Proyecto_1_Arboles.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Proyecto_1_Arboles.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.8.8"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2z0Ix4DYmVZ3"
      },
      "source": [
        "# Proyecto 1\n",
        "\n",
        "Juan Camilo Sánchez – 201519337\n",
        "\n",
        "Zharet Bautista Montes – 201812763\n",
        "\n",
        "Leonel Andrés Polanía Arias - 201720141"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Lg4zvGGrmVZ8"
      },
      "source": [
        "## 1. Importación de librerias"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AIdqKb1k4EMG",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f750b118-5848-4022-aba5-53a19995e53a"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "IS4wpJ4WmVZ9",
        "outputId": "9c308130-3266-410e-ecd5-232dd6042eb2"
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import re\n",
        "\n",
        "#import contractions\n",
        "import nltk\n",
        "nltk.download(\"stopwords\")\n",
        "nltk.download(\"averaged_perceptron_tagger\")\n",
        "nltk.download(\"universal_tagset\")\n",
        "from nltk.corpus import stopwords\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "from sklearn.feature_extraction.text import CountVectorizer\n",
        "from sklearn.feature_extraction.text import TfidfTransformer\n",
        "# Para crear el arbol de decisión \n",
        "from sklearn.tree import DecisionTreeClassifier \n",
        "# Para evaluar el modelo\n",
        "from sklearn.metrics import confusion_matrix, classification_report, precision_score, recall_score, f1_score, accuracy_score\n",
        "from sklearn.metrics import plot_confusion_matrix\n",
        "# Para búsqueda de hiperparámetros\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "# Para la validación cruzada\n",
        "from sklearn.model_selection import KFold \n",
        "#Librerías para la visualización\n",
        "import matplotlib.pyplot as plt\n"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]   Package stopwords is already up-to-date!\n",
            "[nltk_data] Downloading package averaged_perceptron_tagger to\n",
            "[nltk_data]     /root/nltk_data...\n",
            "[nltk_data]   Package averaged_perceptron_tagger is already up-to-\n",
            "[nltk_data]       date!\n",
            "[nltk_data] Downloading package universal_tagset to /root/nltk_data...\n",
            "[nltk_data]   Package universal_tagset is already up-to-date!\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GVF5mBi8mVaA"
      },
      "source": [
        "## 2. Carga de datos"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TazCfLBOmVaA"
      },
      "source": [
        "df = pd.read_csv('drive/MyDrive/Reviews.csv')"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "k8HrTMZXmVaB",
        "outputId": "24a2d730-2e2a-4153-ea84-4685114b58f2"
      },
      "source": [
        "df.info()"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 568454 entries, 0 to 568453\n",
            "Data columns (total 10 columns):\n",
            " #   Column                  Non-Null Count   Dtype \n",
            "---  ------                  --------------   ----- \n",
            " 0   Id                      568454 non-null  int64 \n",
            " 1   ProductId               568454 non-null  object\n",
            " 2   UserId                  568454 non-null  object\n",
            " 3   ProfileName             568438 non-null  object\n",
            " 4   HelpfulnessNumerator    568454 non-null  int64 \n",
            " 5   HelpfulnessDenominator  568454 non-null  int64 \n",
            " 6   Score                   568454 non-null  int64 \n",
            " 7   Time                    568454 non-null  int64 \n",
            " 8   Summary                 568427 non-null  object\n",
            " 9   Text                    568454 non-null  object\n",
            "dtypes: int64(5), object(5)\n",
            "memory usage: 43.4+ MB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1raHy1men24_"
      },
      "source": [
        "## 3. Preparación de Datos\n",
        "\n",
        "- Pasar a minúscula\n",
        "- Expandir las contracciones \n",
        "- Eliminar stopwords\n",
        "- Concatenar"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JhvAO4N1n25A"
      },
      "source": [
        "STOPWORDS = set(stopwords.words('english'))\n",
        "def lower(text):\n",
        "    return text.lower()"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6kQW3m88n25A"
      },
      "source": [
        "df[\"Words\"] = df[\"Text\"] + df[\"Summary\"]\n",
        "df_clean=df.drop(['Id', 'ProductId', 'UserId', 'ProfileName' , 'Time',\"Text\",\"Summary\"], axis=1)\n",
        "df_clean['Words'] = df_clean['Words'].astype(str).apply(lower)"
      ],
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 359
        },
        "id": "h2-_hfmKsq8S",
        "outputId": "5152bc8f-8d8e-466f-c2b3-a1b4e5d8583e"
      },
      "source": [
        "df_clean.head(10)"
      ],
      "execution_count": 8,
      "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>HelpfulnessNumerator</th>\n",
              "      <th>HelpfulnessDenominator</th>\n",
              "      <th>Score</th>\n",
              "      <th>Words</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>5</td>\n",
              "      <td>i have bought several of the vitality canned d...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>product arrived labeled as jumbo salted peanut...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>this is a confection that has been around a fe...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>if you are looking for the secret ingredient i...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>5</td>\n",
              "      <td>great taffy at a great price.  there was a wid...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>i got a wild hair for taffy and ordered this f...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>5</td>\n",
              "      <td>this saltwater taffy had great flavors and was...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>5</td>\n",
              "      <td>this taffy is so good.  it is very soft and ch...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>5</td>\n",
              "      <td>right now i'm mostly just sprouting this so my...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>5</td>\n",
              "      <td>this is a very healthy dog food. good for thei...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   HelpfulnessNumerator  ...                                              Words\n",
              "0                     1  ...  i have bought several of the vitality canned d...\n",
              "1                     0  ...  product arrived labeled as jumbo salted peanut...\n",
              "2                     1  ...  this is a confection that has been around a fe...\n",
              "3                     3  ...  if you are looking for the secret ingredient i...\n",
              "4                     0  ...  great taffy at a great price.  there was a wid...\n",
              "5                     0  ...  i got a wild hair for taffy and ordered this f...\n",
              "6                     0  ...  this saltwater taffy had great flavors and was...\n",
              "7                     0  ...  this taffy is so good.  it is very soft and ch...\n",
              "8                     1  ...  right now i'm mostly just sprouting this so my...\n",
              "9                     0  ...  this is a very healthy dog food. good for thei...\n",
              "\n",
              "[10 rows x 4 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Sz8Aaaq_ov63",
        "outputId": "5aedc19c-bf4a-4b7b-8644-9fd74100c7e4"
      },
      "source": [
        "initialVectorizer = CountVectorizer()\n",
        "BoW=initialVectorizer.fit_transform(df_clean[\"Words\"])\n",
        "allwords = initialVectorizer.get_feature_names()\n",
        "print(len(initialVectorizer.get_feature_names()))\n",
        "tags = nltk.pos_tag(allwords, tagset='universal')\n",
        "noValid = [word[0] for word in tags if word[1] not in [\"VERB\",\"ADJ\"]]\n",
        "delete = set.union(set(noValid), STOPWORDS)\n",
        "print(str(len(noValid)))"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "142082\n",
            "116149\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eWDpIhP7uE2B",
        "outputId": "71b54dbf-50c7-4d81-a121-54119e20ff03"
      },
      "source": [
        "BoW=initialVectorizer.fit_transform(df_clean[\"Words\"])\n",
        "BoW.shape"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(568454, 142082)"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xc7uQMFUukal"
      },
      "source": [
        "def dropcols_coo(M, idx_to_drop):\n",
        "    idx_to_drop = np.unique(idx_to_drop)\n",
        "    C = M.tocoo()\n",
        "    keep = ~np.in1d(C.col, idx_to_drop)\n",
        "    C.data, C.row, C.col = C.data[keep], C.row[keep], C.col[keep]\n",
        "    C.col -= idx_to_drop.searchsorted(C.col)    # decrement column indices\n",
        "    C._shape = (C.shape[0], C.shape[1] - len(idx_to_drop))\n",
        "    return C.tocsr()"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uAbtl2kpHeaz",
        "outputId": "8796d5f5-cdfc-4e92-cb65-ec4c86d2fefe"
      },
      "source": [
        "sum_words = BoW.sum(axis=0) \n",
        "words_freq = [word for idx,word in enumerate(initialVectorizer.get_feature_names()) if sum_words[0,idx]<500]\n",
        "print(len(words_freq))\n",
        "delete = np.unique(set.union(delete, words_freq))"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "137582\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GWgmtaW887ap",
        "outputId": "9f08879f-2333-48a2-9188-f082a56672bd"
      },
      "source": [
        "indexs=[]\n",
        "BoWords=[]\n",
        "for i,word in enumerate(initialVectorizer.get_feature_names()):\n",
        "    if word in delete:\n",
        "      indexs.append(i)\n",
        "    else:\n",
        "      BoWords.append(word)\n",
        "newBoW= dropcols_coo(BoW,indexs)\n",
        "print(newBoW.shape)"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(568454, 142082)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SuYs_g-d9a2F",
        "outputId": "acab1984-0d92-4cfd-a0f7-23c2a9e7f013"
      },
      "source": [
        "newBoW.shape"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(568454, 142082)"
            ]
          },
          "metadata": {},
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1CNDtcQkTOeq"
      },
      "source": [
        "##4.Construccion del Modelo"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OW9qks2hEIFV"
      },
      "source": [
        "# Se selecciona la variable objetivo, en este caso \"Score\".\n",
        "Y = df_clean.drop(['HelpfulnessNumerator','HelpfulnessDenominator','Words'], axis=1)\n",
        "X = newBoW"
      ],
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lgerPZo8EIFV"
      },
      "source": [
        "# Dividir los datos en entrenamiento y test\n",
        "X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.2, random_state=0)"
      ],
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6dY1krucEIFa"
      },
      "source": [
        "# se fijan 3 particiones.\n",
        "particiones = KFold(n_splits=3, shuffle=True, random_state = 0)\n",
        "param_grid = {'criterion':['gini','entropy'],'max_depth':[3,4,6,8,10,15,20],'min_samples_split':[2,3,4,5]}\n",
        "# se crea el modelo vacio\n",
        "arbol = DecisionTreeClassifier(random_state=0)\n",
        "bestModel = GridSearchCV(arbol, param_grid, cv=particiones)"
      ],
      "execution_count": 42,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZCXgpVEwEIFa"
      },
      "source": [
        "# Se empieza el ajuste del modelo\n",
        "bestModel.fit(X_train, Y_train)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0X3S0oAdEIFa",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "494d3c7e-5d7f-4307-92e6-8257a30eadc5"
      },
      "source": [
        "# los parametros serian\n",
        "bestModel.best_params_"
      ],
      "execution_count": 44,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'criterion': 'gini', 'max_depth': 20, 'min_samples_split': 4}"
            ]
          },
          "metadata": {},
          "execution_count": 44
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2jPo_MXGEIFb",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "cf8b3666-9a0e-4050-94f2-bc1672c673bc"
      },
      "source": [
        "# se saca el ultimo modelo\n",
        "arbol_def = bestModel.best_estimator_\n",
        "# se pasa el grupo de prueba.\n",
        "y_pred_train = arbol_def.predict(X_train)\n",
        "y_pred_test = arbol_def.predict(X_test)\n",
        "print('Exactitud entrenamiento: %.2f' % accuracy_score(Y_train, y_pred_train))\n",
        "print('Exactitud test: %.2f' % accuracy_score(Y_test, y_pred_test))"
      ],
      "execution_count": 45,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Exactitud entrenamiento: 0.73\n",
            "Exactitud test: 0.70\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "V0nss-HzEIFV",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "dc9eac78-d828-4393-abaf-915b5e3b6878"
      },
      "source": [
        "# Se genera la matriz de confusión\n",
        "confusion_matrix(Y_test, y_pred_test)"
      ],
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[ 5270,   111,   249,   469,  4168],\n",
              "       [ 1822,   616,   393,   549,  2805],\n",
              "       [ 1463,    85,  1803,  1091,  4008],\n",
              "       [ 1023,    92,   440,  3936, 10738],\n",
              "       [ 2388,   157,   552,  1818, 67645]])"
            ]
          },
          "metadata": {},
          "execution_count": 48
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yOjAmEnAEIFW",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 279
        },
        "outputId": "12b16611-baad-411d-fdfc-2738884892c5"
      },
      "source": [
        "# Se puede visualizar la matriz de confusión\n",
        "plot_confusion_matrix(arbol_def, X_test, Y_test, normalize='true')  \n",
        "plt.show()  "
      ],
      "execution_count": 49,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lh6UnvKQ90nb",
        "outputId": "d2e78b62-04fa-449e-f17c-05ff5db01558"
      },
      "source": [
        "print(classification_report(Y_test, y_pred_test, target_names=['1','2','3','4', '5']))"
      ],
      "execution_count": 50,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           1       0.44      0.51      0.47     10267\n",
            "           2       0.58      0.10      0.17      6185\n",
            "           3       0.52      0.21      0.30      8450\n",
            "           4       0.50      0.24      0.33     16229\n",
            "           5       0.76      0.93      0.84     72560\n",
            "\n",
            "    accuracy                           0.70    113691\n",
            "   macro avg       0.56      0.40      0.42    113691\n",
            "weighted avg       0.66      0.70      0.65    113691\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 279
        },
        "id": "T7KnPJ6nMocd",
        "outputId": "3c91c267-e1af-46b6-83fc-d5eef27359a6"
      },
      "source": [
        "# Se puede visualizar la matriz de confusión\n",
        "plot_confusion_matrix(arbol_def, X_train, Y_train, normalize='true')  \n",
        "plt.show()  "
      ],
      "execution_count": 51,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAATUAAAEGCAYAAAAE8QIHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdd3hUVfrA8e+Zkt57JaF3AUWKIEVpKoLrroX9YVkLCwq6FrCAuqKCuq7rCoqLbXUtKDZQaYJUJVTpGBJCep30Pu38/piYZCjJhEzIEM/neeZJZu65974zmbz3nHvuPUdIKVEURekoNO0dgKIoijOppKYoSoeikpqiKB2KSmqKonQoKqkpitKh6No7gMa0vt5SFxrQ3mE4xP1UdXuH0DJCtHcELSI0F0+8xlDP9g7BYabSIsxVla36cCeO9ZaFRRaHyu47VLteSjmpNftrKZdKarrQAKIX3dfeYTik658PtHcILSLc3ds7hBbRXETxZtzRr71DcFjKB6+2ehuFRRZ2r+/kUFltZFJIq3fYQi6V1BRFcX0SsGJt7zDOSSU1RVFaRCIxScean+1BJTVFUVpM1dQURekwJBKLC99eqZKaoigtZkUlNUVROggJWFRSUxSlI1E1NUVROgwJmNQ5NUVROgqJVM1PRVE6EAkW181pKqkpitIytjsKXJdKaoqitJDAgusOOKCSmqIoLWLrKFBJTVGUDsJ2nZpKaoqidCBWVVNTFKWjUDW1NuR5sIyQD7MQVknZ2GBKpoTbLffdWkjwJ9mYg/QAlE4IpXxsMAA6g5HQtzPQFRpBQM68LphDWz8w4eAxZcx8LhutRrL20yA+X2ofk97NytzX0+nev5qyYh2LZsaRl+kGwC2z85g0rQiLVbBsQRT7tvoRGmVk7r/TCQg1g4Q1HwXzzbuhAHTpW80DL2bi5mHFYhYsfSKGxANe5x37ZaNKmPVMOhqNZN1noXz+VtQZsT/6zxS696ukrETH4tndyMtyxzfAxII3k+lxSSU/fBnCm8/E168z6rpCps3ORqOBXT8G8N5Lsecdn12sI4v46/wUNBrJ+i8iWPm2/XZ1eiuPvpRIt74VlJfoWfxwL/KzPOjRv5w5C5MA22DAHy/txM6NtnEMp96WxcSbchEC1q2MYNWH0U6J9XQj4tJ5bPQOtBrJV0d68+7eS+2W39T/KNMGHMEiBVVGPc9uGk1KURA6jYW/j9tCnzADWo2V1cd78u6eS8+xl7YjEVhceCaANktqQoj3gMlAvpTS+UODWiWh72eS/URXzMF6YhacoPJSf0wxHnbFKoYFYvhLzBmrhy1Lo/iGCKr7+yJqLE4Z7lqjkdy/KIsnbu2CIUfPkjVJJKz3Jz2pIaaJ04qoKNHxlxG9GT21mLsXZLNoZjydutcwZmoJM8b2JCjcxIufpXD3SF8sZsHyhVEkH/bC09vC0nUn2L/Nl/QkD+5ZkM1Hr4azd7Mfl19Vxt0Lspn3p27nH/vCNJ68rSeGXDdeX3WUhI2BpCc3DFU98eYCKkq13DV2AKMnF3LX4xksntMNY62GD1+NIa5HNfE9q+rL+waYuOeJDOZM6UtpkZ5HXjnJwCtKOfCz//l/yHWx3vf0Sebf1Q9DnjuvrTxAwo9BZJz0boj1T7lUlOm4Z+LljLo2n7seOcWLD/cmLcmLB/80CKtFEBhq5I1v9rNrczCxXaqYeFMuD908EJNJw3NvH2H3liBy0p07VLdGWJk/djszvrqe3ApvVkz7ks0p8aQUBdWXWZPYnZWH+wIwpssp5o76mVnfTGZC95O4aa3c+NEteOhMfHP7Z6xN7EZ2mZ9TY3SEKzc/2zLd/hdos7HJ3ZOrMIW7Yw53B52GiuGBeO8rdWhdfWYNWKC6vy8A0kOLdG/9R9FzUBXZqW7kprtjNmnYsiqA4RPtYxo+sZQfVgYCsP27AAaOrAAkwyeWsmVVACajhrwMd7JT3eg5qIqifD3Jh221r+pKLRnJHoREmmxxS/D2tQ3W5+1noShPf/6xD6ggJ82d3AwPzCYNW78NZvj4YvvYxxez8UtbrWb72iAGXlEGSGqrtRzd64up1v6LHtmplqxUD0qLbHEd+MmfEZPst3k+elxSTna6B7mZnphNGratCWX41UV2ZYZdXcjGb2y15B3rQxkwvMQWa40Wq8UWp5ubld/u9ontUkXiId/65Uf2+DNivKHVsZ6uf0Q+6aX+ZJb5YbZqWXuiG2O7ptqVqTS61f/uqTfX/y4ReOpNaIUVd50Fk0VDRa0bF5pEYJRahx7toc1qalLKbUKI+Lbavq7YhDm44Z/YHKTHI7nqjHLee0rw+LUCU6Q7htuisQS7oc+pweqtJfxfp9DnG6nu50PhtCho5WQfwREmCrIbvmSGHD29LrWPKSTCTEG2LW6rRVBZpsUvyEJIpInj+7wbretGcITJbt3wGCNd+1Xz635bknvr6WgWfZrCvU/nIITkoSndWxd7TkPz25DrRs+BFfZlwhvKWC2CynItfoFmyorPnkyzUz2I6VJNeHQtBbluDB9fjM6t9ZeiB4fXYjg91gHl9mXCjHaxVpXr8AswU1aip+clZfzthSTComp45bGeWC2CtCRv7ngoDd8AE8YaDYNHF5F0xKfVsZ4uzLuS3PKGv3NeuTeXROSfUe7WS45w+6UH0Wst3P3lFAB+SOrC2C6p/HjvB3jozfxj6wjKaj3OWLet2S6+/R02Px0lhJgBzADQhrSuWXK6ykv9Kb8iEPQa/DYZCF+WTvaCbggrePxaQeainphD3Ah/PRXfrUX159tckYeXhafeSeWtp6OoqrAdASffUch/nolix5oARl1fwsOvZvD4LV3bOdIGFWU6lj4VzxNLk5FWOLbfl8hONe0dFomH/Jh1/WXEdqni4RcT2bstiIwUL1a+HcPz7x6htkpDynHv+hpde1hxqB8rDvXj2p4nmDFkHws2XE2/8HysUnD1O7fj517Lf2/6hoT0GDLbofnpyh0F7Z5upZTLpZSDpZSDtb7eza9QxxyoR1fYUJPRFZnqOwR+Y/XVgd72FsvGBuN2ylZrMgfpMcZ52pquWkHlYH/cU1s/5V1hrp7QKGP985BIE4Yc+5gMuTpCo2xxa7QSbz8LZUVaDDmnr2ukMNe2rlYneeqdVH78KpCf1jZMITj+piJ2rLEdCLZ960+PgWfWVFsUe2Rtw/4jjBTm2jdtCvMaymi0Em9fC2XFTR8Xd20K5G9/6MtDf+xLZooHWadaX7MozHMn5PRY8+w7eQrz3exi9fI1U1ZiH2tGihc1VVrie1QCsOHLCB784yDm3TaAijI9WanOn/ouv9KbCN/K+ufhvpXkVZ77e782sTtX1TVPr+uVxI7UWMxWLUXVXhzIiaRv+Jm1vLYmpcAiNQ492kO7J7XzVdvVC31uLbr8WjBb8dlZTOVl9kcsbXFD0vPeV4op2qN+XU2VBU2Z7XyF59EKjNGt7/lMPOBFdGcj4bG16PRWxkwtIWGDfe0zYYM/42+ynVe6cnIJB3f4AIKEDf6MmVqC3s1KeGwt0Z2NJP7iBUge/mcGGUkefLU81G5bhXl6Lhlu+wcZOLKC7FPn/x4SD/kQFV9LeIwt9tHXF5Kw0X4O1oSNgYz7o+0805XXFHFwpx80c8T2D7b9DXz8zEyens+6z0KbLO+IE4d9iYqrITy6Bp3eyqhrC0j4MciuzK4fgxl3Qx4AIycWcCghABCER9eg0dqawGFRNcR0qSYv0/a98A+yHVRCI2u4YryBLd+FtTrW0x3JDSMuoIRovzJ0GgvX9Ehmy8l4uzKdAkrqfx/VOY30Ett3KKfcl6GxWQB46kxcEpHHqeJAp8foCCvCoUd7aPfm53nTCgx3xhD5Yortko4xQZhiPAlcmUNtFy+qLvPHf30B3vvKkFqw+ujI/2vdXIUaQeH/RRP1QjIAtZ09Kbuq9U1Pq0XwxvxoFn2SgkYLG1YEkXbCg9vn5nLioCcJG/xZ92kQ815P5/2fjlNeomXRrDgA0k54sO3bAJZvScRiESx9MhqrVdB3SAXjbiom5ZgHb/6QCMD7iyPZ86Mfr82NYdbCbLRaibFWw2tzz+zlbUnsbz4Txwsf/opGAxtWhpKW5MVtD2WSdNibhI2BrPsslHn/Osl7mw9SXqpj8ZyGpu4H2w/g5WNBp5cMH1/M/Nt7kZ7syayn0+jc21aD/OT1aLJOtb72Y7UIlj3XleffPYJGI9nwZTjpyd5Mn5NK0hFfdm0OZv0XETz6ciLvrN9DeamOlx7uBUDfy0q56d5MzGaBtMKbz3alrMRWI57/+nH8AkyYzRreXNiVynLn/3tYpIZFm6/krT98h1ZIvj7ai5NFQdw/bDdH80PZktKZaQOOMKxTJmarhrIad+avvwqATw/24/nxP/L1bSsQwDfHenLCcOFPmdg6Clw3dQjZRoO9CSE+BcYAIUAe8IyU8t2m1nHvEi3VZMZtQ01m3HYyZl5ckxlX52S0qgrVrb+X/OeqHg6VvaHrwX1SysGt2V9LtWXv57S22raiKO3L4sLXqbluHVJRFJf0u72jQFGUjsvaTj2bjlBJTVGUFrHd0O66Sc11I1MUxSVJBCapdejRHCHEJCFEohAiWQjx+FmWdxJCbBZC/CKEOCSEuLa5baqamqIoLSIlTrmwVgihBd4AxgOZwB4hxGop5bFGxRYAn0splwkh+gBrgPimtqtqaoqitJBjF946cPHtECBZSpkipTQCK4Cpp5WRwG9X1fsD2c1tVNXUFEVpEUmLamohQoi9jZ4vl1Iur/s9GshotCwTGHra+n8HNggh5gDewLjmdqiSmqIoLdaCjgJDKy++nQb8V0r5TyHEcOB/Qoh+UspzztKnkpqiKC0iEc4aJDILaDxkcUzda43dTd24jFLKnUIID2x3KZ3zTn51Tk1RlBaxTZGnc+jRjD1AdyFEZyGEG3ArsPq0MunA1QBCiN6AB1DQ1EZVTU1RlBZyzmTGUkqzEGI2sB7QAu9JKY8KIRYCe6WUq4FHgLeFEA9hy6d3ymZuWFdJTVGUFpE4744CKeUabJdpNH7t6Ua/HwNGtGSbKqkpitJirjzyrUpqiqK0iJRC3fupKErHYesoaJ+ZohyhkpqiKC0k2m3+AUe4VFLTlWoI+f7CT/l1PuSIge0dQotoEzOaL+RChKfzJz1pK2av9o7Acc7IRbaOAnVOTVGUDsSVhx5SSU1RlBZx4h0FbUIlNUVRWkzN0K4oSochJZisKqkpitJB2JqfKqkpitKBqDsKFEXpMNQlHYqidDCq+akoSgfjwPwD7UYlNUVRWsTW+6nu/VQUpYNQF98qitLhqOanoigdhur9VBSlw1G9n4qidBhSCswqqSmK0pGo5mcbGdYznYem/IxGI1m9uxf/2zzIbvm0UYeYMuQ4FquG4goPXvh8DLklvgDMvi6BK3qloxGS3UkxvLrqCmjDk5+DB2Yx6y+70Wgk6zZ157Nv+tst7987l5l37qFLXDGLXhvF9oR4AAb0zWHmnXvqy8VGlbLotdH8vKeT02O8bEQhf30sCY1Gsv6rSFa+F2+3XKe38ugLx+jWp5zyUj2L5/YlP9uTsKhq/vPNLjJTbaMlJh7yY+nzvQBYuOwAQSFGtFrJ0f3+vLmoJ1Zr6z/ny4blM+PhY2g0kg2rY1n5YbfTYrXwyDMH6darlPJSN15cMIj8HC+0WisPzD9Et55laLVWNq2NYeUHtnUfXHCQISPyKSl24/4/j251jOcyMiadJ6/YgUZIvvi1N+8cvPSs5cZ3Psnr4zfwp6/+yFFDGFdEZ/DwkAT0Wismi4Z/7BrOruyYNovzXH6359SEELHAh0A4ts9huZTy387avkZYefQPP/HA8uvIL/Xm/Qe+YvvReFLzA+vLJGYFc+e/b6TWpOfG4UeZfV0CCz4eT/+4XC6Jz2X6q38C4D/3r+LSLjnsT4lyVnj2sWqszL47gcefm4ChyIsli79n595Y0jMD6svkG3x45Y0R/GnKUbt1Dx6NZNbcKQD4+tTy/pKv2HfQ+XFqNJL7nkxk/oxBGPLcee3TvSRsCSUjxbu+zMQbs6ko03HP5OGMmpTHXX87yYvz+gGQk+nJnJuHnLHdxY/2o7pSB0jmv3qEkRPy2bYuvNWxzpp7lAVzhmLI9+Bf/91BwvZwMk75NsQ6JYOKcj33/mkso8Zn85f7f+WlBZcy8uoc9G5W7v+/Ubi7W1i2YitbN0SRn+PFxu9i+G5lPA8/c6BV8TUZu7Dy1Mjt3P399eRVevP5H75kc1o8J0uC7Mp56Y3c3u8wB/PC6l8rrvFg1vprKajypntgIW9f+z1jPr69zWJtiisntbZsGJuBR6SUfYBhwP1CiD7O2nifTvlkGvzILvLDbNHyw4FujOqbaldm/8loak16AI6khRMWUAnYMqybzoJea0Wvs6DTWCmqaLvho3t2M5Cd60duvi9ms5atP3XmisH2w2vnFfhwKj0I2cSX5cphaez9JZpao/OPRT36lZGd7kVulidms4Zt68IYPtZ+IuxhYwxsXB0JwI4fQhkwtBjbp3lutoQGWp1Ep7c2V9yxWPuUkJ3pRW62ly3WH6IYNirPrszQUXls+t5Wi9nxYwQDLjfUx+rhYUGjteLmbsFs1lBVF+PRA8GUl+lbH2ATLgnNJ73Un8xyP0xWLWtOduOq+NQzyj04eDfvHBhEraXhb328MJSCKttBJqk4CHetGb3G0qbxns1v16k58mgPbZbUpJQ5Usr9db+XA8eBaGdtP9SvivwSn/rn+aXehPpXnrP89UN+ZeevtibbkbQI9p2M4run/8f3T33ErhOxdjU8ZwsJqqKgsKHGU1DkRXDwuWM9lzEjTrF5R2dnhlYvOLwWQ557/XNDnjvBYbVnlCmoK2O1aKiq0OIXYAIgIrqaJZ/t5qX39tP30hK79Z5bdoBPtuygulLHjh/CaK3gsBoMeQ0HIUO+B8GhNfZlQmsoyPdoFKseP38TOzZFUlOj5aPvN/Hf1T/y1cddqChza3VMjgrzriS3suG7kFfpTbi3/XehT3ABET4VbM2IO+d2JnRO4bghpN2u7LciHHq0hwvShSGEiAcGAbsuxP5ON+nSE/SOKeCjLQMAiAkuJT6shCnPT+f656dzWbcsBnTOaY/QHBYUUEV8p2L2HnTaccFpigrcuWPCCObcMoS3/9GNeS8exdPbXL/8qVkDmX7VCPRuVgYMKW7HSKFH3xKsFsFt113NXX8Yyx/+nEJEVFW7xtSYQPLY8J95aecV5yzTLbCIR4Ym8Mz2tjvv1xQpwWzVOPRoD22+VyGED/Al8DcpZdlZls8QQuwVQuw11zpeeyko8yIsoKL+eZh/JQWl3meUu7x7Jnde9Qtz35+EyWI7qo3ud4oj6WFUG/VUG/Xs/LUT/ePyzljXWQxFXoQ2qpmFBlVRWHhmrE0ZdUUqP+/uhMXSNn+ywjx3QsIbamYh4bUU5rufUSa0roxGa8XLx0JZiR6zSUN5qa3Zlnzcj5wMT2Li7BOFyahl5+YQhp3WpD2vWPM9CAmvbog1rIbCAvtZyAoLPAgNq2kUq4myUj1jJmazLyEUi0VDabE7xw4F0q23fc2yLeVXehPRqGYW7l1JXqOam7feSPegIj68fjUbp33EgLA83py4lr4h+XXlK1gyfh2Pb76KjHL/Cxb36X6XzU8AIYQeW0L7WEr51dnKSCmXSykHSykH69wd/0c/nhFGbEgpkYFl6LQWxg9MZvsx++p6jygDj/1xO3P/O4niyobmSl6JD5d2yUGrsaLVWBjUJZvUvIDTd+E0ickhREeWERFWjk5nYfSIU+zc27Jeq7Ft2PQEOHHUl6i4KsKjq9HprIyalE/ClhC7Mru2hDBuiq1GO3J8AYd2BwICv0AjGo3tfFVEdDVRnarIyfTEw9NMYEhDEhxyZSEZp1qWzM8a63F/omMrCY+sssU6Pptd2+w7H3ZtD+fq6zJtsV6Vy6G9IYCgINeTAYMLAXD3MNOrXwmZaT6n76LNHC4II86/hGjfMvQaC9d2TWZzWnz98gqTO1d8+BfGfTqdcZ9O52B+OPetv4ajhjB83Wp5a9IaXt09jF/yIi9YzKdz9XNqbdn7KYB3geNSyledvX2LVcMr34zk3/euQaORfLe7J6fygrh3wh5+zQxl+7F45kxOwMvNxAu3/QBAXrEPc/87iR8PdeGybtl8/PBKJJCQGMuO4/HODrGe1aph6btDWTR/IxqNlfWbu5OWGcjtt/zCiZPBJOztRI+uBp6ZuxlfbyPDLsvktpsPMOPhGwAID60gNKSSQ8ci2i5Gi4Zli3rw/LIDaLSSDd9EkX7Sh+n3pZB0zJddW0JZ/3Ukjy46xjvf7aS8VMdLdT2f/S8rYfp9pzCbBVLC0ud7UVGmJyDIyDOvH0LvZkVo4NDuQNasbH3PrdWiYdkr/XjuddslMj98G0P6KV+mz0gk6XgAu7aHs2F1LI/+/QBvf7GZ8jI9Ly+wXTbx3RdxPPTUQd78dCtCwA/fxZCa7AfAvOd+of+lhfgFGPng2018vLw7G7517qUzFqnh+Z+u5J1rvkOjkXyV2Ivk4iDmXLabI4ZQNqed+8D1f32P0MmvlFmX7mXWpXsBuGfNZIpqLvzEo011aLU3IaUTuqPOtmEhRgLbgcOAte7lJ6WUa861jndwrOx77d/aJB5n80utab6QC9GpyYzbTMrdzr9msK2k/edVarIyWpWRfHtGyEFv3uZQ2e3jXtknpRzcmv21VJvV1KSUO2jLq1kVRWkXUrr2dWoX9R0FiqK0B4FFTZGnKEpH4srn1FRSUxSlRX63934qitJBSdt5NVelkpqiKC2mhvNWFKXDkC7eUeC6kSmK4rKkdOzRHCHEJCFEohAiWQjx+DnK3CyEOCaEOCqE+KS5baqamqIoLeaM3k8hhBZ4AxgPZAJ7hBCrpZTHGpXpDjwBjJBSFgshmh3mRdXUFEVpEVstTDj0aMYQIFlKmSKlNAIrgKmnlbkXeENKWWzbt8xvbqMqqSmK0mItuKE95LdReOoeMxptJhpofP9eJmeOudgD6CGE+EkIkSCEmNRcbKr5qShKi7Xgkg5DK+/91AHdgTFADLBNCNFfSnnO8aJUUlMUpUUkAqtzej+zgNhGz2PqXmssE9glpTQBp4QQJ7AluT2cg2p+KorSYtLBRzP2AN2FEJ2FEG7ArcDq08p8g62WhhAiBFtzNKWpjaqamqIoLSOd0/sppTQLIWYD6wEt8J6U8qgQYiGwV0q5um7ZBCHEMcACzJVSFja1XZXUFEVpOSfdJlU3vuKa0157utHvEni47uEQldQURWmxi3KUDiHEEprIx1LKB5wdjLbWil9KdfMFXYBIONTeIbRI4fRh7R1CiwTvNbR3CA6zuLnw3d2nc0IukoDVehEmNWDvBYtCUZSLhwQuxpqalPKDxs+FEF5SSteZIFFRlHbjykMPNXtJhxBieF3Pw691zwcIId5s88gURXFdTrqmoy04cp3aa8BEoBBASnkQGNWWQSmK4socu++zvToTHOr9lFJm2KbxrGdpm3AURbkouHDz05GkliGEuAKQdTOuPwgcb9uwFEVxWRKkC/d+OtL8nAncj+3u+WxgYN1zRVF+t4SDjwuv2ZqalNIA/N8FiEVRlIuFCzc/Hen97CKE+FYIUSCEyBdCrBJCdLkQwSmK4qIu8t7PT4DPgUggClgJfNqWQSmK4sJ+u/jWkUc7cCSpeUkp/yelNNc9PgI82jowRVFcl7MmXmkLTd37GVT369q6WV5WYMvRt3DaXfWKovzOuHDvZ1MdBfuwJbHfov9ro2US2wwviqL8DgkX7iho6t7PzhcyEEVRLhLt2AngCIfuKBBC9AP60OhcmpTyw7YKSlEUV9Z+nQCOaDapCSGewTZGeB9s59KuAXYAKqkpyu+VC9fUHOn9/BNwNZArpfwLMADwb9OoFEVxbVYHH+3AkeZntZTSKoQwCyH8gHzsp7VqN4MHZjHrrj1oNJJ1m7rx2df97Zb375PHzL/soUtcMYteHcX2hLj6ZaEhFTw8ayehIVVICQteuJq8Ap/WxzSmjJkLs9BqJGs/DebzN8LtluvdrMz9dzrd+1dRVqxj0aw48jLdAbhldh6Tbi3EYhUseyqafVv96tfTaCRL1p6gMFfP03f8du2z5M7HcrlycglWC3z3YQir3gtt9XsAGNYjnYen/IRGSFbv6c2HWwbZLZ925UGmXv4rZqugpNKT51eOIbfEF4DZ1+xkRO90hJDsTorh1dUjaMtbZi67PJe/3n8AjUayfk1nVq7oZbe8X/8CZtx/kM5dSnnx+aH8tC2mftnCxdvp1aeIY0eC+fv8kW0WY2NXRqczf8hPaIVkZVJvlh8edNZyE+JSWDp2Azd+eyNHCsPqX4/0LmfNDZ+x5MBg3js68ILEbOdiHSSykb1CiADgbWw9ohXAzuZWEkJ4ANsA97r9fCGlfKYVsdrRaKzMvncXjy8cj6HQiyUvrWHnnljSMwPqy+QXePPK0hH8acrRM9afN+cnPv2yP/sPReHhYXLKDboajeT+FzJ5YlpXDDl6lqw5QcIGf9KTGi7rmzitiIpSLX8Z2YfRU4q5e34Oi2bF06l7DWOmFjPjql4EhZt4ccVJ7r6yd/2wyTfcU0BGkjtevg2Hvwk3FxEaZeSeUb2QUuAfbGr1ewDQCCtzb9jBnHcmk1/qzX9nf8X2Y3Gcyg+qL3MiK4Q7Em6k1qTnxmFHmX1tAgs+GU//uFwuic/l//51EwDLZ63i0i7Z7E85feJt59BoJPc98Avz512JocCL197cRMLOKDLSGg4I+flevPryYP5404kz1v/y8x64e1i4dnKTs645L15h5ZmhO/jLhsnkVnnz5eSv2JQex8nSILty3jojd/Q+zIGCsDO28cTlO9mW1emCxHsurtz72WzzU0p5n5SyREr5FjAeuKOuGdqcWuAqKeUAbDfBTxJCOG2g/J7dCsnO9SU3zxezWcvWHfFccXmGXZm8Ah9OpQWeMa5Tp5gStFor+w9FAVBTo6fW2Po5aHoOqiI71Z3cdHfMJg1bVgUyfGKpXZnhE0r5YaXtC7z9+wAGjiwHJMMnlrJlVSAmo4a8DMfFuuUAACAASURBVHeyU93pOcg20HBIpJEhV5ex9tNgu21Nvr2Qj/8VUf/+Sgv1rX4PAH1i88ks9CO7yA+zRcsPB7syqk+qXZl9KdHUmmz7O5IeTph/BWC74NJdZ0GvtaLXWdBprRRVeDklrrPp0auI7CwfcnN8MJs1bNscy/Arsu3K5Od5k5oSgPUstYuDv4RTXXXh5h+6JCSftHI/Mir8MFm1fH+qK+M6pZ5R7sFL9/D2kYHUWrR2r4/rdIrMCl+SSwIvUMTncDHeJiWEuPT0BxAE6Op+b5K0qah7qq97OO1thgRVUWDwrn9eUORFcLBjo43HRJVRUenG03O38OY/vuXe2/ei0bT+BEBwhImC7IbEYsjRExJhX3sKaVTGahFUlmnxC7TYvf7busF16858Not3no9CnhZiZHwto6cUs2RNIs//7yRRnWtb/R4AwvwryStpaIrnl/oQ6l95zvJTLj/OzkRbzeFIegT7UqL5fsGHrFnwPxJOxJCa33b/gMEh1RgKPOufGwo8CQ5x3cl7wr0qya1s+GxzK30I97L/bPsEFRDpVcGWzDi71710Ju7td4ClBwZfkFgvVk0dov7ZxDIJXNXcxoUQWmxN1m7AG1LKXWcpMwOYAeDhdmH6H7QaSf/e+cyaO5n8Am/mP7KNCWNPsm5T9wuy/5YYOq6UEoOO5MNeXDK83G6Z3k1irNUw59qejLimhEf+mc4jN17Y9zBp0Al6xxQw862pAMQElxIfWsz1i24DYMk93zEwPocDqZEXNK6LlUDyxJCfeXzH2DOWzRm4l/8e60+V2Tk18tZw5eZnUxffnvmptpCU0gIMrDsn97UQop+U8shpZZYDywH8fKId/qgMRV6EhjQc4UKDqigsdKyZU1DoxcnUIHLzbCe2f94dS+/urZ+SrTBXT2hUQ80sJNKEIdf+C2ioK2PIcUOjlXj7WSgr1ta/3njdwlw9w8aXMmxCGZdfdRQ3d4mXr4V5r6fx8gNxGHL07FhjOxD8tNafR15Nb/V7AMgv9SY8oKL+eZh/BQWl3meUu7xbJndetZ9Zb03FVNdMGtP3FEcywqk22t73zsRY+sXltllSKzR4EhLaUDMLCa2m0ODZxBrtK6/Kmwjvhs82wruCvKqGz9Zbb6RHQDH/m7QagFDPapZdvY5ZmyYxIDSPifEnmTs4AT83I1YpMFp0fPRrvwv7JiQufZuUI5d0tJqUsgTYDExy1jYTk4OJjiwnIqwcnc7C6JGp7NzrWKfsiZPBeHsb8ferAWBgv1zSMltfS0w84EV051rCY2vR6a2MmVpMwgY/uzIJG/wYf1MRAFdeV8LBn3wBQcIGP8ZMLUbvZiU8tpbozrUk/uLF+y9GMX1wX+4Y1pfF98Vx8CdfXn7A1iz5eZ0/A66w/YNcMryCzBT3Vr8HgOOZYcQGlxIZWIZOa2H8gJNsOx5vV6ZHlIHHb9zG3P9OoriyIYnklvgwqHM2Wo0VrcbCoC45bdr8PPFrIFHRFYRHVKLTWRk1NoOEn123VnjYEEa8XykxPmXoNRau63ySTRnx9csrTO4MXXEnV30xnau+mM6BgjBmbZrEkcIw/rz2hvrXPzjWn7cODbrwCe03LnxOrc3OkAohQgGTlLJECOGJrZPhJWdt32rVsPSdISx6aqOtK//HbqRlBHD7rQc4kRxMwt5YenQ18MxjW/D1NjJscAa33XqAGX+bitWq4e0PLuOlv29AAEkpwazd2Ppmm9UieGNBDIs+SUGjkWz4LIi0E57c/mgOJw56kfCDP+tWBDPv9TTe33GM8hIdi+6zJai0E55s+zaA5Zt/xWIRLJ0f0+yEsZ+9EcZjS9O58d4Cqqs0vDbXOT1iFquGV1aN5PW7v0ejkXy7pyen8oKYMX4PxzND2X48njnX7sTLzcSi6T8AtmQ294Nr+PFwFwZ3y+Ljhz4HKdh5IpYdpyVEZ7JaNSxbMpDnX9pu+8zXxpOe5s/0O4+SlBjIrp1RdO9ZxFPP7sTHx8jQ4TlMv+MYs+6eAMDLr20mNrYcD08zH674ntdeuYz9eyPaLF6L1LAwYSTvjv8erZB8kdyT5JIgHhi4hyOFofzYKMG5MldufgrZRuODCCEuAT4AtNhqhJ9LKRc2tY6fT7QcMmBWm8TjbBfbDO0laob2NpN8W0h7h+CwzKX/oiYzo1VtR/fYWBnzt4ccKpvy6CP7pJQXtGfDkdukBLbhvLtIKRcKIToBEVLK3U2tJ6U8BJz9qkJFUS5uLlxTc+Sc2pvAcGBa3fNy4I02i0hRFJcmpOOP9uDIObWhUspLhRC/AEgpi4UQbm0cl6IorsyFez8dSWqmuuvNJNR3ALTTraqKorgCV+4ocKT5+TrwNRAmhHgB27BDi9o0KkVRXNvFfEmHlPJjIcQ+bMMPCeAGKaWaoV1Rfq/a8XyZIxzp/ewEVAHfNn5NSumcy9cVRbn4XMxJDfiehglYPIDOQCLQtw3jUhTFhQkXPqvuSPPTbuTFuhE67muziBRFUVqhxfd+Sin3A0PbIBZFUS4WTuooEEJMEkIkCiGS6+YXPle5PwohpBCi2bsTHDmn9nCjpxrgUiD7HMUVRenonNRRUHep2BvY7gvPBPYIIVZLKY+dVs4XeBA4Y+iys3Gkpubb6OGO7RzbVMdDVxSlw3FOTW0IkCylTJFSGoEVnD23PIdtMIwaR0JrsqZWl0l9pZSPOrIxRVF+JxyvqYUIIfY2er68bgxFgGig8Rj8mZx2aqvuHH6slPJ7IcRcR3Z4zqQmhNBJKc1CiBGOxa4oyu+BoEW9n4bzHaVDCKEBXgXubMl6TdXUdmM7f3ZACLEaWAnUDzUrpfyq5WEqinLRc97Ft1nYT7cZU/fab3yBfsAW22BBRACrhRBTpJSNa392HLlOzQMoxDYnwW/Xq0lAJTVF+b1yTlLbA3QXQnTGlsxuBf5cvwspS4H6weqEEFuAR5tKaNB0Ugur6/k8QkMyq99fS6NXFKUDcUIGqDu9NRtYj20w2feklEeFEAuBvVLK1eez3aaSmhbw4exTa7dJUhO1JnTJF8fVIpY2GjG4rQR+vr+9Q2iR/NuanYXRZXj0KWnvEBwmPCzO2Y6Tvv5SyjXAmtNee/ocZcc4ss2mklpOc8NvK4ryO+XCx/SmkprrjgKnKEr7kRfvvZ9XX7AoFEW5uFyMNTUpZdGFDERRlIvHRT2emqIoyhlUUlMUpcNox6G6HaGSmqIoLSJQzU9FUToYldQURelYVFJTFKVDUUlNUZQO42KfIk9RFOUMKqkpitKRXKy3SSmKopyVan4qitJxqItvFUXpcFRSc57LrjDw18dOoNFI1n8dzcr34u2W6/RWHn3hKN16l1FeqmfxvP7kZ3sSFlXNf77eSWaqFwCJh/1Z+nxvAEZNzOWWe1LRaCW7t4Xw/mvdzzu+wWPKmPlcNlqNZO2nQXy+NNxuud7NytzX0+nev5qyYh2LZsaRl+kGwC2z85g0rQiLVbBsQRT7tvoRGmVk7r/TCQg1g4Q1HwXzzbuhADz5VioxXWsB8PazUFmm5b7xPc879stGlTDrmXQ0Gsm6z0L5/K2oM2J/9J8pdO9XSVmJjsWzu5GX5c6gkaXcNS8DnV5iNgneWdyJgzv9cPewMP+NZCLjarFaBAmbAnj/5dhz7P38De+WzqPX/oRGSL7Z35sPtg+yW/7HwUe5aehRLFZBtVHPC6tHcaogCH/PGl66dQN9ovL57kBPXv7+SqfHdjbu+yvweycPrJKq8QFU/jHkjDIeO8rwWVEAAszxHpQ8Eg1A4LPpuCVWY+zjRfEC53+Wjvjd31FQN83eXiBLSjm5NdvSaCT3PZnI/L8OwpDnwWuf7CZhSwgZKT71ZSb+IYuKMh33XD+CUZNyuetvybw4rz8AOZmezLllmN02ff2N3PVQEg9MG0pZsRsPP3eUAUOKOLg76Lziu39RFk/c2gVDjp4la5JIWO9PepJHQ3zTiqgo0fGXEb0ZPbWYuxdks2hmPJ261zBmagkzxvYkKNzEi5+lcPdIXyxmwfKFUSQf9sLT28LSdSfYv82X9CQPFs2Mr9/ujKezqSx3ZBrXJmJfmMaTt/XEkOvG66uOkrAxkPRkz4bYby6golTLXWMHMHpyIXc9nsHiOd0oK9LxzD09KMp3I65HFS98kMj04bbE8sXbkRxK8EOnt/Lix78yeHQJe7cGnHecZ8QtrDw2eQf3fzCZvDJvPvzrV2z7NY5TBQ1/v3WHu/Pl3r4AjOqZykOTdvLA/66j1qxl2abL6RZWRNfwCzQojUXi959cip7thCVYT8jcU9QO8cUc615fRJttxOdLA4UvxiN9tGhKzPXLKm8IpqrWiteG9h1tV1hdN6ud/3+B4x4EjjtjQz36lZKd4Ululhdms4Zt68IZPqbArsywsQVsXB0JwI4fwhgwpIim6soRMdVkp3tRVmyrLR3YFcSIcfnnFV/PQVVkp7qRm+6O2aRhy6oAhk8stSszfGIpP6wMBGD7dwEMHFkBSIZPLGXLqgBMRg15Ge5kp7rRc1AVRfl6kg/bapfVlVoykj0IiTSdtmfJqCklbP4m8LziBug5oIKcNHdyMzwwmzRs/TaY4eOL7WMfX8zGL221iu1rgxh4RRkgOXnMm6J82+eXdsITdw8rejcrtTVaDiX4AWA2aUg+4k1IpPG8YzybvjH5ZBT5kVXsh9miZcPhrozulWpXprLWrf53TzdT/behxqTnYHoktWatU2Nqij6pGkukG5YIN9ALqkf64b6r3K6M14ZiKq8NRPrY4rIGNNQ9jAO8kZ4X4t+2CY5OZNxOea9NPx0hRAxwHfCOM7YXHFaLIbeh1mPI9yA4vPaMMgV1ZawWDVUVOvwCbEkgIrqaJZ8l8NK7e+k7yPYPm5PuRUx8FWFR1Wi0VoaPzSc0wqGJoM+ML8JEQXbDP5AhR39GAgqJMFOQra+LT1BZpsUvyEJI5OnruhEcYb9ueIyRrv2q+XW/l93r/YZWUlygI/uUO+crOMJEQU7D+oZcN4Ij7BNQcHhDGatFUFmuxS/QbFdm5DXFJB/xxmS0/2p5+5oZenUJB37yO+8YzybMt5K80oaaen6ZD2F+lWeUu2nIEb752yfMmZDAK9+331S22iIzlpCGJGUN1qMtsv8MddlGdFlGgh9PJXjeKdz3V1zoMJslpGOP9tDWzc/XgHnY5u9rV0UF7twxcSTlpW50613GU68dZOaNw6ko17P0hV488fJhrFbB8YP+RMZUt3e4Z/DwsvDUO6m89XQUVRX2NYuxN5Sw5RvnNenOV1z3Ku56LIP5t9uf19NoJY+/fpJV/w0nN8PjHGu3rZW7+7Fydz8m9k/i7tH7+fvXV7VLHA6xgi7HSOHzcWgLTQQ/mUbBv7vU19xcguu2PtuupiaEmAzkSyn3NVNuhhBirxBir9HadDIpzHcnpFEtKiSshsI89zPK/FbT0mitePmYKSvRYzZpKC+11YSSj/uRk+FJTFwVALu3hvLQ9CE8cvvlZKZ6kZVmXxNyVGGuntCohtpNSKQJQ47erowhV0dolKkuPom3n4WyIi2GnNPXNVKYa1tXq5M89U4qP34VyE9r7ZOXRisZcW0pW1e3LqkV5uoJjWyo9YZEGCnMdbMvk9dQRqOVePtaKCvW1Zd/6j9JvPJIF3LS7RPXg4tOkZ3qwTfvR7QqxrPJL/cm3L+hJhPmV0F+mfc5y2840o0xvVOdHoejLEE6tIaGmpmm0IQlyL5uYQnWUTPEF3QCS7gb5ig3dDnObba3livX1Nqy+TkCmCKESAVWAFcJIT46vZCUcrmUcrCUcrCbxvP0xXZOHPUjqlM14dHV6HRWRk3KI2FrqF2ZXVtCGTclB4CR4/M5tDsQEPgFGtFobJ9yRHQVUXHV5GTa9ucfZPvC+PiauO7mTNZ/bd/r56jEA15EdzYSHluLTm9lzNQSEjb425VJ2ODP+JtsTd8rJ5dwcIdtFsKEDf6MmVqC3s1KeGwt0Z2NJP7iBUge/mcGGUkefLU89Ix9XnplORnJ7hhy3M5Y1qLYD/kQFV9LeIwt9tHXF5Kw0T5RJmwMZNwfDbbYryni4E4/QODta2bhe4m8/1Isx/bZV8rveCQTb18Lby3s1Kr4zuVYVhixQaVEBZSh01qY0P8k236NtysTG9RwUn1kjzTSC/1pL6bunmhzjGjzjGCSeO4oo3aI/WdWO9QXtyO2A64oM6PLNmIO159tc+3Hhc+ptVnzU0r5BPAEgBBiDLaZlae3ZptWi4Zli3vy/LJf0GgkG76JIv2kD9PvO0nSUT92bQ1l/ddRPPrCUd759ifKy/S8NK8fAP0vLWb6/SmYTQIpBUuf70VFme2L8td5iXTpYTvaf7K8M1lp5z7SNx2f4I350Sz6JAWNFjasCCLthAe3z83lxEFPEjb4s+7TIOa9ns77Px2nvETLollxAKSd8GDbtwEs35KIxSJY+mQ0Vqug75AKxt1UTMoxD978IRGA9xdHsudH27mp0VOd0/S0WgRvPhPHCx/+ikYDG1aGkpbkxW0PZZJ02JuEjYGs+yyUef86yXubD1JeqmPxnK4ATLkjj6i4Wv78QDZ/fsA2b+uTt/dEr5dMm51NerIHS787CsC3H4ax7rOwVsf7G4tVwz++H8mS279Hq5Gs3t+TlIIg/nrVHo5nhbItMZ6bhx5hSNcszBYN5TXu/P2rsfXrr37oI7zdTei1Fkb3SmX2h9fZ9Zw6nVZQdm8EQc9mgEVSPS4Acyd3fD4pwNTNg9ohvtQO8sbtQCUhs0+CRlB2ZxjSz/avGvxEKtosI5oaK2F3J1EyOxLjIJ9mdupkLj6blJAXYFLeRkmtyUs6/PVhcnjQn9o8HmewFBQ0X8iFCPfz70RoD4aLaDJj4+SLZzLjkw+/Q3Vydqumv/QJjpX9rnnIobK7Pn5kn5RycGv211IX5OJbKeUWYMuF2JeiKBfABagMna+L7o4CRVHa3+/6jgJFUToYdUO7oigdjSt3FKikpihKi6mkpihKxyFRHQWKonQsqqNAUZSORSU1RVE6it/9IJGKonQwUrr0IJEqqSmK0nKum9NUUlMUpeVcufnZzuMCK4py0ZGAVTr2aIYQYpIQIlEIkSyEePwsyx8WQhwTQhwSQmwSQsQ1t02V1BRFaTknjKdWNynTG8A1QB9gmhCiz2nFfgEGSykvAb4AXm4uNJXUFEVpMSeNfDsESJZSpkgpjdgGk53auICUcrOUsqruaQIQ09xG1Tk1RVFarAW9nyFCiL2Nni+XUi6v+z0ayGi0LBMY2sS27gbWNrdDldQURWmZlo3SYXDGIJFCiOnAYGB0c2VdK6lZrcjKM6c3U1pP1tY2X8iFhHzQ5Hw9LmXdwt3tHYLDhni3ftJm28W3Tun+zAIaTzMfU/ea/f6EGAfMB0ZLKZv9IqtzaoqitJzVwUfT9gDdhRCdhRBuwK3A6sYFhBCDgP8AU6SUDs0y7lo1NUVRLgrOqKlJKc1CiNnAekALvCelPCqEWAjslVKuBv4B+AArhRAA6VLKKU1tVyU1RVFaxokj30op1wBrTnvt6Ua/j2vpNlVSUxSlhdS9n4qidDRqkEhFUToMF5/MWCU1RVFaTtXUFEXpUFw3p6mkpihKywmr67Y/VVJTFKVlJI5cWNtuVFJTFKVFBNJZt0m1CZXUFEVpOZXUFEXpUFRSUxSlw1Dn1BRF6WhU76eiKB2IVM1PRVE6EIlKam3lslHFzFyQikYrWfd5OCv/E223XO9m5ZF/JNO9XwVlxXoWP9id/CwPelxSzgPPpwAgBHz8egw//xB83nEMHlPGzOey0Wokaz8N4vOl4WfEMff1dLr3r6asWMeimXHkZboBcMvsPCZNK8JiFSxbEMW+rX7NbFNy52O5XDm5BKtV8N2Hwax6N7R+Xz0GVPHat0ksmhXHju8D2iV2AI1GsmTdCQpz9Dx9RxcAHvpnBj0uqQIBWSnuvPK3WGqqtC37sOtcNrqUWc+k2/72K0L5fFnkGXE/+moK3ftXUVasY/HsruRluuMbYGbBW8n0uKSSH74I4c2nbTOueXpbeGXl8fr1QyJN/Ph1MP9Z2Om84mvKns2+vPVUNBar4Jpphdwyx37sw7xMPa8+3InSQh2+ARbmLUkjNMoEwDUxA4jvVQNAWLSRZz845fT4HOK6rc+2TWpCiFSgHLAAZmeMVf4bjUZy/99P8eQdfTDkuvHvrw6za1Mg6cle9WUm3JRPRamOu6++lNHXGbhrXjovPtiDtBNePPCHS7BaBIGhRt787iAJPwZhtYjzi2NRFk/c2gVDjp4la5JIWO9PepJHfZmJ04qoKNHxlxG9GT21mLsXZLNoZjydutcwZmoJM8b2JCjcxIufpXD3SF+Ac25zwi3FhEaZuGdUL6QU+Aeb7GK5e34O+7b6tlvsVqvtM7zhHgMZSR54+Vjqt/WfZ6KoqrAlsRnPZDHlLsMZSdThuJ9L48n/64Eh143XVx8jYWMA6UmeDXHfYqCiVMddoy9h9PWF3PV4Botnd8NYK/jwlWjielYT37O6vnx1pZb7r+1X/3zJd0f5aV1gi2NrjsUCbzwZw+IVJwmJNDHn2h4Mm1hKXI+GUarfXhjNuD8VMf7mYg7s8OH9xZHMW5IOgJuHlWUbE50eV0u58nVqF2I477FSyoHOTGgAPQZUkJ3mQW6GB2aThq3fhzBsXLFdmeHjitj4ta0Ws31dMAOHlwKS2hptfQJzc7ciZcuT2W96DqoiO9WN3HR3zCYNW1YFMHxiqX0cE0v5YaXtH2T7dwEMHFkBSIZPLGXLqgBMRg15Ge5kp7rRc1BVk9ucfLuBj/8VXh9zaaG+fj9T7zKwY40/JQbHjlVtETtASKSRIVeXsfaTILtt/ZbQQOLuIeE8P/eeAyvJSXVv+Nt/G8Tw8af97ccXs/HLEFvca4IYOKIckNRWazm61xdT7bm/+tGdawgINnFkt895xdeUxF+8iIqvJTLOiN5NMmZqMTvX+9uVSTvhzoARFQAMGFFxxnKXIKVjj3Zw0c5REBJupCDHvf65IdeN4HD7ORmCw40YcmxNJatFUFWhxS/QDEDPAeW8tfYAy74/yNKnOp9XLQ0gOMJEQbZbQxw5ekIiTXZlQiLMFGTr6+OoLNPiF2QhJPL0dd0IjjA1uc3IOCOjp5SwZO0Jnv8ohajOtfVxXHFNKd994Hgzui1iB5j5bDbvPB+JtJ75mT7yr3RWHDxGbLcaVr0X4nCs9nEbKcg5+77P9t6sFkFlecPfvjmjry9k63dB2KYYca7CXH19UxJszVxDjt6uTJc+Nfy01pbIflrrT1WFlrIi2wHBWKth9qQePDi5Oz+vbadkJyVYrI492kFbJzUJbBBC7BNCzDhbASHEDCHEXiHEXqOsaeNwGiQe9GXmNQN58Mb+3DwzC72bC58kaETvLjHWCuZc04O1HwfxyKu2aRNnPpvFuy9EtqrW6QxDx5VRYtCRfNjrrMv/+VAn/jyoD+lJHoyeUnKBo3PM6ClFbFl1/udYW2vG01kc3unDfeN7cHinDyGRRjR1ldz/7T7G0nUnePyNNN56JprsVLemN9ZWXLim1tYdBSOllFlCiDDgByHEr1LKbY0L1E1suhzAXxvi8KdgyHMjNLKhZhYSYaQwz92uTGGeGyGRRgy57mi0Ei8fC2XF9m8546QX1VVa4ntUkXSk5c0N25HX2BDHWY68hlwdoVEmDDluaLQSbz8LZUVaDDmnr2ukMNe27rm2acjRs2NNw1H8kX/ZklqPAdU8sSwNAP8gC0OuLsdiEexcd+6jeVvEPmxCGcMmlHH51cdwc5d4+dpOdL88J66+rNUq2LIqgJvvy2fDZ/ZNVEcU5roRGnn2z+3092bIrYvb98y//dl07l2FVitJPuLd4rgcYatBNsR6ttpxcISZp99NBaC6UsOONf74+NvOTTausV9yRQUnj3gSFW/kgvu9nlOTUmbV/cwHvsY2zbxTnDjkQ1RcDeExNej0VkZfZyBhk/2J3YRNQYz7QwEAV04q5GCCPyAIj6lBo7X9UcKiaontUk1elvvpu3BI4gEvojsbCY+tRae3MmZqCQkb7BNJwgZ/xt9kO+dz5eQSDu7wAQQJG/wZM7UEvZuV8NhaojsbSfzFq8lt/rzOr/58yyXDK8lMscV9x7De3DG0D3cM7cP27/xZ8kR0kwmtrWJ/f3Ek0wfb4lg8K46DO3zqEpokKv63g5Bk+MQyMk56cD4SD3oT1bm2Pu7R1xeR8MNpf/uNAYz7o8EW97VFHPzZF0eak2OmFLJlddvV0noOrCLrlDu56W6YjIItqwIZNqHMrkxpoZbfrm1dsSSMCbfY5uosL9FirBX1ZY7u8aZTjwvXuqknAat07NEO2qymJoTwBjRSyvK63ycAC521fatFsOzZzjz//nG0WsmGlWGkJ3lx24PpnDjiw65NQaz/PIy5/0zi3U37KS/R8eLfegDQd3A5N//1V8wmgZSCN57pQlmxvpk9njuON+ZHs+iTFDRa2LAiiLQTHtw+N5cTBz1J2ODPuk+DmPd6Ou//dJzyEi2LZtlqLWknPNj2bQDLtyRisQiWPhld33t4tm0CfLY0nMeWpnHjvQaqKzW89mjsOWNrr9jPRgh49N/pePlYEQJSjnmw5PGY8477zac78cKHiba4Pw8hLcmT2x7OIumQFwkbA1n3WSjz/pXCe1sPUV6iY/HsLvXrf7DjIF6+FnR6yfAJxcy/rWd9z+moycU8dWf384rLEVod3P9CJk/+uQtWi2DCrUXE96zhg5cj6DGgiuETyzi004f3FkchhKT/0EruX5QJQHqSO68/FovQgLTCLffn2fWaXjjSFoCLErKNqpFCiC7YamdgS56fSClfaGodf22IHOZ5XZvE42zWqqr2DqFDE/p2Old0HtalXUQztE/MYO/BmladePV3C5dXRExzqOy6jH/vc/aVD81p1bSU+wAABnFJREFUs5qalDIFGNBW21cUpR258Dm1i/qOAkVR2olKaoqidBzqhnZFUToS+f/t3VuMXVUdx/HvDyVQKxfRahqLYkTBhiiSikh1UokSKkTBSDSAL2KwRgExxsQXFZ5MNOKDMV4KMQRFQwpGJWkrUNKWAL3ZYjvV8NBqvCTlKgxiTCY/H/Y6Mowz59J22Jfz+yST2bPPPnv/5iTzn7XX3mttIFMPRUSnpKUWEd3h2oZADSNFLSJGY3CD71NLUYuI0dU0WmAYKWoRMbr0qUVEZ9i5+hkRHZOWWkR0h/H09ODNapKiFhGj6U091FApahExutzSERFdYcBpqUVEZ7jZk0SmqEXEyJp8oWDBZr49HJIeB/58lHf7OuCJo7zPhdSmvG3KCu3Ku1BZ32x7yZHsQNJ6qnzDeML2RUdyvFE1qqgtBEk7Xu7phI9Em/K2KSu0K2+bsjZNax9mHBExlxS1iOiUcShqP647wIjalLdNWaFdeduUtVE636cWEeNlHFpqETFGUtQiolM6W9Qk3SrpkKS9dWcZRNKpkjZJmpS0T9L1dWfqR9LxkrZJ2lPy3lh3pkEkvULS7yX9tu4sg0g6KOkPknZL2lF3nrbpbJ+apAlgCrjN9ll15+lH0lJgqe1dkk4AdgKX2p6sOdqcJAlYbHtK0rHAVuB62w/XHG1ekr4MrABOtH1J3Xn6kXQQWGG7LTcKN0pnW2q2NwNP1Z1jGLb/YXtXWX4O2A+8sd5U83Nlqvx4bPlq7H9HScuAi4G1dWeJhdfZotZWkk4D3g08Um+S/srp3G7gEPA7203O+z3gq0BzR2G/lIGNknZKuqbuMG2TotYgkl4NrAO+ZPvZuvP0Y3va9tnAMuBcSY08xZd0CXDI9s66s4zg/bbPAVYDXyhdKTGkFLWGKH1T64Cf2b6r7jzDsv0MsAl4WQctj2Al8NHST/UL4AJJt9cbqT/bfyvfDwF3A+fWm6hdUtQaoHS83wLst/3duvMMImmJpJPL8iLgw8Af6001N9tfs73M9mnAp4D7bV9Vc6x5SVpcLhYhaTFwIdD4K/hN0tmiJukO4CHgDEl/lXR13Zn6WAl8mqoVsbt8faTuUH0sBTZJehTYTtWn1vhbJVriDcBWSXuAbcA9ttfXnKlVOntLR0SMp8621CJiPKWoRUSnpKhFRKekqEVEp6SoRUSnpKi1iKTpcrvHXkl3SnrVEezrp5I+UZbXSlreZ9tVks4/jGMclPR/Tx2ab/2sbab6vT7H9t+U9JVRM0b3pKi1ywu2zy6zjvwHWDPzRUmH9RxX258dMCPIKmDkohZRhxS19toCnF5aUVsk/RqYLAPNvy1pu6RHJX0OqlELkr4v6U+S7gVe39uRpAckrSjLF0naVeZKu68MsF8D3FBaiR8oIwrWlWNsl7SyvPe1kjaWOdbWAhr0S0j6VRm4vW/24G1JN5f190laUta9VdL68p4tks48Gh9mdEee0N5CpUW2GujdaX4OcJbtA6Uw/NP2eyQdBzwoaSPVzB9nAMup7lqfBG6dtd8lwE+AibKvU2w/JemHwJTt75Ttfg7cbHurpDcBG4B3AN8Attq+SdLFwDCjOD5TjrEI2C5pne0ngcXADts3SPp62fcXqR5Issb2Y5LeC/wAuOAwPsboqBS1dllUpvuBqqV2C9Vp4TbbB8r6C4F39vrLgJOAtwETwB22p4G/S7p/jv2fB2zu7cv2fPPRfQhYXg1ZBeDEMsPIBPDx8t57JD09xO90naTLyvKpJeuTVNME/bKsvx24qxzjfODOGcc+bohjxBhJUWuXF8p0P/9T/rifn7kKuNb2hlnbHc2xpMcA59n+9xxZhiZpFVWBfJ/tf0l6ADh+ns1djvvM7M8gYqb0qXXPBuDzZSojJL29zPawGfhk6XNbCnxwjvc+DExIekt57yll/XPACTO22whc2/tBUq/IbAauKOtWA68ZkPUk4OlS0M6kain2HAP0WptXUJ3WPgsckHR5OYYkvWvAMWLMpKh1z1qq/rJdqh468yOqFvndwGPltduoZjB5CduPA9dQnert4cXTv98Al/UuFADXASvKhYhJXrwKeyNVUdxHdRr6lwFZ1wOvlLQf+BZVUe15nmryyb1UfWY3lfVXAleXfPuAjw3xmcQYySwdEdEpaalFRKekqEVEp6SoRUSnpKhFRKekqEVEp6SoRUSnpKhFRKf8F6Kd4mNpk7djAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pNhEHAirNHem",
        "outputId": "d3c72dfa-19e8-44c7-b155-f19e642d05d0"
      },
      "source": [
        "print(classification_report(Y_train, y_pred_train, target_names=['1','2','3','4', '5']))"
      ],
      "execution_count": 52,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           1       0.49      0.56      0.53     42001\n",
            "           2       0.86      0.17      0.28     23584\n",
            "           3       0.67      0.28      0.40     34190\n",
            "           4       0.62      0.31      0.41     64426\n",
            "           5       0.78      0.95      0.85    290562\n",
            "\n",
            "    accuracy                           0.73    454763\n",
            "   macro avg       0.68      0.45      0.49    454763\n",
            "weighted avg       0.72      0.73      0.70    454763\n",
            "\n"
          ]
        }
      ]
    }
  ]
}