Arbres_de_décision_régresseur_sklearn_ipynb.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Arbres de décision régresseur sklearn. ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyPcb4vSoiUvlol7kT/xJa4j",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/MorganGautherot/Tuto_Tree_ML/blob/master/Arbres_de_d%C3%A9cision_r%C3%A9gresseur_sklearn_ipynb.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "72tI3wHAInwU",
        "colab_type": "text"
      },
      "source": [
        "# Arbres de décision régresseur avec sklearn\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rs0hYKR_Js4v",
        "colab_type": "text"
      },
      "source": [
        "Dans ce notebook, je vais vous présenter la fonction d'arbre de décision de sklearn. Vous allez voir à quoi servent tous les paramètres ainsi que les attributs pour être sûr de choisir les bonnes options, afin de créer les modèles d'arbres les plus performants. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Pvgpb3uzKr8e",
        "colab_type": "code",
        "outputId": "80507fba-cd6f-411b-b605-29f6912ab0b8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 217
        }
      },
      "source": [
        "# Importation des packages\n",
        "from sklearn.tree import DecisionTreeRegressor\n",
        "from sklearn import tree\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "\n",
        "# Importation des données\n",
        "from sklearn.datasets import load_boston\n",
        "\n",
        "dict_data = load_boston()\n",
        "data_boston = pd.DataFrame(dict_data['data'], columns = dict_data['feature_names'])\n",
        "y = dict_data['target']\n",
        "\n",
        "data_boston.head()"
      ],
      "execution_count": 0,
      "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>CRIM</th>\n",
              "      <th>ZN</th>\n",
              "      <th>INDUS</th>\n",
              "      <th>CHAS</th>\n",
              "      <th>NOX</th>\n",
              "      <th>RM</th>\n",
              "      <th>AGE</th>\n",
              "      <th>DIS</th>\n",
              "      <th>RAD</th>\n",
              "      <th>TAX</th>\n",
              "      <th>PTRATIO</th>\n",
              "      <th>B</th>\n",
              "      <th>LSTAT</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.00632</td>\n",
              "      <td>18.0</td>\n",
              "      <td>2.31</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.538</td>\n",
              "      <td>6.575</td>\n",
              "      <td>65.2</td>\n",
              "      <td>4.0900</td>\n",
              "      <td>1.0</td>\n",
              "      <td>296.0</td>\n",
              "      <td>15.3</td>\n",
              "      <td>396.90</td>\n",
              "      <td>4.98</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.02731</td>\n",
              "      <td>0.0</td>\n",
              "      <td>7.07</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.469</td>\n",
              "      <td>6.421</td>\n",
              "      <td>78.9</td>\n",
              "      <td>4.9671</td>\n",
              "      <td>2.0</td>\n",
              "      <td>242.0</td>\n",
              "      <td>17.8</td>\n",
              "      <td>396.90</td>\n",
              "      <td>9.14</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.02729</td>\n",
              "      <td>0.0</td>\n",
              "      <td>7.07</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.469</td>\n",
              "      <td>7.185</td>\n",
              "      <td>61.1</td>\n",
              "      <td>4.9671</td>\n",
              "      <td>2.0</td>\n",
              "      <td>242.0</td>\n",
              "      <td>17.8</td>\n",
              "      <td>392.83</td>\n",
              "      <td>4.03</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.03237</td>\n",
              "      <td>0.0</td>\n",
              "      <td>2.18</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.458</td>\n",
              "      <td>6.998</td>\n",
              "      <td>45.8</td>\n",
              "      <td>6.0622</td>\n",
              "      <td>3.0</td>\n",
              "      <td>222.0</td>\n",
              "      <td>18.7</td>\n",
              "      <td>394.63</td>\n",
              "      <td>2.94</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.06905</td>\n",
              "      <td>0.0</td>\n",
              "      <td>2.18</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.458</td>\n",
              "      <td>7.147</td>\n",
              "      <td>54.2</td>\n",
              "      <td>6.0622</td>\n",
              "      <td>3.0</td>\n",
              "      <td>222.0</td>\n",
              "      <td>18.7</td>\n",
              "      <td>396.90</td>\n",
              "      <td>5.33</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "      CRIM    ZN  INDUS  CHAS    NOX  ...  RAD    TAX  PTRATIO       B  LSTAT\n",
              "0  0.00632  18.0   2.31   0.0  0.538  ...  1.0  296.0     15.3  396.90   4.98\n",
              "1  0.02731   0.0   7.07   0.0  0.469  ...  2.0  242.0     17.8  396.90   9.14\n",
              "2  0.02729   0.0   7.07   0.0  0.469  ...  2.0  242.0     17.8  392.83   4.03\n",
              "3  0.03237   0.0   2.18   0.0  0.458  ...  3.0  222.0     18.7  394.63   2.94\n",
              "4  0.06905   0.0   2.18   0.0  0.458  ...  3.0  222.0     18.7  396.90   5.33\n",
              "\n",
              "[5 rows x 13 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 2
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aKC2mFePJO4K",
        "colab_type": "text"
      },
      "source": [
        "## Les paramètres\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "K8kvxbj-MuvA",
        "colab_type": "text"
      },
      "source": [
        "Dans cette partie, je vais prendre le temps de vous expliquer chaque paramètre de l'algorithme d'arbre de décision. De cette façon, vous pourrez choisir judicieusement les paramètres les plus adaptés à votre problème dans le but d'entraîner le modèle d'arbre le plus performant. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fcfHdsOZJQsc",
        "colab_type": "code",
        "outputId": "e4034744-487d-493d-b495-0ee658533ac4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 123
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "DecisionTreeRegressor()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "DecisionTreeRegressor(ccp_alpha=0.0, criterion='mse', max_depth=None,\n",
              "                      max_features=None, max_leaf_nodes=None,\n",
              "                      min_impurity_decrease=0.0, min_impurity_split=None,\n",
              "                      min_samples_leaf=1, min_samples_split=2,\n",
              "                      min_weight_fraction_leaf=0.0, presort='deprecated',\n",
              "                      random_state=None, splitter='best')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LAqVUlkuQt6u",
        "colab_type": "text"
      },
      "source": [
        "Si vous ne choisissez pas de paramètres, vous aurez les paramètres par défaut choisi par Sklearn. Ces paramètres ne seront pas toujours adaptés à votre jeu de données.\n",
        "\n",
        "Pour choisir vos paramètres, il vous suffit d'ajouter dans la parenthèse le nom du paramètre et d'ajouter '=' et l'option choisit. Comme vous pouvez voir sur l'exemple ci-dessous."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jj5rBBZYRLTP",
        "colab_type": "code",
        "outputId": "f51347dd-50d2-47b8-dac5-6932baa96267",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 123
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "DecisionTreeRegressor(criterion='mae', min_samples_split=\"10\", max_depth='5')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "DecisionTreeRegressor(ccp_alpha=0.0, criterion='mae', max_depth='5',\n",
              "                      max_features=None, max_leaf_nodes=None,\n",
              "                      min_impurity_decrease=0.0, min_impurity_split=None,\n",
              "                      min_samples_leaf=1, min_samples_split='10',\n",
              "                      min_weight_fraction_leaf=0.0, presort='deprecated',\n",
              "                      random_state=None, splitter='best')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dk7ZD4xDJRD4",
        "colab_type": "text"
      },
      "source": [
        "### Criterion\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1hQkyrwHODF0",
        "colab_type": "text"
      },
      "source": [
        "Le paramètre *criterion* est la fonction de coût, c'est-à-dire le critère de performance que l'on choisit pour construire notre modèle et mesurer sa qualité. \n",
        "\n",
        "On a le choix entre trois options : le mse, le friedman_mse et le mae."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pxi6UELlOlup",
        "colab_type": "text"
      },
      "source": [
        "#### Mean Square Error (MSE)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9m39H0ktOw-y",
        "colab_type": "text"
      },
      "source": [
        "Le Mean Square Error (MSE) ou l'erreur moyenne au carré et une fonction de coût qui s'écrit comme cela :\n",
        "\n",
        "$ MSE = \\frac{1}{m} \\sum^{m}_{i=1}(y-\\hat{y})^2 $\n",
        "\n",
        "où :\n",
        "- *m* est le nombre d'exemples de notre jeu de données ;\n",
        "\n",
        "- *y* est la variable cible ;\n",
        "\n",
        "- $\\hat{y}$ est la variable estimée.\n",
        "\n",
        "C'est la fonction de coût la plus utilisé pour entraîner des modèles de régression en machine learning. C'est d'ailleurs pour cela que la fonction la propose en valeur par défaut. Cette fonction à la particularité de pénaliser plus fortement les larges erreurs étant donné que l'erreur est élevée au carré. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u9OoWQIdRZ9d",
        "colab_type": "code",
        "outputId": "9b7922f3-804d-4aa8-fb82-32ca0a31ab2b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 123
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "DecisionTreeRegressor(criterion='mse')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "DecisionTreeRegressor(ccp_alpha=0.0, criterion='mse', max_depth=None,\n",
              "                      max_features=None, max_leaf_nodes=None,\n",
              "                      min_impurity_decrease=0.0, min_impurity_split=None,\n",
              "                      min_samples_leaf=1, min_samples_split=2,\n",
              "                      min_weight_fraction_leaf=0.0, presort='deprecated',\n",
              "                      random_state=None, splitter='best')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FlWqbCoVOlpG",
        "colab_type": "text"
      },
      "source": [
        "#### Friedman MSE\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bAYI5wsqUqwh",
        "colab_type": "text"
      },
      "source": [
        "Friedman à changé le MSE pour créer une fonction de coût plus adapté aux arbres de décision. Si vous voulez en apprendre plus sur cette fonction de coût, je vous conseille de lire [son article](https://statweb.stanford.edu/~jhf/ftp/trebst.pdf)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OiiElMJfRdUU",
        "colab_type": "code",
        "outputId": "e280cbc3-f1a6-4b7b-aae7-cb6170fced91",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 123
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "DecisionTreeRegressor(criterion='friedman_mse')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "DecisionTreeRegressor(ccp_alpha=0.0, criterion='friedman_mse', max_depth=None,\n",
              "                      max_features=None, max_leaf_nodes=None,\n",
              "                      min_impurity_decrease=0.0, min_impurity_split=None,\n",
              "                      min_samples_leaf=1, min_samples_split=2,\n",
              "                      min_weight_fraction_leaf=0.0, presort='deprecated',\n",
              "                      random_state=None, splitter='best')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wDy7ytjBOlam",
        "colab_type": "text"
      },
      "source": [
        "#### Mean Absolute Error (MAE)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "B7oZTt6DW2Zm",
        "colab_type": "text"
      },
      "source": [
        "Le Mean Absolute Error (MAE) ou l'erreur absolut moyenne et une fonction de coût qui s'écrit comme cela :\n",
        "\n",
        "$ MSE = \\frac{1}{m} \\sum^{m}_{i=1}|y-\\hat{y}| $\n",
        "\n",
        "où :\n",
        "- *m* est le nombre d'exemples de notre jeu de données ;\n",
        "\n",
        "- *y* est la variable cible ;\n",
        "\n",
        "- $\\hat{y}$ est la variable estimée.\n",
        "\n",
        "Contrairement au MSE, le MAE va moins pénaliser les larges erreurs, car l'erreur n'est plus élevée au carré. Je vous recommande d'utiliser la MAE, c'est ce que j'utilise le plus souvent."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ROO9aFH6Rg09",
        "colab_type": "code",
        "outputId": "c467398e-6932-4bb4-b6fe-bd5a0ade7945",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 123
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "DecisionTreeRegressor(criterion='mae')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "DecisionTreeRegressor(ccp_alpha=0.0, criterion='mae', max_depth=None,\n",
              "                      max_features=None, max_leaf_nodes=None,\n",
              "                      min_impurity_decrease=0.0, min_impurity_split=None,\n",
              "                      min_samples_leaf=1, min_samples_split=2,\n",
              "                      min_weight_fraction_leaf=0.0, presort='deprecated',\n",
              "                      random_state=None, splitter='best')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gS0XA4hYJTyH",
        "colab_type": "text"
      },
      "source": [
        "### splitter\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hHJz11jXXio8",
        "colab_type": "text"
      },
      "source": [
        "Le paramètre splitter nous permet de choisir comment nous allons séparer notre jeu de données pour construire notre arbre. \n",
        "\n",
        "Nous pouvons choisir aléatoirement la variable sur laquelle séparer notre jeu de donnée ou nous pouvoir itérer afin de choisir la meilleure variable pour séparer notre jeu de données. \n",
        "\n",
        "Je vous conseil de choisir *best* qui choisis la variable la plus importante à séparer de cette façon vous obtiendrez un arbre moins profond et plus performant. \n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sMg5yNMaY5WU",
        "colab_type": "code",
        "outputId": "229b9860-c250-42f3-dee9-935c853c69e5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 123
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "DecisionTreeRegressor(splitter='best')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "DecisionTreeRegressor(ccp_alpha=0.0, criterion='mse', max_depth=None,\n",
              "                      max_features=None, max_leaf_nodes=None,\n",
              "                      min_impurity_decrease=0.0, min_impurity_split=None,\n",
              "                      min_samples_leaf=1, min_samples_split=2,\n",
              "                      min_weight_fraction_leaf=0.0, presort='deprecated',\n",
              "                      random_state=None, splitter='best')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BS5oKJdtY1-t",
        "colab_type": "text"
      },
      "source": [
        "Vous pouvez utiliser *random* lorsque vous savez que toutes vos variables sont importantes pour votre modèle et que vous voulez économiser du temps de calcul. Dans tous les autres cas utilisés *best*."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iTWdvgcVY_An",
        "colab_type": "code",
        "outputId": "92eec37f-6829-422f-9143-449f6d00c325",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 123
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "DecisionTreeRegressor(splitter='random')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "DecisionTreeRegressor(ccp_alpha=0.0, criterion='mse', max_depth=None,\n",
              "                      max_features=None, max_leaf_nodes=None,\n",
              "                      min_impurity_decrease=0.0, min_impurity_split=None,\n",
              "                      min_samples_leaf=1, min_samples_split=2,\n",
              "                      min_weight_fraction_leaf=0.0, presort='deprecated',\n",
              "                      random_state=None, splitter='random')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sq0LMLm2JVhR",
        "colab_type": "text"
      },
      "source": [
        "### max_depth\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Jrcu6bhEi1ot",
        "colab_type": "text"
      },
      "source": [
        "La profondeur de l'arbre est un des paramètres les plus important à définir. \n",
        "\n",
        "Ce paramètre va déterminer la longueur maximum de votre arbre. Vous pouvez aussi spécifier None pour ce paramètre, dans ce cas durant l'entraînement l'arbre va s'approfondir jusqu'à obtenir les feuilles les plus pures possible. \n",
        "\n",
        "Je ne vous conseille pas de laisser ce paramètre à *None*, car plus votre arbre sera profond plus il sera complexe et plus il souffrira du sur-entraînement. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o1bmbDTVk5By",
        "colab_type": "code",
        "outputId": "dfc900d6-3f4e-4834-d6a3-e4f4584d3e75",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 247
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_depth=3)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "# Visualisation de l'arbre\n",
        "tree.plot_tree(model_tree)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TzW3OawHjm-c",
        "colab_type": "code",
        "outputId": "3a7b7c28-152e-479f-ea20-192645739eb4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 247
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_depth=None)\n",
        "\n",
        "# Entraînement du modèle d'arbre \n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "# Visualisation de l'arbre\n",
        "tree.plot_tree(model_tree)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "z5ni4lZ5ovdm",
        "colab_type": "text"
      },
      "source": [
        "Plus l'arbre est profond, plus il sera performant sur le jeu d'entraînement, mais plus vous dégraderez les performances sur le jeu de test. \n",
        "\n",
        "Plus la profondeur est réduite, plus le modèle sera simple et généralisable. Cette meilleure généralisation augmentera les performances du modèle sur le jeu de test.\n",
        "\n",
        "Attention, si vous réduisez trop la profondeur de l'arbre, vous dégraderez aussi les performances global du modèle. Tester plusieurs profondeurs afin de maximiser vos performances sur votre jeu de tests."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EhPqWzLkJXS6",
        "colab_type": "text"
      },
      "source": [
        "### min_samples_split\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GIulOfDCmqE4",
        "colab_type": "text"
      },
      "source": [
        "Le min sample split est le nombre minimal d'exemples que le modèle doit avoir pour pouvoir faire une nouvelle séparation. \n",
        "\n",
        "Vous avez le choix de remplir ce paramètre avec un entier (int) ou un flottant (float). \n",
        "\n",
        "Si vous utilisez un entier, le modèle le comprendra comme le nombre minimal d'exemples que le noeud doit avoir pour pouvoir créer une nouvelle séparation. \n",
        "\n",
        "Si vous utiliser un flottant, le modèle le comprendra comme la fraction de données que le noeud doit avoir pour pouvoir créer une nouvelle séparation.\n",
        "\n",
        "Plus ce paramètre sera grand plus votre arbre sera simple et plus il sera généralisable."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9K9dUpFVmphU",
        "colab_type": "code",
        "outputId": "f7850ea1-1a67-4854-8947-e5befedc1e88",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 247
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_depth=2, min_samples_split=2)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "# Visualisation de l'arbre\n",
        "tree.plot_tree(model_tree)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "g5r1eEYRohwx",
        "colab_type": "code",
        "outputId": "a62a544b-df38-4aad-ba0b-2b14321330da",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 247
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_depth=2, min_samples_split=100)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "# Visualisation de l'arbre\n",
        "tree.plot_tree(model_tree)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vSH8sGcupvhh",
        "colab_type": "text"
      },
      "source": [
        "Plus ce paramètre est petit, plus le modèle sera performant sur le jeu d'entraînement, mais plus vous dégraderez les performances sur le jeu de test. \n",
        "\n",
        "Plus ce paramètre est élevé, plus le modèle sera simple et généralisable. Cette meilleure généralisation augmentera les performances du modèle sur le jeu de test.\n",
        "\n",
        "Attention, si vous augmentez trop ce paramètre, vous dégraderez aussi les performances global du modèle. Tester plusieurs valeurs afin de maximiser vos performances sur votre jeu de tests."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "y5Q8vEzOJY7X",
        "colab_type": "text"
      },
      "source": [
        "### min_samples_leaf\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yqjXADfgtM2C",
        "colab_type": "text"
      },
      "source": [
        "Le min_sample_leaf est le minimum d'exemples requis pour créer une feuille.\n",
        "\n",
        "Seules les feuilles ayant ce minimum d'exemples seront conservées. Pour ce paramètre aussi vous pouvez spécifier un nombre entier ou un nombre flottant.\n",
        "\n",
        "Plus ce paramètre sera bas, plus votre modèle pourra faire du cas par cas et aura du mal à généraliser."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4WaKDB7WvOv0",
        "colab_type": "code",
        "outputId": "bc094963-2afd-4994-e522-daa3c0106ea2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 247
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_depth=2, min_samples_leaf=2)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "# Visualisation de l'arbre\n",
        "tree.plot_tree(model_tree)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rqS6YH-MvPJF",
        "colab_type": "code",
        "outputId": "020d49b9-d662-4e75-a323-9866850c914f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 247
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_depth=2, min_samples_leaf=40)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "# Visualisation de l'arbre\n",
        "tree.plot_tree(model_tree)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAADnCAYAAAC9roUQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOydd1hUR/fHv8viNpqylBULJSgCokEQg2IwIIIGRRODHRSiAYnljRpLEmvEEE30jTUkAhqNWFLAkhAVLIgYRFDEhghiQVQIKiIgcH5/+HJ/rrtLb5L5PM95Hu/MmXNn1rtnh7lnzvCICAwGg8FoHtRaugMMBoPxb4I5XQaDwWhGmNNlMBiMZoQ5XQaDwWhGmNNlMBiMZkS9pTvAaJuIxeJ7JSUlhi3dD0bjIRKJ8p49eyZr6X687vBYyBijKeDxeMSerbYFj8cDEfFauh+vO2x5gcFgMJoR5nQZDAajGWFOl8FgMJoR5nQZDAajGWFOl9GqWbp0KXg8Hng8HjZs2FDrdrdv3+ba2dvbN2EPGYy6wZwuo0WoqKiAg4MDxowZI1d+7949SKVSrF27livr3bs3cnNz4efnx5UNGjSIc6pVEhAQwNUbGRkhNzcXc+bMafrB1JHbt29j/Pjx0NXVhVgshq2tLS5fvqxSv6ysDIsXL4apqSnEYjEcHBxw6tQplfrDhg0Dj8fDgQMH5MpnzpwJOzs7CIVC9kPUgjCny2gR+Hw+IiIiEBUVhX379nHl06ZNg7W1NWbNmsWVqaurQyaTQSKRyNkIDAxEbm4uJ19//TVXp6amBplMBk1NzQb1s7y8HHfv3m2QjZf5559/4OTkBJFIhJiYGKSnpyM4OBhaWloq23z++ecIDw/Hli1bkJaWhhEjRsDd3R05OTkKuhs2bEB1oXp+fn4KP3SMZoaImDBpdHnxaNXMV199Rfr6+nT//n2KiIggDQ0Nun79Ole/ZMkSsrOzU2jn7OxMc+bMqdG+qvY1kZaWRnPmzCGZTEZLly6tc3tVzJ8/n5ycnOrUpmPHjrRlyxa5Mjs7O1qwYIFcWXp6OnXq1Inu3r1LAGj//v1K7dX3M/nf/2mLP1uvu7CZLqNFmTt3LszMzODj44PZs2cjJCQEb7zxRq3abtu2DXp6eujZsyc+++wzPHv2rEF9KSgowIYNG2Bvb8/9yb9u3TrMnz+f0wkODoampma1omwGWkV0dDTs7e3x/vvvw8DAAHZ2dtixY0e1/SotLYVIJJIrE4vFSEhI4K7Lysowfvx4rF27Fh07dqznJ8BoDtg2YEaLwufzsXHjRtjb22PAgAGYPn16rdqNHz8exsbGMDIywvnz57FgwQJkZGRgz549dbp/ZWUlDh06hIiICOzfvx89evSAr68vDh48CENDxV3MAQEB8Pb2rtamkZGRyrobN25g06ZN+PTTT/HFF1/g1KlT8PPzg5aWFry8vJS2cXNzwzfffIOBAwfC2NgYe/fuRUJCAszNzTmdRYsWoWfPnvjggw9qOXJGS8GcLqPFCQsLg0QiwdWrV/HgwQMYGBjU2GbatGncv21sbGBkZARXV1fcvHkTxsbGtb53Tk4Ohg8fDl1dXezevRsjR46sVl9XVxe6urq1tv8qlZWVcHBwwIoVKwAAb775JpKTk7FlyxaVTve7776Dn58fzM3NoaamBltbW4wbNw7JyckAgLi4OOzbtw/nz5+vd78YzQdbXmC0KHFxcQgNDUVMTAy6deuGoKCgetnp168fAOD69et1ate5c2fs2rULffv2xQcffABnZ2ds3boVjx8/Vqrf0OUFmUyGHj16yJVZWlpW28bAwAAHDhzA06dPcevWLSQlJaG8vBympqYAgOPHjyMnJwdSqRTq6upQV38xl/Ly8oK7u3udPg9G08NmuowWo6ioCH5+fpg9ezacnJwQFhYGW1tb7Nu3D6NHj66TrdTUVACo83qmuro6xo4di7Fjx+LOnTvYvn07Vq9ejRkzZsDLyws+Pj4YMmQI+Hw+gIYvL/Tv3x8ZGRlyZdeuXUPXrl1r7KtYLIZYLEZhYSFiYmKwatUqAMD06dMVPi8bGxusX78ew4YNq9Euo5lp6Td5TNqmoBbRCwEBAWRhYUHPnj3jykJCQsjAwIAePnxIRMrftF+/fp2WL19OZ8+epaysLIqKiiIzMzNycXFRuEd939QnJCTQtGnTqH379rR8+fI6t1dFYmIi8fl8+uqrrygjI4MiIiJIKBRSdHQ0pzNp0iS5yISEhAT6/fffKTMzk44ePUp9+vQhe3t7KisrU3kfKIleyMjIoJSUFProo4/IysqKUlJSKCUlpdZ9B4teaJzvRkt3gEnblJqc7pEjR0hdXZ1Onz4tV15eXk4ODg40duxYIlLuNHNycujtt98mXV1dEgqFZG5uTvPnz6fHjx8r3Ke+TreKZ8+eUVZWVr3bK+P3338na2trEgqFZGVlRdu3b5erd3Z2Jl9fX+766NGjZGFhQQKBgPT19emjjz6igoKCau+hzOk6OzsTAAWpLczpNo6wfLqMJqGx8ukuXboUBw4cwNmzZ1ukPeP/Yfl0Gwf2Io3R6klJSYGmpiZCQ0Nr3ebu3bvQ1NREcHBwE/aMwag7bKbLaBIaa6ZbUFCAgoICAIC+vj50dHRq1a6iogJZWVkAAJFIhM6dOze4L/922Ey3cWBOl9EksON62h7M6TYObHmBwWAwmhHmdBkMBqMZYU6XwWgifvrpJ9jY2EAikcDIyAgBAQF48uSJUt2QkBDweDzMnTu31vaVtSkoKMCMGTNgYWEBsVgMY2NjzJ49W+V9Gc0Pc7oMRhNw8uRJTJ48GQEBAUhPT0dkZCRiY2Pl8gRXce7cOWzZsgW9evWqtX1Vbe7evYu7d+9izZo1uHjxIsLDw3Ho0CFMnTq1wWNiNBItHSjMpG0K6hB0/yrOzs40a9YsCgoKIm1tbZLJZLRjxw4qKCig0aNHk4aGBllZWVFiYiLXJjs7mzw9PUlHR4ckEgnZ2NjQsWPHuPq0tDRyd3cniURCMpmM/P39qbCwsN59rInVq1dT9+7d5cqCg4PJ2tparuzp06dkaWlJf/75Z61zBNe1zZ49e0gkElFFRUXdB/ISYJsjGkXYTJfRKgkLC0Pnzp1x9uxZTJkyBf7+/pgwYQJGjRqFlJQUWFlZwcfHp8rBIygoCKWlpTh58iQuXLiA5cuXczloCwsL4eLiAgcHB5w7dw4HDhzAtWvXMHnyZJX3z8nJqTGxTXUxwI6OjsjOzkZMTAyICHfv3kVUVBSGDh0qpzdnzhy88847dUpMU9c2jx49go6ODtTU2Ne9VdDSXp9J2xQ0cKY7aNAg7rqkpIT4fD7NmjWLK0tKSiIAdOvWLSIisrGxUXnCw4oVK2jYsGFyZZcvXyYAlJeXp7TN8+fPKSMjo1rJz8+vdhyRkZGkoaFB6urqBIDGjBkjN9vcv38/de/enZ4+fcqNu6ZZa13bPHz4kLp27UqLFy+u1m5tAJvpNoqwLGOMVsnLa5VCoRA6OjqwtrbmyqoSjN+/fx+dO3fGxx9/jKCgIBw+fBhubm7w9vaGpaUlAOD8+fM4fPiw0vPSMjMzlebvVVdXl0sSXlcuXbqETz75BF9++SVcXFxw69YtzJkzB3PnzsW3336LBw8eYNq0afj9998Vzn5TRV3bPH78GO+++y569uyJL774ot5jYTQyLe31mbRNQQNnuq/O3qRSKYWHh3PXt27dIgCUlJTElWVnZ9OGDRvI09OT2rVrR9u2bSMiIg8PD/rggw+UzlaLi4uV9uHmzZukoaFRraxcuVLlGCZOnMgl7anijz/+IDU1NSoqKqK4uDgCQHw+nxMAxOPxSCgUKrVZlzaPHz8mR0dHcnV1lcvi1hDAZrqNImymy2gzGBsbIygoCEFBQQgMDERERAR8fHxga2uLqKgomJqacnlxa8LIyIjL0auK6k6QKC4uhlgslivj8/ncF69v375IS0uTq58yZQp69uypMmystm0eP34Md3d3CIVCREdHK5yvxmhZmNNltAlmz56NYcOGoVu3bnjw4AFOnToFJycnAC9esoWGhmLChAmYO3cuOnTogKtXr+L3339XmUSnocsLw4cPx7Rp0zBo0CC4urpyywuurq7cMkfPnj3l2mhoaEAqlcoto/j4+KBTp05YtWoVNDQ0amzz+PFjDBkyBMXFxdixYwceP37MnYKhr69f6x8dRtPBnC6jTVBeXo7AwEDcuXMH7du3h6enJ0JCQgAAnTp1Qnx8PBYsWIDBgwejrKwMpqamNZ6H1hAmT56MJ0+e4Ntvv8XMmTPRoUMHuLu746uvvqqTnZycnDpFHZw7dw5nzpwBAIUfjaysLJiYmNTp/ozGhyW8YTQJLOFN24MlvGkcWOAeg8FgNCPM6TIYDEYzwpwug8FgNCPM6TIYDEYzwpwug8FgNCPM6TIY/4PH4+HAgQMt3Q1GG4c5XQbjNcHExAQ8Hk9OXo37vXDhAgYOHAiRSIQuXbpg9erVCnb++ecfBAYGwtDQECKRCJaWljh+/HhzDeNfD9scwWC8RgQHB2PKlCnctZaWFvfvqt1ogwcPxpYtW5CWlgY/Pz9IpVL4+fkBAMrKyuDm5gaZTIbffvsNRkZGuHHjhtKkP4wmoqWTPzBpm4JqEt7s3buXrK2tSSgUklQqpSFDhnApDxMTE8nV1ZV0dXVJR0eHXF1d6eLFi1zbrKwsAkB79+4lR0dHEolE9NZbb9HNmzcpNjaWevbsSZqamuTt7U1FRUVcO2dnZ5o5cyYFBgaStrY26evrU3BwsFy/AND+/fu565ycHBo9ejRpa2uTVCql999/n+7cucPVx8bGkr29PYnFYmrfvj0NHDiQ7t+/r3LcDcXY2JjWr1+vsn7Tpk2kq6tLZWVlXNn8+fPlEqdv3ryZzMzM5HRqC1jCm0YRtrzAaFZyc3Mxbtw4+Pv748qVK4iNjZVL7P3kyRNMmTIFCQkJiI+Ph0wmw/Dhw1FaWipnZ/HixVi2bBmSkpJQUlKCsWPHYuXKlQgPD0dMTAzi4uKwbt06uTZhYWHQ1NTE33//jZCQECxbtgx79+5V2s/nz5/D3d0durq6OHXqFI4fPw4iwogRI1BZWYny8nKMGjUK77zzDi5evIj4+Hj4+PhUO3Zra+tqk6K/muBcGStXroSenh5sbW3xzTffoLy8nKs7ffo0Bg0ahHbt2nFl7u7uSE9P5/IvREdHw9HRkVtesLGxwdq1a6t+KBnNQUt7fSZtU6BippucnEwAKDs7W2n9q5SUlJBAIKCTJ08S0f/PdCMiIjidH374gQBQSkoKVxYUFESurq7ctbOzM/Xq1UvOdlBQEDk6OnLXeGmm+9NPPykcrfPkyRPi8/l05swZys/PJwByRwLVRHZ2drVJ0W/fvl1t+2+++Ybi4uLo/PnztGnTJmrfvj3NmzePq3dzc6PAwEC5Nunp6QSArl27RkREFhYWJBAIaOrUqZScnEw//fQTaWtr07p162rsP9hMt1GErekympXevXtj0KBBsLGxwdChQ+Hu7o7Ro0dDW1sbAJCXl4fPPvsMx48fR15eHiorK1FWVoacnBw5Oy8nOa9KaP5qkvMTJ07ItenXr5/ctaOjI3bu3Km0n+fPn8eVK1cUEp9XVFQgMzMTDg4OmDhxIjw8PODm5oYhQ4bA29u72rVRY2NjlXW14ZNPPuH+3atXLwiFQgQEBGDlypVys9vqqKyshEwmw+bNm8Hn89GnTx/cuHEDW7ZsUXpoJqPxYcsLjGaFz+fj6NGjOHjwIMzNzbF69WpYWVkhLy8PAODr64u0tDSsX78eiYmJSE1NhUQiQVlZmZydl50Mj8dTWlZZWVnvfhYVFaFfv35ITU2Vk4yMDHh6egJ4ccR6fHw8HBwcsG3bNnTv3h3p6ekqbTbG8sLL9OvXD8+fP+d+kGQyGfc5VlF1XfXDJJPJ0K1bN7kUj5aWlgo/aoymg810Gc2OmpoaBg4ciIEDB2LJkiUwMDBATEwMfHx8cOrUKYSGhsLDwwMAcOXKFRQXFzfKff/++2+568TERO5In1extbXFvn37YGhoKBch8Cp2dnaws7PDZ599Bmtra/zyyy9yM+6XOXToEJ4/f67S1qtJz2siNTUVfD4f+vr6AF7M3D///HM8f/6c+wE6fPgwrK2tub8k+vfvj927d6OyspJLGXnt2jV07dq1TvdmNICWXt9g0jYFKtZ0ExMTaeXKlZSUlETZ2dm0fft24vP53HHqtra25OHhQZcvX6ZTp06Ro6MjCQQC7qieqjXdtLQ0zub+/fvp1futWLFCbk3W2dmZNDU1af78+XT16lWKiIggkUhEkZGRnA5eWtMtKiqiN954g1xdXenkyZN048YNOnr0KAUEBNA///xDN27coAULFlBCQgLdvHmT9u/fT5qamrR7926l424oCQkJtHbtWkpNTaXMzEzasWMH6evrk5+fH6dTWFhIhoaGNGnSJLp48SJFRkaSRCKhH3/8kdPJzs4mDQ0Nmj17Nl29epV+//136tChA3333Xc19gFsTbdxvhst3QEmbVNUOd1Lly6Ru7s76enpkUgkIktLSwoLC+Pqz507R/b29iQUCsnCwoKio6PlzkdriNOdOXMmTZ06lbS0tEhPT4++/PJLuTZ4JWTszp07NHHiRJJKpSQUCsnc3JymT59Oz549o3v37pGXlxfJZDISCARkZmZGX331ldIxNwbJycnUr18/0tHRIbFYTFZWVvTVV19RaWmpnN758+fJycmJhEIhderUiUJCQhRsxcfHU9++fbkxrVmzRu6UYlUwp9s4wpKYM5qE1pbEfNCgQbC3t8eaNWtauiuvLSyJeePAXqQxGAxGM8KcLoPBYDQjbHmB0SS0tuUFRsNhywuNA5vpMhgMRjPCnC7jtSI7Oxs8Hg8XL15s6a4wGPWCOV0GowlYtWoV+vbtCy0tLRgYGOC9995DRkYGV19QUIAZM2bAwsICYrEYxsbGmD17Np48eVKt3cmTJyvk1K3aSFLFqFGjYGxsDJFIhI4dO2LSpEm4e/cuV5+RkQEXFxcun66ZmRm3qYLR9LAdaQxGE3D8+HEEBQWhb9++KC8vx6JFizBkyBBcunQJYrEYd+/exd27d7FmzRpYWVnh5s2bCAgIwL179xAZGVmtbU9PT/zwww/ctVAolKsfNGgQPv30UxgZGeHu3buYO3cuvL29ER8fD+DFdumJEyfCzs4O7du3R1paGj788EMAwJdfftnInwRDgZYOFGbSNgVKNkds2LCBunTpQpWVlXLlb7/9Nv3nP/8hotrn063aHBEeHk5SqVTO3vr168nY2FiuLDQ0lLp3705CoVBhQ0ZzcP/+fQJAp06dUqmzZ88eEolE1W5U8PX1pffff79O946KiiI+n1+t3f/85z80aNCgau2AbY5oFGHLC4xmw9vbG7m5udyMCwBu376NkydPYvz48QBqn0+3LuzcuRPLli1DSEgILl++jKVLl2LWrFmIiopS2SY4OLja5DSampp1ShLz6NEjAICurm61Ojo6OlxOBFUcPXoUBgYGsLCwQFBQEAoKClTqFhQUYOfOnXByclJp9/r16/jzzz/h7Oxci5EwGkxLe30mbVOgYhvw0KFD5XK+fv3119S9e3elukSq8+nWZab7xhtv0J49e+R05s+fT0OHDlV53/z8/Gpz32ZkZNDz589Vtn+ZyspK8vT0JGdnZ5U6Dx8+pK5du9LixYurtbVr1y6KioqiCxcu0G+//UaWlpb01ltvKcxiP/30U5JIJASAHB0dKT8/X8GWo6MjCYVCAkABAQEKf4G8CthMt3G+Gy3dASZtU1Q53R07dpCenh7nsN58801aunQpV3/v3j3y9/cnc3Nz0tLSIg0NDQJAO3fuJKK6O92ioiICQGKxmDQ0NDgRCATVOvvGZPr06WRiYiJ31M/LPHr0iPr160fDhg2rtSOvIjMzU2ky9QcPHtDVq1fpr7/+ogEDBtCIESMU2ubk5FB6ejr9/PPP1KlTJ1qzZk2192JOt3GEvUhjNCsjR47EtGnTcPjwYZiamiI1NVXuyBxfX1/8888/WL9+Pbp27QqBQIDevXsr5NOtQk1NrcrJc7z8Fr6oqAgAEB4eDjs7Ozm96hJ/BwcHIzg4uNqxXLp0qcaUiDNmzEB0dDROnDgBIyMjhfonT57Aw8MDmpqa+OWXX6CuXrevpJmZGfT09HD9+nW55QE9PT3o6emhe/fusLS0RJcuXZCUlIS+fftyOl26dAEAWFlZoaKiAoGBgfjkk0+4/MSMpoE5XUazoqGhAS8vL+zatQvGxsbo27cvzM3Nufq65tPV19dHYWEhSkpKIBKJALw49aEKQ0NDdOzYETdu3MCYMWNq3c+AgAB4e3tXq6PMiVZBRJgxYwZ+++03HDt2DKampgo6jx8/hru7O4RCIaKjo7n+14Xbt28jPz8fHTt2rLYvAKpdFycilJWVgYiY021qWnqqzaRtCqo5DXj//v2kpaVFpqamCmdz1TWf7sOHD0kikdDcuXMpIyODQkNDSSqVyq3pbt68mTQ0NGj9+vV09epVOn/+PH3//fe0adMmlX1sKIGBgaSjo0PHjh2j3NxcToqLi4no/5cUbGxs6Pr163I65eXlnB0LCwv69ddfiejFGW1z586l06dPU1ZWFh05coT69OlDPXr04FI8JiUl0X//+19KSUmh7Oxsio2NJScnJzI3N+d0fv75Z4qMjKRLly5RZmYm7dmzhzp16kQ+Pj7VjglseaFxvhst3QEmbVOqc7plZWUklUqJz+dTbm6uXF198unu3buXTE1NSSKR0Lhx42jVqlUKIWPbt2+nXr16kUAgIKlUSi4uLvTnn3+q7GNDAaBUqsYRFxenUicrK0vOTlWb4uJiGjJkCOnr65NAICATExP66KOPKC8vj9O/ePEiubi4kK6uLgmFQjIxMaGAgAC59eS9e/dSnz59SEtLiyQSCVlZWVFwcDA9e/asxjFRK3i2XndhCW8YTQJLeNP2YAlvGgcWp8tgMBjNCHO6DAaD0Ywwp8tgMBjNCHO6DAaD0Ywwp8tgMBjNCNsc0YiIxeJ7JSUlhi3dDwaD0XphTrcRKSkpMWRhUi8wMTFhO5vaGCKRKK+l+9AWYHG6jQiLTW17tPbYVB6Pp4YXmyo2AegB4AiAjkT0cYt2jKEStqbLYLym8Hi8NwDkAwgDsBtAHgBXAO/y2J8ZrRY2021E2Ey37fEazHRlACYCmAJAiBezXmMAvYnockv2jaEcNtNlMF5jiOgeEa0B0BPAOABxePG9DmjRjjFUwma6jQib6bY9WvtMVxk8Hk8AoIKIKlq6LwxF2Ey3FbJ06VLueO0NGzY0qu0dO3ZwtufOnduothmtAyIqYw639cKcbjNSUVEBBwcHhWTa9+7dg1Qqxdq1a7my3r17Izc3F35+flxZaGgoBg0aBG1tbfB4PO5UhCrOnz+PcePGoUuXLpBIJLCyssL69evldEaPHo3c3Fw4Ojo2wQjrT01je5nMzExoaWlBT0+v1vbr06YxEIvF93g8HjFpWyIWi+/V95lgTrcZ4fP5iIiIQFRUFPbt28eVT5s2DdbW1pg1axZXpq6uDplMBolEwpUVFxfDw8MDixYtUmo/OTkZBgYG2LFjBy5evIhFixZh/vz52LJlC6cjEokgk8kgEAgaNJbHjx+jsLCwQTZepqaxVVFeXo4JEybAycmp1rbr06axqIrdZtK2pCGboNjmiGbGysoKy5Ytw/Tp0+Hs7IxDhw4hNjYW58+fr/Ho7dmzZwMAjh07prT+5Vkx8OL8rNOnT+PXX39FQEDD36tUVlYiNjYWERER+O2333Do0KFGO7a7prFVsXz5cpiZmWHIkCFISkqqle36tGEwmgo2020B5s6dCzMzM/j4+GD27NkICQnBG2+80ST3evToEXR1dRtk4/r16/j8889hYmKC9957DwKBAIcOHcLbb7/N6QwdOhSampoqxdrauqFDQUJCArZv346NGzc2aRsGoylhM90WgM/nY+PGjbC3t8eAAQMwffr0JrnP6dOnsWfPHvz55591bltUVITIyEhEREQgMTERrq6uWLVqFUaNGiW35FHFjz/+iGfPnqm0V93Ju7XhyZMnmDRpEn744Qd06NChydowGE0Nc7otRFhYGCQSCa5evYoHDx7AwMCgUe2np6fDy8sLy5Ytg4uLS53b79u3D1OnTkXPnj1x4cIFWFlZVavfqVOn+na1VsycORPvvvsu3NzcmrQNg9HUsOWFFiAuLg6hoaGIiYlBt27dEBQU1Kj2L126BBcXF0ybNg0LFy6slw0vLy+sXbsWfD4fdnZ28Pb2xoEDB1BeXq5Uv6mXF+Li4rBp0yaoq6tDXV0d/v7+yM/Ph7q6On7++edGa8NgNDVsptvMFBUVwc/PD7Nnz4aTkxPCwsJga2uLffv2YfTo0Q22n56eDhcXF/j6+uLLL7+st50OHTpg9uzZmD17NlJTUxEeHo4pU6ZATU0N48aNg4+PD/r06cPpN/Xywl9//YWysjLuOioqCmvWrMHJkyfRpUuXRmvDYDQ5LR160ZYE1Rw7XkVAQABZWFjIHXcdEhJCBgYG9PDhQyIiWrJkCdnZ2Sm0zc3NpZSUFPrhhx8IACUkJFBKSgo9efKEiIjS0tJIX1+fJk6cSLm5uZzcv39fwZazszPNmTOnxv6+TFlZGf366680YsQIateuHZ04caJO7aujprG9Snh4OEmlUrmyX3/9lSwsLFTeQ1mbmkADjx2vzTPRlliyZInSY+UlEgmnU1paSl988QWZmJiQSCSivn37Unx8fLV2fX19ldq1srLidIKDg8ne3p40NTVJX1+fRo0aRdeuXZOzc+3aNRo+fDjp6upS+/btydfXlwoLC+s8zoY8Fy3uqNqS1PQFO3LkCKmrq9Pp06flysvLy8nBwYHGjh1LRKqdrqoHOi4urtp6Y2NjBVv1cbovc//+fe5HojGoaWyvosyBhoeHU3X/B8zpNj1PnjyR+8HPzc0lKysr8vX15XTmzZtHnTt3pj///JMyMjJoxYoVpKGhQTdv3lRpt7CwUM7mrVu3SFdXl5YsWcLpuLu7U7jvOeIAACAASURBVHh4OF28eJFSU1Np2LBhZGJiQsXFxUREVFRURCYmJvTBBx9Qeno6JScn08CBA8nT07PO42ROt5VIY33BVDndxqShTvffQnM6XWdnZ5o1axYFBQWRtrY2yWQy2rFjBxUUFNDo0aNJQ0ODrKysKDExkWuTnZ1Nnp6epKOjQxKJhGxsbOjYsWNcfVpaGrm7u5NEIiGZTEb+/v71mtnVl9TUVAIg91dRx44dacuWLXJ6dnZ2tGDBglrb/e2334jH41F2drZKnfv37xMAOnXqFBERxcTEEJ/Pp6KiIk4nLS2NANCVK1dqfW+ihj0X7EVaKyUlJQWampoIDQ1tVLu7d++GpqYmTp482ah2GY1DWFgYOnfujLNnz2LKlCnw9/fHhAkTMGrUKKSkpMDKygo+Pj5VDh1BQUEoLS3FyZMnceHCBSxfvhwikQgAUFhYCBcXFzg4OODcuXM4cOAArl27hsmTJ6u8f05OTrUvRDU1NREcHFzr8fz444/o3r07Bg4cyJWVlpZyfaxCLBYjISGh1na3bt2KwYMHw9jYWKXOo0ePAICLUy8tLYWamprcbkyxWAwAdbp3g6mvt2bSdDPd/Px8ysjIoIyMjEaflTx58oSz3ZjLA20VNPNMd9CgQdx1SUkJ8fl8mjVrFleWlJREAOjWrVtERGRjY0NLly5Vam/FihU0bNgwubLLly8TAMrLy1Pa5vnz59zzoUry8/NrNZ5nz55Rhw4dKCQkRK58zJgxZGNjQ5mZmVReXk67du0iNTU16t69e63s3rlzh/h8Pu3evVulTmVlJXl6epKzszNXdv/+fdLS0qI5c+ZQcXExFRYWkre3NwGg4ODgWt27ioY8Fy3uqNqSNJbTZbQemtvpzpw5U65MV1eXQkNDueucnBwCQMnJyURE9P3335O6ujoNGDCAli5dSpcuXeJ0R48eTe3atSMNDQ05wf9eVDY1P//8M6mrq9O9e/fkyvPy8ujdd98lHo9HfD6f7O3tacKECdSjR49a2Q0ODiapVEqlpaUqdaZPn04mJiZ0584dufI//viDjI2NSU1Njdq1a0effPIJGRoa0ldffVWnsTXkuWDLCwxGK+LV0DoejydXxvvfKTyVlZUAXiRLun79OsaNG4ezZ8+id+/e2L59O4AX4YkjR45EamqqnGRkZODNN99Uev/GXF748ccf4enpCUND+dwwBgYGOHDgAJ4+fYpbt24hKSkJ5eXlMDU1rdEmESEsLAyTJk1SmbRpxowZiI6ORmxsLIyMjOTqPDw8kJ2djdzcXOTn52P58uV48OBBre7dWLA4XQbjNcfY2BhBQUEICgpCYGAgIiIi4OPjA1tbW0RFRcHU1BR8Pr9WtoyMjJCamlqtTm1yeWRlZSEuLg7R0dEqdcRiMcRiMQoLCxETE4NVq1bVaPf48eO4fv06/P39FeqICDNmzMBvv/2GY8eOVetIq3aAbtu2DUKhsHl3LdZ3isyELS9UcerUKeLz+UojLjZs2EDGxsYkFAqpX79+9Pfff1drKzQ0lJycnKh9+/bUoUMHcnNzo6SkJDkdKAktA0Bff/01pzNy5Ejq2rUrCYVCkslkNHHiRIU/NWsDmnl54dWIEqlUSuHh4dz1rVu3CAD3mcyaNYtiYmLoxo0bdObMGbKxsaHAwEAiIrp9+zZJpVIaM2YMJSUl0fXr1+ngwYM0derUOn8OdeXzzz8nIyMjKi8vV6hLSEig33//nTIzM+no0aPUp08fsre3p7KyMk5nwYIFNGnSJIW2EydOpH79+im9Z2BgIOno6NCxY8fkwsuqQsaIiLZu3UqJiYl07do12rJlC0kkEvr222/rPL6GPBct7qjakvwbne7jx4/J3Nyc3NzcFJxuZGQkCQQCCgsLo/T0dJo6dSp16NCBHjx4oNLe+PHjaePGjZSSkkKXL1+myZMnU/v27enu3buczqtxoGFhYcTj8SgzM5PTWbduHSUkJFB2djYlJCRQ//79acCAAXUeX2t3ukFBQWRmZkZCoZAMDQ3J39+fHj9+zOlfvnyZvLy8SEdHh8RiMVlZWdGiRYvq9iHUkYqKCurcubPK+xw9epQsLCxIIBCQvr4+ffTRR1RQUCCn4+vrK/cSjOhFrK5YLJZb434ZVT/GL39+c+bMIX19fRIIBGRlZUVbt26t1xiZ020lUtsvWFuKx5w0aRItXLhQaWyxg4MDffzxx9x1RUUFGRkZ0erVq2ttv7y8nLS0tGjnzp0qdby8vMjFxaVaO1FRUcTn86mioqLW9yZqXqfLeH1oyHPBXqS1EG0hHnP37t1IS0vD0qVLFerKysqQnJyMIUOGcGVqamoYPHgwTp8+XevPqbi4GM+fP1e5jpiXl4eDBw8qXeOroqCgADt37oSTk1ONieIZjCanvt6aScNmuq97PGZOTg7JZDK6ePEiESnuortz5w4BoDNnzsi1mzdvHvXv31+l3VeZPn06devWTWV4UEhICHXo0EEul0UVn376KUkkEgJAjo6OtY4vfRmwmS5DCQ15Llj0QgvRq1cv7t9CoRA6Ojpy6Q+rwmzu37+Pzp074+OPP0ZQUBAOHz4MNzc3eHt7w9LSEsCLAykPHz4MTU1NhftkZmYqzdWrrq4Oc3Pzevffx8cH8+bNa5QTIVTx9ddfIzIyEsePH1cZHhQWFoYJEyYo7HACgHnz5sHf3x83b97EsmXLMGXKFERFRTVZfxmMWlFfb82kYTPdur4wIXqxrrthwwby9PSkdu3a0bZt24iIyMPDgz744AOls9WX39y+zM2bNxWC5l+VlStXqhwDAOLz+ZzweDyuLD4+nkpLS4nP51N0dLRcOx8fH3rvvfdq/IxWr15NOjo6CpELL3PixAkCQKmpqTXaq/o8a4qeeBWwmS5DCQ15LthM9zWiNcVjpqWlyV1v2rQJJ06cQGRkJMzMzCAQCGBnZ4fDhw9j+PDhAF4E9B89epQ7hFIVX3/9NVauXImYmBjY29ur1Nu6dSvs7OzQu3fvau0BqHKAKC0trVGXoQiPx8P+/fvh6enZ0l15/amvt2bSvDPd1hqPWYWy6IXIyEgSCoUUERFBly5domnTpimEjE2aNEkuu9RXX31FAoGA9u3bJxcW9mpe3UePHpFEIqHNmzcr9CUpKYn++9//UkpKCmVnZ1NsbCw5OTmRubl5tVtHlQE20yWiF5/D/v37W7obSomLi1MZLvbyXzYVFRUUEhJC5ubmJBAIqFOnTrRu3bp63bMhzwWb6b4mlJeXIzAwEHfu3EH79u3h6emJkJAQAC/OJ4uPj8eCBQswePBglJWVwdTUFCNHjmzRPo8ZMwYPHjzA4sWLce/ePbz55pv4888/oaenx+nk5OTIRRRs3rwZZWVlCqdoLFmyRC5KIjIyEkSEcePGKdxXLBYjKioKy5Ytw9OnT9GxY0d4eHhg9+7dKteGGa8v/fv3R25urlzZF198gaNHj8r9pTR79mwcPnwYa9euhbW1NR4+fIinT582d3fZTLcxBW1kVsP4f9DMM929e/eStbU1CYVCkkqlNGTIEC62ODExkVxdXUlXV5d0dHTI1dWVix4hIsrKyiIAtHfvXnJ0dCSRSERvvfUW3bx5k2JjY6lnz56kqalJ3t7ecjllqxLtBAYGkra2Nunr6ytk3cIrM92cnBwaPXo0aWtrk1Qqpffff19ux19sbCzZ29uTWCym9u3b08CBA5WeYNIUlJWVkb6+Pi1fvpwru3TpEqmrq9c5b64qGvJcsKBFBqOVkJubi3HjxsHf3x9XrlxBbGwshg4dytU/efIEU6ZMQUJCAuLj4yGTyTB8+HCFderFixdj2bJlSEpKQklJCcaOHYuVK1ciPDwcMTExiIuLw7p16+TahIWFQVNTE3///TdCQkKwbNky7N27V2k/nz9/Dnd3d+jq6uLUqVM4fvw4iAgjRoxAZWUlysvLMWrUKLzzzju4ePEi4uPj4ePjU+3Yra2tq40Zf/lzqIno6Gjk5+djypQpXNn+/fthZmaG6OhomJiYwMzMDNOnT8eTJ09qbbfRqK+3ZsJmuv8G0Iwz3eTkZAJQ7WkIL1NSUkICgYBOnjxJRP8/042IiOB0qs6cS0lJ4cqCgoLI1dWVu3Z2dqZevXrJ2Q4KCiJHR0fuGi/NdH/66SeytraW03/y5Anx+Xw6c+YM5efnEwC5HZM1kZ2dXW3M+O3bt2tta+jQoTR06FC5so8++ogEAgE5OjpSfHw8HTlyhKysrGjkyJG1tvsyDXku2Joug9FK6N27NwYNGgQbGxsMHToU7u7uGD16NLS1tQG82H332Wef4fjx48jLy0NlZSXKysqQk5MjZ+flGPCqeO9XY8BPnDgh16Zfv35y146Ojti5c6fSfp4/fx5XrlxRiAuvqKhAZmYmHBwcMHHiRHh4eMDNzQ1DhgyBt7e30njxKqo7AaIu3L59GzExMdizZ49cedVntX37di4+ffPmzXB2dkZeXp5C+smmhC0vMBitBD6fj6NHj+LgwYMwNzfH6tWrYWVlhby8PACAr68v0tLSsH79eiQmJiI1NRUSiUTumHkASvPvvlpWlY+3PhQVFaFfv35K8/RWhZT99NNPiI+Ph4ODA7Zt24bu3bsjPT1dpc3GWl4IDw+HVCrFiBEj5MplMpnChqCqzUWv/mg1NWym+y+HxV+2LtTU1DBw4EAMHDgQS5YsgYGBAWJiYuDj44NTp04hNDQUHh4eAIArV66guLi4Ue77999/y10nJiZyTulVbG1tsW/fPhgaGkJLS0ulTTs7O9jZ2eGzzz6DtbU1fvnlF5U7GA8dOoTnz5+rtFV1lll1EBHCw8Ph4+OjkAy+f//+KC8vR1ZWFpdn99q1awCArl271mi7MWEzXcZrQUFBATp37gwej4eioiKuPCEhAQMGDIBUKoVYLEaPHj2wdu1ahfYbN26EiYkJRCIR3nrrLSQlJTVn92vFmTNnEBwcjLNnz+LmzZvYvXs3ioqKYGFhAQDo1q0btm/fjitXriAhIQF+fn6NFgKXmZmJBQsW4Nq1a9i2bRt+/PFHzJo1S6nuhAkToKOjg1GjRiE+Ph5ZWVmIjY1FYGAgCgsLkZWVhYULF+L06dPIycnBwYMHcevWLfTo0UPl/Y2NjWFubq5SOnXqVOMYYmNjkZWVhQ8//FChzs3NDb1794a/vz/Onz+Pv//+Gx9//DHee++9Zl1aAMBepDWm4DV8kYZWHPT+Mu+//z55eHgQALmNEqmpqbRz505KS0ujrKws2rlzJ2loaNAPP/zA6dQnr28VaMYXaZcuXSJ3d3fS09MjkUhElpaWFBYWxtWfO3eO7O3tSSgUkoWFBUVHR8ttqql6kZaWlsa12b9/P73ahxUrVsi9CKsKGZs6dSppaWmRnp4effnllwqfw8vPyZ07d2jixIkklUpJKBSSubk5TZ8+nZ49e0b37t0jLy8vkslkJBAIyMzMrM5nkNWHcePGVZtM6fbt2zRy5EjS0NAgmUxG06ZNo0ePHtXrXg15LlrcUbUlqe4LxuIv68/WrVvJycmJjh49quB0lTFq1CiaPHkyd92QvL7N6XRbCmU7JBnV05Dngi0vNAMs/rL+L0gyMzPx2WefYdu2bbXKhZuSkoKEhAQ4OzsDaLy8vgxGo1Ffb82k9rMaFn9Zv/jL58+fU79+/bilgqo99spmup06dSKBQEBqampyfxo3NK8v2EyXoYSGPBcseqEZYPGX9WPlypUwMDBQ+mLkVU6ePImioiIkJiZiwYIF6NatG7y9vet9738Tx44da+ku/KtgywvNAIu/rN/yQlxcHA4ePAh1dXWoq6vD1dUVANC+fXuFo4RMTU1hY2ODqVOn4j//+Q+WLVsGANDT0wOfz+c+6yry8vIgk8nq9TkxGA2BzXSbCRZ/qZzq4i/Dw8PlskAlJSXBz88Pp0+fhomJicp2RMSthzckr29rJzs7G6ampkhLS0PPnj1bujuMWsJmus0Ai7+sX/ylqakpevbsyUlVULulpSX09fUBvNjKGR0djWvXriEjIwPh4eFYs2YNJkyYwNn55JNPEBoaim3btuHy5csIDAxEcXFxtQd3MhqP0NBQ9OrVC9ra2tDW1oajoyP++OMPrj47Oxs8Hk+pVPfSd/78+bCxsYGGhgY6deqEKVOm4N69e3J6o0aNgrGxMUQiETp27IhJkybh7t27cjo//fQTbGxsIJFIYGRkhICAgKZNhFPfxWAmtX9pwuIvGwdlL9I2btxI1tbWJJFISFtbm2xtbWnTpk0KR62vX7+eunbtSgKBgBwcHBRerKkCrfhFmrLnojWyf/9+OnDgAF27do2uXr1Kn3/+ObVr144uXbpERETl5eVyCetzc3Np2bJlpKmpqTI8sLCwkAYPHky7d++mK1eu0OnTp8nBwYH69esnp7du3TpKSEig7OxsSkhIoP79+9OAAQO4+hMnTpCamhpt2LCBbty4QcePH6du3brRlClTqh1TQ56LFndUbUma8gtWH9hb6YbTVE53w4YN1KVLF6qsrJQrf/vtt+k///kPEdU+frvK6YaHh5NUKpWzt379ejI2NpYrCw0Npe7du5NQKFSYADQXHTp0kIvGeZU333yT/Pz86mTz77//JgByceWvEhUVRXw+n/tRXr16NXXv3l1OJzg4WCGK51Ua8lyw5QUGowXw9vZGbm4u4uPjubLbt2/j5MmTGD9+PIDax2/XhZ07d2LZsmUICQnB5cuXsXTpUsyaNavaU5KDg4OrfRmqqalZ66QxFRUViIyMRHFxMd566y2lOsnJyUhNTYW/v3+dxvbo0SPw+Xzo6OgorS8oKMDOnTvh5OTExXw7OjoiOzsbMTExICLcvXsXUVFRdcrfW2fq662ZsJnuvwE04fLC0KFDuXPuiIi+/vprhVnXy6iK367LTPeNN96gPXv2yOnMnz9fIf/sy+Tn51cba52RkUHPnz9X2Z6I6MKFC6ShoUF8Pp90dHTojz/+UKkbGBhIlpaW1dp7lWfPnpGdnR35+Pgo1H366ackkUgIADk6OlJ+fr5cfWRkJGloaJC6ujoBoDFjxigsT71KQ56LFndUbUlam9NlNJymdLo7duwgPT09zmG9+eabtHTpUq7+3r175O/vT+bm5qSlpUUaGhoEgHbu3ElEdXe6RUVFBIDEYjFpaGhwIhAIqnX2jUFpaSllZGTQ2bNnacGCBaSvr6/06Jzi4mLS0dGhNWvW1Np2WVkZeXl5kZ2dHT1+/Fih/sGDB3T16lX666+/aMCAATRixAiuLj09nYyMjGjt2rV0/vx5OnDgAFlYWHBLPKpgTreVCHO6bY+mdLpFRUUkkUjo0KFDdPnyZQJAGRkZXL27uzs5ODjQH3/8Qenp6ZSRkUESiUTlC9Zt27aRrq6u3D2+/fZbzuneu3ePAFBkZKTCTLW63ZIrV66Uc9LK5ObNm7X8RF/g6upK06dPVyjfvn07tWvXrtb5PMrKymjkyJHUq1cvhRmsMqpO2a46JXjixIk0duxYOZ0//viD1NTU5PKYvEpDngsWp/sawOIx2yYaGhrw8vLCrl27YGxsjL59+8ol2a5r/La+vj4KCwtRUlICkUgE4MUuwyoMDQ3RsWNH3LhxA2PGjKl1PwMCAmrc3WdkZFRre4B8LPXLbN26FSNGjOBCAqvj+fPn8Pb2RkZGBuLi4qCrq1ur+wLg7l1cXKwQK87n81/+0Wx86uutmTTfTPd1CQ06fvw4eXp6UseOHVWmjLx27RoNHz6cdHV1qX379uTr60uFhYXV2s3NzaUJEyaQoaEhaWhokJeXl0LOhpEjR1LXrl1JKBSSTCajiRMnqnyLferUKeLz+WRnZ1fjmNDEIWP79+8nLS0tMjU1pXXr1snV2drakoeHB12+fJlOnTpFjo6OJBAIVM50Hz58SBKJhObOnUsZGRkUGhpKUqlUbk138+bNpKGhQevXr6erV6/S+fPn6fvvv6dNmzbV+FnUl0WLFtGxY8coKyuL0tLSaNGiRaSmpkZHjhyR08vIyCAej6dyvdfCwoJ+/fVXInoxwx0xYgR17tyZUlNT5cLNSktLiYgoKSmJ/vvf/1JKSgplZ2dTbGwsOTk5kbm5OacTHh5O7dq1ox9++IELGbOxsaHBgwdXO6aGPBct7qjakvzbne6hQ4fos88+o19//VWp0y0qKiITExP64IMPKD09nZKTk2ngwIHk6emp0mZlZSX169ePnJ2dKTk5mS5evEjvvfce9e7dm8rLyzm9muIxq3j8+DGZm5uTm5tbq3C6ZWVlJJVKic/nU25urlxdfeK39+7dS6ampiSRSGjcuHG0atUqhZCx7du3U69evUggEJBUKiUXFxf6888/a/ws6su0adPI2NiYBAIB6evrk6urK/31118KegsXLqQuXbqofIkFQGHsyiQuLo6IiC5evEguLi6kq6tLQqGQTExMKCAgQOHH+LvvviNLS0sSi8VkZGREU6ZMoby8vGrHxJxuKxFlX7B/azymMqcbExNDfD5fbq0sLS2NACh9qUJEdPXqVYX6R48eEY/Hq9ZRvBqPWcWkSZNo4cKFtGTJklbhdBmvJw15LlicbhPzb43HVEZpaSnU1NTktjhXraclJCSobAOAW6MEAKFQCDU1NZVtlMVjAsDu3buRlpaGpUuX1nsMDEZDYS/Smhh9fX24ublh165dGDhwIABg165d6NatG+zt7QEAgwcPlmuzdetWaGtrIykpCU5OTvW675IlS7B27VqMHDkSwIs8BufOncP3338PLy8vpW2a4oXJy7z11lsQiURYuHAhVqxYgbKyMixatAgAFPbMV9GjRw906dIFCxcuxObNmyEQCLBw4UJUVFQotJk/fz42bNiA4uJiODo64sCBA1zdrVu3MHv2bBw5cqTR8lowGPWBzXSbgQkTJmDv3r0oLy8HAPz888/cLBd4kWbwww8/RLdu3aCtrQ2pVKo0n25tefr0KTIzM+Hr6ys3S127di0yMzNVttPV1a02OY25uTnU1ev/O62vr489e/Zg37590NTUhL6+Pjp37gxDQ0OVp0K0a9cOv/zyCy5cuIAOHTpAW1sbDx48QJ8+fRTazJs3DykpKfjrr7+gpqaGKVOmcHU+Pj6YN2+eymxoDEZzwWa6zcDIkSMxbdo0HD58GKampkhNTZXLnuTr64t//vkH69evR9euXSEQCNC7d2+FfLpVqKmpVa0XcrycPrHqtNzw8HDY2dnJ6b16NPXLBAcHK+SpfZVLly416MhqDw8PZGdn4/79+xCLxVBTU8O6deu4DGLK6Nu3Ly5evIh//vkHRARdXV3IZDKFNnp6etDT00P37t1haWmJLl26ICkpCX379sWxY8dw8uRJfPrppwBepHckIqirq+P48eMYMGBAvcfEYNQF5nSbgX9zPKYqqk6b2LZtG4RCIdzc3Gps06FDBwDgTtioyo+rjKofpao14bS0NLn6TZs24cSJE4iMjISZmVm9xsBg1AfmdJuJ8ePHY/z48dDT01PIZ1uVT9fW1hYFBQWYO3duteuODg4OEIlE+OKLL/DRRx8hLi4OBw4ckDtmZ/HixZg7dy60tLQwZMgQlJSUIDExERUVFQgMDFRqV1dXt1YB5qooKirC9evXueusrCykpqZCJpNxpzSEhYXB2toaurq6iI2NxSeffIKVK1dyDhV4sY67atUqjBo1CgCwZ88eyGQydOrUCUlJSZgxYwZmzpzJJWM/e/YsEhIS8Pbbb6NDhw64ceMGFi9eDHNzczg4OACAwqYSAwMDiEQittmE0fzUN+yBSd3Cg/4N8ZhV+W5flSVLlnA6c+bMIX19fRIIBGRlZUVbt25VsIOX4jGJiL755hsyMjKidu3akZmZGYWEhMiFgtU2HvNlWMgYoyE05Lng0Strg4z6w+PxiH2ebQsejwci4tW3vVgsvldSUmLYmH1itDwikSjv2bNn9TpkjzndRoQ53bZHQ50ug/EqLGSMwWAwmhHmdBkMBqMZYU6XwWAwmhHmdBkMBqMZYU6XwWAwmhHmdBkMBqMZYTvSGhGRSJTH4/FYTGYbQiQS5bV0HxhtCxan24rh8XghADQAvAfgYwDxAMYCiCeicy3Zt5aCx+NJAQQC2IcXO96OAfgegBsRsaw1jFYPW15o3YwA8D6AXwD4ALgGwA5AQUt2qoUpBqANIA5ABICtAIIA9P6fQ2YwWjXM6bZSeDxeDwA9AGgBeBNANIAuRORLRNkt2beWhIieEdGnALoAWIkXn5EYL/4i+LAl+8Zg1Aa2ptt6qcCLGe5CIspo6c60NoioHMABAAd4PJ4egMUAbrdsrxiMmmFrugwGg9GMsOUFBoPBaEYaZXmBpa9rOzQkZd3LsGei7dFYz8a/nUZZXmApDdsOjZXKkD0TbQ+W5rJxYMsLDAaD0Ywwp8tgMBjNCHO6DAaD0Ywwp8tgMBjNCHO6DAaD0Ywwp1sHjh07Bi8vL3Ts2BEaGhqwtbVFZGSkgt4///yDwMBAGBoaQiQSwdLSEsePH6/Rdp8+fSAUCmFubo7t27cr6GzcuBEmJiYQiUR46623kJSUJFefkZGBESNGQCqVokOHDpg8eTIePXrUsEEzaqSkpASTJ0+GjY0N1NXVMXr0aAWdsrIyLF68GKamphCLxXBwcMCpU6eqtfvrr7/Czc0N+vr60NbWxoABA3D48GE5HRMTE/B4PAUJCgqS00tISMCgQYMgkUigo6ODd955p+EDZ9QL5nTrQEJCAnr16oVffvkFFy5cwOTJkzFhwgT88ccfnE5ZWRnc3Nxw69Yt/Pbbb7hy5Qo2btwIAwMDlXazsrLw7rvv4p133kFqaipmz54NPz8/HDlyhNPZvXs3PvnkEyxZsgTnzp1Dr1694O7ujocPHwIAnj59iiFDhkAkEuHkyZM4evQobty4gYkTJzbdB8IAAFRUVEAsFmPmzJkYPHiwUp3PP/8cBIhbEwAAGehJREFU4eHh2LJlC9LS0jBixAi4u7sjJydHpd0TJ07Azc0Nhw4dQnJyMpydneHp6YkLFy5wOklJScjNzeWkyil/8MEHnM7p06cxdOhQDB06FGfPnsWZM2cwderURho9o84QUYPlhZnqcXZ2plmzZlFQUBBpa2uTTCajHTt2UEFBAY0ePZo0NDTIysqKEhMTuTbZ2dnk6elJOjo6JJFIyMbGho4dO8bVp6Wlkbu7O0kkEpLJZOTv70+FhYU19qUxGTZsGE2dOpW73rx5M5mZmVFZWVmtbXz66afUs2dPubIxY8bQu+++y107ODjQxx9/zF1XVFSQkZERrV69moiIYmJiiM/nU1FREaeTlpZGAOjKlSu17sv//i+b5ZkganvPha+vL73//vsK5R07dqQtW7bIldnZ2dGCBQvqZN/KyopWrlypsn7WrFn0xhtvUGVlJVfWr18/+vzzz+t0H2U01rPxb5dmnemGhYWhc+fOOHv2LKZMmQJ/f39MmDABo0aNQkpKCqysrODj41P1pUVQUBBKS0tx8uRJXLhwAcuXL4dIJAIAFBYWwsXFBQ4ODjh37hwOHDiAa9euYfLkySrvn5OTA01NzWolODi4TmN69OgRdHV1uevo6Gg4Ojpyyws2NjZYu3YtNyZlnD59GkOGDJErc3d3x+nTpwG8mD0nJyfL6aipqWHw4MGcTmlpKdTU1CAQCDgdsVgM4MUMvTXTFp+LVyktLeX6WIVYLK7T/01lZSWePHki97y9TFlZGXbs2AE/Pz/weC/2MNy/fx9nzpyBvr4+HB0dYWhoCFdXV5w9e7b+g2E0jMbw3KjlTHfQoEHcdUlJCfH5fJo1axZXlpSURADo1q1bRERkY2NDS5cuVWpvxYoVNGzYMLmyy5cvEwDKy8tT2ub58+eUkZFRreTn59c4lir27NlDQqGQLl26xJVZWFiQQCCgqVOnUnJyMv3000+kra1N69atU2mnW7duFBISIld28OBBAkBlZWV0584dAkBnzpyR05k3bx7179+fiIju379PWlpaNGfOHCouLqbCwkLy9vYmABQcHFzrMaEFZrpt6blQNdMdM2YM2djYUGZmJpWXl9OuXbtITU2NunfvXiu7RERff/01SaVSevDggdL63bt3E5/Ppzt37nBlp0+fJgAklUopIiKCkpOTKSAggDQ1Nen27du1vjcRm+k2ljRrasdevXpx/xYKhdDR0YG1tTVXZmj4Yqv+/fv30blzZ3z88ccICgrC4cOH4ebmBm9vb1haWgIAzp8/j8OHD0NTU1PhPpmZmUrXUNXV1WFubt4oYzl27Bj8/PywdetWrk/Ai9mITCbD5s2bwefz0adPH9y4cQNbtmzBrFmzGuXeytDX18eePXsQEBCAtWvXgs/nY8aMGTA0NISaWuteum9Lz4UqvvvuO/j5+cHc3BxqamqwtbXFuHHjkJycXKv2u3btwrJlyxAdHQ09PT2lOlu3bsXQoUNhZGTElVVWVgIAAgIC4OvrC+DFC9nDhw9jx44dmD9/fgNHxqgrzfptbNeundw1j8eTK6v6k6jqQZk2bRquX7+OcePG4ezZs+jduzf3Vr+oqAgjR45EamqqnGRkZODNN99Uev/G+jPy+PHjGD58OL799ltMmDBBrk4mk6Fbt27g8/lcmaWlZbUvTGQyGfLy5I/iysvLg66uLtq1awc9PT3w+XylOjLZ/+cf8fDwQHZ2NnJzc5Gfn4/ly5fjwYMHMDU1rXFMLUlbeS6qw8DAAAcOHMDTp09x69YtJCUloby8vFb/N7t374a/vz/27NkDFxcXpTo3b97EkSNH8OGH8nncq56PHj16cGVqamqwsLCo9plkNB2tPom5sbExgoKCEBQUhMDAQERERMDHxwe2traIioqCqampnIOrDiMjI6Sm/l97Zx8V1XH+8S+gu7Ar8rIsCCqIbDBiTSSiBE8JGqNEA5KXFptUYwtNqqFpbSCNiae/qJGgRm2OFJMQeYuamBDTBGMstbyoqIBVsLwpq4hEg9FqUYkCCt/fHx5uvbALCy4rhvmcM3/cmefOnVmefXaYO/Od0i5tjM2XtZOfn4+wsDCsXr3a4BvgKVOm4NNPP0VbW5s0wqyuroanp6fROoOCgmQrIABg9+7dCAoKAgAoFApMnDgRu3fvRnh4OIBbASgnJweLFy/uVF/7aC4jIwNKpRIzZszosk/3Iv3NL0zFzs4OdnZ2aGhoQHZ2NhISErq0/+STTxAVFYVt27Zh9uzZRu3S0tLg6uqKJ554Qpbv7e0Nd3d36PX/08EnCb1ej0ceeeTOOiPoHeaYo4CJc7qxsbGyPI1Gw7S0NOn622+/JQAeOnSI5K03sdnZ2aypqWFRURHHjx/PRYsWkSTPnDlDjUbDuXPn8tChQzxx4gR37twpW0lgbnJzc6lSqfj666+zvr5eSrfP99XW1lKtVnPx4sU8fvw4v/zySzo5OXHDhg2SzZIlSzh//nzpuqamhiqViq+++iqrqqqYlJTEQYMGcffu3ZLNtm3bqFQqmZ6ezsrKSr744ot0cnKSze+lpKSwsLCQ1dXVfP/996lSqbh+/foe9RF3YU73XvcLkqyoqGBJSQnDw8P52GOPsaSkhBUVFVL5gQMH+OWXX/LkyZPMycnhQw89xICAANkql45+sXXrVg4aNIhJSUkyf+u4EqO1tZWenp587bXXDLZt7dq1dHR05Pbt21ldXc3Y2Fja29uLOd27lPp10I2JieHo0aOpVCrp5ubG6OhoXrlyRbKvqqpiREQEHRwcaGdnRz8/P77xxhvdtqW3LFiwgLh1Aq0shYSEyOwKCgo4adIkKpVK6nQ6rl27lq2trbJ6Ot6Tl5fHCRMmUKFQcPTo0UxPT+/0/MTERHp6elKhUHDy5MmdXqzFxsZSq9VSoVDQz8+PKSkpPe7jvRB0+5tfkKSXl1cnv/Dy8pLKc3JypJesWq2Wv/3tb3np0iVZHR39IiQkxKC/LViwQHZfdnY2AfD48eMG29bW1saVK1dy+PDhVKvV/OlPf8ri4uIe91EEXfMkoacrkCH0dAXGEHq65qF/v9YWCASCHxki6AoEAoEFEUFXIBAILIgIugKBQGBBBmTQtbKywtdff323myHoZwi/EFiCARl0+zO1tbWIjo6WdFd9fHywfPly3LhxQ2ZjSEO1sLBQsklPT+9U3lFwRXBvER8fjylTpkClUhncCrxs2TKDfqFWqyUb4Rd3n36/I22gcezYMbS1teGDDz6ATqdDeXk5XnjhBVy/fh2rVq2S2ebn52PMmDHStUajkZU7OzujoqJCum7fTiu4N2lpacHPf/5zBAUFISMjo1N5XFwcFi5cKMubPn06Jk2aJMsTfnGXMcdiXxhZCJ+Zmclx48ZRqVRSo9Fw5syZ0iaBwsJCTp8+nc7OznRwcOD06dNZXl4u3Xvq1CkCYGZmJoOCgmhra8uHH36Yp0+fZm5uLn/yk59wyJAhjIyMlGnIhoSE8Pe//z0XLVrEoUOHUqvVdlLZAsAdO3ZI13V1dfzZz37GoUOHUqPR8JlnnpEpNeXm5jIgIIB2dnZ0dHRkcHAwz58/b7DPfcGaNWt43333Sdftn01ZWZnRe9LS0qjRaHr8LFhgc4TwizvD1L9taWkpAXDv3r09vtcQ5vKNgZ76bHqhvr4ezz77LKKjo3Hs2DHk5uZi1qxZUvnVq1fx61//GgcOHEBBQQGGDRuG8PBwNDc3y+r5v//7PyxfvhyHDh1CU1MTfvGLXyA+Ph5paWnIzs5GXl4e3n33Xdk9qampGDJkCIqLi7F69WosX74cmZmZBtt548YNhIaGwtnZGfv378eePXtAEnPmzEFbWxtu3ryJp556CtOmTUN5eTkKCgrw/PPPd9n3cePGdSmecvvnYAodNXvbmT17Ntzc3BAcHIydO3cavM/T0xMjR47Ek08+icrKyh49ty8QfmE+v+iOTZs2wdfXF8HBwbL8/ugXAwpzRG4YGNUcPnyYAFhbW9upzBBNTU1UKBTct28fyf+NaG7fDvvhhx8SAEtKSqS8mJgYTp8+XboOCQnhAw88IKs7JiaGQUFB0jVuG9Fs3ryZ48aNk9lfvXqVNjY2LCoq4sWLFwlAdjJBd9TW1napzdqTPe8nTpzg0KFDmZqaKuVduHCB69atY2FhIYuLi/mnP/2JVlZW3Llzp2Rz4MABZmRksKSkhPn5+QwLC6Ojo6NspGYI9PFIV/jFnfuFKaPV69ev08nJqZNOc2/9ghQjXXOlPgu6N2/e5NSpU2lvb8/IyEimpKTw8uXLUvm5c+cYHR1NnU5He3t7qtVqAuDWrVtJ/u/LdeTIEemerKwsSdi7nRUrVnD8+PHSdUhISCdxky1bttDR0VG6vv3LFRcXRxsbG6rValkCwI8//pgkOW/ePNra2jI8PJyJiYlGxbDNzdmzZ6nT6fjiiy92azt//nyZGHhHWlpa6OPjY1T8u52+DrrCL+4cU4Luxx9/zEGDBvHcuXNd2pnqF6QIuuZKfTa9YGNjg5ycHOzcuRM6nQ7vvPMO/Pz8JE3YBQsWoKysDImJiSgsLERpaSlUKhVaWlpk9RjSVe2Y166z2hsaGxsRGBhoUH81LCwMALB582YUFBRg8uTJyMjIgK+vr+xFREfM8W/kd999h2nTpiEoKAjvvfdet/aBgYE4ceKE0fLBgwfD39+/SxtLIPzCMtMLmzZtQlhYmCQAb4z+4hcDiT5dvWBtbY3g4GAEBwfjzTffhKurK7Kzs/H8889j//79SE5OxuOPPw7g1lv7a9eumeW5xcXFsuvCwkLZ6Q634+/vj88//xxubm6wt7c3WufEiRMxceJELF26FOPGjcP27dtlpxvczjfffCNb4tWR9rPLjHH27FlMmzYNEydORFpamkknP5SWlsLd3d1oeWtrq3QK7d1G+IVhuvMLUzl16hTy8vKQlZXVrW1/8ouBQp8F3aKiIuTk5GDmzJnQarXYu3cvGhsbpSVO9913Hz766CP4+/vj0qVLiIuLkx2qeCecPHkSS5YsQVRUFA4ePIhNmzYhPT3doO0vf/lLrFmzBk899RSWLVuG4cOH49SpU8jMzERCQgL++9//Ijk5GXPmzMHw4cPx73//G99++61Mib8jXl5evW772bNnMXXqVHh5eWHt2rW4cOGCVNZ+CkBGRgYUCgX8/f0BAF988QVSU1NlfVyxYgUefvhh6HQ6NDQ04J133kFdXR2io6N73TZzIPyi99TV1eHSpUuoq6tDa2urJLzu5+cn+4xSU1Ph7u5ucOTcX/1iQGGOOQoYmL+rrKxkaGgoXVxcaGtry7Fjx8peBh05coQBAQFUKpUcM2YMs7KyZDqqhpZF7dixgx2f9dZbb8leeLQvDXrhhRdob29PFxcXrly5UnYPOiwNOnv2LOfNm0eNRiNp4L700ku8fv06z507x4iICA4bNkzSul21alWn/pqLtLQ0gxqqt/c7PT2dY8eOpUql4tChQxkYGMjt27fL6lm8eLGkvevm5sYnnniCpaWl3T4ffTynK/yi9xjTcz516pRk09rayhEjRhjVD+6tX5BiTtdc6Uenpzt16lQEBARg7dq1d7sp9yQ/Vj1d4Rd3jtDTNQ9iG7BAIBBYEBF0BQKBwIL86KYXBHfGj3V6QXDniOkF8yBGugKBQGBBRNAVCAQCC3LPBN12Ddny8vK73RTBXUT4geBe554JuvcKCQkJmDRpEuzt7eHq6oqnn34aer1eZhMTEwOdTgc7OztotVpERETg2LFjXdb71ltvYezYsVCr1XBycsJjjz2GoqIimc2oUaM6CVR31OAVWIa9e/ciPDwcHh4e3Z5I0dLSggkTJpj0Y6LX6zFnzhxoNBo4OTnhV7/6FS5fviyzMSRkvm3bNplNU1MTXn/9dYwcORJKpRKjR4/GZ5991vsOC0xGBF0zs2fPHsTExKCwsBC7d+9Gc3MzZs6cievXr0s2/v7+SE1NRVVVFf7xj3/AysoKoaGhXWoF6HQ6bNiwAWVlZSgoKIC3tzdCQ0Nx8eJFmd3bb7+N+vp6Kb388st91leBcX744Qc8+OCDSEpK6tb2jTfe6HIL9+11zpw5E7a2tti3bx9ycnJQU1ODefPmdbLdvHmzzA+efPJJWfncuXORn5+PjIwMHD9+HJs3b4a3t7fpHRT0HnPssEAXgtUk+de//pUjR45kW1ubLP+RRx7hH//4R5Kmi1e370QypLSUmJhILy8vWV5ycjJ9fX2pVCo77X6yBOfPnycA7t+/36jN0aNHeyR3SJKXL1/uJC3o5eXFxMTEO2ov+nBH2kD1A3TY6XY7//znP+nr68vy8vJuhemzs7NpY2MjE2cvKysjAB47dsyk55Hkrl276ODgwIsXL/a4H+wHO7ru9WSRkW5kZCTq6+tRUFAg5Z05cwb79u3Dc889B8B08eqesHXrVixfvhyrV69GVVUVli1bhj/84Q/46quvjN7z9ttvd6kENWTIENTV1ZnchvZ//QyJkAO3Ri9paWnw8fHB8OHDTaqzpaUFycnJcHJywvjx42Vl8fHxcHFxgb+/P9atW4ebN2+a3Na+ZiD7gSEuXbqEqKgopKeny84xM0ZzczOsra1lOgvtIjkHDhyQ2S5cuBBarRaBgYGdjvbJyspCQEAAVq1aBQ8PD9x///1YunRpJyU3QR9hjsiNbka6JDlr1iwuWrRIul6zZg19fX2N2hsTr+7JCMfHx4efffaZzOa1117jrFmzjD734sWLXQpN6/V63rhxo9v+kmRbWxvDwsIYEhLSqSwpKUnSZ73//vtZU1PTbX07duygWq2mlZUVPTw8+K9//UtWvm7dOubl5fHo0aPcuHEjHR0d+eqrr5rU1nbQx9oLA9EPYGTk+cwzz3Dp0qUG+2WI8+fP097enrGxsbx27RobGhoYGRlJALKjh1asWMGCggIeOXKECQkJVCqVTEpKkspDQ0OpUCg4Z84cFhcX86uvvqKHhwcXL17cbT/YD0aK93qyWNDdsmULXVxcJEedMGGCTDjZVPFqU79sjY2NBEA7OzuZCLVCoejyS25OXnrpJY4aNcqgKn9DQwOrq6u5Z88ehoeHMyAggE1NTV3W19jYSL1ez4MHDzIqKoqjR4/mhQsXjNqnpKRw8ODBMnHv7ujroDsQ/cBQ0E1PT+dDDz0k/W1MCbrkrakBLy8vWltbc/DgwXzllVfo5ubWpdjOn//8Z44aNUq6njFjBu3s7GTi8RkZGbSzs+s09dOxH+wHQeteTxYLuo2NjVSpVPzmm29YVVVFANTr9VJ5aGgoJ0+ezF27drGiooJ6vZ4qlcqoulRGRgadnZ1lz1i/fr30ZTt37hwBcNu2bZ1GKF3NncbHx3c6LaBjOn36dLf9/d3vfscRI0aYNIJtbm6mSqXqNBrrDp1OxzVr1hgtb58nPHHihMl19nXQHWh+0P6Zdgy6CxYsoLW1NW1sbKQEgDY2NiadFPL999/zypUrbGxspLW1NT/99FOjtl9//TWtrKykH7r58+fTx8dHZlNcXEwAXR6sKYKueZLFjmBXq9WIiIjAJ598Ai8vL0yaNAk6nU4q76l4tVarRUNDA5qammBrawsAOHr0qFTu5uYGd3d31NTUYO7cuSa3c+HChYiMjOzSxsPDw2gZSbz88sv429/+hvz8fJPfCJPs8bxld/eUlpbCxsYGWq22R/X2JQPFD7ojPj4ecXFx0vV3332H0NBQfP7555g8eXK397u6ugK4pa2sVCoxY8YMo7alpaVwcXHBoEG3vu5TpkzB9u3b8cMPP0hzydXV1VCpVHBxcel1nwQmYo7IDRNGuuStOUl7e3t6e3vz3XfflZX5+/vz8ccfZ1VVFffv38+goCAqFAqjI5z//Oc/VKlUjIuLo16vZ3JyMjUajWwu77333qNarWZiYiKPHz/Oo0eP8oMPPuDGjRtNam9vWLRoER0cHJifn8/6+nopXbt2jeStwwlXrlzJQ4cO8fTp0zx48CAjIiLo7OwsmyoYM2YMv/jiC5K3DhlcsmQJDx48yNraWh4+fJi/+c1vqFQqWVlZSfLWgYN/+ctfWFpaypMnT3LLli3UarWMiorqUfthgSPYB4IfXL16lSUlJSwpKSEAbtiwgSUlJayvrzdob2h64cyZMxwzZgyLioqkvJSUFBYWFrK6uprvv/8+VSoV169fL5VnZWXxww8/ZFlZGfV6PTdu3EiVSsUVK1ZINleuXKG7uzvnzp3LyspK5ubm0svLi6+88kqXfTKXbwz0ZNGg29LSQo1GQxsbm07O1xvx6szMTHp7e1OlUvHZZ59lQkJCp6VCH330ER944AEqFApqNBo++uij/Pvf/25Se3sDjAiQt/ejvr6es2fPpqurKxUKBUeMGMHnnntOtuSnvZ72e5qbm/n000/Tw8ODgwcP5rBhw6SXIO0cPnyYgYGBdHBwoJ2dHf38/Lhq1So2Nzf3uP3sY58YCH6Ql5dn0A/efPNNg/aG+tWel5eXJ+XFxsZSq9VSoVDQz8+PKSkpsnp27drFCRMmcMiQIVSr1XzwwQeZnJzcaa62oqKCjz76KG1tbenp6cklS5Z06ysi6JonCZUxgQyhMiYwhlAZMw9iR5pAIBBYEBF0BQKBwIKIoCsQCAQWRARdgUAgsCAi6AoEAoEFEUFXIBAILIhZdqTZ2tp+b2Vl5WaOugR3F1tb2+/NVY/wiR8X5vKNgY5Z1ukKBAKBwDTE9IJAIBBYEBF0BQKBwIKIoCsQCAQWRARdgUAgsCAi6AoEAoEFEUFXIBAILIgIugKBQGBBRNAVCAQCCyKCrkAgEFgQEXQFAoHAgoigKxAIBBZEBF2BQCCwICLoCgQCgQX5f7NrJ4weDvNbAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2MCLKRsVwFuI",
        "colab_type": "text"
      },
      "source": [
        "Plus ce paramètre est petit, plus le modèle sera performant sur le jeu d'entraînement, mais plus vous dégraderez les performances sur le jeu de test. \n",
        "\n",
        "Plus ce paramètre est élevé, plus le modèle sera simple et généralisable. Cette meilleure généralisation augmentera les performances du modèle sur le jeu de test.\n",
        "\n",
        "Attention, si vous augmentez trop ce paramètre, vous dégraderez aussi les performances global du modèle. Tester plusieurs valeurs afin de maximiser vos performances sur votre jeu de tests."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "40NlrCpHJalH",
        "colab_type": "text"
      },
      "source": [
        "### min_weight_fraction_leaf\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "D0JDXihMwMvp",
        "colab_type": "text"
      },
      "source": [
        "Dans l'entraînement d'un modèle d'arbre de décision, on peut ajouter un vecteur de poids qui donne une importance différente aux observations de notre jeu de données.\n",
        "\n",
        "Si vous avez donné un vecteur de poids avec votre jeu de données, ce paramètre vous permet de spécifier le minimum de poids de votre feuille pour sa création. Si vous n'avez pas ajouté de vecteur de poids pour l'entraînement de votre modèle, ce paramètre a la même utilité que le paramètre min_sample_leaf.\n",
        "\n",
        "Pour ce paramètre aussi vous pouvez spécifier un nombre entier ou un nombre flottant."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Hir_3s4EJcbV",
        "colab_type": "text"
      },
      "source": [
        "### max_features\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wQOOX2n4hZbX",
        "colab_type": "text"
      },
      "source": [
        "Le paramètre max_features contrôle le nombre de variables à tester pour trouver la meilleure séparation.\n",
        "\n",
        "Vous pouvez indiquer un entier qui déterminera le nombre de variables à tester.\n",
        "\n",
        "Vous pouvez indiquer un flottant qui sera la proportion de variables à tester parmi les variables du jeu de données.\n",
        "\n",
        "Il y a une option auto qui effectue un test pour chaque variable de votre jeu de données, cette option est équivalente à choisir None. \n",
        "\n",
        "Il y a une option sqrt qui teste la racine carré du nombre de variables du jeu de données.\n",
        "\n",
        "Il y a une option log2 qui teste le log2 du nombre de variables de votre jeu de données.\n",
        "\n",
        "À noter que même si vous spécifiez un maximum de variable à tester l'algorithme peut le dépasser dans le cas où il ne trouve pas de variable valide pour séparer le jeu de données dans le nombre de tests imparti. \n",
        "\n",
        "\n",
        "C'est un paramètre à tuner si vous avez des problèmes de performance et que vous voulez réduire le temps de calcul. Si vous n'avez pas de problème de performance, je vous recommande de laisser None ou auto."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KxQj9IciJd8-",
        "colab_type": "text"
      },
      "source": [
        "### random_state\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7qhHryHRi7yS",
        "colab_type": "text"
      },
      "source": [
        "Ce paramètre permet d'initialiser une seed. C'est-à-dire que les nombres générés aléatoirement seront toujours les mêmes. C'est intéressant à utiliser durant les tests où l'on voudra tomber sur les mêmes résultats et être sûr que les changements de performance sont dû au changement des paramètres et non à une initialisation différente."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BnimrgUTJfXf",
        "colab_type": "text"
      },
      "source": [
        "### max_leaf_nodes\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "j7bUBdzYjRxH",
        "colab_type": "text"
      },
      "source": [
        "Ce paramètre vous permet de choisir le nombre maximum de feuilles de votre arbre, il peut être intéressant de fixer ce nombre pour contrôler la complexité de l'arbre. Il garde bien sûr les meilleures feuilles disponibles.\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yFR-hLv4kLzC",
        "colab_type": "code",
        "outputId": "f6ddb5b5-2932-44c8-a02e-96089870402a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 247
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_leaf_nodes=14)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "# Visualisation de l'arbre\n",
        "tree.plot_tree(model_tree)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YbWfpiA0kLTF",
        "colab_type": "code",
        "outputId": "1dce1964-0cb3-4966-d5af-28ff914a4ae7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 247
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_depth=2, min_samples_leaf=4)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "# Visualisation de l'arbre\n",
        "tree.plot_tree(model_tree)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV0AAADnCAYAAAC9roUQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOy9d1xUx/f//7qICjZMLCigIkovuwgooMaWGDugRkwUTNS3MSbBFtSPyQ97ookxxPhLeatoinmrsWIi1tg7AQyCBjWgAlGxYAMEdl/fP1ZuWHcpKrKwzPPxmMeDO3fm3jOzh7OzM3POSCQhEAgEgsrBxNACCAQCQU1CGF2BQCCoRITRFQgEgkpEGF2BQCCoRITRFQgEgkpEGF2BQCCoRITRFQgEgkpEGF2BQCCoRITRFQgEgkpEGF2BQCCoRITRFQgEgkpEGF2BQCCoRITRFQgEgkpEGF2BQCCoRITRFQgEgkrE1NACCKoe5ubmV/Py8iwNLYexYGZmdi03N7eFoeUQVA0kEcRc8DiSJFHoRcUhSRJISoaWQ1A1ENMLAoFAUIkIoysQCASViDC6AoFAUIkIoysQCASViDC6gnJz+fJl2NnZ4c6dOwCAzMxM2NnZ4datW1i9ejWaNWuGd999FwCwf/9+WFhYQKlUQqlU4osvvgAApKWlQalUonHjxpUi888//wxnZ2e4urpi/PjxestMmzYNbm5ucHd3R3R0tM79wYMHQ6lUytebNm2Cq6srTExMkJCQ8NxkFxgpJEUSSStp1EI/CxYs4Pvvv0+SDA4O5vLly0mSq1at4sSJE+Vy+/btY0BAQInPsbCwKPHe49y5c4eFhYXlLl/EuXPn2KFDB2ZnZ5Mkr127plPm119/5auvvsrCwkLeuXOHSqWSubm5WveHDx9OhUKh9dyzZ8+yW7dujI+PL1OOR/1p8M9VpKqRxEhX8ER88MEH2L17N5YsWYK0tDSMGTPmub3r2LFjGD16NBQKBfLy8p64/ooVK/D+++/DwsICANC8eXOdMsnJyejWrRtq1aqFRo0awc7ODvv37wcA5OXlYdGiRfjoo4+06jg6OsLJyenJGyQQQEwvCJ6QOnXqYMGCBZg6dSqWLVsGSSp5++nhw4ehUCgwYMAApKSklOv5t27dwpdffgmFQoH58+fLdevXrw8AGDp0qDxlUTxt375d51nnz59HUlIS/P394e/vjwMHDuiUcXd3R0xMDPLy8nD16lUcO3YMGRkZAIAFCxZgwoQJ8rsFgopAeKQJnpgdO3agZcuWSEpKgre3t94yHTp0QFpaGho0aIDNmzcjODgY8fHxpT63aI54wIABiImJgZWVlU6ZDRs2lFvOwsJCpKWl4eDBg/j777/xyiuvICUlBXXr1pXL9OnTBydPnoSvry9atGiBzp07o1atWjh//jxOnjyJefPmIS0trdzvFAjKQox0BU/EqVOncOLECRw+fBhz587F3bt39ZZr1KgRGjRoAAAICgpCRkYGCgoKSn22paUl1qxZgwcPHmDgwIFYunQpbt68qVXmSUa6NjY2GDRoEExNTeHg4IAWLVrg8uXLOuUiIiKQkJCAHTt24OHDh3BwcMCpU6fw559/wtbWFl26dEFycjJ69epV3m4SCErG0JPKIlW9hBIW0lQqFX18fHjgwAGS5Pz58zlp0iSSugtpV69elf8+dOgQHR0dtZ5V1kLapUuXOGvWLNrb2/O1115jXl5eqeX1ERMTw9GjR5Mk//nnH7Zq1Yo5OTlaZQoKCnjz5k2S5MmTJ+nl5UW1Wq1VJjU1VWshrQixkCbS0ySDCyBS1UslGd1vv/2WI0aMkK/z8vLo5OTEM2fO6Bjdr776iq6urlQoFOzcuTNjY2O1nlXe3QsqlYoxMTFPZXTVajUnTpxIZ2dnenh4MDo6miR56tQpjhkzhiR5//59Ojs709nZmf7+/jx79qzOcx43ulu3bqW1tTXr1KnD5s2bc+DAgaXKIYyuSMWTCHgj0OFpAt6sXr0aCQkJiIyMLFf5xo0bIzs7+2nEq3aIgDeC4og5XUGFYG5ujm3btsnOESVR5BxhaSkiRwpqJmKkK9BBhHasWMRIV1AcMdIVGA2nT59Gp06doFQq4evrizNnzmjdj4+Ph6mpKbZs2VLiM9RqNXx8fBAYGChfBwQEwNHREW5ubpg5c+ZzbYPA+BFGV2A0zJgxA3PnzkVCQgImTpyI2bNny/fUajVmzJiBV155pdRnfPPNN2jXrp1W3jvvvIO//voL8fHxOHLkCHbv3v08xBfUEITRFTw1aWlpcHNzQ0hICBwcHBAaGopdu3bB398fDg4OiI2NBQAcOHBA3k/r6emJe/fuAQAWLVoEHx8feHh4YPHixc8sjyRJ8r7hu3fvokWLf0/IWb58Ofr27VvqXPK1a9ewadMmjBs3Ts4zMTFBnz59AAC1a9eGUqlEenr6M8sqqMEYevuESFUvoZSAN8VJTU2lqakpk5KSqFKpqFQqOXbsWJLkpk2bOGTIEJLkgAEDePjwYZLkvXv3WFBQwJ07d3LChAlUq9UsLCxkt27dmJiYqPMOPz8/KhQKnaRvf2xycjJbtWpFGxsbtm3bVg5wk5WVxa5du7KgoICjRo3i5s2b9bYnNDSUR48eLTFYT3Z2Nu3s7Jiamlqu/ikCYsuYSMWSGOkKnon27dvDxcUFJiYmcHd3x8svvwwAUCgUsvts586dMXnyZHz55ZfIzs6Gqakpdu3ahe3bt8PT0xNeXl64cuWK3vgMR48eRUJCgk4qHmqxiG+++QZfffUVrly5glmzZmHChAkAgOnTp2POnDkwNS3Z6/3gwYOQJAl+fn5676vVaoSEhOC9996Dra3tE/aSQPAvIvaC4JkoHsfAxMREvjYxMUFhYSEAzVxr//79ERMTgy5dumDHjh1Qq9X46KOPyoxS5u/vj5ycHJ381atX6xjen376CUuXLgUADBs2DP/3f/8HAIiNjcXevXsBADdu3MD27dtRq1YtDBw4UK579OhR7N69G7a2tsjLy8Pdu3cxatQofP/99wCAiRMnwtraGpMnT36i/hEIdDD0UFukqpfwBNMLxT21iv90L37vwoULcpng4GBu3ryZMTEx9PPz4/379+XyRXFvnxYnJyceOnSIJLlt2zZ2795dp0xp0wtFPD698Omnn7J///5PFdOXpJheEEkriZGu4LkTGRmJ33//HaampnBxcUHfvn1Rt25dJCcnw9fXFwBgYWGBjRs3yrFvn4bly5cjLCwMKpUK9evXx3//+99Sy2dmZmLs2LF6g+UUce/ePUyfPh329vbw8vICoBn1vvXWW08tp6BmI5wjBDoI54iKRThHCIojFtIEAoGgEhFGVyAQCCoRYXQFVQ5bW9tKjUCWlpYGc3Nz2YEjPDxcvrd9+3Y4Ojqiffv2+Oqrr+R8kpg+fTocHBzg5OSE1atXV5q8guqNWEgTCKA5bPLx49QLCwsRFhaGffv2oUmTJvDy8kJQUBBsbGywcuVKZGVl4a+//gKg2YomEJQHMdIVlMmDBw/Qv39/KBQKuLm5Yd26dQCAuXPnwsfHB25ubggLC5PLd+/eHVOmTIG3tzdcXFwQGxuLoKAg2Nvby/EQ0tLS4OrqiuDgYDg7OyMkJETvcT4//vgjfHx8oFAoMGXKlFLlqWhOnjwJZ2dntGrVCvXq1cPgwYOxbds2AMB3332HiIgISJIESZLQrFmz5yKDwPgQRldQJjt27ICVlRVOnz6NM2fOyLEI3nvvPZw6dQqJiYnIyMjQOm3X3NwcsbGx+M9//oPAwEB8++23SExMxIoVK+Spg+TkZEyZMgVnz56FSqWSHRGKOHv2LDZv3oxjx47h9OnTuH79On777bcS5SnOwoUL9Z6lNnfuXL1tvHDhAjw9PdGzZ0+cPHkSgGZLmbW1tVzGxsZGPik4LS0NP/zwA7y9vTFo0CBcuXLlGXpYUJMQRldQJu7u7ti9ezemTZuGQ4cOyXtp9+7di44dO8LDwwNHjhxBcnKyXGfQoEEANO7A7u7usLS0hJmZGezs7OSAMW3btkWnTp0AAMOHD8fhw4e13rt3716cOHEC3t7eUCqVOHnyJC5cuFCiPMWZMWOGXvfhiIgInbItW7bEpUuXEB8fjwULFuC1115Dfn5+qX2Sm5sLCwsLxMbGIjg4WHY5FgjKQszpCsrEwcEBcXFx2L59OyIiItCjRw9MmzYNYWFh+OOPP2BlZYUPPvgADx8+lOsUdwd+3FW4yD1Ykv7dulr0M704arUao0ePxrx583Rkelyex43pwoULsXbtWp16gwcP1ilbt25dWUY/Pz80b94cly9fhpWVlTyyBYD09HTY2NgA0Ix6Bw8eLD9z2rRpJXWfQKCFGOkKyiQzMxPm5uYYOXIkpk+fjri4OOTl5QEAmjRpgjt37pQaGLwkUlNTcerUKQDAunXr0KVLF637vXr1wvr16+VFquvXr+Off/7RK8/jPMlINysrCyqVCgCQkpKCzMxM2NjYoGPHjkhOTsaVK1eQm5uLTZs2YcCAAQCAwMBA7N+/HwCwf/9+uLi4PHH7BTUTMdIVlEliYiLCw8NhYmKCOnXq4Ouvv0bjxo0xevRouLq6wsrKSnbnfRKcnZ3x+eef4/Tp0/Dy8kJoaKjWfVdXV0RERKBXr14gCTMzM6xatQrp6ek68jwLBw8exKxZs2BqagpTU1OsWrUKZmZmADQuzL169YJKpUJYWBhatWoFQBO57I033sBnn30GCwsLrFix4plkENQchBuwQIfKcANOS0tDYGCgzjYtY0S4AQuKI6YXBAKBoBIRI12BDiLgTcUiRrqC4oiRrkAgEFQiwugKKpy0tDS9x+k8b7777ju4urpCoVDg5Zdflh0Wjhw5InvOeXl5ybsOHufNN99E27ZtZUeKixcvAgCioqLg7u4OpVKJLl26yPuRi44NUigUUCgU2Lx5c6W0U1DNMXQUdZGqXkI5T44oicdPlKgsDh48yHv37pEkv/vuO77xxhskyTNnzvDSpUskyaSkJLZu3Vpv/ZJOlbh79678d3R0NPv160eSzMnJYX5+Pkny6tWrtLS0pEql0qkPcXKESMWSGOkKymT69OlapzBMnjwZq1atwv3799GrVy906NABCoUCu3bt0qm7evVqTJo0Sb7u3r27vGNh586d8PPzg6enJ0aOHFmmF1hZdO3aFQ0aNAAA+Pj4yJ5vrq6uaN26NQDNNrUHDx7ojfNQEg0bNpT/vn//vuzEYW5ujtq1awOAfI4bKebCBaUjjK6gTIYNG4b169cD0BiV6OhoBAUFwczMDJs3b0ZcXBx27tz5RIc23rhxA4sWLcLvv/+O+Ph4tGnTBsuXL9cpFxYWpjeGQln7YqOiovDqq6/q5G/atAkdOnSQjeXjzJgxAx4eHpg+fbrsOQcAK1asQPv27REeHo4vvvhCzk9ISICbmxvc3NzwzTffoFatWuXtAkFNxdBDbZGqXoKe6QVHR0dev36dhw8fln9eP3z4kOPHj6ebmxsVCgXr1KnDnJwcremFVatWceLEifJzunXrxvj4eG7bto1NmzalQqGgQqGgk5MTp06dqvPep2HdunXs1KkT8/LytPL/+usvtmvXTuugzOJkZmZSrVYzNzeXw4YNY2RkpN5nh4aG6uT/9ddf7NSpE3Nzc3XuQUwviFQsCY80QbkICgrCxo0bcfbsWQwbNgwAsGbNGty7dw/x8fEwNTVF06ZNteIvAICpqSnUarV8XXRfrVajT58++PHHH0t9b1hYGA4ePKiT/95772Hs2LE6+UeOHMGsWbOwf/9+rZgP169fR2BgIKKiotCuXTu972rZsiUAwMzMDKGhofjf//6nU2bYsGF4++23dfIdHBzQqFEjnDlzBt7e3qW2SVDDMbTVF6nqJegZ6cbFxbF79+5s27atfFR6ZGQkJ02aRFJz5DkA3r59W2uke+jQIb700ktUq9W8ePEizc3NGR8fz+vXr7NVq1b8+++/SZJ37tyR/35azp07x/bt2zM5OVkr/8GDB+zUqRN//vnnUutnZmaSJFUqFceNG8ePP/6YJJmSkiKX2b59O728vEiSaWlpfPjwIUny8uXLtLa2ZlZWls5zIUa6IhVLYqQrKBeenp7IyMiAq6urHEpxxIgRGDBgANzd3dGlSxd5sao4nTt3hrW1NVxcXKBQKODh4QEAaNasGVauXCmHUTQxMUFkZCTatm371DJ++OGHyM7Oxuuvvw5AEzpy8+bNiIqKwunTp7Fo0SIsWrQIgCZsZJMmTdCvXz+sWLECVlZWGDlyJLKysqBWq9GpUyd5jnrlypWIjo5GnTp18MILL8hxf0+dOoXZs2ejVq1aqFWrFpYtW4amTZs+tfyCmoHwSBPoIDzSKhbhkSYojti9IBAIBJWIMLoCgUBQiQijKxAIBJWIMLoCgUBQiYjdCwIdzMzMrkmSZGloOYwFMzOza4aWQVB1ELsXBE+MJEnjAcwGMJjkUQOLU+lIktQGwG8A9gGYRFJlYJEE1QhhdAXlRpIkEwALAQQC6EfygoFFMhiSJDUGsAFADoDXST4wsEiCaoKY0xWUC0mSzAGsBeAHwK8mG1wAIJkNoB+AmwAOSJLU0sAiCaoJwugKykSSpGYA9gJQAXiF5E0Di1QlIJkPYDSALQCOSZLkamCRBNUAYXQFpSJJkgOAY9DMX44gmWdgkaoU1DAfwIcA9kmS9LKhZRJUbYTRFZSIJEldARwCsJDkhyTVZdWpqZBcA+A1AGskSXrL0PIIqi5iIU2gF0mSXgfwJYCRJHWPhBDoRZIkRwDbAfwMIEIEsRA8jjC6Ai0kzVk0/wdgPID+JBMNLFK1Q5Kk5gCiAVwAMIbkwzKqCGoQYnpBICNJUm0AywEMBeArDO7TQfI6gB4AzADskiTpRQOLJKhCCKMrAABIkmQBzYb/FgBeIplpYJGqNSRzAQwDcBLAUUmS7AwskqCKIIyuAJIktYJmwewCgECS9w0sklFAUk0yHMBSAEckSfI1tEwCwyOMbg1EkqS2j/beQpKkDtBsCfsewLskC0utLHhiSH4NYCyAaEmShgCAJEl1JUnyMKxkAkMgFtJqIJIkHQKwCAABrAYwnuRGgwpVA5AkyRPANgBfQOPdlwCg9aOpCEENQUQZq2E8Gl21BWALzYb+gSSPG1SoGgLJeEmS/KCZO7cDEAvNvO/3BhVMUKmIkW4NQ5KkbwB4AGgCYAKAOiR3GFaqmsOjed060HzhvQDAhKQ4s70GIYxuDUKSpCYArgG4CyAXgBrAVpLvGVSwGoQkSWHQ7IFuAeAegNYA+pDcaVDBBJWGmF6oWVgDSAbwLYBdAC4Kj6nKheRSAEsfRSXrDuBdAK0MKpSgUhEjXYFAIKhExJYxgUAgqESMYnrB3Nz8al5enjjTq4IwMzO7lpub28LQchgDQjcrFmPQTaOYXpAkSUxNViCSJIGkZGg5jAGhmxWLMeimmF4QCASCSkQYXYFAIKhEhNEVCASCSsSoje7ly5dhZ2eHO3fuAAAyMzNhZ2eHW7duYfXq1WjWrBneffddAEBKSgr8/f1hZmaGyMhI+RkpKSno0qULXF1d4eHhgV9++UW+FxoaihdffBFbtmx57m3p378/GjdujMDAQL33p0yZgsaNG5f6jPj4eJiamlaKvILSeRLd3L9/PywsLKBUKqFUKvHFF18AANLS0qBUKsv83CuCiIgI+f2Ojo6lvvPBgwdo3bo1Jk2aJOcFBARAqVTCzc0Nb7/9NlQqFQBg7Nix8nPbtm0LpVL53NticEhW+wT5fEBdFixYwPfff58kGRwczOXLl5MkV61axYkTJ8rlsrKyeOLECc6cOZNffPGFnJ+WlsakpCSS5D///MOWLVvy7t278v1Ro0Zx8+bNJb6/OGq1mrdv3y5X2cfZu3cvo6OjGRAQoHPv9OnTHDFiBC0sLEqsr1Kp2Lt3b/bp06dMeR/1p8E/V2NIFaGb+/bt0/u5F1Ha5/44d+7cYWFhYbnL6+Obb77hm2++WeL98PBwBgcHa7Xhzp07JDX/A0OHDuX69et16k2fPp2zZ88u9d3GoJtGPdIFgA8++AC7d+/GkiVLkJaWhjFjxugt17RpU3Ts2BG1a9fWym/Tpg1cXFwAAC1atECzZs1w48aNJ5Lh6tWrWLhwIZydnbF9+/anakfPnj3RsGFDnXySCA8Px6JFi0qtv3z5cvTt2xeWlmL3UlWhvLpZERw7dgyjR4+GQqFAXt6zHei8du1aDB8+XO+9pKQkpKamok+fPlr5jRo1AgAUFhYiLy8PmlOhtFm/fn2JzzUmjGKfbmnUqVMHCxYswJAhQ3Dq1Cm9H3Z5OXXqFAoKCmBra1tmWbVajZ07d2L58uVISUlBSEgI9u/fjxYtNFsMf/jhByxZskSnnq+vL7799ttyyxQVFYUePXrA2tq6xDI3btzAmjVr8Pvvv2Ps2LHlfrbg+fIkunn48GEoFAq0atUKS5YsgYODQ5nPv3XrFn788UdERUXBxsYGY8aMwXfffScPLIYOHYoLFy7o1Pv444/Rr18/vc/8559/cO7cOfTq1Uvv/SlTpuDbb7/FgQMHdO4NGjQIhw8fxquvvoqgoCCteydOnICFhQUcHR3LbFd1x+iNLgDs2LEDLVu2RFJSEry9ny6g0/Xr1xEaGoqoqKhyGe7AwEAkJCQgKioKL7/8ss790NBQhIaGPpUsRdy6dQsrV67E/v37Sy03ffp0zJkzB6amNeLjrlaURzc7dOiAtLQ0NGjQAJs3b0ZwcDDi4+NLfW7RHPGAAQMQExMDKysrnTIbNmx4Ynl/+eUXBAUF6dWlH374Ab6+vmjbtq1eoxsdHY2HDx8iJCQEv//+O1555RX53rp162rEKBcw8oU0QDM6PXHiBA4fPoy5c+fi7t27T/yMnJwcDBo0CBEREfDz8ytXnU8++QSBgYF455138P777+v8k/zwww/yAkLxNH78+HLLlZycjIsXL8LBwQG2tra4e/eu3pFCbGws3nrrLdja2mLDhg0YN24ctm3bVu73CJ4P5dXNRo0aoUGDBgCAoKAgZGRkoKCgoNRnW1paYs2aNXjw4AEGDhyIpUuX4ubNm1plhg4dqlcHS5sCK21q4fjx44iKioKtrS0++OADREVF4cMPP9QqU7duXQQGBmLr1q1yHkn88ssvCA4OLrVNRoOhJ5UrIqGExQqVSkUfHx8eOHCAJDl//nxOmjSJpO5iRRGzZs3SWkgrLCzkwIEDuXDhQr3vKGshLS8vj2vWrGH37t3p6enJw4cPl1i2LCpiQaU8C38wgsWKqpIqQjevXr0q/33o0CE6OjpqPausz/3SpUucNWsW7e3t+dprrzEvL6/U8qU9x8bGhiqVqsyyxduQm5vLy5cvk9T8P73xxhv86quv5LIHDx6kr69vuWQwBt006pHu8uXL4eDggJdeegmAZuFix44dSEpK0imblZUFGxsbLFmyBLNnz4aNjQ0ePnyImJgY/Pbbb/jf//4njwQSE8t/MnndunXxxhtvYN++ffjll1/0LoaVh+7du+O1117Dzp07YWNjU+aUwrfffvtEc8OCyuVJdPOXX36Bm5sblEolZsyYgTVr1jzRu1q3bo3Zs2fj3LlzGD169FPLvH79egwdOhQmJv+ajdjY2DLXCR4+fIihQ4fC3d0dCoUCL774otYvupo0tQDU4NgLq1evRkJCgtae3KfhzTffRGBgYIn7Z6sjxuDfXlWoDN1s3LgxsrOzn0a8aocx6KZRj3RLw9zcHNu2bZM3oD8NoaGhOHDgAMzMzCpQMkFNp7y6WeQcIbYBVi9q7EhXUDLGMJqoKgjdrFiMQTdr7EjX0CxduhSSJMk/C0liwoQJaN++PTw9PXH27FmdOqW5JPfr10+ec7a2tpanO6KiouDu7g6lUokuXbogOTm5chooqHaU5JJ7+/ZtBAYGQqFQwM/PDykpKTp1MzMztXZAWFhYIDIyEmq1GgEBAXB0dISbmxtmzpwp17l06RJ69eoFDw8PvPzyy7h+/XqltdWgGHolryISSnG1rIpkZmby1VdfZevWrWW34G3btsk7E/bs2cPevXvr1CvLJbmI4OBgrl69miS17kdHR7Nfv35lygcjWCGuKqm66WYRxV1yP/jgA86fP58kGRcXV6YOqVQqWllZMTU1lSqVijExMSTJ/Px8vvTSS9y1axdJcujQofzpp59Iklu2bOE777xTplzGoJsGF6BCGlGKYqemptLV1ZUjR46kvb09Q0JCuHPnTvr5+dHe3p6nTp0iSe7fv58KhYIKhYJKpVI2VgsXLqS3tzfd3d352WeflfieJ2HEiBH8448/2KZNG9no/uc//+GGDRvkMjY2Nrx//36pz/Hw8ODff/+tlXf//n2++OKLsq97cX7++Wf279+/TPmMQbGrSqpuullE27Ztee7cOZJkv379ePToUfles2bN+ODBgxLr7tu3r8QtYGFhYYyKiiJJuri4MDMzk6RmcGBpaVmmXMagmwYXoEIaUYZim5qaMikpiSqVikqlkmPHjiVJbtq0iUOGDCFJDhgwQN5De+/ePRYUFHDnzp2cMGEC1Wo1CwsL2a1bNyYmJuq8w8/PT/6nKJ7i4+N1yu7Zs4dvv/02SWoZ3f79+/PYsWNyuY4dO/L8+fMltuvkyZN0dnamWq3Wyl+7di0DAwO18pYvX8527drR2tqaKSkpJT6zCGNQ7KqSqpNuFnH8+HEqlUr5evr06QwPDyep+QIAwAsXLpRYf/z48YyMjNTJz87Opp2dHVNTU0lqfpEtW7aMJLl69WpKksT8/PwSn0sah24aXIAKaUQZiu3k5CRfh4SEcO3atSTJixcv0svLiyT5ySef0MfHh5GRkbxy5QpJcurUqbS1tZUV1c7Ojhs3bizxXWWRn59Pf39/3rhxg+TTG91r167RyclJa/RRRFBQkNy+x1m3bh1DQ0PLlNMYFLuqpOqim8WZPHmyljNQdnY2Q0JCqFQqOXr0aLq6usqG83EKCwtpaWkpj2CLUKlUHDhwIJcsWSLnpaenMyAggJ6enpw2bRqbNWvGgoKCUmUzBt2sEc74devWlf82MTGRr01MTFBYWORu+eYAACAASURBVAgAmDFjBvr374+YmBh06dIFO3bsgFqtxkcffVRm9Cd/f3/k5OTo5K9evVorPug///yDlJQUeHl5AQDS09Ph4eGBP/74A1ZWVsjIyJDLZmZmomXLljrPLM0l+e7duzh48CB++uknvXIOGzYMb7/9dqltEVQuVUU3iyA1LrmHDh2S8ywsLPDDDz8A0EQJs7W1hY2Njd737d27F87Ozjq6O3HiRFhbW2Py5MlynrW1tRzb+ebNm9iyZUuNiA9i/C0sJxcvXoS7uzvc3d0RFxeHc+fOoXfv3pg7dy6GDx+O+vXrIy0tDS+88AIsLCy06h49erRc72jdujWysrLka1tbWyQkJKBx48YYMGAAoqKiMGTIEFlx69evr1VfpVJh+PDhCAoKwuuvv67z/K1bt+KVV15BvXr15Lzz58/D3t4eABATE4N27dqVu08EVYPK0M0iDh8+DBsbG61IetnZ2ahfvz5q166NZcuWYfDgwSUaR33eZZ999hlSU1O14i0Amuh3TZo0gSRJmDdvXo0ZEIgtY4+IjIyEq6srFAoFJElC37590adPHwwdOhS+vr5wd3fHyJEjnzkWaUkMGDAALVq0QLt27TB16lT5dIDMzEw5zF5ZLsn6FH7lypVwcXGBUqnEp59+iu+///65yC94flSmburTocTERLi4uMDR0RHHjx/Hxx9/DEBbNwGgoKAA27Ztw5AhQ+S8e/fuYfr06Th//jy8vLygVCqxatUqAJpRsaOjIxwcHFBQUICwsLBnlr86IJwjBDoYwwb0qoLQzYrFGHRTjHQFAoGgEhFGVyAQCCoRYXQFAoGgEhFGt4KwtbU1SHi9jIwMNGrUSA4DqFar0alTJygUCri4uGDWrFly2YsXL6Jjx45o37493nzzTfkYbEHNobL1NDo6WismQ61atZCQkABAc9xUUFAQnJyc4OzsjDNnzlSaXIZEGN1qTnh4OHr37i1fm5iYYM+ePTh9+jT+/PNP7Ny5E6dOnQKgOStt5syZuHDhAlQqFdavX28osQU1hEGDBiEhIQEJCQlYu3Yt2rRpI+8PDgsLQ2BgIM6dO4e4uDi0adPGwNJWDkZrdB88eID+/ftDoVDAzc0N69atAwDMnTsXPj4+cHNz09qi0r17d0yZMgXe3t5wcXFBbGwsgoKCYG9vj9mzZwPQxC91dXVFcHAwnJ2dERISovesqh9//BE+Pj5QKBSYMmVKqfI8Czt37oSlpSXc3Ny08otOp8jPz0d+fn7Rii8OHTqEQYMGAdDEAi7amC4wHDVBT4tYu3atfA7anTt3cPz4cYwaNQqAJobw056qUu0wtEtcRSTocbXcsGGD7MdOalwZSfLmzZskSbVazcGDB3P//v0kyW7dunHmzJkkySVLltDa2ppXr15lbm4ura2tefv2baamphIAjx8/TpJ8/fXXuXz5cpL/uvQmJyczKChIdmccMWIEf/311xLlKc4nn3yi109+zpw5OmXz8vLYuXNnZmdn65zrRpIdOnRg/fr1ZZ/5rKwsrbO1kpOT2blzZ53nksbhallVkj7dLI6x62lxnJycmJCQQJKMj4+nr68vR44cSaVSyQkTJpTr7DZj0E2DC1AhjdCj2H/99RfbtGnD8PBwHjx4UM5fv349fXx86ObmRktLS3799dckNcpcpKR79+5lnz595Dpdu3ZlYmIiU1NT2bZtWzl/69atHDVqFMl/lfmrr76ilZWVrIj29vaMjIwsUZ6nZc6cOXK0Jn1Gl9T8w/Tq1YuJiYnC6FYh3SyOsetpEadPn6azs7N8ferUKZqamjI2NpZqtZrjxo3j4sWLy3yOMeim0boBOzg4IC4uDtu3b0dERAR69OiBadOmISwsTI518MEHH+Dhw4dyneJ+74/7xBf5wUvSv/uyJUnSugY0C1mjR4/GvHnzdGR6XJ6IiAit+wsXLsTatWt16g0ePFin7MmTJ7Fy5UrMmTMH2dnZsszvvPOOXMbCwgI9e/bEjh07MHXqVNy+fRtqtRomJiZIT0+HtbV1mf0oeL4Yu54WUXxqAQBsbGzQpk0bOQ5JYGBgiTFDjA5DW/2KSNAzmsjIyGBOTg5JMiYmhgEBAbx9+zZbtGjBvLw8Zmdns127dvIIsVu3bnK4u8ePOi+6l5qaSkmSePLkSZKan2QrVqwg+e8I4syZM3RwcGBWVhZJTUSwzMxMvfJUFMVHujdv3pSjmOXm5rJbt27ctm0bSU0EsqLj10eOHMk1a9bofR6MYDRRVZI+3SxOTdHTdu3ayfF5i+jatascSW/atGlyoPTSMAbdNNqRbmJiIsLDw2FiYoI6derg66+/RuPGjTF69Gi4urrCysoKvr6+T/xcZ2dnfP755zh9+jS8vLwQGhqqdd/V1RURERHo1asXSMLMzAyrVq1Cenq6jjzPg5s3byI4OBiFhYVQq9UYNmwYBgwYAAD49NNPMXz4cEydOhWdO3fGsGHDnosMgvJTE/Q0NjYWDRs2hKOjo1b+l19+ieHDhyM/Px9OTk5yTAZjR8ReeALS0tIQGBgo7zM0VozBv72qYIjYC8asp8agm0a7ZUwgEAiqImKkK9DBGEYTVQWhmxWLMeimGOlC83NMXxT9501KSgr8/f1hZmYmu/EWURRD1dXVFd99953e+qdPn4afnx/c3d0xZMgQ+YSAmJgYKBQKKJVKeHt74/Dhw1r1Hjx4gNatW2PSpEnPp2GCCsNQulmSDsXGxmq59Zqbm+s42ZR27HpUVBTc3d2hVCrRpUsXJCcnA9DE4g0NDYWHhwdcXFywbNmyymtsZWPolbyKSHjGY65TU1OpUCie6RlPQ1ZWFk+cOMGZM2dq7bNNTEykh4cHc3Jy+PDhQ/r7+zMjI0Onvre3t3xgYWRkJBctWkRSc3ihSqUiSf755590cXHRqhceHs7g4GBOnDhRr1wwghXiqpKqq26WpUOkZh/4iy++qHMycGnHrhedZEyS0dHR8nHua9eu5fDhw0mSd+7cYfPmzfWeaG0MummUI93p06fjv//9r3w9efJkrFq1Cvfv30evXr3QoUMHKBQK7Nq1S6fu6tWrtUaA3bt3lxckdu7cCT8/P3h6emLkyJHIz89/JjmbNm2Kjh07onbt2lr5Z8+eha+vL8zNzVGnTh34+voiOjpap/7FixfRuXNnAEDPnj2xadMmAECDBg1gYqL5aO/fv6+1RzMpKQmpqano06fPM8kueDqqi26WpkNFbN68Gb1799Y6HgrQ7Bcu0q/atWtDqVQiPT0dALRcfYs/V5Ik5OTkQKVSIScnB40aNdLag2xMGKXRHTZsmBzMhSSio6MRFBQEMzMzbN68GXFxcdi5c6fWIXllcePGDSxatAi///474uPj0aZNGyxfvlynXFhYmNbPr6K0YsWKcr/L1dUVBw4cwO3bt3Hv3j3s2bNH69DKIhwcHPDbb78BADZt2qRVZvv27XByckK/fv205JwyZQo+/fTTcssiqFiqk26WpENF6Dva53Hu3LmDX3/9FT169JDzVqxYgfbt2yM8PFw+liooKAjm5uZo2bIl2rdvj3nz5hmt0TXKfbpeXl5IT09HVlYWUlJS4OTkhMaNGyM/Px/Tp0/H4cOHUatWLVy4cAG5ubnleubx48eRmJgon8D78OFD9O/fX6fc0qVLn1l+FxcXTJkyBb169UKjRo3g4+ODWrVq6ZSLiorCxIkTMWvWLAwcOFCrTL9+/dCvXz8cOXIEERER2L17N3744Qf4+vqibdu2OHDgwDPLKXhyqpNu6tOhIm7evInY2Fj07du3xPpqtRohISF47733tA66HDt2LMaOHYv169dj/vz5+P7773Hy5EnUqVMHmZmZyMrKQrdu3dCjRw9YWlo+kczVAaM0uoDmm3Pjxo04e/as7ASwZs0a3Lt3D/Hx8TA1NUXTpk213CsBwNTUFGq1Wr4uuq9Wq9GnTx/8+OOPpb43LCwMBw8e1Ml/7733MHbs2HLLP27cOIwbN05+poODg04ZFxcX+R8hLi4OR44c0SnTuXNnXLp0CTdu3MDx48exbds2fP/997h//z7y8/NRv359LFiwoNxyCZ6d6qabxXWoadOmAICNGzdiwIABqFOnTon19B27Xpxhw4bJJwD//PPP6Nu3L0xNTdGyZUt4enri9OnTWmFLjQZDTypXRIKexYq4uDh2796dbdu2lSMlRUZGctKkSSTJbdu2EYAclaloseLQoUN86aWXqFarefHiRZqbmzM+Pp7Xr19nq1at+Pfff5PUTPYX/f2s6AtYc+3aNZLkhQsX6OzsLLtm6itTWFjIIUOGcOPGjXKdokWQhIQEWltbU61Wa9VdtWqVWEgTulkiZelQz549uWPHjhLrf/rpp+zfvz8LCwu18lNSUuS/t2/fTi8vL5LkwoUL5ehmd+7cYbt27fS2wRh00+ACVEgjSlghtre354ABA+TrrKwsdurUiW5ubhw/fjxbt26to9hqtZqvv/46nZycGBwczE6dOsm+7rt27aKXlxfd3d2pUCi4b98+ve8tL9evX6e1tTUbNmxICwsLWltby+Ht/Pz86OzsTE9PTx45ckSuM2bMGJ46dYokuXjxYjo4ONDe3p6zZ8+WyyxdupQuLi5UKBT09fXloUOHdN4tjK7QzdIoTYeuXr1KS0tLOSxkEX379mVGRgbv3r1LSZLo4OAgRzEriog3ffp0Ojs7U6FQsHv37jxz5gxJzW6JIUOG0NXVlS4uLly2bJleuYxBN4VzhEAHY9iAXlUQulmxGINuGuXuBYFAIKiqCKMrEAgElYgwugKBQFCJCKMrEAgElYhR7NM1MzO7JkmS8e2iNhBmZmbXDC2DsSB0s2IxBt00it0LlYEkSfMBtCU5wgDvNgVwGsBMklsr+/2Cqo0kSe0BHAfgRvKqAd4/FUAvkv0q+93VEWF0y4EkSbYA/gCgIJluIBleAfAtABeSD8sqL6g5SJK0BcAxkosM9P46AM4AmERyuyFkqE6IOd3y8RmASEMZXAAguRuPFNtQMgiqHo++jN0ARJZV9nlBMh/AZABLHhlgQSmIkW4ZSJLUHcBqAM4kyxeB5PnJUvQz0p3kP4aURWB4qtK0k6SJ0bgdwC6SXxhSlqqOMLqlIElSLQBxAOaT/MXQ8gCAJEmLADQn+ZahZREYFkmS3gMQCOCVquD2JkmSM4CDAFxJXje0PFUVYXRLQZKktwGMANCtKig1AEiS1AjAOQABJE8ZWh6BYZAkqQmAswB6kjxjaHmKkCQpEoA5ybcNLUtVRRjdEpAkqTE0xq0vyXhDy1McSZJGAxgLoHNV+TIQVC6SJH0Fzf/ve4aWpTiSJL0Azf/NqySN7wz4CkAY3RKQJGkJgAYkxxlalseRJMkEwEkAS0j+bGh5BJWLJEluAH6HZp3hpqHleRxJksYDGA6ghxgU6CKMrh4kSXICcAhVeG5KkqTOANYCcCL5wNDyCCqHRwtWuwFsIVklj8wtthYyj+QGQ8tT1RBbxvSzBMAnVdXgAgDJI9B8MUwztCyCSmUQgJbQ7NmukpBUQbO1cbEkSeaGlqeqIUa6jyFJUj8AX0CzLevZjlR9zkiS1ApAPAAvkpcMLY/g+SJJUl0ASQDeebRvu0ojSdIGAAkk5xtalqqEMLrFeLSx+08AU0n+Zmh5yoMkSbOg8VILNrQsgueLJEnToFk8DTC0LOVBkqS2AGIBeJDUPc66hiKMbjEkSZoMoDeAftVlAUCSpHrQbB0KIal76qDAKJAkqQU0Hom+JC8YWp7yIknSAgCtSYYYWpaqgjC6j5AkqTk0P91eInnW0PI8CZIkBQOYDsDn0XyawMiQJCkKwA2S1WoOX5KkBgD+AjCU5DFDy1MVEEb3EZIkfQcgh6T+86KrMI9WtA8C+J7kCkPLI6hYJEnyARANwJHkXUPL86RIkhQC4H1oRunqssobO8LoApAkSQlgJzTbr24bWp6nQZKkDgB+g6YNdwwtj6BiePSFegTACpJRhpbnaXi0r/wogG9Ifm9oeQxNjd0yJhUDwJcAZlVXgwsAJOOgMbr/H/Bv+wwrleBpKfbZvQ6gDjRBl6olj0a3EwF8LElSQ0CrfTWOGjvSfbRoZgogFRpD1aG6z4c+OqEgCYA/gHkA/n+xuFb9eORKexhAR2hcal8nediwUj07kiT9ACAdwB4Ao0mONLBIBsEojut5SmwAZAFYDOAtaEb91droArgFYCGAzwHkALA2rDiCp6QZNKPbadAY3xOSJEnVZUeNPiRJqg3g/6DZkpmEGqybNXZ6AYAFAF8ACQCGAdhnWHGeDUmSzABcgeaL1BFAQ2jaKKh+WADIBfAegAMALgNwN6hEz87nAFYCWAFgDGqwbtZko9sCQB8A3gBqAxhgWHGeDZJ50Ewr9AZAAF0BvGBQoQRPS2MArQBchcaddgjJPw0r0jMzFRq39dEAOkDz/1cjqclG1x1AHjRzS2NJZhtaoGeF5N8AXobGjdkcgJ9hJRI8JW7QGN4dADxJHjWwPM8MyQKSCwB0h2Zar5lhJTIcNXkh7RUA8SRvGFqW54EkSa4ATEmeNrQsgidDkqRm0LjO7jW0LM+DR8cMBZDcaGhZDEGNNboCgUBgCGry9IJAIBBUOmVuGTM3N7+al5dnWRnCGDNmZmbXAED0ZcVQt25dPHz40NBiGA2iPysWMzOza7m5uXoXC8ucXqjm2wOrDEUOOKIvKwZJkkRfViCiPyuWR/2p1+tOTC8IBAJBJSKMrkAgEFQiwugKBAJBJWI0Rnf8+PFwdHSEu7s7xowZg8LCQvnenj174OHhAVdXVwwcOFBv/e3bt8PR0RHt27fHV199JedfvHgRHTt2RPv27fHmm29CpdKEZ7h06RJ69eoFDw8PvPzyy7h+vcqeYflUbNq0Ca6urjAxMUFCQoKcn5+fjzfffBPu7u7w9PTEsWO6calv3ryJV155Bc7OznBzc8OyZf8eWjt27FgolUoolUq0bdsWSqVSvleez6m6s3TpUkiShOxsjS9OSkoK/P39YWZmhsjISL11SuvP06dPw8/PD+7u7hgyZAhycnLke8ban/Pnz4eHhweUSiV69+6Nf/75R75XnjbfunULQUFBcHJygrOzM86cOQO1Wo2AgAA4OjrCzc0NM2fOlMtfvnwZ3bt3h6enJzw9PXHgwIFnawDJUpOmSNUnJiaGKpWKarWab7zxBv/73/+SJG/dukVnZ2deuXKFJHnt2jWdugUFBWzXrh0vX77MBw8e0MnJSS4/ZMgQbt68mSQ5cuRI/vzzzyTJoUOH8qeffiJJbtmyhe+8806p8kHjmlsxja0Ezp07x7Nnz7Jbt26Mj4+X85ctW8axY8eSJK9cucKOHTvq1L116xaPHTtGkrx37x4dHBz4119/6ZSbPn06Z8+eLdcp63MqTnXqyyIyMzP56quvsnXr1rx9+zZJMisriydOnODMmTP5xRdf6K1XWn96e3vz8OHDJMnIyEguWrRIrmOs/Xnnzh3576VLl3LChAkky9/mESNGcPXq1STJnJwc3r17lyqVijExMSTJ/Px8vvTSS9y1axdJcvz48fz2229JkkePHqWXl1eZMj7qT7029alHumlpaXBzc0NISAgcHBwQGhqKXbt2wd/fHw4ODoiNjQUAHDhwQB7ZeHp64t69ewCARYsWwcfHBx4eHli8ePGzfXMA6NOnD0xMTCBJEry9vZGeng4A+PnnnzFs2DDY2NgAAJo3b65T9+TJk3B2dkarVq1Qr149DB48GNu2bQNJHDp0CIMGDQIAhIaGYsuWLQCA5ORk9OzZEwDQs2dPbNq06Znkr2r96ejoCCcnJ5384u22sbFBYWEhzp07p1XmhRdegK+vLwCgQYMGcHJyQkaG7rmE69evx/DhwwGU73N6EqpafwJAeHg4Pv74YxQPJdu0aVN07NgRtWvXLrFeaf158eJFdO7cGYC2HlZkf1a1vmzUqJH894MHD+T+LE+b79y5g+PHj2PUqFEAAHNzczRs2BAmJibo06cPAKB27dpQKpWyDZEkCXfvag7suHv3Llq0eMawESVZY5Yx0k1NTaWpqSmTkpKoUqmoVCrlEdCmTZs4ZMgQkuSAAQPkb+J79+6xoKCAO3fu5IQJE6hWq1lYWMhu3boxMTFR5x1+fn5UKBQ6qfjI63Hy8/PZoUMHHjlyhCQ5ceJETpgwgV27dqW3tzc3bNigU+eXX37h22+/LV9//fXX/PDDD5mVlUVHR0c5Pzk5mZ07dyZJBgcHc9myZSTJ1atXU5Ik5ufnlygXyhjpVtX+fHyk+8033zA4OJiFhYU8e/Yszc3NuWfPnhLrp6WlsXXr1rx3755W/vHjx6lUKuXr8nxOxSmtL8mq15979uyRdaxNmzbySLeIWbNmlTjSLc7j/dmpUyf++uuvJMnZs2fTxsaGZMX2Z1XrS5KcM2cOW7VqRRcXF16/fr3cbY6Pj6evry9HjhxJpVLJCRMmMC8vT6tMdnY27ezsmJqaSlLzC8XNzY02NjZs1qwZk5OTS+1LsvSR7jMZXScnJ/k6JCSEa9euJUlevHhRHoJ/8skn9PHxYWRkpDzsnzp1Km1tbeWOtbOz48aNG8tsSHmYMGECw8LC5Ot3332X/v7+fPDgAa9du0Y7OztmZGRo1Xkao5uens6AgAB6enpy2rRpbNasGQsKCkqUqzxGtyr25+NGNz8/n++//z4VCgWHDBnCnj17ct++fXrrPnjwgB07dtQry+TJk7lw4UL5ujyfU3HKY3SrSn/m5+fT39+fN27cIPn0RldffyYlJfHll1+ml5cXZ8+ezTZt2pCs2P6sSn35OIsWLWJERATJ8rX51KlTNDU1ZWxsLNVqNceNG8fFixfL91UqFQcOHMglS5bIeZ9//jkjIyNJkvv27aO3t3eZcpVmdJ8piHndunXlv01MTORrExMTeSFrxowZ6N+/P2JiYtClSxfs2LEDarUaH330EcaMGVPq8/39/bUWBopYvXq11gJMEZ9//jkuXbqErVu3ynk2NjawtLREvXr1UK9ePXh7eyMpKQlWVlZyGSsrK62fv+np6bCxsUGTJk1w+/ZtqNVqmJiYID09HdbWmtjL1tbW8lTDzZs3sWXLFpiaPltM+KrWn/qoXbs2li5dKl8rFArY29vrlFOr1XjjjTcQFBSEwYMHa90jiV9++QWHDh2S88rzOT0pVaU///nnH6SkpMDLywuARr88PDzwxx9/oFmz8gXbKqk/XVxcsHv3bgBAXFwcjhw5AqDi+7Oq9OXjjBw5Er1798acOXPK1WYbGxu0adNG/iwCAwPx008/yfcnTpwIa2trTJ787/m0K1euxJ49ewAA3bt3x+XLl5Gfn486deqU2qaSeO4nR1y8eBHu7u5wd3dHXFwczp07h969e2Pu3LkYPnw46tevj7S0NLzwwguwsNCOa3z0aPkj2q1fvx7r1q3Dvn37UKtWLTk/ICAAkyZNgkqlQm5uLhISEuDg4KBVt2PHjkhOTsaVK1fQtGlTbNq0Cbt27YIkSejcuTOio6MRGBiIH374AQEBAQCAGzduoEmTJpAkCfPmzcPbb7/9DL1UfiqrP0uiaA6tXr162LJlC9q0aSN/ERXn/fffR8uWLTFjxgyde4cPH4aNjQ1sbW3lvPJ8Ts+DyujP1q1bIysrS762tbVFQkICGjduXG45S+rP69evo3nz5lCpVPj4448xfvx4AIbpz8rSzfPnz8tf9Fu3bpXXHsrT5hYtWsDKygoXLlxA+/btsX//fri4uAAAPvvsM6SmpmoN2gDN57d3716MHDkSp0+fRsOGDZ/a4AJ4tukFhUIhX48aNUpe5S9+77333qOLiws9PDw4fPhwef7k888/p5ubG93c3Ni5c2devXq1zCF7aTRp0kTrZ8z8+fPle4sWLaKzszNdXV25fPlykmRGRgb79u0rl4mOjqa9vT3t7OzknxIkef78eXp5edHOzo4hISHyFMLatWtpb29Pe3t7TpgwodSpBbJ80wtVqT+3bt1Ka2tr1qlTh82bN+fAgQNJkhcuXKCDgwOdnJz46quvMjMzU65TJGNiYiIB0M3NTf48duzYIZd79913tfq4CH2fU0mU1pdk1evP4hSfXrh+/Tqtra3ZsGFDWlhY0NraWpahPP25ePFiOjg40N7eXt4JUkRF9WdV68uRI0fK7xkwYIA8lUGW7389Li6OXl5edHd352uvvcb79+/z7t27lCSJDg4Och9HRUWR1Ezh+Pn50cPDg56enty/f3+ZMqKU6QURe6GSELEXKhYRK6BiEf1ZsYjYCwKBQFBFEEZXIBAIKpEqb3RtbW1ll8nKYMuWLVAoFHB3d4e/vz8SExO1ZHF3d4dSqUSPHj3k/DfffFN2aVUqlbh48WKlyfskVHZf3r59Gz169ECDBg0wadIkrXtFfaVUKtGsWTP5/uzZs2FtbS3fK77DoSpT2X1bREZGBho1aqTlQqzPzbW6Udn9mZCQAKVSCYVCAYVCgc2bN8v3SgoR8LQ8990L1Q1ra2vs3bsXTZs2xa5duzB+/Hh5Gw4AHDp0SO+q8xdffIHAwMDKFLXKU7duXcydOxdJSUk6XmvF4zn4+voiKChIvg4PD9cx0gL9hIeHo3fv3lp5YWFhCAwMxObNm5Gbm6sVh0SgH0dHR5w6dQq1a9fGtWvXoFAoEBAQALVajbCwMOzbtw9NmjSBl5cXgoKCZK+3p+GJR7oPHjxA//79oVAo4ObmhnXr1gEA5s6dCx8fH7i5uSEsLEwu3717d0yZMgXe3t5wcXFBbGwsgoKCYG9vj9mzZwPQuBm6uroiODgYzs7OCAkJQUFBgc67f/zxR/j4+EChUGDKlCmlyvO0+Pj4oGnTpvLfRa6AzwNj78t69eqha9euMDMzK7FMWloa0tPT0bVr12d61+MYe98CwM6dO2FpaQk3Nzc5ryQ312fF2PvT3NxcnhxdmgAAD5lJREFUdsUu2i9MssQQAc9ESdsaWMKWsQ0bNsgugKTGZY4kb968SZJUq9UcPHiwvK2iW7dunDlzJklyyZIltLa25tWrV5mbm0tra2vevn2bqampBMDjx4+TJF9//XV5u0fR9prk5GQGBQXJW7NGjBjBX3/9tUR5ivPJJ5/odTGcM2dOqds+Fi9ezP/85z/yta2tLT09Penl5cXvv/9ezh81ahQdHR3p7u7OadOm6d0+Bj1bxmpKX65atYoTJ07Ue2/hwoWcNGmSfD1r1iza2trS3d2d48aN4927d/XWe7wvH8fY+zYvL4+dO3dmdna2ljdbedxcRX/q19X4+Hi6urqyXr163LRpE8mSvVXLAhXpBvzXX3+xTZs2DA8P58GDB+X89evX08fHh25ubrS0tOTXX39NUtP5RZ26d+9e9unTR67TtWtXJiYmMjU1lW3btpXzt27dylGjRpH8t/O/+uorWllZyR1nb2/PyMjIEuV5Vg4dOkQnJyfZdZOk7FKYmZlJZ2dnxsXFyddqtZq5ubkcNmyY3j2o+oxuTenL0oyup6en3CaSvHr1KgsLC1lYWMgpU6ZoGeTilGUkjL1v58yZI+8jLW50y3JzLYma3p+Pt7VTp07Mzc19Lkb3ied0HRwcEBcXh+3btyMiIgI9evTAtGnTEBYWhj/++ANWVlb44IMPtA65K+4y+Lg7YdF8U/HIS5IkaV0DGjfI0aNHY968eToyPS5PRESE1v2FCxdi7dq1OvUGDx6sUxbQxDh96623EB0djSZNmsj5Re6ELVu2RP/+/fHHH3/A09MTLVu2BACYmZkhNDQU//vf/0roPW1qQl+WRkpKCrKzs9GpUyc5z9Ly33M7R48e/dSefsbetydPnsTKlSsxZ84cZGdnyzIHBQWV6ub6tBh7fz7e1kaNGuHMmTMlhgh4JkqyxixhpJuRkcGcnBySmhi2AQEBvH37Nlu0aMG8vDxmZ2ezXbt28jdv8YAp+/btY0BAgPysonupqamUJIknT54kqfkJsWLFCpL/fuOdOXOGDg4OzMrKIqmJlZmZmalXnmfh2rVrdHZ21vE6KfJaITURlDw9PXn06FGSlL2yVCoVx40bx48//ljnudAz0jX2viyipJHu3LlzOWPGDK284h5uCxYs4Lhx4/Q+8/G+fJya0rekbrCcrl278vz58yTJadOmaXlnlkRN78+0tDQ+fPiQJHn58mVaW1szKyuLBQUFtLOz4+XLl5mTk0MnJydevny5zOehIke6iYmJCA8Ph4mJCerUqYOvv/4ajRs3xujRo+Hq6gorKys59ueT4OzsjM8//xynT5+Gl5cXQkNDte67uroiIiICvXr1AkmYmZlh1apVSE9P15HnWfj000+RkZGBiRMnAtB8W584cQLXrl3D4MGDQRIqlQqjR4+Gn58fAE3QjaysLKjVanTq1EkrWEZpGHtfAkD79u1x69YtFBQUYMOGDThw4ADatWsHAFi3bh3WrFmjVX7atGlISEj4f+3df0yUdRwH8DecTdxY1EJXIEgFduO45wHOJHWWAZoSKOzIE7PU2Q8hB6s2z7+ittZWy0VlJVpaS5yygU7nD2wZcdDcMGD4A3dJwXIFJ+RKSDvl+fTH7fnuHu4HiNejl5/Xxsb9eO6e+9zD547n+/1+PoiIiEBKSgqqq6sn9Lx3QmwD+fDDD7FixQq43W4YjUbs3Lnzph/z/x7P1tZWvPnmmzAYDDAYDNiyZYsYUK+qqkJOTg5GRkZQXl6OhISEm3qu22IZcE9PDwoLCzXTiP5v9FoGfCfEErg1y1b/z7HleIYWLwNmjLHbxG3xTfdOwAVvQosLtIQWxzO0+JsuY4zdJnRLuj09PePuThBK1dXVMJlMkGUZubm5+PXXXwF4qvlbLBakp6cjLS0NW7du9bt9aWkpJEmCJEkoLi7G8PAwAE9F+6lTp4oaAaFYYTSWWxXDsdqEK4qCRx99NOAy6EBtwhsbGxETEyNi+MEHH4htenp6kJ2djdTUVJhMJgwMDIT0Nd2qWB45cgSyLCM9PR2zZs1Cc3MzAKClpUWs7LJYLGhsbPS7/WeffYbk5GRNG3cg8HukDu7KsozU1FRUVlb+J6/rVsVT1d7ejkmTJoluLgBgMBjEsfXss8/63a67uxuzZ89GcnIy1qxZg5GREQBAb28vcnJyIEkScnNz4XK5AAAHDhzQ1A0xGAw3fk460LQG9Qchas08uhCyXpqamkQTv+rqalq5ciURefpWqVNOhoaGKCkpyW9xZe92z6+99hq99957RBR8wr8/CEEL9lsVw7HahG/ZsoVsNlvAaTuB2oSPnkrkbf78+XT8+HEi8qw2Gr2qKlxjefnyZRoZGSEios7OTkpNTSUiotOnT1Nvby8ReYpmJyYm+t2+o6ODfvnlF58+a8HeI3Wq47Vr1ygrK0tM0fIWrvEk8kzVXLRoES1evFgUVyciiomJGXNbq9Uqtlm1ahXt3r2biIiKi4tp165dRES0f/9+Ki0t9dm2q6tLs7jDG/6LFux2ux3btm0Tl1999VXs3LkTQ0NDyMnJQWZmJmRZxrFjx3y2/fLLLzUFTRYsWCA+LRoaGjBnzhxkZGRg1apVcLvdE91FAMD8+fMRHR0NQFtL4a677sKUKVMAAFevXhWfcKOp7Z6JCH///bfP5O2bES4xDNYmvL+/H/X19XjppZcCbh+oTXggZ86cQWRkpKjkFhMTo5lc70+4xDI6OhqRkZ4/u6GhIXE8mUwmJCYmAvBMoxoeHvZbh0CWZU2bI1Ww90itveB2u+F2u8d1DIdLPAFg+/btWLJkiWZhzXgQERwOB5YuXQoAeP7558U35bNnzyI7OxtA4GN2z549sNlsN7y/E066y5cvR21trdj5AwcOoKioCFFRUdi3bx/a2trQ0NAw7jmrgKfv2Lvvvovjx4+jvb0dM2bMwPbt233uV15ervmKr/58/vnnQR9/x44deOqpp8Rll8sFSZKQkJCAjRs3BnzT1q9fjwceeABdXV0oKysT19fW1kKSJJSUlKCvr2/cr1MVjjEcbePGjXj77bdFIvFn5syZOHToEACgvr5es8KnubkZsiwjPz8fTqcTgKcH1t13341ly5YhIyNjXCvdwimWhw8fhtFoRF5ent/Hq6+vR2Zmpt8EOlEWiwXTpk1Dbm4uZs2aNeb9wyWeAwMDqKmpwYYNG3xuGx4ehsViwdy5c3HkyBGf2wcHB3HvvfeKY3f69Oni2DSbzSLR1tfXw+Vy+XwI7t27FytWrBj361dNuLSjxWLBhQsXcPHiRTidThiNRtxzzz1wu92w2+1obm6GwWDA+fPnceXKlXE95okTJ3Dq1Cmx6OCff/7B008/7XM/726041VbW4vW1lZ8//334rpp06ahs7MTfX19KCoqwjPPPOM38W7duhWffvopKioqsHfvXqxduxYFBQUoKSnB5MmTUVVVhdLSUk0NzvEItxiO1tTUhIiICMyZMyfgOUjA82FXUVGByspKFBQUiMahmZmZ6OnpQXR0NPbt2webzYb29nZcv34dDocDHR0diIuLQ2FhIerq6mC1WgM+RzjFMi8vD3l5eWhpacEbb7whuvkCnnOzdrsdDQ0NN/SYY/nxxx/x559/wmq14vTp05rKZP6ESzztdjveeustv524e3t7ERcXB6fTiYULF+LEiRNiyf5YNm/ejFdeeQVffPEFFi5ciNjYWM1/CJ2dnYiIiIAsy+PeV9VN1dMtKipCXV0durq6sHz5cgBATU0NLl++LE5sx8bGatZjA8CkSZOgKIq4rN6uKAoWL16Mr7/+OujzlpeXo6mpyef6DRs24IUXXvC5vqWlBZWVlWhsbPT7b+r9998Ps9kMh8OB4uJiv88ZGRmJkpISvPPOO1i7dq2mJsOLL76I999/P+g+BxIuMfTnhx9+wDfffIOkpCRcvXoVf/31F1avXo2vvvpKc79AbcLVUzdqHF5++WVcu3YN06dPR2ZmJmbMmAEAyM/PR1tbW9Ckqz5GOMVy3rx56O3txcDAAGJjY+FyuVBYWIgdO3aIVXuhFBMTg+zsbBw9enTMpAuERzxPnjyJb7/9FoDnW+/hw4dhMBhQUFAgaqXMnDkTs2fPxpkzZzRJ97777sOlS5egKAoiIyNx4cIF0dk6Pj5enGoYHBzE/v37NYl9oqcWANzcQFpbWxstWLCAHnzwQVFaraqqSlSGOnjwIAEQZdzUE+0Oh4Mef/xxUhSFuru7acqUKdTe3k4ul4sSEhLo559/JiLPIJb6+0SdO3eOkpOT6ezZs5rrf/vtNzHAdunSJTIajdTZ2emzvdPpFL/b7XZ6/fXXiYjo999/F9fv2rWLFi1aFHQ/EGAgLRxiqBq9xt9bsEGx/v5+IiK6fv06Wa1WqqurIyLSDFw6HA565JFHxP3S09NpYGCAFEUhm80mBjVU4RrL8+fPi4G0jo4Oio+PJ0VRaHh4mLKyssRAzlhGD6SpRr9Hg4ODolLelStX6IknnqCDBw/6bBeu8fTm3aX4jz/+EIOvfX19lJSUJAYqvRUVFWkG0mpqaojIMzCpKAoREVVUVNDmzZs12z388MN07ty5gPuCUJZ2HC0lJYXy8/PF5YsXL1JWVhalpaXR+vXrKTEx0edNURSFSkpKyGg0ks1mo6ysLFEc49ixY6I9sizL9N133wV9/rFYrVaKjY0VpeEKCwuJyHNgmM1mkiSJzGYzVVdXi22WLFkiyjg++eSTon30ypUrxWyGTZs2kclkIkmSKDc3l7q7u4PuR6CkS3T7xzBYm3DV6KS7bt06am1tJaLAbcI//vhjMplMJMsyzZs3j06ePClua2hoILPZLGKgJipVuMbyo48+otTUVJJlmR577DFyOBwiFlFRUZq6r2qy9D4eP/nkE4qPjyeDwUBxcXFUVlZGRIHfI6fTSRkZGWQ2m8lkMgWsJRuu8fTmnXRbWlooLS2NJEkiSZI0H9re8fzpp5/IYrHQQw89RM8995yo27tnzx5KSUmhlJQUKisr09TIbm1tpfT09KD7Eizp8oo0nfCKtNDiFVShxfEMLV6RxhhjtwlOuowxpiNOuowxpiNOuowxpqMx5+lGRUX1R0RE3Nj6OuYjKiqqHwA4lqExefLkkC7JvtNxPENL/Xv3Z8zZC4wxxkKHTy8wxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiOOOkyxpiO/gV3I4morjABpQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lN-fUYfakt5R",
        "colab_type": "text"
      },
      "source": [
        "Plus ce paramètre est grand, plus le modèle sera performant sur le jeu d'entraînement, mais plus vous dégraderez les performances sur le jeu de test. \n",
        "\n",
        "Plus ce paramètre sera petit, plus le modèle sera simple et généralisable. Cette meilleure généralisation augmentera les performances du modèle sur le jeu de test.\n",
        "\n",
        "Attention, si vous réduisez trop ce paramètre, vous dégraderez aussi les performances global du modèle. Tester plusieurs valeurs afin de maximiser vos performances sur votre jeu de tests."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wsAvSTZOJgwg",
        "colab_type": "text"
      },
      "source": [
        "### min_impurity_decrease\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Pm71vNUqkqkT",
        "colab_type": "text"
      },
      "source": [
        "Ce paramètre permet de contrôler si le gain de pureté est suffisant pour continuer la construction de l'arbre dans cette voie. \n",
        "\n",
        "Ce paramètre peut être intéressant, car il est possible d'avoir des règles supplémentaires qui complexifie le modèle pour ne faire gagner que peu de pureté.\n",
        "\n",
        "Plus la valeur de ce paramètre est bas plus l'arbre sera complexe et moins il sera généralisable.\n",
        "\n",
        "Plus la valeur de ce paramètre est élevée plus l'arbre sera simple et généralisable. "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LoQzAQQLJiGI",
        "colab_type": "text"
      },
      "source": [
        "### min_impurity_split\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aITyOyK4la8i",
        "colab_type": "text"
      },
      "source": [
        "Ce paramètre n'est plus maintenu ne l'utilisait pas préférez utiliser min_impurity_decrease."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "c-e34RVsJjyd",
        "colab_type": "text"
      },
      "source": [
        "### presort\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6MufklbsllDq",
        "colab_type": "text"
      },
      "source": [
        "N'utilisez pas ce paramètre, il n'est plus maintenu et il sera supprimé dans la version 0.24."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hqkbcJtIJlOG",
        "colab_type": "text"
      },
      "source": [
        "### ccp_alpha\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HhKc1TNklr3d",
        "colab_type": "text"
      },
      "source": [
        "Ce paramètre initialise le *alpha* de l'algorithme du minimal Cost-Complexity Pruning (CCP).\n",
        "\n",
        "Le CCP est un algorithme qui permet de supprimer des parties de l'arbre qui ajoute de la complexité au modèle sans apporter beaucoup de valeur supplémentaire. Le but de cette suppression est de rendre l'algorithme plus simple et plus généralisable tout en conservant une performance acceptable.\n",
        "\n",
        "Le paramètre alpha est un nombre positif. Plus le paramètre est élevé plus le modèle sera simplifié. \n",
        "\n",
        "Attention si *alpha* est trop élevé votre modèle deviendra trop simple et cela dégradera la performance de votre modèle. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DRqUfB5IpZp1",
        "colab_type": "code",
        "outputId": "60f42abe-c2b8-4ab4-a0f7-0583cb404742",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 247
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_leaf_nodes=30, ccp_alpha=0)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "# Visualisation de l'arbre\n",
        "tree.plot_tree(model_tree)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tSkAduQOpdDH",
        "colab_type": "code",
        "outputId": "2ccd4b44-a8a0-4760-d656-436f356350b8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 247
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_leaf_nodes=30, ccp_alpha=0.2)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "# Visualisation de l'arbre\n",
        "tree.plot_tree(model_tree)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4qfcQf6Wpzds",
        "colab_type": "code",
        "outputId": "53bbf578-6b5b-4bd7-ec32-7e2790df80a0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 247
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_leaf_nodes=30, ccp_alpha=0.4)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "# Visualisation de l'arbre\n",
        "tree.plot_tree(model_tree)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7sh2Tbu-p1g5",
        "colab_type": "code",
        "outputId": "082a6511-ef3f-48a2-d115-df83148b4936",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 247
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_leaf_nodes=30, ccp_alpha=0.6)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "# Visualisation de l'arbre\n",
        "tree.plot_tree(model_tree)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "X1l_h3dQp3pW",
        "colab_type": "code",
        "outputId": "c3ffc1a7-b3b4-4b15-dfdf-94e13635573c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 247
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_leaf_nodes=30, ccp_alpha=0.8)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "# Visualisation de l'arbre\n",
        "tree.plot_tree(model_tree)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vgLuFCEcJmi1",
        "colab_type": "text"
      },
      "source": [
        "## Les attributs\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JWcArqkopwvv",
        "colab_type": "text"
      },
      "source": [
        "Les attributs sont les différentes informations que l'on peut obtenir du modèle une fois qu'il est entraîné."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KnlPnIFtJn6R",
        "colab_type": "text"
      },
      "source": [
        "### feature_importances_\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BSGqLMBqq6rj",
        "colab_type": "text"
      },
      "source": [
        "L'attribut feature_importances permet de retourner l'importance des variables dans votre modèle. L'importance de chaque variable peut être facilement évaluée lorsque votre modèle est simple. Mais plus votre modèle sera complexe, plus cet attribut vous sera utile. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nAeAyf-HqZwi",
        "colab_type": "code",
        "outputId": "91a8bd4b-4d57-4091-99d9-acfde51869e9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 496
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_leaf_nodes=4)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "# Visualisation de l'arbre\n",
        "tree.plot_tree(model_tree)\n",
        "plt.show()\n",
        "\n",
        "feature_importances = pd.Series(model_tree.feature_importances_, index=data_boston.columns)\n",
        "feature_importances"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "CRIM       0.000000\n",
              "ZN         0.000000\n",
              "INDUS      0.000000\n",
              "CHAS       0.000000\n",
              "NOX        0.000000\n",
              "RM         0.753912\n",
              "AGE        0.000000\n",
              "DIS        0.000000\n",
              "RAD        0.000000\n",
              "TAX        0.000000\n",
              "PTRATIO    0.000000\n",
              "B          0.000000\n",
              "LSTAT      0.246088\n",
              "dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rrE2hPKjrDTV",
        "colab_type": "code",
        "outputId": "f49fa1e5-c09c-44cf-dedb-b3c9bdf26ddc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 496
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_leaf_nodes=None)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "# Visualisation de l'arbre\n",
        "tree.plot_tree(model_tree)\n",
        "plt.show()\n",
        "\n",
        "feature_importances = pd.Series(model_tree.feature_importances_, index=data_boston.columns)\n",
        "feature_importances"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "CRIM       0.038564\n",
              "ZN         0.001232\n",
              "INDUS      0.001508\n",
              "CHAS       0.000764\n",
              "NOX        0.049319\n",
              "RM         0.588685\n",
              "AGE        0.013729\n",
              "DIS        0.073138\n",
              "RAD        0.000409\n",
              "TAX        0.013890\n",
              "PTRATIO    0.007930\n",
              "B          0.005165\n",
              "LSTAT      0.205667\n",
              "dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ge-XJmfeJo-P",
        "colab_type": "text"
      },
      "source": [
        "### max_features_\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yjslz9rarwKi",
        "colab_type": "text"
      },
      "source": [
        "L'attribut max_features permet de vous retourner la valeur du paramètre max_features lors de l'entraînement du modèle."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "h10DDa5prSuS",
        "colab_type": "code",
        "outputId": "6f49e91e-a2b5-4c27-ffcb-78d58db54151",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_leaf_nodes=4, max_features=None)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "print(data_boston.shape)\n",
        "print(model_tree.max_features_)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(506, 13)\n",
            "13\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7H9C9gl4r96y",
        "colab_type": "code",
        "outputId": "266b7a09-5ac7-4fec-ed34-35b2766f6583",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_leaf_nodes=4, max_features=5)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "print(data_boston.shape)\n",
        "print(model_tree.max_features_)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(506, 13)\n",
            "5\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TQoQ4DWpqV6E",
        "colab_type": "text"
      },
      "source": [
        "### n_features_"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-mA2pEZOs931",
        "colab_type": "text"
      },
      "source": [
        "L'attribut n_features_ permet de connaître le nombre de variables utilisé pour entraîner le modèle. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "54RHFxBnrlPl",
        "colab_type": "code",
        "outputId": "049fbb78-e548-4490-cd70-27c17504d03d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_leaf_nodes=4)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "print(data_boston.shape)\n",
        "print(model_tree.n_features_)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(506, 13)\n",
            "13\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KMuSAnHLJqCp",
        "colab_type": "text"
      },
      "source": [
        "### n_outputs_\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ed6Hpaiauong",
        "colab_type": "text"
      },
      "source": [
        "L'attribut n_outputs retourne le nombre de sorties du modèle. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Hsw0FL2QsHul",
        "colab_type": "code",
        "outputId": "c7b60f03-aff2-4c79-af16-75606f9f29b4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_leaf_nodes=4)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "model_tree.n_outputs_"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PZWWAD4eJrf6",
        "colab_type": "text"
      },
      "source": [
        "### tree_"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zoxi1hq4u-KF",
        "colab_type": "text"
      },
      "source": [
        "L'attribut tree_ retourne l'objet de l'arbre créé pendant l'entraînement. Vous pouvez vous en servir pour étudier l'arbre plus en profondeur et en retirer les règles de décisions."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SDNokSoisRHJ",
        "colab_type": "code",
        "outputId": "ee44bd90-479a-4eaf-8eb9-c3fd8665be50",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "# Initialisation des paramètres du modèle\n",
        "model_tree = DecisionTreeRegressor(max_leaf_nodes=4)\n",
        "\n",
        "# Entraînement du modèle d'arbre\n",
        "model_tree.fit(data_boston, y)\n",
        "\n",
        "model_tree.tree_"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<sklearn.tree._tree.Tree at 0x7f3f059123d8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RY2h0NZeu-vm",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}