Y1hVaUWqKUZN2PeAdSCVzI4amyUGx7\nR2lgcJGYAkrkHfnDrHif5eCMV3Eox8N4pLgLIRTwa8CfBX4I/LYQ4jsppe9eOebHgL8C/HRK6UwI\ncftZLXhi4rNQMjs6ttZhlCSJxHnrmVf6sqKmKjTBJhARozT3NgOjjQiZ/dmtzXYFrs/VG+ed5ax3\n+JQIKbHpLJu+zceGREpkiwELZxbGa74GEjgg32jevqURKEY38pWDPaKSnK06AhBJ1IWgKjR7ZUFT\nZauFu5sOAbx1sKSpFEZJyl2DUWN0bgJTOcHsY2TTO5a1odoNqDYKUsoxdhsiezs/98/iWVoKTHYF\nmce5An8K+MOU0vcAhBB/E/g54LtXjvn3gF9LKZ0BpJTuPu2FTkx8HiHGPAAiRtyulro0gsoojJaE\nkEMGvffEGOltYNMF6koghWI7eMzOu30zjoSUiCFxum7pbSCRk7btAPdXCQEUBlbx+kX9gjmwv4SD\nmeCNwz3u7FecbAZEEphCY5IAkcMUy7okeuiEoy4Ne7XBxZLRRQ5nJaVWjC7gUySlvLsvtM7hq5jQ\nUjIEz+gkVaGYlfpTce4QI/mWM3EdPI64vwG8e+XrHwLffuCYbwAIIf4vcujml1JKf/eprHBi4jFw\nPhITSAW10pRG5WHW8eOu1d4GSCCEYNVZhuDwnSImDyS6FLNVgSeLmkg4F1h1PbOyphsGVl0ubXTA\nyt2MMseGHF9vKphXkkXdMCsNwxi5s98gRb5pSSnwuzLRNw/m3FqWCCkYfR588dXDOS5ErI/EmJhV\nikIVjD7Qjo69ushPQSHgfKBQikJLBHwqYfmqVqjcJJ7Ws4vm/2fvzYMsS8/yzt+3nu0uudTSawlt\njZAQYNQgYowDwsEie9hscMBgjccxA/I4gogJE2aZYNhEjEfGY2PC4cAIhWb+8WAbzCJArMMANhgh\nAUJSC1lLS91d1UtV5XqXs33L/PHdzM6qrlVdVV2VfX4RHV158+TNk3d57nfe732fB14LfCXwEPAH\nQog3xhgvMbsTQrwNeBvAmTNnbtGvHng5crndQN2lTpnSGBZdz7KNyUbWC5xK4RsQkTJ1eAQRsVLx\n9N6CxnkmJiMQWNSePNO0dWB72ZJnGSetZD5v2W8h9qkT5qUaSlKkLhxF2iitgHKUPtRwoGSkUDKF\nSkuJUSpFHEXB6UlG0xlOTgoe3KjofEQKKDPJvOm5b63ErzYgR4VZdclExrnFL1vqzjHKDRaFIMXj\ndd6jgsREBo+Yu4wbEfdzwMNHvn5oddtRzgLvizH2wKeFEB8nif37jx4UY3wn8E6ARx999G7yTRq4\nh/AhMm8cPqSVeONS2UREgRTJonbZ9zSdZ61Kdd9lm7zYLSCFZr92PLu7pPM9Igouzpds1x3eeSIR\nKaHrQKpIN1+JVEzWvC/ltKknlV9KBcFDNUor9kKl4aO1MuO+zREnRhYXUi7sQxtjJqWlaR1Pb8/Z\nHOWpFzz0tA7GeY4jriLuFHXviCtxPmhVXCszLsxqYh0JREKA3Oq0CRvi4BFzF3Ijj/77gdcKIV5J\nEvVvAy7vhPkl4L8D/i8hxAlSmebxW3miAwMH1J2jcw4tJVHArHGrDT3BODdYLamsQWap7Q7SJttB\nNqoLacWaZxLhDHvLjt1lSx8CSxdou54YJVvzGbO6ZX8BWzPYeQn/5jGpxTFXsDFOG7laQ2ElGok2\nmnGmmU5KTk5Sj/qa0XTOYG0y/Koyy5mTY6pMoaVIvjqrwa/1MieEmOx+y+ySKU9Wv7uyJg0s9SlV\n6uDbRktKNL0PCMHLukPlbuK64h5jdEKI7wJ+k3Q1+O4Y42NCiLcDH4gxvmf1va8RQnyUtLj4nhjj\n1u088YGXL8lGVhKBRevSNKRKXSAhpl5cLeUlPdZWp03TrXnLrHU8uTWn63sKa8iMYnOas7Vfc26r\nwSq4sJwzX7Rc3IfdJvnDvBRI0nRpDim7NUsTpEaAlqCF5MG1CbZQWKXIVepbr6xh3jpGlWFSWryP\nRAJnTqyzXPXyAxRWASmhSggOV9uX94rXvadcmXvFlQd+WJXG8lW6lRQczhEMvPTc0HVTjPG9wHsv\nu+2Hjvw7At+9+m9g4LYiogCZNlE77xFCMGt6ok+j8F0K+uHEJD/8GR8iO4uWJ7ZmNB3JZqBxbC16\neufIteHsTk3rekJQXNxrWC5hu0kDSS/s1r4z5MC0SCv2zKT/iixHS0HrOzSKopRsVJZMGU6s5Ty4\nXrHoHHlMfi5t7ygzRW4sWsFGlSFExMWAjMmvflJe2rZ4ea94puVhsIaS8nAS9WDTdKix330MRbGB\ne46DZKTGOeo+IAUEB1FG9pZ9GmjSl6Y4XZw1XJw1KKmw2hMbuLhbI7VESVg2C/ZnNU0f2A5z9uaw\n28Ly4D7u8N8oWLU2ZvDQpkWR2g+FNhgFuc1YLiVaS6ZFweY4Y73IeWCtIrPJRK13qUwyKSylVYQY\nWbaOEyONj6lmLwUURt1AtJ07FHCjJXXnCYRD47ahxn73MTwbA/cchdWrTVWAQNOmPuwqT97si64j\n+EgMsFYa6t7z7G5D7Txtn6ZQZ01LD8TWEZWk7hwRwfayYdmD654X9juNBDaBooT71wyn1it6l1wb\nlZUoqXC9Z21csFlqTk4KXnNqQmUNG5XFGknXB/JMkWsFEaKITDNL2weWnUNKKExyY5y17hIztStx\ndKxfrVbybZ8cGA/KOUON/e5iEPeBu4YbDVk4WFUq1SJ6WPQ9k9zS9J6l69FRII1ga9lgtGR33tKT\nhpfO7S5ou56ntpfsLRq0lIwyxd6yY1bXdAH2Zi+NjYAFJiRxFxZOrsHJtQqtJaNMk6kIyhBDxOdw\nqtTkqwnctnfJ0ExL7l8rybVi0bnDjFgtU0ll2TVYJVb7EQIlU+5p0z9/bXK15+FoqUYrQZllg6Df\nxQziPnBXcDRkASGo27TCrjLzgqCFdGxaNWqVAjPO7iywWjCylsxKrBY0fWBn3rDoHN7Bp8/vc3Zr\nzva8pe16BBIXI5+5OKPpHef3oW5TvumdHk4yJGGvRhAdjEt41akTPHyiYrbo2Bzl1K4n1xqtktVC\nCIFca4wWZFYBkWXv8CGS22Qf4EI4FOkQoO57QlSYIDE67VtYLZPHL9cPuxjG+u8dhmdq4K7gIGTh\nYCBJSbAi1YKBQ3HxIbIzb9lrOrxP3TGjzNB6j0QghOD8XjIDm+QWaz14ydaiYb/ucD7gfBJ9j2e2\nisLzpHCNO1mKKYFKwGScauwhgpFQlnByWpIrRQieE+OCjXFOqccoIdipW2rpmTXJT/7+SclGZel8\ncq7snE9llOCwWh363e8tW8YrA7Wm98ya/vBxHWXmkudhCLu49xmerYG7goOQhWYl7GkFD97Hw5Se\ng7pv7XqUEEQVWSw9UaSWwHnbrWruPUYaApGnLi6RIrKz7NjvPK3zRCFw0bPsYXs/ea633LnhJEHq\nghmJtFJPgReQa0FVFqyXGZ9zcoIgYqXgofUSrSWnJxXLrsdYQYhpMEuRzLxcDBRGp8DvEF9QRnE+\nkFtJjuK53QZjkidM8svxnJ7qS56HowxWAvcmg7gP3BUc9FP7ENArcTmwFuhcoHMOKVKPddMFGpei\n3HyMaCWYlpatxYIQoLSWQGR32XFxVtOFnsUyZaA+uzOj855lB8/s3tnaugI2Bfi48oEpYWNagAvM\n6paT45KHTkwxUjIqDVZBrjRSSYRIfvU+RnKdDNLyKscoidYgkUwKgxbycNV9SRml6REitYCujw11\n53EuIIDNScaBC/4QdnF8GMR94K4gDRn1tH2gdg7F8z3VkHxSdpuOGCAE2K9blFDECK3vkSsbXo1C\na0HfR7YXDX0fqH3k/LzhifM77NWpbXLR3Flh18DDE4irVKSNkUAiqYxCWoPVitEoZ2OcM8o1a0VO\niAFBRCA5NS7wIYIWWJkMu3yAKAJGSqo8+b0k98YXduVLKdBR0ruINZKsVKt6vGCUmzSdyhB2cZwY\nnrGBl4TLOzKSiAuMFtQdCBUIXqBWXuu9j9Rt8n0xUmGVYbeu8UFilaALnkwlS9q683R9ZHfWstW0\nXNhesLOomXfg+5SWtHu9E7yFWGANODWSaKGIRE5MS+ZtoHeR0gSmhYYgiNGR6QIhA6VSFNowGWnu\nmyRx3296fAyoCKfXc5reUfcOjaTM1As2nw/PYVWDL6zC+UgXPWK12heAuGy1P4Rd3PsM4j5wx7lS\nR8b+siM3inFhyU1g0fbsdy0+atbKjIuLmtxIXIj4GCisYrZU9LFHy4zWB/oQWXSOeeup25aPPb2V\nwjdI32tbuNjeWSsBDZzUsLYOm5MxmdEYIRASxpWk6T2lUgipWK8UmTZMS0uu07CQCx4ZLX71dysh\nIEq0WcULKs04t4zy5zdEr9RKerREs2h7MmnIrFpt5F66Mh+6Yo4HwzM4cMe5UkdGJOJCIDrYqzsE\nkdJqIpFF2yNIBlXRBbou4EJg4VpOVgVVYdl+rmbetPSd58KyYdl4ntqpWdRQZFB3d8YjRpI2SzUr\nW14N1sKD05JpkZEbveqrb1l2kbHVnF4v8A7GhWZc5IwySWkNSkHfp5U+RLLVQNJ+3WOFIvh4uFoH\nrtnCCEm0R7k5XJmHEBHDyvzYMoj7wB3nSh0ZRila53DeIQVopQi9p+4CGI9AsDvvqb1jkluECBTa\nsN84np3VPLW7oO0deBBesqwXdB7aFpYrf5jbLew5q171Kgn6RmnQEtaKnI1JRZYJNipL20Wq3KB9\nYHOUsTEqiAG0kjxy3wQlBWuFQUpB3Xt8jAjB4fDQ5jhnlOnD1TpwSeA0XLuFcViZvzwYnuGBO86V\nOjKUFPQuEghYqVh2aXMVYHveU/cOIVKmZ+89u/Pk6Hh+VvPJZ3c5t1OjRaSLEt91PDdrEUAf74xV\n7wS4fw0yK5nmGhEF03FBYTWvOT1lXvf44Dk1qTi7PSMAD62XlLklV4plcEggxsiyC+Qroy4BTKzF\nx5QqZZSisprL19lDC+PA5Qzi/jLmRsf9X8x99y5ll6qV4ZTV6oodGUJAlWn26p4n9xYIAURQSrK1\n7KiMxjvY73oWbQ8B+hh5bm/JXuuYLWv26p6ud/gAF2ew06ce9ttNTirBzBsY28Aoz1DasFkZJqOc\nSGBaGcCilaTMDOPSsFEW1F0PQlFqjRKw3/SsFRYk7Cz6ZKubRUa5pspsamfsPGWmLnn+OheGNKSB\nSxjE/WXK9cbMb8V9xwidjwhS2LQUYtWxoSmsZt70zOc9Dk9l7MqfvU8uhT6w33TM2+Td3sbITrOg\nUBajFdvzhmd2F1xYNpy7sMB1HuccWzO40N4ZUYf0BsqAUQ5lAaMqx1rFicoyLi1agBYpyQgRyYyk\ntIppWVBayaSwaA3epfd4Lp0AACAASURBVIGtaW7RWjJvkk+9iMleN8ZI03uskgiRkpCOPn86MqQh\nDVzC8My/TLmdY+YH9935sJo2lTgfmLc9ejUKn1tN03lyq9DK4Hzg/F6DD5Ei02zvzAFB6wLbbcO8\n6dmoMhQOKxUX9pec217wzPYMD3S+Z3uZaux3StgPgqkRyT9+nCnu36jIrCFGxalpiRGSpevx3rNW\nFZRGI8epi3+9shQ29eovm4DWkWlpkFLSa4kQsN+4g/kiYgy4kCx8l11Pqczh8zekIQ1cziDuL1Nu\nZ422d4EQYb9pybRGrYKUQ4wU1tI5z/n9GhcCnUvGVoVV9CEQAiyaVFrxIaJlKkNEH7gwa2hax96y\n49mdGXXv2W16MgXn92GvSaZft4uKZCgmSL3rmYG1CjYriwuRwhgkklNlwelJwXRkuThv2ahy1vKM\ncakprEo2u03PJDcgoNCa0iRv9Kb3VLlcXfEIrJAYI1BSkJv0IQgpsOTyksuQhjRwlEHcX6bcrjFz\nHyKtSwEamdb4GFi0HqNSvb33gc5Fnt2rcS6yPrIYI1k0nu15R2YURgla5zm7O8O5yIX9lt55tpaO\n0LXsLJZIlSxtt7ZSF0zN7V2xG9IqXQDapG6Y3MA0l4e5o2uVpbSWvJBMq4z7pgWlNeiVxW6mk/Wu\nFpITlcJogfOpZ38kDYvW0YWeKoJA0PWBzEiUTFc+QSfBDzGSWTXYBAxck0HcjzlX2zS9XWPmnfMU\nRtE6j5IC50TqX0dgDezXjhADjXP0PlA5jVKKNHvqeXavITeanUXH9rxnUTtCjFzcW7LoerZmLfMO\noncsati+RY/TtdDAlGTwtTFNWaalkWhpicrhnUNJy7QsGJeakbHJ79ymHvSm99Stw+tIppJ1QLrd\nkSlNbvXh1YsUOa1PhmBSK6pMr/x1PF0fqHJ9SV/7YBMwcDWGV8Ix5ka8uW/1mHkIEa2TeVXnPE4k\nL5jOp9qxECBkErkYYdknL8ZF39H5iAyC3bplr2koraJpOmado3We3WXLxe1UGtnjzkTfGVJtXUqw\nGZyelmSZxaZ5ImaLlvGa5YFpxam1go2RobSWIkshGkZLKqsJRep0yYwkXwmw1RolOXz8R7lhlEPd\nqUP747bzuOCZFPYFm92DTcDAtRjE/RhzvU3Tqw2zXG21fyOtk0fLPUYrXIDNUUbdB/bqJrU3SomW\nirVC00fP9rIhN4oYIkLBNLPEGDm7M+fZvYZl29F0jrqDqKFxt1fYJbCuIHpW8XOQFzA1cHqz4nSR\nIZRkv+1ZKwwnxgWn1kpOjwo2JhmL1jHODFKCiDA6sto++viN8iuL8YFoxxApMsVU2yseNwwjDVyL\n4ZVxjPlsNk2vttq3Wh1+WFyrdfJouad3ydUwIDBKIFEgIzGmacy6S6v2tvMsmp5ndpb4mPzZt2c9\nZ3f22d5fsGgibYCdPdjn9vmuV6SVulWgJeRl2jQ9OTKsjUZsjC33jctUQoqeRyYTrJX0DhCB3nt6\n7xmZlGVarQIw2s7T9I7c6BtqNR1Ee+BWMLyCjjGfzabp1Vb7i6YnM+q6rZNHyz111xMiLLpkKWC1\nJABCRKSMNL1gtuy4OFuCkAgJu/sdz+w0LNqavabl/F5kuYpHmnH7hH0CjAwEAZUFoWC9glFWcubk\nhFecGGG0oPWBqbFMCoUUkkXriCIythalJAqJNcnlcr/u0DJtnBIFdZe6gK62Yh8YuJUM4n6M+Ww2\nTa+22u+9p8jSz/kQV5OnASIvKM8cbNhKIemDQ8vULdL2ga6PCBGJRKQAT+TEuOL8/pLZsmXWOLbq\nBdt7HSGkvvUdbn0ZRpE6XyQrYZ9C30Imk9HYyAg2JyO+4MwJ7luriCFglGJaZoQY0QqUUGTaM287\nxMrpscw0pUmbpUpIjBaHqVIaiQ+ezg0r84Hbz/AKO8Z8NpumV1vtG6UOk5HqI1F4UVy5PNM5j5SA\nSx8MnYMoUmqSD8nZsOtTrmcIsOx6Ogfn5y3zRcfuPAnv3N96YR8D0wK8BythVAES8jJtdooAD56a\ncGZzxMMbEyalZpRZCqOYVoa9RU+IkTyT1K2ncZqtWYfVgkwrqlwzqzuyPG2MHn1svY+HwRgDA7eT\nQdyPOTdbv73aar/KzWE7npJAXPmAH9koPPp7DgRMiFV9vQ+I1YDShVnNwnWIKNlZNDy7X+N8ZGfe\nce7CPh8/4vR1q4W9AM6cSIlInYNxJsitTilHAmKITMcFD68V3FeVPLxRgQiUNg0gKSHZHGXsr2yI\npZQUFs5sVigh0AokgnFukSL1pB8IfFj9jqEXfeBOMIj7wCVca7WfJigbZBCHtx8IlffhcDO27Tzz\nrmdWO4xO96eVZGu/46ndGXXr0EryxMU9LuzVPL27ZNk2+Ahdf/s6YQ5CqfsAYyuoTGRjUtF26Sqj\n0Bk2E5yelpyYjnBErFFsVDkSQZnrw43lCLiQpko7l9KgtBJkKoVUH3xItn3AaCCm463WV4zBGxi4\n1QziPvACrrbaV1JQZSblgMrnV6NN73A+Mm8dIYDVqW1yd5VzWuWBZRs4tzvj4qxmtnR0ITCrO87v\nNiyWDfM+BWpcvA2m65Yk7OMM7ltP4mqkYJpnTCaWuOrDX69yTo4L1sqcKtNkRrKzaMl0ChM5iANM\nzpaB2EOeKbxJH2pEgTXysESlpEHJ9GEXRTw0TBs2UwfuBDck7kKItwA/SdqHeleM8R1XOe6bgZ8H\nviTG+IFbdpYDLylH7Xs7F+h9Gos3KwuApnP0LtB4fziBGSOsVZa69ewuep7ZnfPUzhIRIxKYzVv2\nm54uOGY9nJ/BIh56ZL1ocpJbo1VQGogC7p9KHjgxYVZ3WG3YmGjGRcbnnJzw+HNzxqXk1KSkD55l\n4xnlht15y7TUlNYwazv2655T04JR/rxwIyLrVfYC4T5IPhoNXi8DLwHXFXchhAL+NfDVwFng/UKI\n98QYP3rZcWPgfwHedztOdOCl4aDUQoSm9/Q+0PqAC5HGdUgBSkikEggvmHcdsz3HyUlOphRbbcui\nczyz3zBvW9o+cGG34fxsTgyRWQtnZ88bct0KRsBGcWDDK9DCEKOjynJQgklpOTHOeGhjhNWKzVGO\nAJ6b1WzPaqZVxsYkS334MtL1kUkhMCqVZC7OGh5YLwfhHriruZGV+5cCn4wxPg4ghPh3wDcCH73s\nuB8D/inwPbf0DAfuGFeaQD3oe5+3Pft1TxSp28MRaHpPZjRVKam7SOs8zkfqLrC/7InAvHEs+o6L\n+0vObc25uFgyr1OL416X4u8gXRK+2Fr7QVtjYWAySuWhSZ5R5TlaQFlmTIzGxcg0T6t2qwTzpmOt\nyNCrqdppkRFiYHvZMy1zpBSHm6Kpf71lZyGpMnNLA04GBm4lNyLuDwJPHfn6LPDmowcIIb4YeDjG\n+GtCiEHc70GuNpkaQkQrye6yQ6kUKAHQh9SrvrNsmRSp26TtI3Xn2W9antrdJ/SRRdcBku15xzM7\nS+oO5i3srlocV4FLL1rYK1IZpsjAWJhmGaVRVHnOdGToY2CUKaaFYWNUEmLABUdlMk5OcyprqPsM\n71MnTUDwqhNTtE4bpfUqKONgWlciiJFbFnAyMHCredEbqkIICfwL4O/fwLFvA94GcObMmRf7qwdu\nIVebTO1DEkGpwIWIOBjIESCQtH2g7h2dS5YGe8uW3bqFAMvOszV3NF3HfFnTOdhpYd8nUb8Vq3UB\nbABKpySk9ZHk5KhiUliigElmODHJUo29zFj2DhcDo8xwalywMbZMMkPjHNOyorIquVhG8DHy9M6S\nIAJKKvbrDiUllU0hGbcy4GRg4FZzI6/Ic8DDR75+aHXbAWPg84HfE+nS9T7gPUKIb7h8UzXG+E7g\nnQCPPvroMMlxB7jRnNSrTaYqKai7gFWK8/MWpVMfd5lpghBsVBl167kwa9mpG+Z9x2aV03aOc7sL\nms4xbx3n9lueWqTAakkS9hdjJaBIXTAnMphUICRslAVr44IH1gqsUUBEITBSM84NPgaMkWyWGQ+u\nVSidjL1mbU+mFLmR+AjOB4xOto+bo5z9Om2k1q3n9DQHkR7HsHpsO5c+ooYSzcDdxI2I+/uB1woh\nXkkS9W8Dvv3gmzHGPeDEwddCiN8D/vHQLfPSczM5qQcDSxEOrQUEqbUvt4p64VAago8EEdlddIxy\nQ2E0s6bnwrxm1nTM5h07oqft+1Rvbzsee2bJ2VkS9YMX3Gcr7JZUfpkUqba+VgkmZcU4U4yqjJNV\nTpVr1ssspUEtu1T3kUm0tZIoIQkExjpb2QNIqlwjgN4HINkVG6WYFAajBL0PNNajZZrgOvjQhGTd\nO5RoBu42rivuMUYnhPgu4DdJC6Z3xxgfE0K8HfhAjPE9t/skBz47biYn1WrFvOlXQzdpA9GFgA9p\nNV9aQ6YV52dpmhQifR/4zGwGMvmw151ja9YQtWRnVnN+v+PD5xvqPpKTSjAvZrVuSBOmlYYHp1Bl\nGZtrJY/cv5YEXMDpccWJcU7rPPPG0ReBUmlcDEzznHFlmK56I5UCqxSZfX7FbZCH/5YiibySAik1\n48LSrZwuF22PUZKISPF2Q4lm4C7jhl6FMcb3Au+97LYfusqxX/niT2vgVnAzlr8HE6hapRF8KQX5\nygDr4qxBqVSe6UOktAqtJE9vL3lmf0HrHPOlo+4dc+fousgHzy04u+8wEk5l0LWw/Cz/DkNarStg\nbQSnxoppVfDgiYpXbkyoSsU0L4l4MqmpMsN6mTHPetZ6gwsgRGCtLFirNFYqSquxWl7x8Yk+kq+m\ndDvnsTqFbhw8Rr0L7DcduckPb7/WYzsw8FIwLDGOMTdr+SuAcuVBHkJkr+7YW/bs1j1GC7ZnPYWV\nyVcmpIzTWdNzcbehJ9J2kU9fWPDY+ZbOw8kSNjQ8t59W7BlpgX0zq/eMtFpXwHgMkxI2q4L7TxY8\nuDbmtfetITW0XeDkpMIqSYgpv3VSWnrvqXtH30e0BisVo9xgtbzi43Hw9dEp3aMTuUoKhJasFRn2\nYMV+A4/twMCdZhD3Y8jRidLWBQqj0Fpe1/L36IdB3TkWjaMLLnnKdBEXPb2H3d0lLoQk/MuWnUWP\nE5E//swe5+ceBbxyAsHDzjwNKEH6f8bNifsJC+MKskyxnhkmRc5DJysenI5QShFFJJOarFSM8+Tc\nuOx6lEjReEZp2i6wVkkisOh6nI+cmhZYLa9riXw9I7Uhw3TgbmV4JR4zjm6iWqMQQtD0HhsjRssX\nWP4e7aZJQcwRYmR70dI4T9dHMq3IDXS95+z2jHnb0feR7UXD9rzjqb2Wjzxb4wKMFRQBlvMUfFGH\nNKhkSJuhS57vbYck9u2R8y9I4l+x8oJZk1RlTvCeqiw5vZGxPsqoCk2VaUqjCDGyllvMyne+sqm1\nsXUeokBrwbL1q551SZGl/YWNUXZdS+TrGakNGaYDdyuDuB8zLt9EPagJpyzQS5/uy7tpehdoeodR\nyaddC0kdPT4ElJQ4H+i8JwrBrOt4/Pw+H36uZbeJ5Bo2VXJ1REOIoMRhowoKKGQKy+5XtxmgkuAD\n1KtjMp0CM0YVFBqK3HBqUlLmllEOpycjTk0KeufRQuFipPOeeZP+zhhgOs3SIFJwBCIhQiDSu0iR\nWzKt6Ff19AMzr2txLSO1YfN04G5leGUeM25mE7XuHHXr6fxqqiiusk6FwAjFdrtkZ+boQo8SkmXX\nkxvN/qLhfY/v8tHnGmJMvea5hGWd7sY50AaCg4kB20MgreIPSjOVgqpMxl5WgxPpeCVgLTcEFaky\nyytOrLE5tkzzjLUqY1JkGAWLxifB9oHSGJa9J3eKQit8iLgQkULSOkfjHAJB0zt8SFcw2apMNTBw\nXBnE/Zhxo5uonQuc32vooyfTCqMks67HSMls2TNve2Ztx6L17Cwbms6hpGRn4fiFv3iGraVjYgUj\nFZnX6T7nJOEWgHFpYCn2kCnIc2ga0AoyCyfXMkZFxqKuaV3kZJnaDJVUWCspM8MD44IzpyZIGTmz\nMebB9QolBTvLjlERWbYpwi8zqftFC0Fm0hVGaRW9j7gYaOYRawKTwtI7z96y4/S0GDY/B441g7gf\nM24kN9WHyN6yI4qAlRIhktiLAI9f3OeZvQW9C8zrjt5HpJL44Pgvj+/xwXMLpIBXrylU9MyX6UV0\n1Ia9J3WYHNxeeDAaVAF5lsKnRfTkWmNHJfvLjlxrlIL1sWWzKBmXglGeMS2T86JVinnbMcosLniM\nkuQmhV4EH8izNIU6LuxhDbx1PblW3LeWsWw9zoeVF7ugc2FwdBw41gzifsy42gYgcGgE1rlAiCHV\nnn2g7gLb84Ynt2Y8eXHJvGvxUbBY9kQivY/88RNztmvHei54eCzog2dvDnX/fNj0aPV/e/B/k75f\nrcG4kHR9QCswApRRdF1HpjUPbZQUhcVIKLTGWhgXGfdPKnyIPLxRIYBZ7dhvWkpjSFOinmpVMw8h\nYFZTuAc2CwKXBN5olBC0zqOVQAlFjJGmc9e0ZBgYuJcZxP0YcbmPTL4ahb9849R1KTkpBVMHPvns\nPhcWCy7sdpzf22d73pLbjMxqPvz0gs/sdGgJj2xqhHPsNBHfpdV48CmLVJJW5ZBq6bkEpECLSJll\ntH1glGs0kkgk04ZJaehC5PS45MS0wBPJpSLKgJEKrQWnqpxcp46YE+OMECO9C2ilmBYWFyI+JMsA\nLeXhVYqSybe92fO0fXJyzI3FBYgxBXVfz5JhYOBeZhD3Y0DnAvvLbjU1mfq9xRGvkxd00CiFC5EL\ns5YntvZ4ZndJ6z3ztiMIRe9ha6/h07uORRcYGzgzAakctQMdQNlkuiUV4JLPy7RK+aRWQJlrCmOI\nURJVEu1JaRBRoGIy5jo9qbBW8qrNKUKlktIkTxOmANYojEqdPplKq2urFbOmo7QGHyKhd4QgqDKN\nUpeuwq2WnJoW7C87IhElJSEGEJCvrmYG24CB48rwar7H6VxgZ97iQiAziqb37NVzTlQFuVWHw0wh\nxiPTl5K6C7jg2asdS+fZWXTMm5667fn0bs8z80Cm4IEqlV0ARACtU+dLdLBsQUXIDGR5GhrayCVS\nSgqbs1YapJIsFx37bc3eIjLNDdU4Y620TCvN/dMx66McqwQxRjaqNPlJFEgJpdVIIQgxHgZLV5lJ\nfi8CMmOvWVaxWrI+yg6vaFyIlCvL3gMG24CB48gg7vc4i6ZHK0HnI85HlBKYKLm4qKl6g1ESKdLq\nNwJ9n0KsdxYtzsOy9QSXBO/jzy74s7Nz+gAbOZwapV70g85Kaw1ZiMyjY+mhzFOiUpnBJJcYGxjl\nOZW1eOcIglUXTGRNQ+0CaxPDw+tjHphW7C9bRpkmU5LcSgiQG4WWkjJLL82m8xj9fI9+iPGmSigv\nKFVdYYJ0sA0YOI4M4n6P0/sUdRciICIxCnxIXitEx17TYZUCAWtl6jqpe8f2sllttKaa+4fP15zb\n68kVPDCCkU2rcmMFVWZwTmCUZBk7YkibpWMNSwOlFUyK1FpolKDMDUsXyIUgRHhgvcC7nMZ7tJJY\nLdhrWzYmGafGJRsjS4ygFIclGUFaURdW40M4FOCbmQK9kuWxDxFP+tAabAMGjjPDK/oe4WqhG0Yp\nnA9ICd4LOp+yTeedQ2qJjBJkoO9JK/WmQylJ1wX+8txzvO+Jff7k0/u4APeXsFFCsapzG2OJwlPZ\njGysWdaeeduQGYFV0AETbRkXilFpWdYdRmpKIzlRjbFGs986To4LtIQQBNvzBiEkKio+Z3NMkScj\nLyG4JLP1QHgPauGfzWbnlSyPDWnTVQgG24CBY80g7vcA1wrdqHLDzrxFIsi0YNk69uqOtTK1FsZM\nQJQIk2ruUcDTWws+cX6Xn/+z5zi71zEygvurSGZSG2MTIyrCurUr24GekTQYDdOypDCa1gecC6yN\ncqal4cJug1KKcWlYH2dsVDl154hEikwxzSwX5ktec98a09KwUVk2xxbvI84HJqW9If/5G02WgqtP\n60Yfh83TgWPP8Aq/B7ie6K2PssNumdIqhMgY5Zp53TPODDt1R9N4LvYNhMjP//lZfu/j27gQed3J\njEx1uB4Q0HsgQFFqWg/TXIAzNG3PJC84c0LjouTC3pK661h2PVYLTK5Yy3LuX8tYH5eUVhNiIKJQ\nQVDkivvlmPWxRUvBJDcYpZkUCmJErVbq17JOuJlkqYOfvRnL44GB48Qg7vcA1xM9qyUnJjnro4y6\nczy5tWBW90ghmLcpSOO52ZydpePff+AZHr+45IFpxpecGSEinNuaE0VL45K4hwBr4wytBUJIJqWk\n957NqaFQht2mZVJkSJEsBopMM0UQhMDoNDCUa8N6IXC55+Q4Y5xn9FkEGZkUZmXYpRCAWIltBJZt\nDyv/9IPumAMxvplkqfS4XH9ad2DguDK8yu8BbnQFejCdulnlnNtd0vQdF2cNi7bnjz69y288dh4B\nvPkVY151ImexdCw6T2YFEUkXAr5N7Y5Ej0IxzTMyK9mZLVl0nnKsMUqT6UhpSvrQMy0zpiOD66Ew\niklhsUqgcs1GOWJzZMmtRApJ7wNKPr9xeiC2frXZ6TwYLYgRFm2P1ZpRrg//5hs1RTv6eAy2vAMv\nRwZxvwdQUrK37BBExKrnOwSo8iSKR8Wq7hxRRMaFwgXPkzs1P/v+s5zbbXnkVMXffONJmrpjt+lZ\nth0L5wk+id/pkSVOBPvLjrbrGFeGeduy10BpFVVuKLTBVJ66DeS55KHJGutVTowh2QpnCoXEETk9\nLliv7KH3eQiRzCRxjvHS7pe6cxgl0SvrYR8DSkiU5JIYu5stswy2vAMvV4ZX/V3OwQZibhR179mZ\ntUQR2ChzODKFeiCATe9w3uNc5Bc/+DT/4f3nKK3im77wNF/40AglNRd6j+tqOu9SCEam2cgteWlY\nNAFdCrwQtE0A4xjninFumFrDKDNMqzGLtsUqzSg3rFeGcW5YtD2l0WyMMqxWh6LrfbiuwB5dlSur\nOBidOroqH8osAwM3zvCuuMs5qDPHlXPjuNRIBD6GZIQlJfvL7jDPMwT4syd2+Oe/80me2Fry116z\nybe+6UGe3N5HScn+sqVzjo1pSZYpnt1esjEuKKxCS8HUQi9zXO9pXCAGGJcZpyYlVqeyyuYoI1MS\nF5JHi2DlsphZCivJrSaEZMzlfAr6yC+7wricG1mVD2WWgYEbZxD3u5wQIj7C9rxlt+koV97lSoDz\nka15TSSVTZad56d//3F+8YNPc3Kc8fZv+DwemuTsti1GSbb2alwEFwSFFZBZTq0lD5e1yhI8mAJ8\nlBA9PoAxAhcFQkKpLWtjixABrQSBuPJ0ORg6AqLAufTBA4BIpZbrmXPd6Kp8KLMMDNwYw7vkLseF\nyPa8w8Vkb+uCZ3veMikMTe/Yq1uqzPB7H9/iX/3u41yYtbzl80/zltefojAaR8S7AFEwaz1Cpvtc\n+kAuNA+ujam7nt7DKNNMS8vOoiF4ySg3VKVm2TimpSXXyYmx6TxrpaW0miJTSCFwIeAd5DbZCIcY\nMUodxvwdHUi6EsOqfGDg1jKI+11O7wJKRUSQBCLOCYSMzBvHvOtp+si7/vMn+e2/PM+Dazn/5Jve\nwGZu2GlrQghkViJIwdjro5RE1PaBro+UU41RMK4KpBKUymC1ZJJbfAgs2h6InJ6UIMBKRZkptNRs\nlFnKXfUBF9LmpzagpQApKNSlL60bMecaVuUDA7eO4Z10F+NDZNn1SClwMdC7iDWSQiie3lvyp5/Z\n5d1/9AR7dc/fedODfN3n30fTO7ro2RwVLFvHzrwj05Iq15RZxazt2K97fB8RIuK9ZDwynJwUZFpi\nlKbznq6PrJeBrXlDFIIqU7xic8Q4M+RW4UNAwCoNKfmtx1VJxYd4zX71gYGB288g7ncpB9OYetVB\nUmaaRfR0vefpvSX/5vc/zfuf2OXVJyt++Otex4NrBduLlqd3F0yLDGKa4hwVJvWMNylybmQyms6z\n9J7ooMgFY2sorYYIgohVisIkO4MTk4ymdyihWCsskzK1Ns4bR+ccGpkmW11qc1RS0jl3zX71gYGB\n28/wbruLOOqb0rmAVpJxbtlZtASXyiS/87HzvPsPP0PnIm998xm+5U0PYJTi3PaSedMzLS1Kw7Jz\n+ACZEtSuZ9F5PJFxZhhlFq08k9KglSJImC17Tk0KEAJrUsKSNZKxlJhxTmH1JZmjo1xTd6n1UgRB\nmanDmvn1+tUHBgZuP4O43yVc7pviO4ePUFjFepXxgc9s8ZP/7yf54Nk9vvjhNX7o61+PFKlUE4JH\nSZjkSay3Fy1SCDweFwTOB0QI4DXztqPzASEEbRfYmOYYpVgfWaQS2NWVghce7yNVaV5QCz/4EBKk\nSdOj5l030q8+MDBw+7khcRdCvAX4SdI79V0xxndc9v3vBr4DcMAF4H+MMT5xi8/1WHO5b4pWEh8i\nTef5uT87y7/8nY8jheB7v/YR/tZfeQAfIjuLHqk840ynwI4+UPcOJZNfi4+pRXGzKnG5Z+kd3ktO\njjO6PlAay/rYQoDWBZQSuBCpMoUOaWp0VvecmhaH4n09867BrGtg4O7guuIuhFDAvwa+GjgLvF8I\n8Z4Y40ePHPbnwKMxxqUQ4h8CPw586+044ePK5b4pVis+fHaXH/u1v+QjT+/zV1+9yXf99VezWWXU\nnV8NEEF08jC9aDe2xA7WipzWOUarWvqzewsyq8nQCBERCKxSSAESSY8nucuknnWjJEpEstKmK4AQ\nSBHY1zfvGqZIBwbuDm7kHfelwCdjjI8DCCH+HfCNwKG4xxj/vyPH/zHw1lt5ki8Hjq54Oxd45x88\nzr/5/U8xyjX/7JvfyF9/3SnObi/pnCNqeej6uFnltH2g9Z5ZnerohVFpJS7lKshDMskNkZSm9Nz+\nnBgFVWaZlJrepV71eddRZvmhIKdQ6rQ6P+B65l1Dv/rAwN3BjYj7g8BTR74+C7z5Gsf/T8Cvv5iT\nejlysOL90JO7QLMjaQAADlVJREFU/OAvP8Ynzs/5b994Pz/yDW/g5DhLxmECOu/pfRJUEQX7TUdm\nJGuFpXUe10eigUwrtIKdzjPKDEpKpAQpBGtlTt16poUFYJwbpEibsCEEhJZkShERh6WWA27GJmBg\nYOCl45a+A4UQbwUeBb7iKt9/G/A2gDNnztzKX33P0zrPv/jtj/N//9FnODXO+Km/+8V8zRvuQ636\nxncXHaNCA5reB+pWsFs3CJmGjJreY4RiMlZkWhIjaVI0CDZGGd6DNanEUlqNQvHAWgEIpAQtJK8+\nNaFZpScpmSZLD+LvDhjKLgMD9wY38o48Bzx85OuHVrddghDiq4AfAL4ixthe6Y5ijO8E3gnw6KOP\nxisdc9y4kVi4P/zkRb7/Fz7EU9s1b/2yM3zfW17H+EjbYec8cjWlut/0aC1ARNreI6U8rIGXmUYA\nde8oTKp/p/uJSJuE2CiB85FTY0u12gQ1Wh6eV2HVNc93KLsMDNwb3Ii4vx94rRDilSRR/zbg248e\nIIT4K8BPA2+JMZ6/5Wd5j3K9zpK9uuef/Npf8u8/8BSvPFHxs9/5ZbzxwUmaMu08mU294yFEKmPZ\nXjbkRhIRLLsUdH16WpAbRW7SlGjbeUplMFrSucjGyNK5QB/S1OnmKD8cKLqSIN9ISWUouwwM3P1c\n9x0aY3RCiO8CfpPUCvnuGONjQoi3Ax+IMb4H+GfACPg5IQTAkzHGb7iN531PcK3Okj/4xEV+8Jc+\nwtai43/+ilfxHV/+SnoX2Fm0FFYjlWDZ+sMwjmTAJREiQoRcSyqTkxtFHwI56jAAe1pZuj5gslRj\nN0rS9gKjkkXw1YR9YGDg+HBDy68Y43uB91522w8d+fdX3eLzOhZcqbNka9HxY7/6UX7ro8/x+vsn\n/Mzfe5RXn6zo+gAiIgTs1x1WSZQSLJrksNgFn+rmWoKMaKlIdx2RUdI5T4ywUaZp0kVbY0Xy4hWk\nn5taS4zX9lUfGBg4HgzX1reRo50lMUZ+8c/P8Y5f/xhNH/hHX/Va/sFXvJoYIzFCJGWI9j7F1fUh\nsuw85+c1J0Y5J0Y5y65nZ9FRZZr7pjl172l7DzqSaY3Vmmq1Kq8ys6qLPx9nd3BOAwMDx59B3G8j\nB50l57ZqfuRXHuMPP7XFFz405X//W2/kVScqnA+EELEm9ZP3q15xIQSL1tF7T2nSJmnnPOMstS46\nl+reoyzF3oUYXlBuOaiJH5SFhq6WgYGXF8M7/TbzHz5wlv/zt/4rAN//tZ/L3/2yV2D0qlQToA9J\n4I2WCAQuRpzzSCno2+QGqVXqiOm8x0gJOlAd6aaB5N3y2XS13Eg3z8DAwL3HIO43yY2K4Seem/F9\n//FD/NmTu3zl557kB/7G63h4s7rkGCnTkFCIESlECpnuHIvGU1pJlWm0FhglDqdAAxEt1CX3czXv\nlut1tVyvm2dgYODeZRD3m+BGxLBzgZ/+/U/xr373k1SZ4ie+9Qv5pi96kKb3V5zsPOgxP7DKLY2m\nWHveCGx70SLNqqySZkbJrTy8rxdTbrmeT8zAwMC9y/AOvgmuJ4YfOrvL9/78h/jYszO+/gsf4Ie/\n/vWcGGXAtSc7D1bYhX3+A2TROIwW3L9Wsmwdy94xyZ4Py7gVQ0TX84kZGBi4dxnE/Sa4mhjO656f\n+J1P8K7/9Dgnxxk/8/ce5atff/qS4260Bq6kYJSbw2NDiKyPLFYXlxx7K1bWgz3vwMDxZRD3m+BK\nYvhfPnWRH37PYzy5XfPtbz7D937t55JpyaLpX1CTv5nJzjsxBTr4xAwMHF+Gd/FNcFQMF53jx3/j\nv/Jzf3qWV2yW/Ox3fhlf+soN6s4RI/fEBuXgEzMwcHwZxP0mOBDD3/jIM/zor3yUi/OW7/xrr+K7\nv/oRCqsON1vvpQ3KwSdmYOB4Mryrb4KL85Yfec9j/OqHnuF194151//wKF/w0Nrh94cNyoGBgbuF\nQdwv40p97FLAL/75Od7+qx9l2Xr+8dc8wj/4ildjriDkwwblwMDA3cAg7ke4Uh/74+dn/Nh7P8Yf\nfPwCb3rFOv/0m9/Ia06Nr/jzwwblwMDA3cKgOkc42sceQuRn3/8k//y3Pg7Aj37DG/jvv+wV11yF\nDxuUAwMDdwuDuB/hoGb++IU5P/jLj/GnT+zw5a85wQ9/3efx2vsmN3QfwwblwMDA3cCgQkfwMfLO\n3/sUP/X7n6Iwinf87Tfy9V9w/1AzHxgYuOcYxH3FR87t8T0//xf85TMzvvYNp/nBr3s9G6UlxHhJ\nQPTAwMDAvcDLXtyb3vMvf+cT/Mx/epzNyvJTb/1ivvKRk4SQUpGGmvnAwMC9yMta3N/3+Bbf/wsf\n5tMXF3zblzzM//o3P49pYa7/gwMDAwN3OS9LcZ81Pe/49Y/xb9/3JGc2Sv7td7yZv/qaEy/1aQ0M\nDAzcMl524v67H3uOH/jFj/DcfsN3fPkr+e6veYRy6G4ZGBg4ZrxsVG1r3vL2X/0ov/zBp/nc02N+\n6q1v4oseXrv+Dw4MDLzsuRfjKI+9uMcYec9fPM2P/spHmTU9/+irHuEffuWrsVpe/4cHBgZe9tyr\ncZTHWtyf3q35337pI/zux87zRQ+v8ePf8gU8cvrK1gEDAwMDV+JejaO8e8/sRRBC5P/5kyd5x69/\nDB8iP/h1r+fv/zefc1d/yg4MDNyd3Ktur8dO3B+/MOf7f+HD/Mmnt/ny15zg//jbb+ThjfKlPq2B\ngYF7lHvV7fXYiLvzgXf950/zE7/9cTIt+fFv+QL+zpseQoi7+wkYGBi4u7lX3V7v7rO7QR57eo/v\n+48f4iPn9nnLG+7j7d/4Bk5N8pf6tAYGBo4B96rb6w2JuxDiLcBPAgp41//f3r2GWFGHcRz//nJZ\nSymTLJB000op0RfZYkV0v1AG7YtuRpHRUljUm3pRYETYmwrqRSWVUJRC9yAOdIOyEqIthcpKuqyb\n1laU3YQKS92nFzPCYXM9s7vnzJyZfh848J+Z/7LPszPn2Tn/+Z+ZiLh72PaJwGrgeOAX4LKI2NLc\nUP9rx87dPLj2Kx55Z4Cpkzp5+IqFnL9geqt/rZmNUhmnEtYr491eG0YraQKwEjgHGATWS6pFxKa6\nbr3AbxFxtKQlwD3AZa0IeI/1W37l1hc3MrDtTy45fga3XzCPKZN86wCzdlPWqYRll+Vf0SKgPyIG\nACQ9A/QA9cW9B7gzbb8APCRJERFNjBWAP/7exb2vfc7q97YyY+oBrOldxClzDm32rzGzJinrVMKy\ny/KXPRz4tm55EDhhpD4RsUvSduAQ4Of6TpKuA64D6OrqGlPAj76zmTV9W7nm5Nnccu5cJk/0wWHW\nzso6lbDscq2MEbEKWAXQ3d09prP6ZacdxZnHHMZxXVObGpuZtUZZpxKWXZbv4H8HzKxbnpGu22sf\nSR3AFJILq003eWKHC7tZiXR2TGAogqGh5Hxuz1RCPwSntbIU9/XAHEmzJXUCS4DasD41YGnavhhY\n24rxdjMrnz0zTSTYvXsoeQiOL6a2XMNhmXQM/UbgdZKpkI9HxGeSVgAbIqIGPAaskdQP/EryD8DM\nDCjnVMKyy/TXjohXgFeGrbujrr0DuKS5oZmZ2Vj5vrdmZhXk4m5mVkEu7mZmFeTibmZWQS7uZmYV\n5OJuZlZBLu5mZhWkor5IKmkbsHWMPz6NYTclqyDnWA3OsRraKccjIqLhrXALK+7jIWlDRHQXHUcr\nOcdqcI7VUMYcPSxjZlZBLu5mZhVU1uK+qugAcuAcq8E5VkPpcizlmLuZme1bWc/czcxsH9q6uEs6\nT9IXkvol3baX7RMlPZtuf1/SrPyjHJ8MOd4saZOkjZLelHREEXGOR6Mc6/pdJCkklWpWAmTLUdKl\n6b78TNJTecc4XhmO1S5Jb0n6MD1eFxcR51hJelzST5I+HWG7JD2Q5r9R0sK8YxyViGjLF8mDQTYD\nRwKdwMfAvGF9bgAeSdtLgGeLjrsFOZ4BTErb11cxx7TfgcA6oA/oLjruFuzHOcCHwNR0+bCi425B\njquA69P2PGBL0XGPMsdTgYXApyNsXwy8Cgg4EXi/6Jj39WrnM/dFQH9EDETEP8AzQM+wPj3Ak2n7\nBeAsSWV6dlfDHCPirYj4K13sI3mGbZlk2Y8AdwH3ADvyDK5JsuR4LbAyIn4DiIifco5xvLLkGMBB\naXsK8H2O8Y1bRKwjeZLcSHqA1ZHoAw6WND2f6EavnYv74cC3dcuD6bq99omIXcB24JBcomuOLDnW\n6yU5cyiThjmmH29nRsTLeQbWRFn241xgrqR3JfVJOi+36JojS453AldKGiR5cttN+YSWm9G+Xwvl\nhxqWhKQrgW7gtKJjaSZJ+wH3A1cXHEqrdZAMzZxO8ulrnaQFEfF7oVE11+XAExFxn6STSJ6rPD8i\nhooO7P+onc/cvwNm1i3PSNfttY+kDpKPgr/kEl1zZMkRSWcDy4ELI+LvnGJrlkY5HgjMB96WtIVk\nLLNWsouqWfbjIFCLiJ0R8TXwJUmxL4ssOfYCzwFExHvA/iT3ZKmKTO/XdtHOxX09MEfSbEmdJBdM\na8P61IClaftiYG2kVz5KomGOko4DHiUp7GUbp4UGOUbE9oiYFhGzImIWyXWFCyNiQzHhjkmWY/Ul\nkrN2JE0jGaYZyDPIccqS4zfAWQCSjiUp7ttyjbK1asBV6ayZE4HtEfFD0UGNqOgrug2uXi8mOcPZ\nDCxP160gefNDcvA8D/QDHwBHFh1zC3J8A/gR+Ch91YqOudk5Duv7NiWbLZNxP4pk+GkT8AmwpOiY\nW5DjPOBdkpk0HwHnFh3zKPN7GvgB2EnySasXWAYsq9uHK9P8P2n349TfUDUzq6B2HpYxM7MxcnE3\nM6sgF3czswpycTczqyAXdzOzCnJxNzOrIBd3M7MKcnE3M6ugfwGpCYusxrHPYAAAAABJRU5ErkJg\ngg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IzKDpjNHdtyA",
        "colab_type": "text"
      },
      "source": [
        "## Salvar e carregar a RNA\n",
        "### _Save and load ANN_"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FWZ7Q9F5Py_a",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from google.colab import files"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Mu9yecImQCex",
        "colab_type": "text"
      },
      "source": [
        "### Save validation and test results"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rUU-YFF0FpPz",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "files.download(filename_validation)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3Cg9WScnGQUQ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "files.download(filename_test)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KzDv9U5vQEXA",
        "colab_type": "text"
      },
      "source": [
        "### Zip ANN"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Lup_qBUGE7Iz",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!zip -r /content/ann.zip /content/ann"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "t0y9s62qQHiZ",
        "colab_type": "text"
      },
      "source": [
        "### Load ANN\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fJMrzKylQL9N",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "uploaded = files.upload()\n",
        "\n",
        "for fn in uploaded.keys():\n",
        "  print('User uploaded file \"{name}\" with length {length} bytes'.format(\n",
        "      name=fn, length=len(uploaded[fn])))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QmM5B0Vs5huR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!unzip ann.zip \n",
        "!mv content/ann ."
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ogpBeuFrRiX-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "EHF = tf.estimator.DNNRegressor(\n",
        "      feature_columns=construct_feature_columns(),\n",
        "      hidden_units=[50,20],\n",
        "      model_dir='ann'\n",
        "      )\n"
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}