07. naive_bayes_donors_choose.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "WGlKPQnEbkEn"
      },
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')\n",
        "\n",
        "from sklearn.compose import make_column_transformer\n",
        "from sklearn.feature_extraction.text import CountVectorizer\n",
        "from sklearn.preprocessing import OneHotEncoder\n",
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "from sklearn.preprocessing import KBinsDiscretizer\n",
        "\n",
        "from sklearn.pipeline import make_pipeline\n",
        "from sklearn.naive_bayes import MultinomialNB\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "from sklearn.metrics import roc_auc_score\n",
        "from sklearn.metrics import roc_curve\n",
        "from sklearn.metrics import confusion_matrix\n",
        "from sklearn.metrics import ConfusionMatrixDisplay\n",
        "\n",
        "from google.colab import drive\n",
        "\n",
        "from prettytable import PrettyTable"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "drive.mount('/content/drive')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qS7RdwPQh_jj",
        "outputId": "8bbbcbf1-6cfc-4e42-e853-b704b426293d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "- Data contains various categorical, numerical and text features.\n",
        "\n",
        "- We have to predict whether the project application will get accepted or not.\n",
        "\n",
        "- https://www.kaggle.com/c/donorschoose-application-screening"
      ],
      "metadata": {
        "id": "x4WYF7f5Htt3"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "data=pd.read_csv('/content/drive/MyDrive/6_Donors_choose_NB/preprocessed_data.csv')\n",
        "data.head(3)"
      ],
      "metadata": {
        "id": "cVKKBLaHiH05",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 450
        },
        "outputId": "ed2d89a4-ff30-4a7f-c436-45df7436fbf3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "  school_state teacher_prefix project_grade_category  \\\n",
              "0           ca            mrs          grades_prek_2   \n",
              "1           ut             ms             grades_3_5   \n",
              "2           ca            mrs          grades_prek_2   \n",
              "\n",
              "   teacher_number_of_previously_posted_projects  project_is_approved  \\\n",
              "0                                            53                    1   \n",
              "1                                             4                    1   \n",
              "2                                            10                    1   \n",
              "\n",
              "    clean_categories                 clean_subcategories  \\\n",
              "0       math_science  appliedsciences health_lifescience   \n",
              "1       specialneeds                        specialneeds   \n",
              "2  literacy_language                            literacy   \n",
              "\n",
              "                                               essay   price  \n",
              "0  i fortunate enough use fairy tale stem kits cl...  725.05  \n",
              "1  imagine 8 9 years old you third grade classroo...  213.03  \n",
              "2  having class 24 students comes diverse learner...  329.00  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-817fa4a3-8627-4e39-9ee2-f3b874effbe2\">\n",
              "    <div class=\"colab-df-container\">\n",
              "      <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>school_state</th>\n",
              "      <th>teacher_prefix</th>\n",
              "      <th>project_grade_category</th>\n",
              "      <th>teacher_number_of_previously_posted_projects</th>\n",
              "      <th>project_is_approved</th>\n",
              "      <th>clean_categories</th>\n",
              "      <th>clean_subcategories</th>\n",
              "      <th>essay</th>\n",
              "      <th>price</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>ca</td>\n",
              "      <td>mrs</td>\n",
              "      <td>grades_prek_2</td>\n",
              "      <td>53</td>\n",
              "      <td>1</td>\n",
              "      <td>math_science</td>\n",
              "      <td>appliedsciences health_lifescience</td>\n",
              "      <td>i fortunate enough use fairy tale stem kits cl...</td>\n",
              "      <td>725.05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>ut</td>\n",
              "      <td>ms</td>\n",
              "      <td>grades_3_5</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>specialneeds</td>\n",
              "      <td>specialneeds</td>\n",
              "      <td>imagine 8 9 years old you third grade classroo...</td>\n",
              "      <td>213.03</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>ca</td>\n",
              "      <td>mrs</td>\n",
              "      <td>grades_prek_2</td>\n",
              "      <td>10</td>\n",
              "      <td>1</td>\n",
              "      <td>literacy_language</td>\n",
              "      <td>literacy</td>\n",
              "      <td>having class 24 students comes diverse learner...</td>\n",
              "      <td>329.00</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-817fa4a3-8627-4e39-9ee2-f3b874effbe2')\"\n",
              "              title=\"Convert this dataframe to an interactive table.\"\n",
              "              style=\"display:none;\">\n",
              "        \n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
              "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
              "  </svg>\n",
              "      </button>\n",
              "      \n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      flex-wrap:wrap;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "      <script>\n",
              "        const buttonEl =\n",
              "          document.querySelector('#df-817fa4a3-8627-4e39-9ee2-f3b874effbe2 button.colab-df-convert');\n",
              "        buttonEl.style.display =\n",
              "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "        async function convertToInteractive(key) {\n",
              "          const element = document.querySelector('#df-817fa4a3-8627-4e39-9ee2-f3b874effbe2');\n",
              "          const dataTable =\n",
              "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                     [key], {});\n",
              "          if (!dataTable) return;\n",
              "\n",
              "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "            + ' to learn more about interactive tables.';\n",
              "          element.innerHTML = '';\n",
              "          dataTable['output_type'] = 'display_data';\n",
              "          await google.colab.output.renderOutput(dataTable, element);\n",
              "          const docLink = document.createElement('div');\n",
              "          docLink.innerHTML = docLinkHtml;\n",
              "          element.appendChild(docLink);\n",
              "        }\n",
              "      </script>\n",
              "    </div>\n",
              "  </div>\n",
              "  "
            ]
          },
          "metadata": {},
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y = data['project_is_approved'].values\n",
        "X = data.drop(['project_is_approved'], axis=1)\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, stratify=y)"
      ],
      "metadata": {
        "id": "nmj6xdCniKtt"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "### **Set-1:**"
      ],
      "metadata": {
        "id": "X09XQLrMpycb"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "<h3>Encoding features using column transformer.</h3>"
      ],
      "metadata": {
        "id": "KszC7uNlm94V"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "transform_set1 = make_column_transformer(\n",
        "    ( CountVectorizer(max_df=0.85,stop_words='english') , 'essay' ),  #BoW on essay\n",
        "    ( OneHotEncoder(handle_unknown='ignore') , (0, 1, 2, 4, 5) ),     #One Hot encoding on categorical features\n",
        "    ( KBinsDiscretizer(n_bins=4,strategy='quantile') , [3,7] )        #categorizing numerical features based on quantile(have tried bining instead of normalizing)\n",
        "    )"
      ],
      "metadata": {
        "id": "GnpUuugZiTIG"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "a=transform_set1.fit_transform(X_train,y_train)\n",
        "print('No. of Features:',len(a[0].toarray()[0]))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dleoJPH6j4gS",
        "outputId": "863b8b3c-18a5-4616-f179-be6577f47234"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "No. of Features: 48308\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "<h3>Doing Grid-Search over various alpha values.</h3>"
      ],
      "metadata": {
        "id": "DP_3Xw1m8Ggt"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "pipe_set1 = make_pipeline( transform_set1 , MultinomialNB() )\n",
        "\n",
        "pipe_set1.get_params().keys()"
      ],
      "metadata": {
        "id": "5WtI_3mfio3e"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "params = {'multinomialnb__alpha':[0.00001,0.0005, 0.0001,0.005,0.001,0.05,0.01,0.1,0.5,1,5,10,50,100]}\n",
        "                    \n",
        "clf_set1 = GridSearchCV( pipe_set1, params, cv=3, scoring='roc_auc', return_train_score=True, refit=False) \n",
        "clf_set1.fit(X_train,y_train)\n",
        "\n",
        "cv_set_1 = pd.DataFrame(clf_set1.cv_results_)"
      ],
      "metadata": {
        "id": "BnO4nltmivof"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "sns.set_style(\"whitegrid\")\n",
        "plt.figure(figsize=(12,7))\n",
        "sns.lineplot(np.arange(len(params['multinomialnb__alpha'])), y=cv_set_1.mean_train_score, label='mean train auc score')\n",
        "sns.lineplot(np.arange(len(params['multinomialnb__alpha'])), y=cv_set_1.mean_test_score, label='mean cv auc score')\n",
        "plt.xticks(np.arange(len(params['multinomialnb__alpha'])), labels=params['multinomialnb__alpha'])\n",
        "plt.yticks(np.arange(0.45,1,0.025))\n",
        "plt.tick_params(labelright=True)\n",
        "plt.xlabel('MultinomialNB - alpha')\n",
        "plt.ylabel('mean train and cv auc score')\n",
        "plt.title('Train & CV curve for hyperparameter selection')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 458
        },
        "id": "4CBXFfi3vx9A",
        "outputId": "b3e24479-d8ce-47ea-ec70-f968cbeea591"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 864x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Observation:**\n",
        "\n",
        "- alpha = 5 and alpha = 10 both seems reasonable.\n",
        "\n",
        "<h3>Again doing Grid search with alpha=[5,10] and various values of max_features</h3>"
      ],
      "metadata": {
        "id": "Q9GnTDPMBMWb"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "params = {'multinomialnb__alpha':[5,10],\n",
        "          'columntransformer__countvectorizer__max_features':[1000,2500,5000,7500,10000]}\n",
        "                      \n",
        "clf_set1 = GridSearchCV( pipe_set1, params, cv=3, scoring='roc_auc', return_train_score=True, refit=False) \n",
        "clf_set1.fit(X_train,y_train)\n",
        "\n",
        "cv_set_1 = pd.DataFrame(clf_set1.cv_results_)"
      ],
      "metadata": {
        "id": "mNbd7W-SA3iL"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "sns.set_style(\"whitegrid\")\n",
        "plt.figure(figsize=(12,7))\n",
        "sns.lineplot(np.arange(10), y=cv_set_1.mean_train_score, label='mean train auc score')\n",
        "sns.lineplot(np.arange(10), y=cv_set_1.mean_test_score, label='mean cv auc score')\n",
        "plt.xticks(np.arange(10))\n",
        "plt.yticks(np.arange(0.45,1,0.025))\n",
        "plt.tick_params(labelright=True)\n",
        "plt.xlabel('index of gridsearch result')\n",
        "plt.ylabel('mean train and cv auc score')\n",
        "plt.title('Train & CV curve for hyperparameter selection')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 458
        },
        "id": "lD8KCghTCS5O",
        "outputId": "94a13b27-80d4-44f5-f3ad-688264561bcc"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 864x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Observation:**\n",
        "\n",
        "   Index no. 2 seems reasonable. getting hyperparameter value below."
      ],
      "metadata": {
        "id": "idG4SEtvENLg"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "cv_set_1.iloc[2,[4,5,10,16]]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wmbWHxNZDdwj",
        "outputId": "2f1d5080-215f-4264-d1b4-b9dee38232bd"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "param_columntransformer__countvectorizer__max_features        2500\n",
              "param_multinomialnb__alpha                                       5\n",
              "mean_test_score                                           0.692022\n",
              "mean_train_score                                          0.719535\n",
              "Name: 2, dtype: object"
            ]
          },
          "metadata": {},
          "execution_count": 196
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Observation:**\n",
        "\n",
        "We get,\n",
        "\n",
        "max_features = 2500\n",
        "\n",
        "alpha = 5\n",
        "\n",
        "\n",
        "<h3>Training on whole X_train.</h3>"
      ],
      "metadata": {
        "id": "XBaT_1OSJgLw"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "param={'multinomialnb__alpha':5,\n",
        "   'multinomialnb__class_prior':None,\n",
        "   'columntransformer__countvectorizer__max_features':2500}\n",
        "\n",
        "pipe_set1.set_params(**param)\n",
        "pipe_set1.fit(X_train,y_train)\n",
        "y_predict_test = pipe_set1.predict_proba(X_test)[:,1]\n",
        "y_predict_train = pipe_set1.predict_proba(X_train)[:,1]\n",
        "auc_score_set1_test = roc_auc_score(y_test, y_predict_test )\n",
        "auc_score_set1_train = roc_auc_score(y_train, y_predict_train )\n",
        "print('Test AUC Score:', auc_score_set1_test )\n",
        "print('Train AUC Score:', auc_score_set1_train)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gRxlnrMkxd41",
        "outputId": "f516ef61-0e15-4b0d-84e9-72f4f98a4322"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Test AUC Score: 0.70131306385036\n",
            "Train AUC Score: 0.7128906734395434\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "confusion_matrix_train_set1 = confusion_matrix(y_train,pipe_set1.predict(X_train),normalize='true')\n",
        "confusion_matrix_train_set1 "
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HQ479N-k5Vyt",
        "outputId": "19f6b6dd-5d13-4492-f820-d6fcda91d8db"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.53018136, 0.46981864],\n",
              "       [0.23011286, 0.76988714]])"
            ]
          },
          "metadata": {},
          "execution_count": 198
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "confusion_matrix_test_set1 = confusion_matrix(y_test,pipe_set1.predict(X_test),normalize='true')\n",
        "confusion_matrix_test_set1 "
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4ripDLDiI2aU",
        "outputId": "2db88de3-6807-48bd-a7f0-5ce6166f9916"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.51767723, 0.48232277],\n",
              "       [0.23171967, 0.76828033]])"
            ]
          },
          "metadata": {},
          "execution_count": 199
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "<h3>Trying class prior = [50,50]</h3>"
      ],
      "metadata": {
        "id": "JrsVd-f9G6S6"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "param={'multinomialnb__alpha':5,\n",
        "   'multinomialnb__class_prior':[50,50],\n",
        "   'columntransformer__countvectorizer__max_features':2500}\n",
        "\n",
        "pipe_set1.set_params(**param)\n",
        "pipe_set1.fit(X_train,y_train)\n",
        "y_predict_test = pipe_set1.predict_proba(X_test)[:,1]\n",
        "y_predict_train = pipe_set1.predict_proba(X_train)[:,1]"
      ],
      "metadata": {
        "id": "HrP5UgpaHG_2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "auc_score_set1_test = roc_auc_score(y_test, y_predict_test )\n",
        "auc_score_set1_train = roc_auc_score(y_train, y_predict_train )\n",
        "print('Test AUC Score:', auc_score_set1_test )\n",
        "print('\\nTrain AUC Score:', auc_score_set1_train)\n",
        "\n",
        "confusion_matrix_train_set1 = confusion_matrix(y_train,pipe_set1.predict(X_train),normalize='true')\n",
        "print('\\nTrain Confusion Matrix:\\n\\n',confusion_matrix_train_set1)\n",
        "confusion_matrix_test_set1 = confusion_matrix(y_test,pipe_set1.predict(X_test),normalize='true')\n",
        "print('\\nTest Confusion Matrix:\\n\\n',confusion_matrix_test_set1)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oFAbBMobGqff",
        "outputId": "544f3ce6-47d5-4285-8abc-78aafa898ceb"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Test AUC Score: 0.7013130698381262\n",
            "\n",
            "Train AUC Score: 0.7128906807027795\n",
            "\n",
            "Train Confusion Matrix:\n",
            "\n",
            " [[0.6716593 0.3283407]\n",
            " [0.3521324 0.6478676]]\n",
            "\n",
            "Test Confusion Matrix:\n",
            "\n",
            " [[0.6644074 0.3355926]\n",
            " [0.3538718 0.6461282]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Observation:**\n",
        "\n",
        "- By using class prior of [50,50] we get no change in auc score but we get good balance between FP & TP.\n",
        "- So, we continue using class prior of [50,50].\n",
        "\n",
        "**Note:**\n",
        "\n",
        "- With change in class prior we can tradeoff between TP & FP as per requirement.\n",
        "\n",
        "- We want our model to have high TP and low FP, so now we threshold probability at max. value of TPR(1-FPR)."
      ],
      "metadata": {
        "id": "fBW6WsA5IbO_"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "<h3>Finding optimal probability threshold</h3>"
      ],
      "metadata": {
        "id": "uBtXrNzyjrfa"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "fpr_train, tpr_train, thresholds = roc_curve( y_train,  y_predict_train )"
      ],
      "metadata": {
        "id": "hXxVumRpIZtF"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "fpr_test, tpr_test ,threshold_test = roc_curve( y_test ,  y_predict_test )"
      ],
      "metadata": {
        "id": "oROQTIaNRpz6"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "sns.lineplot(fpr_train , tpr_train , label='Trian ROC Curve')\n",
        "sns.lineplot(fpr_test , tpr_test , label='Test ROC Curve')\n",
        "plt.title('Train and Test ROC Curve')\n",
        "plt.xlabel('False Positive Rate')\n",
        "plt.ylabel('True Positive Rate')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "Iy8VAhCJOEdi",
        "outputId": "c873680d-d4f5-4353-ff21-b7d510f2275d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Train AUC Score: 0.71\n",
        "\n",
        "Test AUC Score: 0.70"
      ],
      "metadata": {
        "id": "nXo4GkR_p9HM"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "threshold =  thresholds[np.argmax(tpr_train*(1-fpr_train))]\n",
        "\n",
        "y_train_after_thresholding = np.copy(y_predict_train)\n",
        "y_train_after_thresholding = np.where( y_train_after_thresholding > threshold , 1, 0)\n",
        "confusion_matrix_train_set1_thresh = confusion_matrix(y_train,y_train_after_thresholding,normalize='true')\n",
        "print('Train Confusion Matrix after thresholding probability:\\n',confusion_matrix_train_set1_thresh)\n",
        "\n",
        "y_test_after_thresholding = np.copy(y_predict_test)\n",
        "y_test_after_thresholding = np.where( y_test_after_thresholding > threshold , 1, 0)\n",
        "confusion_matrix_test_set1_thresh = confusion_matrix(y_test,y_test_after_thresholding,normalize='true')\n",
        "print('\\nTest Confusion Matrix after thresholding probability:\\n',confusion_matrix_test_set1_thresh)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YlwwazhiLr_y",
        "outputId": "da724690-7f45-4899-db0a-1ea28d767e61"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Confusion Matrix after thresholding probability:\n",
            " [[0.6764414  0.3235586 ]\n",
            " [0.35585143 0.64414857]]\n",
            "\n",
            "Test Confusion Matrix after thresholding probability:\n",
            " [[0.66660561 0.33339439]\n",
            " [0.35792502 0.64207498]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Observation:**\n",
        "\n",
        "- After thresholding we see slight chances in TP & FP compared to using class prior [50,50].\n",
        "\n",
        "- We continue with new threshold.\n"
      ],
      "metadata": {
        "id": "nLLWxiPhOPuS"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "<h3>Confusion Matrix(normalized) after threshold</h3>"
      ],
      "metadata": {
        "id": "0fejE6bDlpP_"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "sns.set_style('white')\n",
        "ConfusionMatrixDisplay.from_predictions(y_train, y_train_after_thresholding, normalize='true')\n",
        "plt.title('Confusion Matrix-Train set')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "eNHu5gS2Wr8u",
        "outputId": "f9970672-ad89-4835-e548-e8aaeb381558"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "ConfusionMatrixDisplay.from_predictions(y_test, y_test_after_thresholding, normalize='true') \n",
        "plt.title('Confusion Matrix-Test set')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "aZWIJY7bWL9e",
        "outputId": "27ee0ec7-d667-4495-cb50-e500e6ea93de"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "<h3>Getting Top 50 Features:</h3>"
      ],
      "metadata": {
        "id": "3UGlsCDJHLXF"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def top_features(clf,n):\n",
        "  features = clf.named_steps.columntransformer.get_feature_names_out()\n",
        "  prob = clf.named_steps.multinomialnb.feature_log_prob_\n",
        "  prob_0 = prob[0,:]\n",
        "  prob_1 = prob[1,:]\n",
        "  sort_index_prob_0 = np.argsort(prob_0)[::-1] # https://stackoverflow.com/questions/6771428/most-efficient-way-to-reverse-a-numpy-array\n",
        "  sort_index_prob_1 = np.argsort(prob_1)[::-1]\n",
        "  return features[sort_index_prob_0][:n] , features[sort_index_prob_1][:n]\n"
      ],
      "metadata": {
        "id": "uTd6kfsBmUza"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "top_50_features_0 , top_50_features_1 = top_features(pipe_set1,50)\n",
        "print('top 50 features for predicting class 0:\\n', top_50_features_0,'\\n' )\n",
        "print('top 50 features for predicting class 1:\\n', top_50_features_1 )"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-i6DJ1b3nOpY",
        "outputId": "83180601-c298-48dd-eb49-75dbad97deb0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "top 50 features for predicting class 0:\n",
            " ['countvectorizer__learning' 'countvectorizer__classroom'\n",
            " 'countvectorizer__learn' 'countvectorizer__help' 'countvectorizer__need'\n",
            " 'countvectorizer__work' 'countvectorizer__come' 'countvectorizer__love'\n",
            " 'countvectorizer__reading' 'countvectorizer__able'\n",
            " 'countvectorizer__materials' 'countvectorizer__skills'\n",
            " 'countvectorizer__day' 'countvectorizer__use' 'countvectorizer__class'\n",
            " 'countvectorizer__year' 'countvectorizer__want' 'countvectorizer__make'\n",
            " 'countvectorizer__new' 'countvectorizer__student'\n",
            " 'countvectorizer__technology' 'onehotencoder__teacher_prefix_mrs'\n",
            " 'kbinsdiscretizer__teacher_number_of_previously_posted_projects_0.0'\n",
            " 'countvectorizer__time' 'countvectorizer__math' 'countvectorizer__grade'\n",
            " 'countvectorizer__children' 'countvectorizer__provide'\n",
            " 'onehotencoder__project_grade_category_grades_prek_2'\n",
            " 'countvectorizer__project' 'countvectorizer__different'\n",
            " 'countvectorizer__supplies' 'countvectorizer__world'\n",
            " 'countvectorizer__teach' 'countvectorizer__best'\n",
            " 'countvectorizer__education' 'onehotencoder__teacher_prefix_ms'\n",
            " 'countvectorizer__books' 'countvectorizer__allow' 'countvectorizer__high'\n",
            " 'countvectorizer__create' 'countvectorizer__science'\n",
            " 'countvectorizer__needs' 'countvectorizer__like'\n",
            " 'countvectorizer__activities' 'countvectorizer__life'\n",
            " 'kbinsdiscretizer__teacher_number_of_previously_posted_projects_1.0'\n",
            " 'kbinsdiscretizer__price_3.0'\n",
            " 'onehotencoder__project_grade_category_grades_3_5' 'countvectorizer__way'] \n",
            "\n",
            "top 50 features for predicting class 1:\n",
            " ['countvectorizer__learning' 'countvectorizer__classroom'\n",
            " 'countvectorizer__learn' 'countvectorizer__help' 'countvectorizer__need'\n",
            " 'countvectorizer__work' 'countvectorizer__reading' 'countvectorizer__use'\n",
            " 'countvectorizer__love' 'countvectorizer__day' 'countvectorizer__able'\n",
            " 'countvectorizer__come' 'countvectorizer__class'\n",
            " 'countvectorizer__technology' 'countvectorizer__skills'\n",
            " 'countvectorizer__books' 'countvectorizer__new' 'countvectorizer__year'\n",
            " 'countvectorizer__make' 'countvectorizer__time' 'countvectorizer__want'\n",
            " 'onehotencoder__teacher_prefix_mrs' 'countvectorizer__student'\n",
            " 'countvectorizer__grade' 'countvectorizer__allow' 'countvectorizer__read'\n",
            " 'countvectorizer__math' 'countvectorizer__provide'\n",
            " 'countvectorizer__children'\n",
            " 'kbinsdiscretizer__teacher_number_of_previously_posted_projects_0.0'\n",
            " 'countvectorizer__materials'\n",
            " 'onehotencoder__project_grade_category_grades_prek_2'\n",
            " 'countvectorizer__world' 'countvectorizer__teach' 'countvectorizer__high'\n",
            " 'countvectorizer__different' 'countvectorizer__project'\n",
            " 'countvectorizer__like' 'onehotencoder__teacher_prefix_ms'\n",
            " 'countvectorizer__learners'\n",
            " 'onehotencoder__project_grade_category_grades_3_5'\n",
            " 'countvectorizer__create' 'countvectorizer__group' 'countvectorizer__way'\n",
            " 'countvectorizer__best' 'countvectorizer__needs' 'countvectorizer__home'\n",
            " 'countvectorizer__education' 'countvectorizer__free'\n",
            " 'kbinsdiscretizer__teacher_number_of_previously_posted_projects_1.0']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Observation:**\n",
        "\n",
        "- Most of the top features are from Countvectorizer.\n",
        "- Most of them are same for both classes."
      ],
      "metadata": {
        "id": "ygvnplDsOaNO"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## **Set-2:**"
      ],
      "metadata": {
        "id": "9OycAIjbrVMq"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "<h3>Encoding features using column transformer</h3>"
      ],
      "metadata": {
        "id": "AZqCTOKDmEqT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "transform_set2 = make_column_transformer(\n",
        "    ( TfidfVectorizer(max_df=0.85,stop_words='english') , 'essay' ),\n",
        "    ( OneHotEncoder(handle_unknown='ignore') , (0, 1, 2, 4, 5) ),\n",
        "    ( KBinsDiscretizer(n_bins=4,strategy='quantile') , [3,7] )\n",
        "    )"
      ],
      "metadata": {
        "id": "ZFMBMh25pjrj"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "a=transform_set2.fit_transform(X_train,y_train)\n",
        "print('No. of features:',len(a[0].toarray()[0]))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YJ42n7iKgqx-",
        "outputId": "439cfcf6-a2fb-4629-fb72-0c47fabbd2ab"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "No. of features: 48308\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "<h3>Doing Grid-search with various values of alpha.</h3>"
      ],
      "metadata": {
        "id": "w0mwEUa4_3vm"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "pipe_set2 = make_pipeline( transform_set2 , MultinomialNB() )"
      ],
      "metadata": {
        "id": "KWs6CZqNpjVv"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "params = {'multinomialnb__alpha':[0.00001,0.0005, 0.0001,0.005,0.001,0.05,0.01,0.1,0.5,1,5,10,50,100]}\n",
        "                      \n",
        "clf_set2 = GridSearchCV( pipe_set2, params, cv=3, scoring='roc_auc', return_train_score=True, refit=False) \n",
        "clf_set2.fit(X_train,y_train)\n",
        "\n",
        "cv_set_2 = pd.DataFrame(clf_set2.cv_results_)"
      ],
      "metadata": {
        "id": "2uDaJsuUrgTV"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "sns.set_style(\"whitegrid\")\n",
        "plt.figure(figsize=(12,7))\n",
        "sns.lineplot(np.arange(len(params['multinomialnb__alpha'])), y=cv_set_2.mean_train_score, label='mean train auc score')\n",
        "sns.lineplot(np.arange(len(params['multinomialnb__alpha'])), y=cv_set_2.mean_test_score, label='mean cv auc score')\n",
        "plt.xticks(np.arange(len(params['multinomialnb__alpha'])), labels=params['multinomialnb__alpha'])\n",
        "plt.yticks(np.arange(0.45,1,0.025))\n",
        "plt.tick_params(labelright=True)\n",
        "plt.xlabel('MultinomialNB - alpha')\n",
        "plt.ylabel('mean train and cv auc score')\n",
        "plt.title('Train & CV curve for hyperparameter selection')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 458
        },
        "id": "Po_Tph9rrp2W",
        "outputId": "66f8faa8-f028-4e53-9953-3024b39cefff"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 864x504 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Observation:**\n",
        "\n",
        "- alpha = 1 and alpha = 5 both seems reasonable.\n",
        "\n",
        "<h3>Again doing Grid search with alpha=[1,5] and various values of max_features.</h3>"
      ],
      "metadata": {
        "id": "-XQO09FNr2f8"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "params = {'multinomialnb__alpha':[1,5],\n",
        "          'columntransformer__tfidfvectorizer__max_features':[1000,2500,5000,7500,10000]}\n",
        "                      \n",
        "clf_set2 = GridSearchCV( pipe_set2, params, cv=3, scoring='roc_auc', return_train_score=True, refit=False) \n",
        "clf_set2.fit(X_train,y_train)\n",
        "\n",
        "cv_set_2 = pd.DataFrame(clf_set2.cv_results_)"
      ],
      "metadata": {
        "id": "RlVszfBKr7qt"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "sns.set_style(\"whitegrid\")\n",
        "plt.figure(figsize=(12,7))\n",
        "sns.lineplot(np.arange(10), y=cv_set_2.mean_train_score, label='mean train auc score')\n",
        "sns.lineplot(np.arange(10), y=cv_set_2.mean_test_score, label='mean cv auc score')\n",
        "plt.xticks(np.arange(10))\n",
        "plt.yticks(np.arange(0.45,1,0.025))\n",
        "plt.tick_params(labelright=True)\n",
        "plt.xlabel('index no. of gridsearch result')\n",
        "plt.ylabel('mean train and cv auc score')\n",
        "plt.title('Train & CV curve for hyperparameter selection')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 458
        },
        "id": "jBXMKNwtsIVd",
        "outputId": "c541a055-011e-4b98-e956-46cb4abe3a9f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 864x504 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAwEAAAG5CAYAAADWGTgkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzde1xUdf4/8NcMMICpgSaDFVIJKoYkhoh35aIiIihi6i/XVMTES2pq5iqa9wuJeMtYFbe1b1u6EOF4QfGC7notbValVUkSUMYUUK4zMJzfHz484wlktBgZ4/V8PPbxYGY+53M+nxe4fd5zbjJBEAQQEREREVGDIa/vARARERER0bPFIoCIiIiIqIFhEUBERERE1MCwCCAiIiIiamBYBBARERERNTAsAoiIiIiIGhgWAUT01CIiIpCUlFTfw/jTKS8vx/vvv4+3334b06ZN+8P9nT59Gr169aqDkZG5aNu2LX755Zc679fT0xPZ2dl13i8RmS/L+h4AET0bnp6e4s9lZWVQKBSwsLAAAHzyyScYPHjwE/e1devWPzSWbdu2ISEhAeXl5XBzc8Pf/vY32NjY1LqNWq3Ghg0bcP78ecjlcrRq1QojR45Ejx490LdvX+zfvx+tWrWSbDN58mS0atUKH3300R8a77Oyf/9+3LlzB6dPn4alJf/v2VyMHj0agwcPRnh4eH0PpU7UNJ/z58/X44iIqD7wvzJEDcSj/5H39fXF0qVL0a1bt2rtKisrTboAzczMxLp167B79260bt0a33//PeTy2g9Knj9/HuPGjcOkSZOwatUq2Nvb49KlS/jb3/6GsLAwdO3aFcnJyZg6daq4TWFhIY4dO4Z//etfJptLXWd18+ZNvPbaa7+rT1P/3v6Iuh6bOc+1Jnq9Xiy4iYjMBU8HImrgHp4yEh8fj+7du+Pjjz/GvXv3MHHiRPj4+KBz586YOHEi8vLyxG1Gjx6NXbt2AQASExMxcuRIrFq1Cp07d4avry+OHTv22P1ZWlrCwsICr7zyCiwtLdGlSxcoFIpax7h69WqEhoYiMjISzZo1g0wmg7u7O+Li4gAAoaGhSE5OlmyjUqng4uKCtm3b1tjnuXPnMGLECHh5eaF3795ITEysNrdH5/dQ27Zt8eWXX6Jfv37o168fFi5ciFWrVkn6njRpEhISEgAAGo0GU6dOhY+PD3x9ffHFF1/UOJ7169dj8+bN2LdvHzw9PbFr1y5UVVVh8+bN6Nu3L7p27Yo5c+agqKgIAJCTk4O2bdti165d6NOnD8aMGfPY/LZv346uXbuiR48eYlGkVqvRrVs36PV6sV1qaqp4RGjDhg2YNm0apk+fDk9PTwwZMgQ//fST2La2eT3cdtasWejUqROSkpKM9hcfHw9/f394enpi4MCBOHjwoOR3MGLECCxfvhxdunTBhg0bcOPGDfzlL39Bly5d0KVLF3z44Ye4f/++uI2vry+2bt2K4OBgdOzYEfPmzcOdO3cQEREBT09PvPfee7h3757Y/sKFC+Lfw+DBg3H69GkAQGxsLM6dO4fFixfD09MTixcvBvCgmB07diy8vb3Rv39/7N27V+xr7ty5WLhwISZMmICOHTuKfT0qMTERfn5+8PT0hK+vL7777jvxs927dyMwMBCdO3fG+PHjkZubW+PvVafTYdWqVejTpw+6deuG6OholJeXi58fOnQIISEh6NSpE/z9/ZGenv7Y+Tx6mlFRURHmzJkDHx8f9O3bF5s3b0ZVVZU47qf5905EZkwgoganb9++wr///W9BEATh1KlTgpubm7B69WpBq9UKZWVlQn5+vrB//36htLRUKCoqEqZOnSpMmjRJ3P7dd98VvvnmG0EQBOFf//qX0L59e+Hrr78WKisrhS+//FLo3r27UFVVVeO+i4qKhL59+wpjx44VysvLjY61tLRUaNeunXDy5MnHtikrKxM6deoknD17Vnxv+PDhQkJCQo3tc3JyhI4dOwopKSmCTqcT8vPzhcuXL1eb28P5jRgxQnzdpk0b4b333hMKCgqEsrIy4cyZM0KvXr3E+RYWFgodOnQQ8vLyBL1eLwwZMkTYsGGDoNVqhRs3bgi+vr5Cenp6jeNav3698OGHH4qvd+3aJfj7+ws3btwQiouLhcmTJwuzZs0SBEEQsrOzhTZt2gizZ88WSkpKhLKysmr9Pfzdrlu3TtDpdMLRo0cFDw8PobCwUBAEQQgMDBSOHj0qto+KihK2bdsmjqV9+/bCvn37BJ1OJ2zdulXo27evoNPpjM7r4bYHDx4U9Hq9UFZWVmt/giAIe/fuFTNTqVTCW2+9JWg0GvF34ObmJnzxxRdCRUWFUFZWJmRlZQknTpwQtFqtcPfuXWHUqFHC0qVLxbn07dtXCA8PF3799VchLy9P8PHxEUJDQ4VLly4J5eXlwujRo4UNGzYIgiAIeXl5gre3t3D06FFBr9cLJ06cELy9vYW7d+/W+DdRUlIi9OrVS9i9e7dQUVEhXLp0SfD29hauXr0qCIIgfPTRR0KnTp2Ec+fOCXq9vtrfeUlJieDp6SlkZmYKgiAIGo1GuHLliiAIgnDw4EHB399fuHbtmlBRUSFs2rRJeOeddyR/f1lZWYIgCMKyZcuEiRMnCgUFBUJRUZEwceJEISYmRhAEQfjxxx+FTp06CSdOnBD0er2Ql5cnXLt2rcb5/Lbf2bNnC++//75QVFQkZGdnC/369fvd/96JyHzxSAARQS6XY9q0aVAoFLCxsYG9vT369+8PW1tbNG7cGJMmTcLZs2cfu/3LL7+M4cOHw8LCAkOGDMGvv/6KO3fu1Nj2gw8+wDvvvANnZ2dERUVBq9UCAGbNmoV//OMf1drfv38fVVVVaNGixWP3b2NjgwEDBohHA7KysnDp0iUEBwfX2H7Pnj3o1q0bBg0aBCsrK9jb28PNze2x/f9WZGQk7OzsYGNjAy8vL8hkMpw7dw4AcODAAXTs2BFKpRL//e9/kZ+fjylTpkChUMDJyQnDhw+XfGtcm5SUFLz33ntwcnLCCy+8gJkzZ2Lv3r2orKwU20ydOhWNGjV67DUVlpaWmDx5MqysrNC7d280atQI169fB/DgCMrDb6ALCwtx4sQJDBo0SNz2zTffxIABA2BlZYWxY8dCp9Phxx9/fKJ5dezYEf7+/pDL5eLYHtcfAAQGBkKpVEIul2PgwIFwdnaGWq0W+3NwcMDo0aNhaWkJGxsbODs7o3v37lAoFGjWrBnGjh1b7W/03XffxUsvvQSlUgkvLy94eHigffv2sLa2RkBAAC5fvgwASE5ORq9evdC7d2/I5XJ0794d7u7uj/2G++jRo3jllVcQFhYGS0tLtG/fHv3798f+/fvFNn5+fnj77bchl8thbW1drQ+5XI6rV6+ivLwcDg4OcHV1BQD885//RGRkJFq3bg1LS0u8//77yMjIqHY0QBAEfPPNN5g3bx7s7OzQuHFjTJw4ESqVCsCDowlhYWHo3r075HI5lEolWrduXeN8HqXX67F37158+OGHaNy4MV599VWMHTtWcqTiaf69E5H5en5OqiQik7G3t5csVMrKyrBixQocP35cPGWipKTksec2v/TSS+LPtra2AIDS0tJq7X7++Wf88MMP2LJlCywsLDB79mxERUVh48aNuHDhAiZOnFhtm6ZNm0Iul+PXX3+tdREzZMgQTJo0CfPnz0dycjJ69OiB5s2b19j21q1b1S4ifhotW7YUf5bJZBg4cCD27NmDzp07IyUlRTylJjc3F7dv34aXl5fYXq/XS17X5vbt23jllVfE16+88goqKytx9+5d8T1HR8da+7Czs5OcP29rayv+bkJCQhAYGIjS0lLs27cPXl5ecHBwqLHvhwvJ27dvi2OrbV41jau2/r799lskJCSIi93S0lIUFBQ8tr87d+5g2bJlOHfuHEpKSiAIApo2bSpp8+jfpbW1teS1jY2NmMPNmzexf/9+HDlyRPy8srISXbp0qTYH4MHvVa1WV5v/oxfXP/o38luNGjVCbGwstm/fjr/+9a/o1KkTPvroI7Ru3Ro3b97E8uXLJaeYCYIAjUYj+VvIz89HWVkZhg4dKmn38LSdW7duoXfv3o8dw+MUFBSgoqICL7/8svjeyy+/DI1GI75+0n/vRGTeWAQQEWQymeT19u3bcf36dXzzzTdo0aIFMjIyEBoaCkEQ/tB+9Ho99Ho9BEGAXC7HypUrMXnyZISGhqJ169bit6GPsrW1RceOHZGamgofH5/H9v3222/jxRdfRFpaGr777jvMnj37sW1btmwp+Zb5t/srKysTX9f0Dedv8xo0aBDGjRuHyMhIqNVqbNq0SdzPq6++itTU1MeOpTYODg6Sb4Bv3rwJS0tLNG/eXLxG47djeRpKpRKenp5ITU1FcnKy5NoHAJLrQKqqqqDRaODg4AALCwuj86ppXI/rLzc3F/Pnz8eOHTvg6ekJCwsLhISE1Nrf2rVrIZPJkJKSAjs7Oxw6dEg8v/1ptWzZEiEhIVi6dOkTt+/cubN43cfv0bNnT/Ts2RPl5eVYt24dFixYgP/7v/9Dy5Yt8f777xu9W5e9vT1sbGygUqmgVCprHOONGzeeelz29vawsrLCzZs34eLiAuBBQVHTPojo+cbTgYiompKSElhbW6Np06YoLCzExo0b66TfN954A87Ozvjkk09QVFSEyspKdOvWDVlZWWjUqNFji4zZs2cjKSkJW7duFb8d/umnnzBjxgyxjUwmQ2hoKGJiYlBUVARfX9/HjiM4OBj/+c9/xFNrCgoKkJGRAQBwc3PDwYMHUVZWhl9++QW7d+82Oq/27dvD3t4e8+fPR48ePcRvpD08PPDCCy8gPj4e5eXl0Ov1uHLlymMLkN8aNGgQ/v73vyM7OxslJSWIjY1FYGBgnd4ZJyQkBNu2bcOVK1fQr18/yWeXLl1CamoqKisr8fe//x0KhQJvvfXW757X4/orKyuDTCZDs2bNAAD/+te/cPXq1Vr7KikpQaNGjdCkSRNoNJo/dNvawYMH48iRIzh+/Dj0ej20Wi1Onz4tFi0vvfSS5B76ffr0QVZWFr799ltUVFSgoqICarUamZmZT7S/O3fu4NChQygtLYVCoUCjRo3EO2SNGDEC8fHx4vyLioqwb9++an3I5XKEh4dj+fLl4pEhjUaD48ePAwCGDRuGxMREnDx5Uiy4Ho7vt/N5lIWFBQYMGIDY2FgUFxcjNzcXCQkJT3ULYSJ6PrAIIKJqxowZA61WCx8fH7zzzjvo2bNnnfRrYWGBzz//HEVFRfD390evXr3www8/IDExEZcvX8a6detq3K5Tp074+9//jlOnTsHf3x/e3t5YsGBBtdMdQkJCcPPmTQQGBtZ6x6GXX34Zf/vb35CQkABvb2+EhoaKd6oZM2YMrKys0K1bN3z00UePva7gtwYNGoT//Oc/knPqLSwssGXLFvz000/w8/ODj48P5s+fj+Li4ifqMywsDIMHD8a7774LPz8/KBQKLFiw4Im2fVIBAQHIzc1FQECAeGrHQ35+fti7dy86d+6M5ORkbNiwAVZWVr97Xo/rz8XFBePGjcOIESPQrVs3XLlyBZ06daq1rylTpuDy5cvw8vJCZGRktQLmabRs2RKbN2/G559/jq5du6J3797Ytm2beGrNX/7yFxw4cACdO3fG0qVL0bhxY2zbtg179+5Fz5490aNHD8TExECn0z3R/qqqqrBjxw707NkT3t7eOHv2LBYtWgTgwe8jIiICM2fORKdOnTBo0CCkp6fX2M/s2bPh7OyM4cOHo1OnTnjvvffE6z08PDywYsUKLF++HG+//Tbeffdd3Lx5s8b5/NaCBQtga2sLf39/jBo1CoMGDUJYWNjTxkpEZk4m/NHj+0RE9Fzz9/fH4sWLJc+N2LBhA3755RfExMTUyT7quj8iIvpjeCSAiKgBO3DgAGQyWa3XWxAR0Z+PSS8MTk9Px7Jly1BVVYXw8HBERkZKPs/NzcW8efOQn58POzs7rFmzBo6Ojjh16hRWrFghtvv5558RGxsLf39/zJ07F2fOnEGTJk0AACtXrnyqW/sREdEDo0ePxrVr17B69WqjT20mIqK6YS7rY5OdDqTX69G/f38kJCRAqVRi2LBhWLt2rXi3AQCYNm0a+vbtiyFDhuDkyZNITEzEmjVrJP0UFhaiX79+OHbsGGxtbTF37lz06dMHAwYMMMWwiYiIiIhMwpzWxyb76ketVsPZ2RlOTk5QKBQICgpCWlqapE1mZqZ4CNrHx6fa58CDQ9U9e/asdsEaEREREdHzxJzWxyY7HUij0Uge7qJUKqvdPq5du3ZITU3FmDFjcPDgQZSUlKCgoAD29vZiG5VKhbFjx0q2i42NxaZNm9C1a1fMmjWr1ruAAMAPP/zwh+6l/XtVVVXxEPsjmIcU8zBgFlLMw4BZSDEPKeZhwCyk6isPmUyGjh07PvZzc1of1+vDwubMmYMlS5YgKSkJXl5eUCqVkqeR3r59G1euXEGPHj3E92bOnIkWLVqgoqICCxYsQHx8PKZMmVLrfmxtbevluoGMjAxer/AI5iHFPAyYhRTzMGAWUsxDinkYMAup+srj4TNn/ohntT42WRGgVColT4fUaDTVnjioVCrFhxCVlJQgNTVV8tj3ffv2ISAgAFZWVuJ7Dx9pr1AoMHToUGzfvt1UUyAiIiIiqjPmtD422XGSDh06ICsrC9nZ2dDpdFCpVNWe4Jmfny8+jCU+Pr7aw0hUKhWCgoIk792+fRsAIAgCDh06BFdXV1NNgYiIiIiozpjT+thkRwIsLS0RHR2NiIgI6PV6hIWFwdXVFXFxcXB3d4efnx/OnDmDtWvXQiaTwcvLCwsXLhS3z8nJwa1bt+Dt7S3pd9asWSgoKIAgCGjXrh0++eQTU02BiIiIiKjOmNP6uEE8Mbg+zwvj+XkGzEOKeRgwCynmYcAspJiHVF3mUVFRgZycHJSXl9dJf89aRUWF5PSQhs7UedjY2ODVV1+tto/n6d9ovV4YTERERGQOcnJy0KRJE7z22mv1ckfBP6qsrIy3U3+EKfMQBAF3795FTk4OXn/9dZPs41ngvaSIiIiowSsvL0fz5s2fywKAni2ZTIbmzZs/t0eNHmIRQERERASwAKAn9mf4W2ERQERERETUwLAIICIiImrgEhMTodFonnq7r776Ct9++60JRkSmxguDiYiIiBq4pKQkuLq6VntwFQDo9XrJE2sfNXLkSFMP7ZmrrKyEpeWff4n8558hERERkZnLyclBREQEOnbsiPPnz8Pd3R1hYWFYv3498vPzERMTAw8PD5SWlmLJkiW4evUqKisrMWXKFPj7+yM3NxfR0dEoKysDACxYsACdOnXC6dOnsXHjRtjb2+PKlSt48803ERMTIzmnff/+/bh48SJmzZoFGxsbfP311xg4cCACAwPxn//8BxERESgpKcHXX3+NiooKODs7Y/Xq1bC1tcWGDRvQqFEjjB8/HqNHj4aHhwdOnz6NoqIiLFu2DF5eXpJ5lpSUICoqCvfv30dlZSU++OAD+Pv7IycnB++//z727NkDANi2bRtKS0sxdepU/PLLL1i4cCHy8/NhYWGBuLg4tGrVSuyztLQU06dPR15eHqqqqhAVFYW+fftCrVZj+fLlKC0thUKhwI4dO2BlZYVFixbh4sWLsLCwwNy5c+Hj44PExESkpqaitLQUVVVViI+PrzHnPxMWAURERESP+Nf3OfjmXHad9jncywlhb79aa5sbN24gLi4Oy5cvx7Bhw5CSkoKvvvoKaWlp2LJlCzZv3owtW7bAx8cHK1aswP379xEeHo5u3bqhWbNmSEhIgLW1NbKysjBz5kwkJiYCAC5fvgyVSgUHBweMHDkS33//vWRxPmDAAHz55ZeYM2cOOnToIL5vZ2eHpKQkAEBBQQGGDx8OAIiNjcXu3bsxevToanPQ6/XYvXs3jh07ho0bN2LHjh2Sz62trbFp0yY0btwY+fn5eOedd+Dn51drLrNmzUJkZCQCAgKg1WrFp+k+dPz4cTg4OCA+Ph4AUFRUhIqKCsyYMQOxsbHw8PBAcXExbGxs8MUXXwAAUlJSkJmZifHjx+PAgQNiTt999x3s7Oywdu3aGnNu1KhRrWN9nrAIICIiIjIDr776Ktq2bQsAcHFxQdeuXSGTydC2bVvk5uYCAE6cOIHDhw9j+/btAACtVotbt26hSZMmmD9/Pn766SfI5XJkZWWJ/Xp4eMDR0REA0K5dO+Tm5lb7hr4mAwcOFH++evUq1q1bh6KiIpSUlKBHjx41bhMQEAAAePPNN8UxP0oQBKxduxZnz56FXC6HRqPBnTt3HjuG4uJiaDQasV9ra+tqbdq0aYNVq1ZhzZo16Nu3L7y8vKBWq9GiRQt4eHgAABo3bgwA+P777/Huu+8CAFq3bo2XX34Z169fBwB0794ddnZ2AB6fc+vWrWtJ7Pli0iIgPT0dy5YtQ1VVFcLDwxEZGSn5PDc3F/PmzUN+fj7s7OywZs0a8Y/Uzc0Nbdq0AQC0bNkSW7ZsAQBkZ2dj5syZKCwsxJtvvonVq1dDoVCYchpERETUgIS9/arRb+1N4dH1jFwuF1/LZDLo9Xrxs/Xr1+ONN96QbLt27Vq89NJLSE5ORlVVlbj4/W2/FhYWkr5q8+jDtubOnYvNmzejXbt2SExMxJkzZ2qdg1wur3E/KSkpyM/PR2JiIqysrODr6wutVgtLS0vJN/xarfaJxggAr7/+OhITE3Hs2DGsW7cOPj4+6NWr1xNv/9BvHy5WU851wVzWxya7O5Ber8fixYuxdetWqFQq7NmzB9euXZO0WbVqFUJDQ5GSkoKoqCh8+umn4mc2NjZITk5GcnKyOEEAiImJwXvvvYeDBw+iadOm2L17t6mmQERERGRWevTogZ07d0IQBAAPTmEBHnxj3qJFC8jlciQnJz/xQv+hF154ASUlJY/9vKSkBC1atEBFRQVSUlJ+9/iLiorQvHlzWFlZ4dSpU+LRgubNm+Pu3bsoKCiATqfD0aNHATz4Bt/R0RGHDh0CAOh0OvG6h4c0Gg1sbW0REhKC8ePH4/Lly3jttdfw66+/Qq1WA3iQT2VlJby8vMTxX79+Hbdu3apxof+4nP8oc1ofm6wIUKvVcHZ2hpOTExQKBYKCgpCWliZpk5mZCR8fHwCAj49Ptc9/SxAEnDp1Cv379wcADBkyxOg2RERERH8WUVFRqKysxODBgxEUFIS4uDgAwPDhw5GUlITBgwfj559/fupz14cMGYKFCxciJCSkxifhfvDBBwgPD8fIkSP/0LfjwcHBuHjxIoKDg5GcnCz2ZWVlhcmTJyM8PBxjx46V7GP16tX44osvEBwcjBEjRlQ7fejKlSsYNmwYQkJCsHHjRkyaNAlWVlaIjY3F0qVLMXjwYIwbNw5arRajRo2CIAgIDg7GjBkzsGLFihq/MX9czn+UOa2PZcLDEqeO7d+/H8ePH8eyZcsAAN9++y3UajWio6PFNh9++CE8PDwwZswYpKamYurUqTh16hTs7e3Rvn17tGvXDpaWloiMjIS/v794AcnBgwcBALdu3cKECRPEK8kf58KFCzWeQ2Zq5eXlsLGxeeb7NVfMQ4p5GDALKeZhwCykmIdUXeZRUVEBV1fXOumrPgiC8Kd4im1deRZ5XL16FVZWVtXed3Nze+w25rQ+rtcLg+fMmYMlS5YgKSkJXl5eUCqV4n1ojxw5AqVSiezsbIwZMwZt2rQRL+p4WtbW1rX+QkwlIyOjXvZrrpiHFPMwYBZSzMOAWUgxD6m6zCMjI6PaOeHPk7Kysud6/HXtWeRhZWVV7e8vIyPjD/f7rNbHJisClEol8vLyxNcajabaAyiUSiU2btwI4MG5ZqmpqWjatKn4GQA4OTnB29sbly9fRv/+/cX7ylpaWiIvL6/Gh1oQEREREZkbc1ofm+yagA4dOiArKwvZ2dnQ6XRQqVTw9fWVtMnPzxevBI+Pj0dYWBgA4N69e9DpdGKbH374AS4uLpDJZOjSpYt4P9ekpKRqfRIRERERmSNzWh+b7EiApaUloqOjERERAb1ej7CwMLi6uiIuLg7u7u7w8/PDmTNnsHbtWshkMnh5eWHhwoUAHlwQsXDhQshkMgiCgAkTJsDFxQUAMHv2bMyYMQPr1q2Dm5sbwsPDTTUFIiIiIqI6Y07rY5NdGGxO6uscSp67KcU8pJiHAbOQYh4GzEKKeUjV9TUBz3O2vCZA6lnkUdPfzPP0d2Sy04GIiIiIiMg8sQggIiIiImpgWAQQERER0Z9WZWVlfQ/BLNXrcwKIiIiICMjJyUFERAQ6duyI8+fPw93dHWFhYVi/fj3y8/MRExMDDw8PlJaWYsmSJbh69SoqKysxZcoU+Pv7Izc3F9HR0SgrKwMALFiwAJ06dcLp06exceNG2Nvb48qVK3jzzTcRExNT7UFav/zyCxYuXIj8/HxYWFggLi4OsbGxCAkJQZ8+fQAAc+fORZ8+fTBgwABxu5KSEkRFRYm3qPzggw/g7++PnJwcvP/+++IDq7Zt24bS0lJMnTq1xn21atVK7LO0tBTTp09HXl4eqqqqEBUVhYEDB0KtVmP58uUoLS2FQqHAjh07YGVlhUWLFuHixYuwsLDA3Llz4ePjg+TkZBw9ehSlpaWoqqpCfHx8jbk1ZCwCiIiIiB514Svg/M667dPzXaDjyFqb3LhxA3FxcVi+fDmGDRuGlJQUfPXVV0hLS8OWLVuwefNmbNmyBT4+PlixYgXu37+P8PBwdOvWDc2aNUNCQgKsra2RlZWFmTNnIjExEQBw+fJlqFQqODg4YOTIkfj+++/h5eUl2fesWbMQGRmJgIAAaLVaVFVVYeDAgdi3bx/69OkDnU6HkydPYtGiRZLtrK2tsWnTJjRu3Fh8cq2fn1+t86xpX486fvw4HBwcEB8fDwAoKiqCTqfDjBkzEBsbCw8PDxQXF8PGxgZffPEFACAlJQWZmZkYP368eKvMy5cv47vvvoOdnR3Wrl1bY26NGjWq/ff2J8YigIiIiMgMvPrqq2jbti0AwMXFBV27doVMJkPbtm2Rm5sLADhx4gQOHz6M7du3AwC0Wi1u3bqFJk2aYLPYmmUAACAASURBVP78+fjpp58gl8uRlZUl9uvh4QFHR0cAQLt27ZCbmyspAoqLi6HRaBAQEADgwcIeAHr16oVly5ZBp9MhPT0dXl5esLGxkYxZEASsXbsWZ8+ehVwuh0ajwZ07dx47x8ft61Ft2rTBqlWrsGbNGvTt2xdeXl743//+hxYtWsDDwwMAxKfkfv/993j33XcBAK1bt8bLL7+M69evAwC6d+8OOzu7WnNr3bp1Lb+RPzcWAURERESP6jjS6Lf2pqBQKMSf5XK5+Fomk0Gv14ufrV+/Hm+88YZk27Vr1+Kll15CcnIyqqqqxMXyb/u1sLCQ9FUba2treHt74/jx49i3bx8GDhxYrU1KSgry8/ORmJgIKysr+Pr6QqvVwtLSUvINv1arfaJ9AsDrr7+OxMREHDt2DOvWrYOPj49YNDyN394itKbcGjJeGExERET0nOjRowd27tyJh495unz5MoAH37C3aNECcrkcycnJT7zQBx58q+7o6IhDhw4BAHQ6nXhtwcCBA5GYmIhz586hZ8+e1bYtKipC8+bNYWVlhVOnTolHLJo3b467d++ioKAAOp0OR48eNbqvhzQaDWxtbRESEoLx48fj8uXLeP311/Hrr79CrVaL862srISXlxdSUlIAANevX8etW7dqXOg/LreGjEUAERER0XMiKioKlZWVGDx4MIKCghAXFwcAGD58OJKSkjB48GD8/PPPT32u++rVq/HFF18gODgYI0aMEE/p6d69O86ePYtu3bpJjig8FBwcjIsXLyI4OBjJycniAtzKygqTJ09GeHg4xo4dK1mYP25fD125cgXDhg1DSEgINm7ciEmTJkGhUCA2NhZLly7F4MGDMW7cOGi1WowaNQqCICA4OBgzZszAihUrahzn43JryEz6xOD09HQsW7YMVVVVCA8PR2RkpOTz3NxczJs3D/n5+bCzs8OaNWvg6OiIjIwMLFq0CMXFxZDL5Zg0aZJ4CGru3Lk4c+YMmjRpAgBYuXKl0Sez8YnB5oF5SDEPA2YhxTwMmIUU85DiE4MN+MRgKXN+YrC5rI9Ndk2AXq/H4sWLkZCQAKVSiWHDhsHX1xcuLi5im1WrViE0NBRDhgzByZMn8emnn2LNmjWwsbHBqlWr8Nprr0Gj0SAsLAw9evRA06ZNAQBz5syR3J6KiIiIiMjcmdP62GSnA6nVajg7O8PJyQkKhQJBQUFIS0uTtMnMzISPjw8AwMfHR/z89ddfx2uvvQYAUCqVaNasGfLz8001VCIiIiIikzOn9bHJigCNRiPejgp4MFiNRiNp065dO6SmpgIADh48iJKSEhQUFEjaqNVqVFRUSB4iERsbi+DgYCxfvhw6nc5UUyAiIqIGxIRnSNOfzO/9WzGn9XG93iJ0zpw5WLJkCZKSkuDl5QWlUgkLCwvx89u3b2P27NlYtWoV5PIH9crMmTPRokULVFRUYMGCBYiPj8eUKVNq3Y9Wq0VGRoZJ51KT8vLyetmvuWIeUszDgFlIMQ8DZiHFPKTqMo/KykrcunULdnZ21Z6m+zwQBKHaXXYaMlPmIQgCCgsLUVlZaZJ/j89qfWyyIkCpVCIvL098rdFooFQqq7XZuHEjgAePnU5NTRXPayouLsbEiRMxY8YMdOzYUdzGwcEBwIN73g4dOlR86ENtrK2teWGwGWAeUszDgFlIMQ8DZiHFPKTqMo+Kigrk5OTgl19+qZP+nrWKigpYWVnV9zDMhqnzsLGxQbt27artw1hRYE7rY5MVAR06dEBWVhays7OhVCqhUqnw6aefSto8vOpZLpcjPj4eYWFhAB7cM3by5MkICQmpdoHD7du34eDgAEEQcOjQIbi6uppqCkRERNRAWFlZ4fXXX6/vYfxuLBClzDUPc1ofm6wIsLS0RHR0NCIiIqDX6xEWFgZXV1fExcXB3d0dfn5+OHPmDNauXQuZTAYvLy8sXLgQALBv3z6cO3cOhYWFSEpKAmC41dGsWbNQUFAAQRDQrl07fPLJJ6aaAhERERFRnTGn9bFJnxNgLvicAPPAPKSYhwGzkGIeBsxCinlIMQ8DZiHFtZ9xfGIwEREREVEDwyKAiIiIiKiBYRFARERERNTAsAggIiIiImpgWAQQERERETUwLAKIiIiIiBoYFgFERERERA0MiwAiIiIiogaGRQARERERUQPDIoCIiIiIqIExaRGQnp6O/v37IyAgAPHx8dU+z83NxZgxYxAcHIzRo0cjLy9P/CwpKQn9+vVDv379kJSUJL5/8eJFBAcHIyAgAEuXLoUgCKacAhERERFRnTGX9bHJigC9Xo/Fixdj69atUKlU2LNnD65duyZps2rVKoSGhiIlJQVRUVH49NNPAQCFhYXYuHEjvvnmG+zatQsbN27EvXv3AACLFi3CkiVLkJqaiqysLKSnp5tqCkREREREdcac1scmKwLUajWcnZ3h5OQEhUKBoKAgpKWlSdpkZmbCx8cHAODj4yN+fuLECXTv3h12dnZ48cUX0b17dxw/fhy3b99GcXExOnbsCJlMhtDQ0Gp9EhERERGZI3NaH1vW/fQe0Gg0cHR0FF8rlUqo1WpJm3bt2iE1NRVjxozBwYMHUVJSgoKCghq31Wg01d53dHSERqMxOhatVouMjIw6mNXTKS8vr5f9mivmIcU8DJiFFPMwYBZSzEOKeRgwCylzzcOc1scmKwKexJw5c7BkyRIkJSXBy8sLSqUSFhYWdb4fa2truLm51Xm/xmRkZNTLfs0V85BiHgbMQop5GDALKeYhxTwMmIVUfeVRF4XHs1ofm6wIUCqVkgsZNBoNlEpltTYbN24EAJSUlCA1NRVNmzaFUqnEmTNnJNt6e3tX6zMvL69an0RERERE5sic1scmuyagQ4cOyMrKQnZ2NnQ6HVQqFXx9fSVt8vPzUVVVBQCIj49HWFgYAKBHjx44ceIE7t27h3v37uHEiRPo0aMHHBwc0LhxY1y4cAGCIODbb7+Fn5+fqaZARERERFRnzGl9bLIjAZaWloiOjkZERAT0ej3CwsLg6uqKuLg4uLu7w8/PD2fOnMHatWshk8ng5eWFhQsXAgDs7OwQFRWFYcOGAQAmT54MOzs7AMDChQvx8ccfo7y8HL169UKvXr1MNQUiIiIiojpjTutjmdAAbrRfn+eF8fw8A+YhxTwMmIUU8zBgFlLMQ4p5GDALKa79jOMTg4mIiIiIGhgWAUREREREDQyLACIiIiKiBoZFABERERFRA8MigIiIiIiogWERQERERETUwLAIICIiIiJqYFgEEBERERE1MCwCiIiIiIgaGEtTdp6eno5ly5ahqqoK4eHhiIyMlHx+8+ZNfPTRRygqKoJer8esWbPQu3dvfPfdd9i2bZvY7n//+x+SkpLg5uaG0aNH4/bt27CxsQEAbN++Hc2bNzflNIiIiIiI6oS5rI9NVgTo9XosXrwYCQkJUCqVGDZsGHx9feHi4iK2+eyzzxAYGIhRo0bh2rVriIyMxOHDhzF48GAMHjxYnODkyZMlj2COiYlBhw4dTDV0IiIiIqI6Z07rY5OdDqRWq+Hs7AwnJycoFAoEBQUhLS1N0kYmk6G4uBgAUFRUBAcHh2r9qFQqBAUFmWqYRERERETPhDmtj012JECj0cDR0VF8rVQqoVarJW2mTJmC8ePHY+fOnSgrK0NCQkK1fvbu3YvNmzdL3ps3bx7kcjn69euHqKgoyGSyWsei1WqRkZHxB2bz+5SXl9fLfs0V85BiHgbMQop5GDALKeYhxTwMmIWUueZhTutjk14TYIxKpcKQIUMwbtw4nD9/HnPmzMGePXsglz84QPHjjz/C1tYWbdq0EbeJiYmBUqlEcXExpk2bhuTkZISGhta6H2tra8nhkmclIyOjXvZrrpiHFPMwYBZSzMOAWUgxDynmYcAspOorj7ooPJ7V+thkpwMplUrk5eWJrzUaDZRKpaTN7t27ERgYCADw9PSEVqtFQUGB+HlNhzoe9tG4cWMMGjSoWvVERERERGSOzGl9bLIioEOHDsjKykJ2djZ0Oh1UKhV8fX0lbVq2bImTJ08CADIzM6HVatGsWTMAQFVVFfbt2yeZZGVlJfLz8wEAFRUVOHr0KFxdXU01BSIiIiKiOmNO62OTnQ5kaWmJ6OhoREREQK/XIywsDK6uroiLi4O7uzv8/Pwwd+5czJ8/Hzt27IBMJsPKlSvF85fOnj2Lli1bwsnJSexTp9MhIiICFRUVqKqqQteuXTF8+HBTTYGIiIiIqM6Y0/pYJgiCYLKZmon6PC+M5+cZMA8p5mHALKSYhwGzkGIeUszDgFlIce1nHJ8YTERERETUwLAIICIiIiJqYFgEEBERERE1MCwCiIiIiIgaGBYBREREREQNDIsAIiIiIqIGhkUAEREREVEDwyKAiIiIiKiBYRFARERERNTAmLQISE9PR//+/REQEID4+Phqn9+8eROjR49GaGgogoODcezYMQBATk4OPDw8EBISgpCQEERHR4vbXLx4EcHBwQgICMDSpUvRAB54TERERER/EuayPrasuylJ6fV6LF68GAkJCVAqlRg2bBh8fX3h4uIitvnss88QGBiIUaNG4dq1a4iMjMThw4cBAK1atUJycnK1fhctWoQlS5bgrbfewoQJE5Ceno7evXubahpERERERHXCnNbHJjsSoFar4ezsDCcnJygUCgQFBSEtLU3SRiaTobi4GABQVFQEBweHWvu8ffs2iouL0bFjR8hkMoSGhlbrk4iIiIjIHJnT+thkRwI0Gg0cHR3F10qlEmq1WtJmypQpGD9+PHbu3ImysjIkJCSIn+Xk5CA0NBSNGzfG9OnT4eXlVa1PR0dHaDQao2PRarXIyMiog1k9nfLy8nrZr7liHlLMw4BZSDEPA2YhxTykmIcBs5Ay1zzMaX1ssiLgSahUKgwZMgTjxo3D+fPnMWfOHOzZswcODg44cuQI7O3tcfHiRUyePBkqlep378fa2hpubm51OPInk5GRUS/7NVfMQ4p5GDALKeZhwCykmIcU8zBgFlL1lUddFB7Pan1sstOBlEol8vLyxNcajQZKpVLSZvfu3QgMDAQAeHp6QqvVoqCgAAqFAvb29gAAd3d3tGrVCtevX6/WZ15eXrU+iYiIiIjMkTmtj01WBHTo0AFZWVnIzs6GTqeDSqWCr6+vpE3Lli1x8uRJAEBmZia0Wi2aNWuG/Px86PV6AEB2djaysrLg5OQEBwcHNG7cGBcuXIAgCPj222/h5+dnqikQEREREdUZc1ofm+x0IEtLS0RHRyMiIgJ6vR5hYWFwdXVFXFwc3N3d4efnh7lz52L+/PnYsWMHZDIZVq5cCZlMhrNnz2L9+vWwtLSEXC7HJ598Ajs7OwDAwoUL8fHHH6O8vBy9evVCr169TDUFIiIiIqI6Y07rY5nQAG60X5/nhfH8PAPmIcU8DJiFFPMwYBZSzEOKeRgwCymu/YzjE4OJiIiIiBoYFgFERERERA0MiwAiIiIiogaGRQARERERUQPDIoCIiIiIqIFhEUBERERE1MCwCCAiIiIiamBYBBARERERNTAsAoiIiIiIGhhLU3aenp6OZcuWoaqqCuHh4YiMjJR8fvPmTXz00UcoKiqCXq/HrFmz0Lt3b/z73//Gp59+ioqKClhZWWH27Nno2rUrAGD06NG4ffs2bGxsAADbt29H8+bNTTkNIiIiIqI6YS7rY5MVAXq9HosXL0ZCQgKUSiWGDRsGX19fuLi4iG0+++wzBAYGYtSoUbh27RoiIyNx+PBh2Nvb47PPPoNSqcSVK1cwfvx4HD9+XNwuJiYGHTp0MNXQiYiIiIjqnDmtj012OpBarYazszOcnJygUCgQFBSEtLQ0SRuZTIbi4mIAQFFRERwcHAAA7du3h1KpBAC4urpCq9VCp9OZaqhERERERCZnTutjkx0J0Gg0cHR0FF8rlUqo1WpJmylTpmD8+PHYuXMnysrKkJCQUK2fAwcOoH379lAoFOJ78+bNg1wuR79+/RAVFQWZTFbrWLRaLTIyMv7gjJ5eeXl5vezXXDEPKeZhwCykmIcBs5BiHlLMw4BZSJlrHua0PjbpNQHGqFQqDBkyBOPGjcP58+cxZ84c7NmzB3L5gwMUV69eRUxMDLZv3y5uExMTA6VSieLiYkybNg3JyckIDQ2tdT/W1tZwc3Mz6VxqkpGRUS/7NVfMQ4p5GDALKeZhwCykmIcU8zBgFlL1lUddFB7Pan1sstOBlEol8vLyxNcajUY8hPHQ7t27ERgYCADw9PSEVqtFQUEBACAvLw9TpkzBqlWr0KpVK0m/ANC4cWMMGjSoWvVERERERGSOzGl9bLIioEOHDsjKykJ2djZ0Oh1UKhV8fX0lbVq2bImTJ08CADIzM6HVatGsWTPcv38fkZGR+PDDD/H222+L7SsrK5Gfnw8AqKiowNGjR+Hq6mqqKRARERER1RlzWh+b7HQgS0tLREdHIyIiAnq9HmFhYXB1dUVcXBzc3d3h5+eHuXPnYv78+dixYwdkMhlWrlwJmUyGnTt34saNG9i0aRM2bdoE4MGtjmxtbREREYGKigpUVVWha9euGD58uKmmQERERERUZ8xpfSwTBEEw9YTrW32eF8bz8wyYhxTzMGAWUszDgFlIMQ8p5mHALKS49jOOTwwmIiIiImpgWAQQERERETUwLAKIiIiIiBoYFgFERERERA0MiwAiIiIiogaGRQARERERUQPDIoCIiIiIqIExWgSUlZVh06ZNmD9/PgAgKysLR44cMfnAiIiIiIjINIwWAR9//DEUCgUuXLgAAFAqlVi3bp3JB0ZERERERKZhtAi4ceMGJkyYAEtLSwCAra0tnvQhw+np6ejfvz8CAgIQHx9f7fObN29i9OjRCA0NRXBwMI4dOyZ+9vnnnyMgIAD9+/fH8ePHn7hPIiIiIiJzZS7rY0tjDRQKBcrLyyGTyQA8KAoUCoXRjvV6PRYvXoyEhAQolUoMGzYMvr6+cHFxEdt89tlnCAwMxKhRo3Dt2jVERkbi8OHDuHbtGlQqFVQqFTQaDcaOHYsDBw4AgNE+iYiIiIjMkTmtj40WAVOnTkVERARu3bqFDz/8EOfPn8eKFSuMTlKtVsPZ2RlOTk4AgKCgIKSlpUkGJJPJUFxcDAAoKiqCg4MDACAtLQ1BQUFQKBRwcnKCs7Mz1Go1ABjtk4iIiIjIHJnT+rjWIqCqqgr37t3Dhg0b8OOPP0IQBPz1r39Fs2bNjE5So9HA0dFRfK1UKsWBPjRlyhSMHz8eO3fuRFlZGRISEsRt33rrLcm2Go0GAIz2WROtVouMjAyj7epaeXl5vezXXDEPKeZhwCykmIcBs5BiHlLMw4BZSJlrHua0Pq61CJDL5di6dSsGDhyIPn36GJ/ZU1KpVBgyZAjGjRuH8+fPY86cOdizZ0+d78fa2hpubm513q8xGRkZ9bJfc8U8pJiHAbOQYh4GzEKKeUgxDwNmIVVfedRF4fGs1sdGLwzu1q0btm3bhlu3bqGwsFD8nzFKpRJ5eXnia41GA6VSKWmze/duBAYGAgA8PT2h1WpRUFDw2G2fpE8iIiIiInNkTutjo0XA3r178eWXX+L//b//h6FDh2Lo0KEICwsz2nGHDh2QlZWF7Oxs6HQ6qFQq+Pr6Stq0bNkSJ0+eBABkZmZCq9WiWbNm8PX1hUqlgk6nQ3Z2NrKysuDh4fFEfRIRERERmSNzWh8bvTD48OHDv2uSlpaWiI6ORkREBPR6PcLCwuDq6oq4uDi4u7vDz88Pc+fOxfz587Fjxw7IZDKsXLkSMpkMrq6uCAwMxMCBA2FhYYHo6GhYWFgAQI19EhERERGZO3NaH8sEIzf9r6iowFdffYVz584BALy9vfHOO+/AysqqDqJ4NurzvDCen2fAPKSYhwGzkGIeBsxCinlIMQ8DZiHFtZ9xRk8HWrRoES5duoSRI0di5MiRuHTpEhYtWvQMhkZERERERKZg9HSg//73v/juu+/E1127dsXgwYNNOigiIiIiIjIdo0cCLCwscOPGDfF1dna2eP4RERERERE9f4weCZgzZw7+8pe/wMnJCYIg4ObNm1i+fPmzGBsREREREZmA0SKga9euSE1Nxc8//wwAeOONN6BQKEw+MCIiIiIiMg2jpwN9+eWXKC8vR7t27dCuXTuUlZXhyy+/fBZjIyIiIiIiEzBaBHzzzTdo2rSp+PrFF1/Erl27TDooIiIiIiIyHaNFQFVVFR59lIBer0dFRYVJB0VERERERKZj9JqAHj16YPr06RgxYgQA4J///Cd69uxp8oEREREREZFpGC0CZs+eja+//hpfffUVAKBbt24IDw9/os7T09OxbNkyVFVVITw8HJGRkZLPly9fjtOnTwMAysvLcffuXZw7dw6nTp3CihUrxHY///wzYmNj4e/vj7lz5+LMmTNo0qQJAGDlypXPzZPZiIiIiKhhM5f1sdEiQC6Xi08LLiwsRF5e3hM9J0Cv12Px4sVISEiAUqnEsGHD4OvrCxcXF7HNvHnzxJ//8Y9/4PLlywAAHx8fJCcnAwAKCwvRr18/dO/eXWw7Z84cDBgwwOgYiIiIiIjMhTmtj41eEzB69GgUFxejsLAQQ4cOxYIFC57oOQFqtRrOzs5wcnKCQqFAUFAQ0tLSHttepVJh0KBB1d4/cOAAevbsCVtbW6P7JCIiIiIyV+a0PjZ6JKCoqAiNGzfGrl27EBoaimnTpiE4ONhoxxqNBo6OjuJrpVIJtVpdY9vc3Fzk5OTAx8en2mcqlQpjx46VvBcbG4tNmzaha9eumDVrltHnFmi1WmRkZBgdc10rLy+vl/2aK+YhxTwMmIUU8zBgFlLMQ4p5GDALKXPNw5zWx0aLAL1ej9u3b2Pfvn2YPn26sea/i0qlQv/+/audZnT79m1cuXIFPXr0EN+bOXMmWrRogYqKCixYsADx8fGYMmVKrf1bW1vXy3UDGRkZvF7hEcxDinkYMAsp5mHALKSYhxTzMGAWUvWVR10WHqZeHxs9HSgqKgrjx49Hq1at4OHhgezsbLz22mtGB65UKpGXlye+1mg0UCqVNbbdu3cvgoKCqr2/b98+BAQEwMrKSnzPwcEBMpkMCoUCQ4cOxX//+1+jYyEiIiIiqm/mtD42WgQEBgYiJSUFixYtAgA4OTlhw4YNRjvu0KEDsrKykJ2dDZ1OB5VKBV9f32rtMjMzcf/+fXh6elb7TKVSVZv87du3AQCCIODQoUNwdXU1OhYiIiIiovpmTutjo6cD/V6WlpaIjo5GREQE9Ho9wsLC4Orqiri4OLi7u8PPzw/Agypn4MCBkMlkku1zcnJw69YteHt7S96fNWsWCgoKIAgC2rVrh08++cRUUyAiIiIiqjPmtD6WCY8+DvhPqj7PC+P5eQbMQ4p5GDALKeZhwCykmIcU8zBgFlJc+xln9HQgIiIiIiL6czFaBKxduxb3798XX9+7dw+xsbEmHRQREREREZmO0SIgPT0dTZs2FV+/+OKLSE9PN+mgiIiIiIjIdIwWAXq9HjqdTnxdXl4ueU1ERERERM8Xo3cHCg4OxpgxYzB06FAAQGJiIkJDQ00+MCIiIiIiMg2jRUBkZCTatm2LU6dOAXjw8LCePXuafGBERERERGQaT/ScgN69e6N3796mHgsRERERET0DRosAT09P8UEFFRUVqKyshK2tLX744QeTD46IiIiIiOqe0SLg/Pnz4s+CICAtLQ0XLlww6aCIiIiIiMh0nuh0oIdkMhn8/f2xceNGzJo1y2j79PR0LFu2DFVVVQgPD0dkZKTk8+XLl+P06dMAHtx16O7duzh37hwAwM3NDW3atAEAtGzZElu2bAEAZGdnY+bMmSgsLMSbb76J1atXQ6FQPM00iIiIiIjqhbmsj40WAampqeLPVVVVuHjxIqytrY1OUK/XY/HixUhISIBSqcSwYcPg6+sLFxcXsc28efPEn//xj3/g8uXL4msbGxskJydX6zcmJgbvvfcegoKCEB0djd27d2PUqFFGx0NEREREVJ/MaX1s9DkBR44cEf934sQJvPDCC9i8ebPRSarVajg7O8PJyQkKhQJBQUFIS0t7bHuVSoVBgwbV2qcgCDh16hT69+8PABgyZEitfRIRERERmQtzWh8bPRKwYsUKo53URKPRwNHRUXytVCqhVqtrbJubm4ucnBz4+PiI72m1WgwdOhSWlpaIjIyEv78/CgoK0LRpU1haPhi2o6MjNBqN0bFotVpkZGT8rnn8EeXl5fWyX3PFPKSYhwGzkGIeBsxCinlIMQ8DZiFlrnmY0/rYaBGg1Wqxe/duXL16FVqtVnz/9xYHNVGpVOjfvz8sLCzE944cOQKlUons7GyMGTMGbdq0QePGjX9X/9bW1nBzc6ur4T6xjIyMetmvuWIeUszDgFlIMQ8DZiHFPKSYhwGzkKqvPOqy8DD1+tjo6UCzZ8/Gr7/+ihMnTsDb2xsajQYvvPCC0Y6VSiXy8vLE1xqNBkqlssa2e/fuRVBQULXtAcDJyQne3t64fPky7O3tcf/+fVRWVgIA8vLyHtsnEREREZE5Maf1sdEi4MaNG5g+fTpsbW0xZMgQfP755489bPGoDh06ICsrC9nZ2dDpdFCpVPD19a3WLjMzE/fv34enp6f43r1796DT6QAA+fn5+OGHH+Di4gKZTIYuXbrgwIEDAICkpKQa+yQiIiIiMjfmtD42ejrQw/OLmjZtiitXruCll17C3bt3jXdsaYno6GhERERAr9cjLCwMrq6uiIuLg7u7O/z8/AA8qHIGDhwoPpDs4cQXLlwImUwGQRAwYcIE8arp2bNnY8aMGVi3bh3c3NwQHh5udCxERERERPXNnNbHMkEQhNoa7Nq1C/369cOVK1cwd+5clJaW4oMPPsCIESP+SAbPVH2eF8bz8wyYhxTzMGAWUszDgFlIMQ8p5mHALKS4Xf+XEAAAIABJREFU9jPO6JGAh5VE586deTtOIiIiIqI/AaPXBBARERER0Z8LiwAiIiIiogaGRQARERERUQNjtAgIDg7Gli1bcOPGjWcxHiIiIiIiMjGjRcCWLVtgYWGB6dOnIywsDNu2bcPNmzefxdiIiIiIiMgEjBYBr7zyCiZMmIDExER8+umn+N///ifew5SIiIiIiJ4/Rm8RCgC5ubnYu3cv9u3bB7lcjtmzZ5t6XEREREREZCJP9JyAyspKBAYGIi4uDk5OTs9iXEREREREZCJGi4BVq1bh/7N3nmFRnF0DvpfeOyyKVAHpigU1YlfssUdTNDExvl8sMTHGaIymF001amLMG80bY4ym2IItoom9ICrSbIhSZFF6L7vz/RgEsS0qsLs493XtBczOzD5z9pnhnOc0Ly+vBzr5vn37+PDDD1GpVIwdO5YpU6bUef+jjz7i6NGjAJSVlZGdnU10dDSJiYm88847FBUVoaenx0svvcTgwYMBmDt3LseOHcPS0hKATz75RGc6s0lISEhISEhISDzaaIt+rNYI2LRpE5MnT8bKygqA/Px8Vq1axauvvnrP45RKJe+99x6rV69GLpczZswY+vTpg7e3d80+b775Zs3va9asISEhAQATExMWLVqEh4cHCoWC0aNHEx4eXjOGOXPmMHDgQHVDl5CQkJCQkJCQkNAatEk/VpsYvG/fvpqTA1hbW7Nv3z61J46NjcXd3R1XV1eMjIwYMmQIUVFRd90/MjKSoUOHAuDp6YmHhwcAcrkcOzs7cnJy1H6mhISEhISEhISEhLaiTfqxWk+AUqmkoqICIyMjQHRLVFRUqD2xQqHA2dm55m+5XE5sbOwd901PTyctLY0uXbrc9l5sbCyVlZW4ubnVbPvyyy9Zvnw5Xbt2Zfbs2TVjuxvl5eUkJiaqHXNDU1ZWppHP1VYkedRFkkctkizqIsmjFkkWdZHkURdJHrVIsqiLtspDm/RjtUbAsGHDePbZZxk1ahQAf/75JyNGjFB32H0RGRnJgAED0NfXr7M9KyuL119/nUWLFqGnJzotZs2ahaOjI5WVlSxYsICVK1cyffr0e57f2NhYI3kDiYmJUr7CTUjyqIskj1okWdRFkkctkizqIsmjLpI8apFkURdNyaMhDY/G1o/VhgNNmTKFl156ieTkZJKTk5k6dSovvvii2oHL5XIyMzNr/lYoFMjl8jvuu23bNoYMGVJnW1FREf/5z3949dVXadeuXc12JycnZDIZRkZGjBo1ijNnzqgdi4SEhISEhISEhISm0Sb9uF59Anr06EGPHj3qs2sNwcHBpKSkkJqailwuJzIyks8///y2/S5evEhBQQGhoaE12yoqKpg2bRrDhw+/LcEhKysLJycnBEFg9+7d+Pj43Ne4JCQkJCQkJCQkJDSBNunH9TICHgQDAwMWLlzI5MmTUSqVjB49Gh8fH5YsWUJQUFBN1+Ft27YxePBgZDJZzbHbt28nOjqavLw8Nm7cCNSWOpo9eza5ubkIgoCfnx/vvvtuY12ChISEhISEhISERIOhTfqxTBAEoXEuU3vQZFyYFJ9XiySPukjyqEWSRV0kedQiyaIukjzqIsmjFkkWdZF0P/WozQmQkJCQkJCQkJCQkGhe3DUcaNiwYfc8cOvWrQ0+GAkJCQkJCQkJCQmJxueuRsCKFSsAWLt2LQDDhw8HJOVfQkJCQkKiqckqKOOT7UlcvZ7LklaeOFmaaHpIElqEUiVwvbhK08OQ0DHuagS4uLgAcOjQITZt2lSzvU2bNowcOZLZs2c3/ugkJCQkHjGqlCoegVQtiXqiVAn8fOQyn+08S7lShUwQGLzkAEvGt6Obt4OmhyehBSgKypj560mOJOcwK9+E6b290dOTqT9Q4pFHbU6AIAicOHGi5u+YmBhUKlWjDkpCQkLiUeOcopB5f8YS+PZOpv+VzvGUB28FL9E8iE3LY8Tyg7y9JZ52bjbsfKUHS4a4YGtmyDM/HOWLXWdRqiSD8VFmb1IWg5bs53RqPu1bmvLF3+eYujaGonLJKyChHrUlQj/88EPefPNNioqKEAQBKysrPvroo6YYm4SEhESzRqUS+PfcNVYdvMT+89cxNtBjSHAL9p3NZOyKw4xo15K5g/xxtpZCPx4l8ksr+WznWX4+ehlHC2OWPhnK0JAWyGQyyq4ZsXl6N97eHM/Xey5w9FIOXz8ZitxKmiOPEhVVKj7dmcT3+y/h52zJsqfaU3H9CoeyTfhoWyKjvini+4kdcbc31/RQJbQYtUZAUFAQW7ZsobCwEABLS8tGH5SEhIREc6a4vIo/YtL48WAKydeLkVsZ8/qANjwZ5oaduREnY+PZk2nAd/uS2ZWgYEYfH54P98DYQF/9ySV0FkEQ2Hwqgw8iE8kpLufZrh7MivDFysSwzn5mRgZ8OrYtXVvbM39jHIOX7OeLce3o6euooZFLNCWXs4uZse4ksWn5TOjizvwh/pgY6pOYLWNydy/8nK2Yvi6GYUsPsOyp9vSQ5oXEXVBrBFRUVLBz507S09Opqqp1L02fPr1RByYhISHR3EjLLeGnw5dZd+wKhWVVtHW1Ycn4dgwOboGhfm10pomhHq9FtGFsB1fej0xg0Y4kNkSnsnBoAL39nDR4BRKNxcVrRSzYFMehi9m0bWXNj5M6EeRifc9jRrVvRUgrG6atjeHZVcd4qVdrXuvvi4G+VP27ubL1dAbz/jyDngxWPNOegUEtbtsn3MeBLdPCmbImmudWH2PuID9e7O5Vp+mUhATUwwh46aWXsLS0JDAwECMjo6YYk4SEhESzQRAEoi/nsvrgJXbEZSKTyRgU5Mzz4Z60d7O957Fu9mZ8P7Ej/5zN4r2tCUz68Th9/JxYODQADwfJzd8cKKtU8s3eC6z4NxljQz3eHxHEU2Fu6NczsdPbyYLN07vx7tZ4vv3nItEpYnhQC2vTRh65RFNSWqHk3a3x/Ho8lfZuNiwZH4qrndld93ezN+OPlx7j9d9P89G2JOIzCvhkVAimRpI3UaIWtUaAQqHghx9+aIqxSEhISDQbKqpURJ7JYNWBFM6k52NtasiUHq2Z2NWdljb3p6D1auPEY60d+PHQJZbsPk/El/t4obsn03t7Y26s9jEuoaX8czaLhZvjuZJTwshQF94c7I+jpfF9n8fEUJ+PR4XQxcueN/88I4YHPdFO8ho1E85mFjL9lxguXCtiaq/WvNrft47n8G6YGxuw/Kn2fPPPRT7bdZYLWUV8N6EDrWzvbjxIPFqonUWhoaGcPXv2gU6+b98+BgwYQP/+/Vm5cuVt73/00UcMHz6c4cOHM2DAADp27Fjz3saNG4mIiCAiIoKNGzfWbI+Li2PYsGH079+fDz74QCqlJyEhoVVkF5WzNOo84Yv28Or605RUVPHBiCAOz+vD3EF+920A3MDIQI8pPVqzd3YvhrZtwbf/XKTP5/+w+VS69BzUMTLzy5i69gTPrT6Ogb6MXyZ35stx7R7IALiZ4e1c2DojHGdrUyb9eJyPtyVSqZSq+ekqgiCw7tgVHl92gNySCn56Pow5A/3qZQDcQCaTMa23Nz8825Er2SU8vuwghy9mN+KoJeqD1ujHghoGDRokBAYGChEREcLQoUNrXuqoqqoS+vbtK1y5ckUoLy8Xhg0bJpw/f/6u+//000/C3LlzBUEQhNzcXKFPnz5Cbm6ukJeXJ/Tp00fIy8sTBEEQRo8eLZw8eVJQqVTCCy+8IPzzzz9qx5KQkKB2n8ZAU5+rrUjyqIskj1qagywSr+YLc347LfjM3ya4v/GXMOGHo8LeJIWgVKru+1z1kUd0So4w9Ov9gvsbfwljvj0oxKXnPciwtZ7mMDduUFmlFL7fd1EIWLBd8J2/TVgadU4oq6y6r3PURx6lFVXC/I2xgvsbfwkjlh8QUnOKH3TIWk9zmh83k19aIUxde0Jwf+Mv4Zn/HhEUBaVqj1EniwtZhUKfz/YKXvMihR8PXhJUqvt/NukS2qr7aZN+rNaP/P3339fTrqlLbGws7u7uuLq6AjBkyBCioqLw9va+4/6RkZHMmDEDgAMHDtCtWzdsbGwA6NatG/v37ycsLIyioiLatWsHwIgRI4iKiqJnz54PNEYJiaYmu6icvWevcfJKLo76JTi5lmNv8XCrfxKaQ6US2JOUxepDlzh4IRsTQz3GdmjFpG4eeDs1biW1Du62bJrWjd+iU1m88yzDlh7gqc5uvNa/DbbmUv6WthFzJZf5G+NIvFpArzaOvPd4EG72jROWYWKozwcjguniZc/cP84w5OsDfDomhIhA50b5PImG5VRqHjPWxZCRV8acgW34vx6tG6T5V2tHCzZO68as9ad4e0s88Rn5vD8iSKo61sRok36s1gi40Tk4Ozub8vLyel6imEvg7Fz7wJHL5cTGxt5x3/T0dNLS0ujSpctdj1UoFLdtd3Z2RqFQqB1LeXk5iYmJ9R57Q1FWVqaRz9VWHkV5CIJASl4lR1OLOZZWQtK1cgTAWF9GuVLg68O7CXA0oaubGV1czWlpZaj2nM0RXZsbJZUq/r5QyJbEfDIKq3Aw02dSezsG+VpiaaxPZXYaiQ/hcb8febS1hBXDWvDzqVx+OXqFzSfTmNhOHEt9k0u1GV2bG7dSWK5kdUwOO84VYmemz/xeTnRzM6c46zKJWfd/vvuRR2tDWDK4BZ/sUzBlzQlG+FvxfAd7DPV1f17cQNfnx82oBIE/4/P5MSYHezMDPh3QAn+nSs6eTarX8fWVxaudzHE0tGFddBqnU66xoLcce7Pml1ukrXNDm/Rjtd96VFQUixYtIisrCzs7OzIyMmjdujWRkZFqT15fIiMjGTBgAPr6jWONGhsb4+/v3yjnvheJiYka+Vxt5VGRR1mlkiPJ2exJyiIqMYv0vFIAQlpZ80o/d/r6OxHQwopth09zvsSMXQkKvo/O4fvoHNrILYkIlBMR4EyQi9UjU9JNV+ZGak4J/zuUwvrjaRSWV9HezYZ5Qz0ZGOR8X3G66ngQeYS1g6mZBbyzJZ7lR6+z50oF7z4eSJinXYONSxPoyty4FUEQ+DMmnY+2JZJXWskL4Z680t8Xi4dM5L5fefgD4R2UfLwtiR8PpXCpUMayp9rfs7KMLqGr8+NWrheV89qG0/x7LocBgXIWj26Ltdn9LQrdjyw+DoCeIVeZteE0r+5QsOKZDnRwv3e1Ml1DU3OjIQ2PxtaP1T6NlixZwvr165k0aRKbNm3iyJEjbNmyRe2J5XI5mZmZNX8rFArkcvkd9922bRsLFy6sc+yxY8fqHBsWFnbbOTMzM+96TgmJpuRaYTl7k7LYnajgwIXrlFQoMTXUJ9zHgRl9vOnj54TTLR09W9sZM7SbL6/29yU1p4RdCQp2xWeyfO8Flu65QAtrEyIC5EQEOhPmadegSqZE/REEgeMpuaw6cIldCZnoyWQMDm7BpG4ehKop8dnU+Dlbse7FLmw7k8mHkQk88d1hHm/bkjcHS12Hm5LzikLe2hTH0Us5tHezYc2IYAJaWmlsPMYG+rzzeCBdvOx4/fdYBn+9n0/HhNyxxrxE03PownVeWX+KvNJK3h8eyDNd3JtkAWhgUAs8HSx48adonlx5hPdHBDKuk1ujf+6jjjbpx2qNAAMDA2xtbVGpVKhUKrp06cJHH32k9sTBwcGkpKSQmpqKXC4nMjKSzz///Lb9Ll68SEFBAaGhoTXbwsPD+eKLL8jPzwfEGKhZs2ZhY2ODhYUFp06dom3btmzatIkJEyaoHYuEREMjCAIJVwvYk5jF7qQsTqfmAdDS2oRR7V3o6y+nq5c9Job1s95d7cx4IdyTF8I9ySmuYE9SFrviM1kfncr/Dl/GysSAPn5ORAQ609PXUSoL2QSUVyn56/RVVh+6RFx6ATZmhvxfz9ZM6Oqu1TXYZTIZQ0Ja0NvPkRX/XGTFvmR2JyqY3sebF8I9pfjfRqS0QsnXe87z/b5kzI0N+HhUMOM6ujZIPHdDMDCoBYEtrZn+Swz/93MMz3Z1580h/tKc0BBVShVLos6zbO8FPB3M+XFSWJMbi22cLdkyvRsz1p3kjT/OEJdewIKhARgZSItOjYU26cdqNQkrKyuKi4vp1KkTs2fPxs7ODjMz9W5EAwMDFi5cyOTJk1EqlYwePRofHx+WLFlCUFAQffv2BUQrZ/DgwXWsXhsbG6ZOncqYMWMAmDZtWk0SxNtvv828efMoKyujR48e9OjRQ+1YJCQagrJKJYcvZrM7UcGepCyu5pchk0HbVjbMjvClj58c/xaWD72CY2duxJgOrRjToRWlFUr2n7/GrgQFUYkKNp3KwMhAj3BvByIC5PT1lz90WUGJulwvKmftkSusOXKZ60XleDtZ8NHIYEaGuuhUox0zIwNmRbRhTAdXPohMYPGOs6w/LnYd7usveVAbmqhEBQs3x5OeV8ro9q14c7CfVib9u9qZ8dv/PcaiHUn8cOASJ67ksuzJ9lLzuSYmI6+Umb+e5HhKLmM6tOK94YGYGWlmccfGzIjVz3Vi8c6zrNyXzNnMQr55pj0OWjh/mwPapB/LBOHehURLSkowMTFBpVKxdetWCgsLGTZsGLa22uUGvxeajAtrDrGKDYUuyiOroIyo6tj+gxeuU1qpxNxIn+4+jvTxd6J3G6cHVsLvVx5VShXRl3P5O0HBzvhM0nJLkcmgg5st/avDhjx19B+5NsyNhIwCVh+8xOZTGVQoVfRu48jz4Z6Eezs0eW5GY8hj37lrvLM1nuRrxfRu48jCYYE6MV+0YW7ci/S8Ut7dEs+uBAU+ThZ8MCKIzl72jfZ5DSmPvxMUzP7tNEqVwCejgxka0rJBztuUaPv8uBN/Jyh4/ffTVFap+GBkECNDWzXIeRtCFptPpTPn91jszY34bkJHgltZN8jYNIGk+6lHrdl5Y9VfT0+PkSNHNvqAJCQ0iSAIxKUXEJWkICoxizPposvNxcaUJzq2oq+/nM5edhpxnxvo69HFy54uXva8NcSfpMxCdsUr2JWQycfbk/h4exI+ThY1icXBLtZaE4agrShVAlGJClYdvMSR5BxMDfUZ18mV57p50NrRQtPDa1B6+DqyY2YP/ncohSVR54n48l9eCPdiRh+p6/CDUKlUserAJb7afR4BgTcG+vFCuKdOhVH0D5CzbWZ3ZvwSw/RfTnL4YjYLhgbUO4xR4v4or6pN0A5ysWLpk+21zhAf3s6F1o4W/GfNCcasOMSi0SGMCHXR9LAkGgnpyS/xyFNaoeTghetEJYlhPoqCcmQyaO9my5yBbejrJ8dXbqFVlXpkMhn+Lazwb2HFzH4+pOWW8HeCgl3xClb8m8zyvRdxtjKp9hDI6expr1PKSWNTWFbJb9Fp/HgohSs5JbjYmDJvkB/jO7ndd0UOXcLIQI8Xe3gxPLQli7afZcW/F9l4Mo15g/wZ3q6lVs1xbeZ4Sg5vbYzjrKKQfv5y3nk8gFa2ulltx8XGlPX/6cpnu87y3b/JnLicy/Kn2zc7I1jTXLpezPRfYojPKGBSNw/mDvLT2lyMIBdrNk/vxtS1Mbyy/hRx6fnMHeSHgVScotkhGQESjyRX80trSngevHCd8ioVFsYG9PB1oK+fnF5tHLUynvdutLI1Y1I3TyZ18ySv5EZisYLfT6Sx5shlLE0M6N3GiYhAOT19HbE0ab6K7r24kl3Cj4dS2BCdSlF5FR3dbZk7yI+IAPkj9Q/OydKEz59oy1Od3XhnSzyvrD/Fz0cu887jgQS56K77v7HJKa7gk+2JbIhOw8XGlJUTOjSLBlyG+nrMG+RPF097Zm04xbClB/hoZLC0AtxAbDyZxlsb4zA00OP7iR3pH6D9OTkOFsasndyZD/5K4L8HLpGUWcjSJ0OlRoTNDMkIaCSiEhXM+OUSrnbXaO1kTmtHC7ydLGjtaIGng7nkfm9iVCqBM+n5RCUq2J2YRcLVAgDc7Mx4qrMbff3khHnaNYvVchszI0a1b8Wo9q0oq1Ry4Px1diVksjsxiy2nMzDS1+Mxb3siApzpF+CEk2XzLh0pCAJHL+Ww6sAl/k5UoC+TMTSkBZO6edLW1UbTw9MoHdxt2TytG7+dSGXxjrMMW3aAp8LcmB0hdR2+GZVK4LcTqXy8PYmisir+09OLmX19NJbI2Vj09nNi28zuvLzuJK+sP8WR5GzeHhaoUwnx2kRxeRULN8fzR0wanTxsWTI+lJY22ltZ7FYM9fV4d3gQgS2teWtTHI8vP8D3Ezvi56y5crcSDYvaJ9iJEydYtmwZGRkZVFVVIQgCMpmMqKiophifztLW1YbBvlbkKo1JyChgR1wmqptSsFtam9C62iho7SgaCa2dLHCyNJZc8g1ESUUV+89fZ09iFnvOZnGtsBw9GXR0t2PuID/6+TvR2lG7wnwaGhNDffoFyOkXIEepEjhxOZdd8ZnsSlDw5sYzzN8Eoa42RAQ60z9A3qxCAMoqlWw9ncGqgykkXi3A1syQab28mdDVHblV8zZ87gc9PRnjOrkxMKgFX/59jjVHLvNX7FVmR/jyZJjbI+UhuRNJmQW8tTGO6Mu5dPKw5YMRwbRxttT0sBqNFtamrHuxC1/uPsc3/1zk5JU8lj8dirdT873mxiAho4Dp62K4dL2Yl/t483JfH529l57o5Iq33IL/W3OCUd8c4rOxbRkcLPWYaA6oNQLmz5/PvHnzCAoKQk9PNyewJnCwMOaFjvY1GeLlVUouZ5dwMauIi9eKuHitmIvXivgtOpXiCmXNcRbGBnWMghu/u9ubN4tV6sYmPa+UPdWr/YeTs6moUmFpYkBPX0f6+YuhMI/qCqe+nowwTzvCPO2YP8Sfc4qiGoPgk+1JfLI9idaO5kQEOhMRIKdtKxudTCzOKixj7ZErrD16metFFbSRW/LJKDG0QUp4vDvWpoa883ggT4aJIUILNsfzy7FU3hkW0KjVbrSV4vIqlkSd54cDl7AyMWDxmBDGtG+lk/fE/WKgr8frA/zo7GnPq+tPMWzpQd4fEcSYDg1TxaY5IwgCa45c5oPIRGxMDVk7uTOPtXbQ9LAemvZutvw1I5z//HyCqWtjmNa7Na/1b/NI3A/NGbVGgKWlJT179myKsTRrjA308ZVb4iuvu5oiCAKKgvJqw6Co2kgo5nByNn+eTK/ZT19PhpudWa2B4GhRE2ZkY/ZoKrUgVnc5nZZHVKJYzScpsxAATwdzJnZxp4+/E508pG67tyKTyWjjbEkbZ0tm9PUhPa+U3QlipaHv9yXz7T8XcbI0rik92tVL+xOL49LzWX0wha2nxRKfff2ceD7ck8da2zdrb09D08bZkl9e7Mz2uEw+jExk3MojDGvbkjcH+2l1k7SGQhAEdsYreG9rPBn5ZYzv5MobA/0eycWDHr6ObJvZnZm/nmT2b6c5fDGb90dorp69tpNfUsmcP06zM15BrzaOfDa2bbOqte9kZcKvU7qwcFM8y/deJPFqIV+Nb4fVI5pj1hxQeyd37tyZRYsWERERgZFR7UMwMDCwUQf2qCCTyXC2NsHZ2oRu3nVXC4rKq7hU7TGoNRKK2XfuOhVKVc1+9uZGdYyCGy8XW1P0m6GVXlRexf5z14hKymJvUhbZxRXo68no5GHL/MH+9PV3wqsZhbU0BS42pjz7mAfPPuZBfkkle89msSshk40n01l79AqWxgb0bONIRKAzvdo4as1DX6kS+DtBLPF57FIOZkb6PBnmyrOPeUhz4CGQyWQMDm5B7zZOfPvvRVb8e5HdCbVdh5urRyU1p4S3t8SzJykLP2dLvn4ylI4edpoelkaRW5mwdnIXlkSdZ+me85xOy2P5U+2bdUjUg3Dicg4vrzuFoqCM+YP9eSHcs1mukhsb6PPJ6GCCXKx4d2sCI5YfZOWEjng7Sc9bXUStEXD69GkA4uLiarbJZDJ++umnxhuVBCCGBgW3sr6tWYdSJZCWW1JjFNwwEHbEZZJbUlmzn7GBHp4O5rflHng5muvcSk5qTom42p+UxZHkbCqVAtamhvRq40gfPyd6+To169KOTYm1mSEjQl0YEepCWaWSQxevsytewe5EBX/FXsVQX0bX1mLH4v4Bco3E1xeUVbLheCo/HkohLbcUFxtT3hriz9iOrlibSvOgoTA10mdWf1/GdmjFB5EJfLrzLBuiU1kwJIC+/k7NxsNSUaXi+/3JLN1zHj2ZjPmD/Xmum4fkQaxGX0/GrP6+dPa0Y+avpxi+/ADvPh7IEx1dm80ceFBUKoFv/73IF3+fo6WNCb+/9BjtmnnBAZlMxoSuHvjKLZm6NoaRyw/y1fh2UidyHUStJrhmzZoHPvm+ffv48MMPUalUjB07lilTpty2z7Zt21i2bBkymQw/Pz8+//xzjhw5wscff1yzT3JyMl9++SX9+vVj7ty5HDt2DEtLcRXik08+0ZnObA2Fvp4Md3tz3O3N6eNX972c4oqbworE0KK49Hy2n7laJzHZxcYUr1tyD7wdLXDUksRkpUrg5JVcdidmsSdJwTlFEQCtHc2Z1M2Tvn5OdHC31dlEK13BxFCfPn5y+vjJa76TGx2L39oUx1ub4mjnalPToKyxV4NSrhfz46GUmlyaMA873hriTz//R6vEZ1PjamfGdxM6sv/8Nd7ZEs/kn6Lp1caRhUMDdN7jcvhiNgs2x3Ehq4iBgc4sHBagUxVcmpJu3g5smxnOq+tP8cYfZzh8MZsPRgZj8YhWu8sqLGPW+tMcuHCdISEt+HhUsNZ4SZuCzl72bJkRzn/WRDP5p2hm9fNlWm/vZukBaWi0RT++6527efNmhg8fzurVq+/4/qRJk+55YqVSyXvvvcfq1auRy+WMGTOGPn364O3tXbNPSkoKK1euZN26dVhbW5OdnQ1Aly5d2Lx5MwB5eXlERETQrVu3muPmzJnDwIED7/n5jyp25kbYmdvR6RYXdllldWLyLQbChuhUSm5KTLY0NsAG6Rz2AAAgAElEQVTrpoRksbSpOW52jZ+YXFBWyb5z19iTmMXes1nkllRiUJ3MOq6TG339nPDQsu6KjxL6ejI6etjR0UOsrnQhq4hdCQp2xWeyeMdZFu84i5eDOf2rDYJQ14ZJLBYEgcMXs1l18BJRSVkY6MkYFtKSSd08dbqlvS7S3ceRHa+IXYe/2n2eAV/t4/lwT2b08dE5RfB6UTkfbUvkz5h0Wtmasuq5jvTxk1Yy1eFkacJPz3dm+d4LfLX7HLFp+Sx7qj0BLR+tspH7zl1j1oZTFJVX8cmoYMZ1ejS9Ii42pvz+f48x949YPv/7HPEZBXz+RFupDPo90Cb9+K7fUmlpKQDFxcX3d3XVxMbG4u7ujqurKwBDhgwhKiqqzkVu2LCBp59+Gmtr8R+5vf3tFSh27txJ9+7dMTWVVmYeBhND/ZpE0JsRBIHMgrI6YUUXrxVx6EI2f8bUTUx2tzPD65bcA29Hi4cKw7mcXczuxCyiEhUcu5RDlUrA1syQ3m2c6OPvRA9f7Yk/l6hFJpPhI7fER27JtN7eXM2/kVis4If9l/ju32QcLIxrOhY/1tr+vrtjllUq2XIqg1UHxUY1duZGzOjtzTNd3HGSSnxqDEN9PSZ39+Lxdi1ZvEPsMrsxJp15g/0Y0c5F6xUhlUpg3fErLN5xlpKKKqb1bs303j5SLfz7QF9Pxst9fQjztOPldScZ8c1B3h4WwFNhblr//T8slUoVX/x9jm//uYiv3IJfXuxyW8GPRw0TQ32+HNeOwJbWfLw9kVHfFLNyYgfc7aVFuzuhTfqxTBAEQf1u98+OHTvYv38/H374IQCbNm0iNjaWhQsX1uwzdepUPDw8iImJQaVSMX36dHr06FHnPBMnTmTSpEn07t0bgLlz53Ly5EmMjIzo2rUrs2fPrpOwfCdOnTqFsXHTZ+iXlZVhYqK7ykpJpYq0/ErS8itIza8ktUD8Pb2gkqravGRsTPRpZW2Iq7UhrtZG1T8NcTQ3QO+mfwhlZWUYGhmTkFXGsbQSjqaVkJov5jC42xgS1sqMzq7m+DkYN8uE5lvR9flxN4oqlESnlXI4tZjjaSWUVgmYGsjo6GJGVzdzOrUyxeIWhetmWeSUVPHX2QK2nSsgv0yFp60RI/yt6eVljtEjEvKjS3Mj6VoZ3x7N5lx2OQGOxrzU2QFv+4Z73jakLC7mlLP08HXOXi8nxNmEaZ0dcLPRrao/2jY38kqVfHogi5iMUnp4mPNyV0fMjZruPm1KeSiKKvlkXxZJ18oZ6GPJf8LsMdGiqmnaMDdiMkr4+N8sAOb1dKJ9SzONjUWT8rhXGI426cdq/TXl5eX8/vvvnD9/nvLy8prtN8ckPShKpZLLly+zZs0aMjMzeeaZZ9i6dStWVqJbMSsri3PnzhEeHl5zzKxZs3B0dKSyspIFCxawcuVKpk+ffs/PMTY21kjeQGJios7nK3S4w7YqpYq03NI6FYsuXivicFoR288V1uxnbKAneg4czfFyMCf2UgEnM8vJL63EUF9GFy97XujhRB8/OW72mntQaIrmMD/uRqe28BJif4xDF7PZFa/g7wQF+y+L4TxdW9sTUd3ErIW1KYmJiVRZtmTVwUv8FZtBlUqgr5+c58M96Or16JX41KW54e8Pw8MFfj+RxqIdSbwcmc6T1V2H7RqgrGZDyKKovIovdp3jx0Pp2JoZ8cUTbRkZqv1eizuhjXPj93YCK/Zd5PNd57hcmMXyp9oT5NI0oXpNJY/tZ67yRmQsggBLnwxlWNuWjf6Z94s2zA1/f+jeroQXf4pmwe5M5g7y48XuXhq51zQlj8TExIc+R1Ppx2qNgNdffx0vLy8OHDjAtGnT2Lp1K15eXmovQC6Xk5mZWfO3QqFALpfftk/btm0xNDTE1dUVDw8PUlJSCAkJAWD79u30798fQ8PacBAnJycAjIyMGDVqFKtWrVI7FomGxUBfDw8HczwczG+rBnCnxOQz6flsO3MVS2M9+ge2pK+fE919HXUuhlji/jE20Kd3Gyd6t3HiwxFBnEzNY1dCJn/HK1iwWWxI1baVNZUVZSRkJWNupM/Tnd157jEPKf9Dh9DTk/FEJ1cGBDmzZPd5/nc4hcjYq7wW4ctTGuw6LAgC285k8t5f8WQVlvNUmBtzBvhJlcQaGD09GVN7edPJQwwPGvXNIeYP8WdiV3edNLRupqxSyQeRCfx85AptW1mz9Mn2j+Si1f3gZm/Gn1MfY/Zvp/loWxIJGQV8Mjqk2ZYWvl+0ST9Wq4VduXKFr7/+mqioKEaOHMnQoUN5+umn1Z44ODiYlJQUUlNTkcvlREZG8vnnn9fZp1+/fkRGRjJ69GhycnJISUmpiZECiIyMZNasWXWOycrKwsnJCUEQ2L17Nz4+PmrHItF03C0xubxKyYVzZwkMCNDQyCQ0jZ6ejA7utnRwt2XeIP/qxOJMdsYrKCxXsWBoAGM7tpJyQCrLQFCp308LsTY1ZOGwAMaHufLOlngWbo7nl6NXeOfxQLo0cdfhy9nFLNgcz75z1whoYcWKZzoQ6mbbpGN41OjkYUfky92Z/dtp3t4Sz5HkbD4ZHaKzZXsvZBUy/ZeTJGUWMqWHF7Mj2mh900RtwdzYgG+ebs/yvRf4/O9zXLhWxHcTOuIiVd7SKv1YrRFgYCDuYmVlxblz53BwcKjJUlZ33MKFC5k8eTJKpZLRo0fj4+PDkiVLCAoKom/fvnTv3p2DBw8yePBg9PX1mTNnDra24kM6LS2Nq1evEhYWVue8s2fPJjc3F0EQ8PPz491331U7FgnNY2ygXyc/4JFBEKAkG66fq36dF3/mp+Fi3ALy+oFrZ3AOBgPdik1+WLydLPB28mZqL+9qt62npofU9AgC5F2B1GOQegRSj4IintbmLSHvJWj3FJjpXrMqX7klayd3ZkdcJh9EJjJ+5RGGhrTgzcH+jV5+s7xKyYp/kln+zwWM9PVYODSAiV3dpRKyTYSduRH/ndiR/x5IZvGOs8Rl7GfZk+1pq0O18wVB4LcTaby9OR5TI31WP9eJ3n5Omh6WziGTyZjexwf/Fla88uspHl96gG+ebk/nJl4Q0Da0ST9Wmxj822+/ERERwblz55g7dy4lJSXMnDmT8ePHP4QImhZNxoVpOj5Pm2jW8lBWQW5KrbKffb5W4S/Nrd3PwATsfcCqJZXppzEsyazd3jIUXMOgVZj40+LR+afTrOfGzSgr4WqsqOynHhGV/8Kr4ntGFtCqI7QMpSRpD2bXT4vzIngMdJoszg8dpLRCyYrqrsN6MhnTerdmcneveocG3M/cOHjhOgs2xZF8vZghIS1YMCQAZ2vtSaJtCHTpXjlxOZeX150kq7CMuYP8eb6bR4OHBzW0PIrKq3hr4xk2ncqgq5c9X41vp5GGiA+CNs+NC1lFTFkTzZXsEhYOC2BCl8YPFZN0P/Xc0xOgUqkwNzfH2tqaTp06ERUV1VTjkpDQTsry4fqFm1b2q1f3c5JBVdutGQs5OPhC4Ejxp4OPqPxbu4KeuCJ5ITERfxfr6lXgY5B2DA5/A6ol4jlsPUQvQatO4k+nANCX8ih0ipKcm1b5j0F6DFSJ5ZexdgOPcPG7veX7vewyDn/bKjj+A8RugJM/g0sH0RgIHAWGuqGUgNh1+NX+vozp0IoPIxP5bNc5NkSnsWBoAP0aqOtwVmEZH/yVyJbTGbjbm/G/58Po6evYAKOXeBg6uNsS+XI4s3+L5f2/EjiSnM2nY0KwMdNOr2dcej7Tf4nhSk4Js/qLja8ehUp1TYG3kwWbpnXj1V9PsXBzPHHp+bw/Iui+S0dLNCz31Cj09PT473//y+DBg5tqPBISmkelgoK0agX/Ql1lv6g2mQc9A7DzEpV8v8HVyr4v2HuDaT1d31YtIXCE+AIxHvzq6eqV4qNwcS/ErhffM7IAl/bVhkGYuGqsg6EizRZBEOfIzav818+J7+kZQIu20HFStdIfJn7398I5GIZ9Bf3fhdPr4fh/YdNLsHM+hD4DHZ8HO90JoXK1M2PFhA4cOH+dd7bG8+JP0fT0dWThsABaP2DXYaVKYO3Ry3y68yzllSpe7uvD1F6tm2cCoiDA1VOYZCcDurHKCGBjZsT3Ezuw6mAKn2xPZMjXB1j6VCjttSg/QxAEVh9M4ePtiThYGPPrlK6EeUrP1obGysSQ7yd25Mvd51i65wLns4pY8UwHnfG0NEfUhgN99tln2NraMnjw4DoNCWxsdCe+TyOumcuHKPtzOibWcrCUg4Vz9c/ql6WzuM3MDh6RWHmtc5FVlkL2hbqx+jcU/xurtQAm1uDQplrJ965V9m09QP/BE97qJQ9BgLzLkHq81jBQxINQ3eXZoQ24dqpdTbb3qfE06BJaNzfqQ0UJZJysVfhTj9aGfpna1ir7rl3EUB6j+lcUuaM8BAFS9ovGQOJfYvKwdz8Ie1H8qac7im+lUsX/DqWwZPd5yqqUPN/Nk+l9vLG8Q1L43eZGbFoeb22KIzYtn3BvB94bHojXAxoTWk1uCsT+Ji4GZJ8Xt3l0h95vgvtjGh3a/XI6NY9pv8SQmV/GnIFtmBzu9dBdxR/22ZFbXMHrv59md2IW/fzlfDomBNsGKG2rCXTpObr9zFVe++00FsYGrJjQoVGMQikcSD1qYwu2bdsGwNq1a2u2yWQyKTRIHRZyyq1bY0KRGANc9DdUFN2+n55htVFwq4Fw008LuRgf/hAK5yOLIEDxtdsTc6+fg7xU4IYNLAMbN1G59+ghhu/cUPbNHTRnqMlkorFh6wEhY8Vt5UWQEVMbRpQUKYaLgGiw3MgpcA0TQ0iMH+1ulg1GwdVaQyz1qOixUVWJ7zn4gt8QUeF37SzOn4aeMzIZePYQXwUZcOJ/cOJH+OUJce52fB5CJ4jzVcu50XV4eDsXFu9I4rt9yfx5Mp15g8Suw/dSDAvKKvls51nWHLmMg4UxXz8ZyrCQFjpfirIOJTkQv1EMBUs9Im5zD4duL5N55SLOF9bB6kHg2RN6zQP3rpodbz1p62pD5MvdeeP3WD7alsSR5Bw+G9u2QfpJPAhHk7OZ+espsovLWTg0gEmNkLMgcWcGBbfA09GcKT+dYPx3R3h/RCDjOrlpeliPHGo9AeXl5bd1273TNm1Ga6zB8iIoUkBhphhWUpRV/bui9meRQqwmcxsyMLO/s4FQx9PgfF8rjk1Jo34PykrIuXQHZf88lOfX7mdoVhuffyNW38EX7FuDYdOWLmsweQiC6NFIPVprGFxLAgSQ6YFTYK1R4BoGtp5a533SupUTlVL0uNys9OddEd8zMBGNqxur/K06gXnDVruotzyUlaIRePy/opdA30jMGeg0WQwX07Lv+W6cSs3j7c1xnE7Lp72bDe8ND6ppNnVDFoIgsOV0Bh9EJnK9qJyJXdx5bUCb5lNStrIMzu8UQ7/O7xJzjBz9IGQcBI8FG7E8YGJiIv6t3eHEajjwFRRngVcv6PUmuHXW6CXUF0EQ+OnwZT6MTMTewoivnwy9rax0fXmQZ4dSJbBszwWWRJ3Dzc6MpU+2J7hV0zQ3a0y07jlaD/JKKpix7iT7z19nYld3FgwNwLCBKnlpje6nxag1AkaOHMnGjRvVbtNmdG4iVFWID/ZChWgsFFYbDEWZN21TiPvcWIm8GWMr0XNws2Fwp5AkU9smVRIa5Hsozb09Tv/6Oci9VFcWli3qrubf+N2ypdaEyzTqvCzNg/Tom5KOo6GiupuzuWNdb0HL0CY3gG5F4w/NsnxRRjcU/rToWs+dhbOoXLl2FpX+Jijn+kDyyEqC6B/g1Drxu3YOEUOFgsZo7cLAzahUAr/HpLF4RxLZxRWM7+TG7AhfslKTMXZwZcHmOA5eyCbYxZoPRwYR0kp3QlLvikoFVw6JoT7xm8UFCwtnsSJUyDhxrt3yjK4zNypKxO/8wFdQch1a9xGNAddOGriY++dMWj7T18WQllvKaxG+/F+P1vcdHnS/94qioIyZv57kSHIOI9q15IORwbrfuFIQ4NI+sk7twKnfy2DVQtMjui+qlCoW7Uji+/2XCPO045un2+Ng8fALzTqn+2mAu878a9euoVAoKCsrIyEhgRu2QlFREaWlpXc7TKIhMDAC61bi616oVFCaU+tZuGEg3OxhSI8Rf1aW3H68vrH6UCRLZzBzaNqqNCol5KfeEqdf/Xvxtdr99AzFJFwnPwh4vG4VHhOrphuvNmJqI8aJe/cT/1YpRe/Azd6Cs5Hie3oGosJYE8Mepn7u6TKCIMZZ31y1RxFPjedEHghtx1eH9oSJoTa6sKLu5AeDP4W+C8UwkuM/wJYZsOstaPc0dHxBzGnRUvT0ZDzR0ZWB1V2HfzyUQmRsBp1bmfDvpRSMDfR4b3ggT3d21/2KLVlJEPurGOtfkAaG5uIzLOQJMcSnvvkdRmbw2AwxFOz4f+HgEvih+r7vNU/0Bmkxwa2s+WtGOHP/PMPiHWc5mpzDF0+0xb4BFMA7sTcpi9d+O01phZJPx4QwpkMr3Q7/qaqAuD/g8HJQnMEJIGGV6AkMn9XgHsrGwkBfj/lDAghsac0bf8Ty+NIDrJzYscYbKNF43NUTsHHjRv7880/i4uIIDg6uMQIsLCwYOXIkERERTTrQh+GRtwYFAcoLb/Im3CEE6YYBcXNN+xvI9ERD4DavgvPtHod7lC68TR4VxdWJubco+9kXoKqsdj9T2+rE3FtW9m3cdbpkpsbnR/F1SLuRcHwc0k/UJkRbudSWJm2CZmaNKouq8ura/EdqjaAihfiekWXdxGqXDlphQDaIPAQBrhwRlcOEzWJ4iVdvUUHwHaj19855RSHvbI3n4IVshrdryfwh/jhZ6nAVkcJMOPO7uOqfGQsyfXHlvu14aDMIjMzrdZp7zo3yIvH7PvS1GFbq3b/aGOjQgBfS8AiCwNqjV3jvrwRszQxZMj603h2m63OvVFSp+HSnuNLs52zJsqfa4+2kw0nkJTliONjRleL/bUd/6DqNi5UOtM7YIhqYhmbQZSo8Nl3MFdMR4tLzmfJTNNnFFSwaHcKIUJcHPtcjr/vVA7XhQDt37mTAgAFNNZ5GQZoI90FVebVhkHW7h6FQUWs0FGXVVqi5GRNr0RiwcLrFq+DM1ctnaWFQWKvs56fWHifTE5X6m0N3ahJzdWM1437RuvmhrITMM3UNg/yb4uBbht5kGDRsM7MGlUXx9dtr8yvLxfdu9F6oqc3vr5VVdRp8bhQq4ORPEL0aCtLBqhV0fA7aP6vVTekEQeDQyTi6tQ/W9FAejPIiSPoLTv8Kl/4VKzq1bC+G+gSNeiDZ12tulBfBsZVwaKnoLfYZAL3miiWGtZj4jHym/3KSy9nFvNKvfnX61cnjSnYJM9bFcDotnwld3Jk/xF93S8hmX4Qj38KptaJ336u3qOS37gsyWa0srp2FvR9BwiYwsYFuM6Hzf+ptaGqa60XlTF0bw7FLObzY3ZM3Bvo9UMdvSfdTj1oj4GHYt28fH374ISqVirFjxzJlypTb9tm2bRvLli1DJpPh5+fH559/DoC/vz++vr4AtGjRghUrVgCQmprKrFmzyMvLIzAwkMWLF2NkdO8VSmkiNAIqpbjSVMebcBcPw83lNo0sblLyb1L27bzAQHeSzRsCnZgfBRnVOQXHayviKCvE9xqwmdkDy0KlEo3Km8t0Zl8Q39MzhJbtblL6w0SDVAdotLmhrIJzO8TV4uS9oowCHhe9A25dtTLsSSfuk5tRVomyjV0vJm1XloghZSHjxJeDz0Od/r7kUV54kzGQK3qAes3V6u7TReVVzN94hs2nMujmbc9X40JxtLz7/4Z7yWPr6Qze/PMMMhksHhPCwCDdipUHqj16h8WQn6RIsUpg8BPQdaoYungTt8ni6mnY86GYcG7uBN1fE3uV6MD/2kqlivf/SuCnw5fp7uPA0idD77vJnDbrftqiHzeaEaBUKhkwYACrV69GLpczZswYvvjiC7y9a2NSU1JSeOWVV/jf//6HtbU12dnZ2NuLq76hoaGcPHnytvPOnDmTiIgIhgwZwsKFC/Hz8+Opp56651i0eSI0ewQBygugUMH5K1fxad9DKxUNTaCT8+PmZmZpx24JrXnwZmb1lkVFsbiyX6P0H4OyPPE9M/u6q/wt22k84flBaZK5cf0CRK+CUz+LidFOgdDpBTEuXYvKyurEfSIIYs+I2A0Q97uYu2RiI672h4wT52MDPfceSB5lBXDsOzi0TLxf2gwWjYEWbRtkTA2NIAhsiE5l4eZ4LE0M+Xp8Ox7zvnPp2zvJo7RCyXt/xbPuWCrt3WxYMj4UVzvtT46vg7JKXMk/vFwsCW1qKxrrnV4Uw2/vwF3nxpWjsOd9sYKYVSvoOQfaPaUTZcfXH7/Cgk3xOFubsHJiB/yc6x+uqa26nzbpx40WFBobG4u7uzuurmJZsyFDhhAVFVXnIjds2MDTTz+NtbUYr3bjAu+GIAgcOXKkxhoaOXIky5YtU3uREhpEJhNDhEysqbqulAwAXcfQRKyUc6MUoSCIpTNvrMKnHYP9X9y5mVmrMNHrcz/VmfLTb6nNH1t7bsfqhPAbtfntW0vz635w8IaBH0Gft0TF9dj3EDkL/n5bjFPvNFlMNpa4O7c28tI3ElfbQ8aBT3/tWXE1sYIer0PYFDj6HRxeBt/1AL+h0PMNaBGi6RHWQSaTMa6TG+1cbZm69gRP/3CUGX18mNnXR2140NnMQqb/EsOFa0VM7dWaV/v7NljJySahLB9ifhK/p/xUsfjFkC+g7ZMPXuXLrTM89xck/wNR78PWl+HgV2IlqaDRWlMx706M6+SGt5MlL/18glHfHOLzsW0ZFKyDHp2b0Cb9uF5GQExMDOnp6SiVtTHgI0aMuOcxCoUCZ+da17tcLic2NrbOPikpKQCMHz8elUrF9OnT6dGjByD2Ihg1ahQGBgZMmTKFfv36kZubi5WVFQYG4rCdnZ1RKBRqx19eXk5iYmJ9LrVBKSsr08jnaiuSPOrSrORhGAReQeD1ArKqUkxzEjC9fgbT7DOYxm/BoLqZmdLQklKHIErtgylxCKbMLgCVobkoi/gzmORdwPR6LKbZZzC7fgbDkkwAVPrGlNoHUuo/gRL7EErtA1EZ35Tsdq2yujdC86DJ54ZpJ+jREZOceGzP/4HViR/RO/49xU7tyfUeTaFLT7GKlAbQtvtErzwfq9Q9WF/egdn10wAUO4ZS0HEeBa59UBlVe1HOJzfK5z+0PJyGoTeoN3bn12N3dh36SX9R4NKT60GTKbd5uFClxuCz/o4sPyrj66jz/BOfypzuTtib1c7FG/IQBIEd5wtZcSwbM0M9PujnTPuWAhfOndXg6OuPQfFV7M6txyZ5C/pVJRQ7tScn/GWKWnYTc+YuXlZ7DvVzQw7dlmKRcQDHM99h8udkyqI+5lrQFIpctNdLbwZ8MVDOB/8oeGltDE+G2PBMO1v01IxX254dN9Am/VjtU/31118nNTUVPz8/9PXFZBqZTKbWCKgPSqWSy5cvs2bNGjIzM3nmmWfYunUrVlZW7N27F7lcTmpqKs8++yy+vr5YWDxYNr+xsbFWuoQeNSR51KV5y+OmBMSaZmbH0E89ikXacSzivufmZmbFgiHmuWehslg8xrIleHatKdOp5xyMub4hupHW9vBobm4EQPhYMbn65M+YR/+A+aH5YrJ/h+egw7Ng1bJJR6QV98mNRl6xG+DcTrHSkkMbsRxr8FjMbdwwB5pifbLB5NG2E5S+BUe+xerIN1jtnAD+j4thQrfEmmua/4bAb9XhQTO3ZfLluHb08HUERHm4eHoz788zRMZeJ9zbgS/GtdWdSlJpJ+DwUrGCl0xPbPbXdSrmLUPv+3lX77kREAB9JkPCRkz2foTrwTfEPJE+b9UkGWsjm9sGsGBTHOui08iqMOLL8e3u2SxQk+FAD0tT6cdqjYC4uDi2bdt237V05XI5mZmZNX8rFArkcvlt+7Rt2xZDQ0NcXV3x8PAgJSWFkJCQmn1dXV0JCwsjISGBAQMGUFBQQFVVFQYGBmRmZt52TgkJCS1DJqtOAveB0KfFbTXNzMSEY73cTPG9G/H81d1RJTSEuQOEvyLWoL+wW0wk/ncR7PsU/IaIoUKe2rty2CCoVGJCZux6MTa7LF+sdtb5P2LehHOI7l+/qQ30ngdd/g8OfyNWnkncAgEjxDAheYCmR1jD2I6utHO1YdovMTy7+pgY6tPPl7PXypiydT8ZeWXMGdjmgRqONTkqpZjke3i5mN9kbC3ea2FTmq5Hi56eGArkP1wsKfrPIvh5NLh3gz4LwL1r04zjPjA20GfR6BCCXKx5b2sCI5Yf5PuJHWntqFvlXrVJP1YbCObj48O1a9fU7XYbwcHBpKSkkJqaSkVFBZGRkfTp06fOPv369ePYsWMA5OTkkJKSgqurK/n5+VRUVNRsj4mJwdvbG5lMRufOndm5cycg9jK49ZwSEhI6wI1mZr3nwcRNpESsFhtdBY+RDABtQk8ffAfA07/Byyeh6zQxufCnx2F5mBi3XJav6VE2LFlJsPtdWBICPw4Wa/v7DoIJG2FWIgz4UEyo1XUD4GZMbaHPfHglVswduBAF3z4Gvz0nykNL8JFbsnlaOE90cGX53osMXXqA17ZnoFLBhv90YWovb+02AMqLxHtmaXvYMAEKr8LARTArHvq/p5kmjfoGEPoMzIiGwZ+JXtvVA0WDIOP25FNNI5PJmNjVg58ndya/pJIRyw6yJ0l92Is2oU36sVpPQG5uLkOGDCEkJARDw1q3y42SRHc9sYEBCxcuZPLkySiVSkaPHo2Pjw9LliwhKCiIvn370r17dw4ePMjgwYPR19dnzpw52NraEhMTw9tvv41MJkMQBF588cWahInXX3+dV199la+++gp/f3/Gjh2r9iIlJCQkJB4SO0+IeB96vwnxm0TvwPY5osIc8oRYWchZR/IlrNQAACAASURBVOv5F2aKnVdj14vVr2408ur7NvgN1pn66g+NmZ0YEtJlqpg8fPQ78bsOGiV6BhzbaHqEmBrps2hMCF1b2/PWpji6uJrxzXPhWJtpcaWbggxRlidWi0ZzqzBR6fcbqj19SgyMIexFsbv48e/hwJewshf4D4Pe88WeKlpEFy97tswIZ8pP0bzwv2he6y/2ldCFDtDapB+rLRF6wxK5lbCwsAe4dM2grWWiHjUkedRFkkctkizqojPyyDgpGgNnfhe7fLt2EUOFAh5vsMo4jSaLG428YteLVVMElRgXHTL+gRt5NQVNOjeKs8V49aMrxX4HwWOgxxxw9G2az1dDpVLFhXNntfdeuXpaDPmJ+0OcX/6Pi94018bRnxp0bpQVwJFvxLKyFUWiod/zDbEKmxZRWqFk7p+xbD6VweBgZz4d0xZzY3F9W9L91KPWE6BLyr6EhISERBPSMhSGL4f+78OpXyD6B/hzMux0hPYTocMk7QrvUlaJCn/setEAuNHIq/trYgMmLVFutQZze+j3DnSdAYe+FsvIxv0BQWNEhdDBW90ZGhWtLP2pUsH5XaInJWW/2D8lbIqYS2LroenR1R8TKzFJPGyKWE706ErR2A99RuwzoInQpTtgaqTPV+PaEdTSmo+3J5J8rZiVEzriZq9jfSE0hFoj4NSpU7z//vskJydTWVmJUqnE1NSUmJiYphifhISEhIS2Y2YHj00Xw0iS98LxH8RwggNfinXzO00Gr96aqUd+t0ZeNzr4unVpXvH9jYG5PfR/V0xePbhE9P7E/S4aTj3naN3qsEaoLIXT68QE6+zzYOUiGscdnhV75egqZnZi6FKXqbD/c4heDad/FcP/wmeBhaOmR4hMJuPFHl60cbZkxrqTDFt2gOVPtefelfUloB5GwHvvvceXX37JzJkz+eOPP9i0aVNN/VIJCQkJCYka9PTAu6/4yrsCJ36EE/+Ds9vAzgs6viB2Kq1nJ+mHIvcynNkgKv/Xz2lvIy9dwtxBzAt5rLrZ1PEfRBmHjBMTih9FY6AoS/SQRP8AJdnQoh2M/gEChutER956Y+ksFm94bIZYKezoCvHe7vJ/4jZTW02PkB6+jmyZ3o0pP51g4qqjzA53QkeicjRGvZZl3N3dUSqV6OvrM3r0aPbv39/Y45KQkJCQ0GVs3MQ6+rMSRKXIQg675sMX/rB5WuNUHinNFVcqVw0Uq/vs+QDMHWHYEph9DsatAf+hkgHwsFg4ilWSXokVV4jjN8KyTrBpKuQ0TpM0rUORIM7jLwPF0rmuXeC5bTDlHzF3ojkZADdj4yaGAE47Dm0Git6Br9rCv59CeaGmR4e7vTl/Tn2McZ1cyStTqj/gEUetJ8DU1JSKigr8/f1ZvHgxTk5OqFSqphibhISEhISuY2AsKkXBYyDzjLh6HLsBTv4MLh2g04sQOBIMH7C5U1W52MArdr0Yi62sqNPICxu3hr0eiVosnERj4IZnIHqVGCrS7knoPlusKNWcEAS4uEdM9r0YBQamYu5L55c0nh/R5Dh4w5hVYkjQ3g9h7wdw9Fvx704vgKGpxoZmbmzAx6NCtLJbsLah1ghYvHgxgiCwcOFCfvzxR65evcrSpUubYmwSEhISEs0J52AY9pUYX376VzG2fNP/wc43xYTDjs/XT3FUqcQmS6d/rW3kZe4kGhQhTzS/Ov7ajqUcBn4M3WaKeSA34sbbPimGCdm6a3qED0dVOZz5TVT+sxJEr1afBeJ8bYrQNm3GOQieXAdp0bDnfdHbd3g59HwdQic0X49IM0GtEeDi4kJZWRlZWVlMnz69KcYkISEhIdGcMbEWq6WETRErqBz/r6g4HFoqxut3miw2k7u1hvq1s+KKf+xvkH8FDM3FOuYhT4BnT7HxkYTmsHSGQYug2yuiMXDiRzFZtt3T0GO27nllirNF78axlVCcBfIgGPGt2GlXCimrS6uOMHEzXNovGgN/vSomkfeaJ3rktKUfgkQd1D4x9+zZw6JFi6isrGTPnj0kJiayZMkStc3CJCQkJCQk7olMBp49xFdBhphoeOJH+OUJUWHs+DxGhm3g8F6I/bW6kZdedSOvBeA35NFp5KVLWLWAwYsh/BX4//buPq7q+v7/+OMIHrJUEMuDS/IKnCQotWqapnEQMFFRgS5Wmiu0lebUzCQnplbk9+vWXLU1tnLpmn61NHLo9Cfk0C68zNBkWhYNLQ6beBEqV4fz++Osc76fLyZWHD8nz/N+u+1245zP1ev9lFt7v/hcbf0V7HnF/QjZ6+5xP47Vnx4bey7//sj9jPy9K6DhLEQluZ/v3+NWnWFqTvdb4L6N8NH/czcDax9w/w4kPO5+T4IZTwiTr9VsE/D888/z2muvMW7cOABiYmI4evSozwsTEZEA0v4HkJDt/ovxP/7qvndg8xN4njfzg+tg2DPQZ6z78hPxf+1/AKmLYdB02PYr2LPMfS/I9ePczYCfPGsecF/vX7bNfUbq0AYICoF+d0D/ydCpt9nVfb9YLNAr2X02r/RNeOtpWH0vRPR1X0YVnaRmyk802wQEBwfTrl27b7Xz4uJinnrqKRobG8nMzGTSpElN1lm/fj3PP/88FouF3r1788tf/pLS0lKeeOIJqquradWqFQ8++CDDhw8HYPbs2ezYscNT0zPPPPO9eTObiIg0I6i1+0bhPmOg8h9UvLeKiAF3wFU/NLsy+bZCr4bUX7qbga2/hD3L/9MMjHffSBp6tXm1OevdTzd693n3mabLO8KQ2e6bW/30rdHfG61aQZ/R7kv2SlbBllz4S6b7SUr2X7jPGgQof5kfN9sEREVFsW7dOpxOJ2VlZSxfvpzrrruu2QE6nU4WLFjA0qVLsdlsZGRkYLfbiYry3kFfVlZGXl4eK1asIDQ0lGPHjgFw2WWXsWjRIrp164bD4SA9PZ1BgwbRvn17AGbNmsWwYcOarUFERL7HOvXmeHQmEWoALg2hXWDEs95mYPef3GcHfjTB/V37H1y8Ws4ed19+tv338OXncGUv96Nk+95h6pNtLkmtgtxPjIpNh/eXux+p+soI9+VV9hzo8iOzK7yo/Gl+3OzFWXPnzuXjjz/GarUyY8YM2rZty5w5c5rdcUlJCV27diUyMhKr1UpqaiqFhYWGdVatWsXdd99NaKj7bXodO7rf79a9e3e6desGgM1mIzw8nKqqqgselIiIiPipsGvcE+6H97ifILTrZVgSD+tnwakvfHvsqk9hw2Pwqz6weR5cGQ0/WQ0PbXc3I2oAfCfY6j7DMvV9SH7K/cjgP9phxU+gYr/Z1V00/jQ/vqD3BEyfPp3p06d/ox07HA4iIiI8n202GyUlJYZ1vnrz8J133kljYyNTpkxh8ODBhnVKSkqor6/nmmu8TxV49tlneeGFFxgwYAAzZ87EarWet5ba2lpTnhdbU1Oj59T+L8rDSHl4KQsj5eGlLIwuuTyiH6R15zSuPLCU0J1/xLX7T5zoMZpjMeNoaHNls5tfaB5t/l1C+MEVtDv6d7C04uQ1yVT1upPaDr3ACRw82AKDMdf36nejQyKthvWnw6H/oePBv9Dq4HpOXTOUf8dOpK5dyzxFyl/z8Kf5cbNNwL59+/j973/P0aNHaWho8Hy/bt265jZtltPp5LPPPmP58uVUVFRwzz33sG7dOs9pjcrKSh599FEWLVpEq//cUT5jxgyuuuoq6uvrmTt3Lnl5ec0+ujQkJMSU+wZKS0t1v8L/ojyMlIeXsjBSHl7KwujSzCMGbkyEqk+xFC8m/IMVhH+a734O/8Bp570R/Lx5OBvgH+vcN/se2QmXhbn3d9Mkwtp3JsxHozHL9/J3o+8NcCYb3nmO0O0vElpe5L50aMhj3/mRsmbl0RKNx8WaHzfbBMycOZNZs2bRq1cvz4EuhM1mo6KiwvPZ4XBgs9marNOvXz9at25NZGQk3bp1o6ysjL59+1JdXc0DDzzA9OnTiY+P92zTqZP7Rh2r1crYsWN5+eWXL7gmERER8VPh3WH0C3DLDPc9A9t/737x2I33u19EdqE36tZ+6b75ePvv4MQ/IbwHDF8M8T/RI2X90eXhMHQe9H/Q/TjRXf95q/iPJrjfPH2JPQ3Mn+bHzc7qw8PDSUxMJDIykquvvtrzv+bExcVRVlZGeXk5dXV1FBQUYLfbDesMHTqUHTt2AFBVVUVZWRmRkZHU1dUxefJk0tLSmtzgUFlZCYDL5WLz5s1ER0c3W4uIiIh8T3TsCaN/C1N2up8u895v4dd9YdMvoPpfX7/diXL3Or+6FjZmQ/sucOdfYMouuGmiGgB/17YT3PaM+56Bfne5HxO8pB/8vxw4c+ncF+pP8+NmzwRMnTqVOXPmMGDAAMO1RcnJyeffcXAwOTk5ZGVl4XQ6SU9PJzo6miVLlhAbG0tiYiK33HILb7/9NsOHDycoKIhZs2bRoUMH8vPz2bVrFydOnGDt2rWA91FHM2fO5Pjx47hcLnr37s38+fObHaSIiIh8z3TsCWNedP81uPi/3Jf17HzJ/UbpgT+HK/5zz8DRPe5lH7rnC/QZ7X6519WB9dSZS0ZoFxj1G/e/8ZZn4O3fuM8IDZgM/R+Cy9qbXeF34k/zY4vL5XKdb4WZM2fyySefEB0dbbgcKDc39zvGcPGYeV3Y9+76PB9SHkbKw0tZGCkPL2VhFNB5/Psj+Pt/wb7V0PpyuH48Zz55l8v/tRdC2rvfO/Djn/n/G4l95JL93XAcgLeecr9EsE24+03UN04E6+Xn3Uxzv+Zd0I3BGzduvBi1iIiIiJzbldGQ/gcY/Cj8fRFsf5Hgy22Q8jRcN+57/xdi+Rq2a+HOV91nfIqedF8e9O5v3W8Xv/5e96NH5Vtp9p6A66+/no8//vhi1CIiIiJyflf1goyX4LFPOZz6mvsyETUAl76rr4dxa2DCevfN3utnwnM/cr992tnQ/PbSRLNnAvbu3cvo0aO5+uqrDfcEtMQjQkVERES+lTYdoFVF8+vJpaXbQPjpejhcCIULIX8ybPs1JGTDtWPgGzzJMtA12wT88Y9/vBh1iIiIiIg0z2KBqKHQM9F9r0DRU/DafWB7FuxzoNew5vchzTcBF/I4UBERERGRi8pigZiR8MPhsP91eOtpWHEndLmRkGunAt+PG3TNonMmIiIiIvL91SoI+t7ufrfEyCXwZQVhn+qy9eY0eyZARERERMTvBbV2v2n4RxNwlJYSbnY9fk5nAkREREREAoyaABERERGRAOPTJqC4uJiUlBSSkpLIy8s75zrr169n+PDhpKam8sgjj3i+X7t2LcnJySQnJ3tejQywf/9+Ro4cSVJSEk8++STNvPBYRERERMRv+Mv82Gf3BDidThYsWMDSpUux2WxkZGRgt9uJioryrFNWVkZeXh4rVqwgNDSUY8eOAXDixAmef/55Xn/9dSwWC2PHjsVutxMaGsoTTzzBwoUL6devHxMnTqS4uJghQ4b4ahgiIiIiIi3Cn+bHPjsTUFJSQteuXYmMjMRqtZKamkphYaFhnVWrVnH33XcTGhoKQMeOHQHYtm0bAwcOJCwsjNDQUAYOHMjWrVuprKykurqa+Ph4LBYLo0ePbrJPERERERF/5E/zY5+dCXA4HERERHg+22w2SkpKDOuUlZUBcOedd9LY2MiUKVMYPHjwObd1OBxNvo+IiMDhcDRbS21tLaWlpd9xRN9cTU2NKcf1V8rDSHl4KQsj5eGlLIyUh5Hy8FIWRv6ahz/Nj019RKjT6eSzzz5j+fLlVFRUcM8997BuXcs/1zUkJISYmIv/wojS0lJTjuuvlIeR8vBSFkbKw0tZGCkPI+XhpSyMzMqjJRqPizU/9tnlQDabjYqKCs9nh8OBzWZrso7dbqd169ZERkbSrVs3ysrKvnbb//t9RUVFk32KiIiIiPgjf5of+6wJiIuLo6ysjPLycurq6igoKMButxvWGTp0KDt27ACgqqqKsrIyIiMjGTRoENu2bePkyZOcPHmSbdu2MWjQIDp16kTbtm3Zu3cvLpeLN954g8TERF8NQURERESkxfjT/NhnlwMFBweTk5NDVlYWTqeT9PR0oqOjWbJkCbGxsSQmJnLLLbfw9ttvM3z4cIKCgpg1axYdOnQA4KGHHiIjIwOAyZMnExYWBsC8efPIzs6mpqaGwYMHM3jwYF8NQURERESkxfjT/NjiCoAH7Zt5XZiuz/NSHkbKw0tZGCkPL2VhpDyMlIeXsjDS3K95emOwiIiIiEiAURMgIiIiIhJg1ASIiIiIiAQYNQEiIiIiIgFGTYCIiIiISIBREyAiIiIiEmDUBIiIiIiIBBg1ASIiIiIiAUZNgIiIiIhIgPFpE1BcXExKSgpJSUnk5eU1Wb5mzRr69+9PWloaaWlprF69GoD33nvP811aWhpxcXFs3rwZgNmzZ2O32z3LSktLfTkEEREREZEW4y/z4+CWHZaX0+lkwYIFLF26FJvNRkZGBna7naioKMN6w4cPJycnx/Bd//79yc/PB+DEiRMkJyczcOBAz/JZs2YxbNgwX5UuIiIiItLi/Gl+7LMzASUlJXTt2pXIyEisViupqakUFhZ+4/1s3LiRW265hTZt2vigShERERGRi8Of5sc+OxPgcDiIiIjwfLbZbJSUlDRZb9OmTezcuZPu3buTnZ1N586dDcsLCgr46U9/avju2Wef5YUXXmDAgAHMnDkTq9V63lpqa2tNuWyopqZGlyv9L8rDSHl4KQsj5eGlLIyUh5Hy8FIWRv6ahz/Nj33WBFyIhIQERowYgdVqZeXKlTz22GMsW7bMs7yyspJDhw4xaNAgz3czZszgqquuor6+nrlz55KXl8eUKVPOe5yQkBBiYmJ8No6vU1paaspx/ZXyMFIeXsrCSHl4KQsj5WGkPLyUhZFZebRE43Gx5sc+uxzIZrNRUVHh+exwOLDZbIZ1OnTo4OlSMjMz+fDDDw3LN2zYQFJSEq1bt/Z816lTJywWC1arlbFjx7Jv3z5fDUFEREREpMX40/zYZ01AXFwcZWVllJeXU1dXR0FBAXa73bBOZWWl5+eioiJ69uxpWF5QUEBqauo5t3G5XGzevJno6GgfjUBEREREpOX40/zYZ5cDBQcHk5OTQ1ZWFk6nk/T0dKKjo1myZAmxsbEkJiayfPlyioqKCAoKIjQ0lNzcXM/2R44c4YsvvuCmm24y7HfmzJkcP34cl8tF7969mT9/vq+GICIiIiLSYvxpfmxxuVyuFh+hnzHzujBdn+elPIyUh5eyMFIeXsrCSHkYKQ8vZWGkuV/z9MZgEREREZEAoyZARERERCTAqAkQEREREQkwagJERERERAKMmgARERERkQCjJkBEREREJMCoCRARERERCTBqAkREREREAoyaABERERGRAOPTJqC4uJiUlBSSkpLIy8trsnzNmjX079+ftLQ00tLSWL16tWdZTEyM5/uf/exnnu/Ly8vJzMwkKSmJadOmUVdX58shiIiIiIi0GH+ZHwe3zHCacjqdLFiwgKVLl2Kz2cjIyMButxMVFWVYb/jw4eTk5DTZ/rLLLiM/P7/J94sXL2bChAmkpqaSk5PDa6+9xk9+8hNfDUNEREREpEX40/zYZ2cCSkpK6Nq1K5GRkVitVlJTUyksLPxO+3S5XLz33nukpKQAMGbMmO+8TxERERGRi8Gf5sc+OxPgcDiIiIjwfLbZbJSUlDRZb9OmTezcuZPu3buTnZ1N586dAaitrWXs2LEEBwczadIkhg4dyvHjx2nfvj3Bwe6yIyIicDgczdZSW1tLaWlpC43swtXU1JhyXH+lPIyUh5eyMFIeXsrCSHkYKQ8vZWHkr3n40/zYZ03AhUhISGDEiBFYrVZWrlzJY489xrJlywB46623sNlslJeXc++999KrVy/atm37rY4TEhJCTExMS5Z+QUpLS005rr9SHkbKw0tZGCkPL2VhpDyMlIeXsjAyK4+WaDwu1vzYZ5cD2Ww2KioqPJ8dDgc2m82wTocOHbBarQBkZmby4YcfGrYHiIyM5KabbuLAgQN06NCBU6dO0dDQAEBFRUWTfYqIiIiI+CN/mh/7rAmIi4ujrKyM8vJy6urqKCgowG63G9aprKz0/FxUVETPnj0BOHnypOeu5qqqKvbs2UNUVBQWi4Uf//jHbNy4EYC1a9c22aeIiIiIiD/yp/mxzy4HCg4OJicnh6ysLJxOJ+np6URHR7NkyRJiY2NJTExk+fLlFBUVERQURGhoKLm5uQAcPnyYefPmYbFYcLlcTJw40XPX9KOPPsr06dP59a9/TUxMDJmZmb4agoiIiIhIi/Gn+bHF5XK5fDpaP2DmdWG6Ps9LeRgpDy9lYaQ8vJSFkfIwUh5eysJIc7/m6Y3BIiIiIiIBRk2AiIiIiEiAURMgIiIiIhJg1ASIiIiIiAQYNQEiIiIiIgFGTYCIiIiISIBREyAiIiIiEmDUBIiIiIiIBBg1ASIiIiIiAcanTUBxcTEpKSkkJSWRl5fXZPmaNWvo378/aWlppKWlsXr1asD9trU77riD1NRURo4cyfr16z3bzJ49G7vd7tmmtLTUl0MQEREREWkx/jI/Dm65IRk5nU4WLFjA0qVLsdlsZGRkYLfbiYqKMqw3fPhwcnJyDN9ddtllLFq0iG7duuFwOEhPT2fQoEG0b98egFmzZjFs2DBflS4iIiIi0uL8aX7sszMBJSUldO3alcjISKxWK6mpqRQWFl7Qtt27d6dbt24A2Gw2wsPDqaqq8lWpIiIiIiI+50/zY5+dCXA4HERERHg+22w2SkpKmqy3adMmdu7cSffu3cnOzqZz586G5SUlJdTX13PNNdd4vnv22Wd54YUXGDBgADNnzsRqtZ63ltraWlMuG6qpqdHlSv+L8jBSHl7Kwkh5eCkLI+VhpDy8lIWRv+bhT/NjnzUBFyIhIYERI0ZgtVpZuXIljz32GMuWLfMsr6ys5NFHH2XRokW0auU+aTFjxgyuuuoq6uvrmTt3Lnl5eUyZMuW8xwkJCSEmJsanYzmX0tJSU47rr5SHkfLwUhZGysNLWRgpDyPl4aUsjMzKoyUaj4s1P/bZ5UA2m42KigrPZ4fDgc1mM6zToUMHT5eSmZnJhx9+6FlWXV3NAw88wPTp04mPj/d836lTJywWC1arlbFjx7Jv3z5fDUFEREREpMX40/zYZ01AXFwcZWVllJeXU1dXR0FBAXa73bBOZWWl5+eioiJ69uwJQF1dHZMnTyYtLa3JDQ5fbeNyudi8eTPR0dG+GoKIiIiISIvxp/mxzy4HCg4OJicnh6ysLJxOJ+np6URHR7NkyRJiY2NJTExk+fLlFBUVERQURGhoKLm5uQBs2LCBXbt2ceLECdauXQvAM888Q0xMDDNnzuT48eO4XC569+7N/PnzfTUEEREREZEW40/zY4vL5XL5dLR+wMzrwnR9npfyMFIeXsrCSHl4KQsj5WGkPLyUhZHmfs3TG4NFRERERAKMmgARERERkQCjJkBEREREJMCoCRARERERCTBqAkREREREAoyaABERERGRAKMmQEREREQkwKgJEBEREREJMGoCREREREQCjE+bgOLiYlJSUkhKSiIvL6/J8jVr1tC/f3/S0tJIS0tj9erVnmVr164lOTmZ5ORkz6uRAfbv38/IkSNJSkriySefJABeeCwiIiIilwh/mR8Ht8xwmnI6nSxYsIClS5dis9nIyMjAbrcTFRVlWG/48OHk5OQYvjtx4gTPP/88r7/+OhaLhbFjx2K32wkNDeWJJ55g4cKF9OvXj4kTJ1JcXMyQIUN8NQwRERERkRbhT/Njn50JKCkpoWvXrkRGRmK1WklNTaWwsPCCtt22bRsDBw4kLCyM0NBQBg4cyNatW6msrKS6upr4+HgsFgujR4++4H2KiIiIiJjJn+bHPmsCHA4HERERns82mw2Hw9FkvU2bNjFy5EimTp3KF198cd5t/+/3ERER59yniIiIiIi/8af5sc8uB7oQCQkJjBgxAqvVysqVK3nsscdYtmxZix+ntraW0tLSFt9vc2pqakw5rr9SHkbKw0tZGCkPL2VhpDyMlIeXsjD6PudxsebHPmsCbDYbFRUVns8OhwObzWZYp0OHDp6fMzMz+e///m/Ptjt27DBse9NNNzXZZ0VFRZN9nktISAgxMTHfeizfVmlpqSnH9VfKw0h5eCkLI+XhpSyMlIeR8vBSFkZm5dFc4+FP82OfXQ4UFxdHWVkZ5eXl1NXVUVBQgN1uN6xTWVnp+bmoqIiePXsCMGjQILZt28bJkyc5efIk27ZtY9CgQXTq1Im2bduyd+9eXC4Xb7zxBomJib4agoiIiIhIi/Gn+bHPzgQEBweTk5NDVlYWTqeT9PR0oqOjWbJkCbGxsSQmJrJ8+XKKiooICgoiNDSU3NxcAMLCwnjooYfIyMgAYPLkyYSFhQEwb948srOzqampYfDgwQwePNhXQxARERERaTH+ND+2uALgQftmnhLSqTkv5WGkPLyUhZHy8FIWRsrDSHl4KQsjzf2apzcGi4iIiIgEGDUBIiIiIiIBRk2AiIiIiEiAURMgIiIiIhJg1ASIiIiIiAQYNQEiIiIiIgFGTYCIiIiISIBREyAiIiIiEmDUBIiIiIiIBBg1ASIiIiIiAcanTUBxcTEpKSkkJSWRl5f3tett3LiRH/7wh+zbtw+AN998k7S0NM//evfuTWlpKQDjxo0jJSXFs+zYsWO+HIKIiIiISIvxl/lxcMsMpymn08mCBQtYunQpNpuNjIwM7HY7UVFRhvWqq6tZtmwZ/fr183w3atQoRo0aBcDBgweZPHkyMTExnuWLFy8mLi7OV6WLiIiIiLQ4f5of++xMQElJCV27diUyMhKr1UpqaiqFhYVN1luyZAkTJ04kJCTknPspKCggNTXVV2WKiIiIiFwU/jQ/9tmZAIfDQUREhOezzWajpKTEy1OCJAAAEtpJREFUsM6HH35IRUUFt956Ky+99NI597N+/Xp++9vfGr57/PHHadWqFcnJyTz00ENYLJbz1lJbW+s5XXIx1dTUmHJcf6U8jJSHl7IwUh5eysJIeRgpDy9lYeSvefjT/NhnTUBzGhsbeeaZZ8jNzf3adT744APatGlDr169PN8tXrwYm81GdXU1U6dOJT8/n9GjR5/3WCEhIYbTJRdLaWmpKcf1V8rDSHl4KQsj5eGlLIyUh5Hy8FIWRmbl8V0bj4s5P/bZ5UA2m42KigrPZ4fDgc1m83w+ffo0hw4dYvz48djtdvbu3cuDDz7oufkBzn2q46t9tG3blhEjRjTpnkRERERE/JE/zY991gTExcVRVlZGeXk5dXV1FBQUYLfbPcvbtWvH9u3bKSoqoqioiPj4eH73u995bmhobGxkw4YNhkE2NDRQVVUFQH19PVu2bCE6OtpXQxARERERaTH+ND/22eVAwcHB5OTkkJWVhdPpJD09nejoaJYsWUJsbCyJiYnn3X7nzp107tyZyMhIz3d1dXVkZWVRX19PY2MjAwYM4Pbbb/fVEEREREREWow/zY8tLpfL9Z1H5OfMvC5M1+d5KQ8j5eGlLIyUh5eyMFIeRsrDS1kYae7XPL0xWEREREQkwKgJEBEREREJMGoCREREREQCjJoAEREREZEAoyZARERERCTAqAkQEREREQkwagJERERERAKMmgARERERkQCjJkBEREREJMD4tAkoLi4mJSWFpKQk8vLyvna9jRs38sMf/pB9+/YBcOTIEfr27UtaWhppaWnk5OR41t2/fz8jR44kKSmJJ598kgB44bGIiIiIXCL8ZX4c/N2Hcm5Op5MFCxawdOlSbDYbGRkZ2O12oqKiDOtVV1ezbNky+vXrZ/j+mmuuIT8/v8l+n3jiCRYuXEi/fv2YOHEixcXFDBkyxFfDEBERERFpEf40P/bZmYCSkhK6du1KZGQkVquV1NRUCgsLm6y3ZMkSJk6cSEhISLP7rKyspLq6mvj4eCwWC6NHjz7nPkVERERE/I0/zY99dibA4XAQERHh+Wyz2SgpKTGs8+GHH1JRUcGtt97KSy+9ZFh25MgRRo8eTdu2bZk2bRo33HBDk31GRETgcDiaraW2tpbS0tLvOKJvx6zj+ivlYaQ8vJSFkfLwUhZGysNIeXgpCyMz8qitrT3vcn+aH/usCWhOY2MjzzzzDLm5uU2WderUibfeeosOHTqwf/9+Jk+eTEFBwbc+Vnx8/HcpVURERETE5y7m/NhnTYDNZqOiosLz2eFwYLPZPJ9Pnz7NoUOHGD9+PAD/+te/ePDBB/nd735HXFwcVqsVgNjYWK655ho+/fTTJvusqKgw7FNERERExF/50/zYZ/cExMXFUVZWRnl5OXV1dRQUFGC32z3L27Vrx/bt2ykqKqKoqIj4+HjPAKuqqnA6nQCUl5dTVlZGZGQknTp1om3btuzduxeXy8Ubb7xBYmKir4YgIiIiItJi/Gl+7LMzAcHBweTk5JCVlYXT6SQ9PZ3o6GiWLFlCbGzseYvbuXMnv/nNbwgODqZVq1bMnz+fsLAwAObNm0d2djY1NTUMHjyYwYMH+2oIIiIiIiItxp/mxxaXHrQvIiIiIhJQ9MZgEREREZEAoyZARERERCTAmPaI0EtdcXExTz31FI2NjWRmZjJp0iSzSzJNdnY2W7ZsoWPHjvz1r381uxxTffHFF8yaNYtjx45hsVi4/fbbuffee80uyzS1tbXcfffd1NXV4XQ6SUlJYerUqWaXZaqvrhG12Wz8/ve/N7scU9ntdq644gpatWpFUFAQa9asMbskU506dYpf/OIXHDp0CIvFwtNPP811111ndlkX3SeffML06dM9n8vLy5k6dSoTJkwwryiT/elPf2L16tVYLBZ69epFbm7uBb1k6lL0yiuvsHr1alwuF5mZmQH9e9Esl7S4hoYGV2Jiouuf//ynq7a21jVy5EjXRx99ZHZZptmxY4dr//79rtTUVLNLMZ3D4XDt37/f5XK5XF9++aUrOTk5oH83GhsbXdXV1S6Xy+Wqq6tzZWRkuN5//32TqzLXyy+/7JoxY4Zr0qRJZpdiuoSEBNexY8fMLsNvzJo1y7Vq1SqXy+Vy1dbWuk6ePGlyReZraGhw3Xzzza4jR46YXYppKioqXAkJCa6zZ8+6XC6Xa+rUqa7XX3/d5KrMcfDgQVdqaqrrzJkzrvr6ete9997rKisrM7ssv6XLgXzgQl8JHShuvPFGQkNDzS7DL3Tq1Ik+ffoA0LZtW3r06HFBb/W7VFksFq644goAGhoaaGhowGKxmFyVeSoqKtiyZQsZGRlmlyJ+5ssvv2Tnzp2e3w2r1Ur79u1Nrsp87777LpGRkVx99dVml2Iqp9NJTU0NDQ0N1NTU0KlTJ7NLMsXhw4fp27cvbdq0ITg4mBtvvJFNmzaZXZbfUhPgA+d6JXQgT/Tk3I4cOUJpaSn9+vUzuxRTOZ1O0tLSuPnmm7n55psDOo+nn36aRx99lFat9J/mr9x///2MHTuW//mf/zG7FFMdOXKE8PBwsrOzGT16NHPmzOHMmTNml2W6goICRowYYXYZprLZbNx3330kJCQwaNAg2rZty6BBg8wuyxS9evVi9+7dHD9+nLNnz1JcXGx4iZYY6f9pRExw+vRppk6dyuOPP07btm3NLsdUQUFB5Ofn8/e//52SkhIOHTpkdkmmeOuttwgPDyc2NtbsUvzGihUrWLt2LX/4wx949dVX2blzp9klmaahoYEDBw5w11138cYbb9CmTRvy8vLMLstUdXV1FBUVMWzYMLNLMdXJkycpLCyksLCQrVu3cvbsWfLz880uyxQ9e/YkKyuL+++/n6ysLHr37q0/qpyHkvGB5l4JLYGtvr6eqVOnMnLkSJKTk80ux2+0b9+eH//4x2zdutXsUkyxZ88eioqKsNvtzJgxg/fee4+ZM2eaXZapvvrvZseOHUlKSqKkpMTkiswTERFBRESE50zZsGHDOHDggMlVmau4uJg+ffpw5ZVXml2Kqd555x26dOlCeHg4rVu3Jjk5mffff9/sskyTmZnJmjVrePXVVwkNDaVbt25ml+S31AT4QHOvhJbA5XK5mDNnDj169OCnP/2p2eWYrqqqilOnTgFQU1PDO++8Q48ePUyuyhyPPPIIxcXFFBUV8atf/Yr+/fuzePFis8syzZkzZ6iurvb8/PbbbxMdHW1yVea56qqriIiI4JNPPgHc18L37NnT5KrMVVBQQGpqqtllmO4HP/gBH3zwAWfPnsXlcgX878axY8cA+Pzzz9m0aRMjR440uSL/pUeE+sDXvRI6UM2YMYMdO3Zw/PhxBg8ezMMPP0xmZqbZZZli9+7d5Ofn06tXL9LS0gB3PkOGDDG5MnNUVlYye/ZsnE4nLpeLYcOGkZCQYHZZ4geOHTvG5MmTAfd9IyNGjGDw4MEmV2WuuXPnMnPmTOrr64mMjCQ3N9fskkxz5swZ3nnnHRYsWGB2Kabr168fKSkpjBkzhuDgYGJiYrjjjjvMLss0Dz/8MCdOnCA4OJh58+bpBvrzsLhcLpfZRYiIiIiIyMWjy4FERERERAKMmgARERERkQCjJkBEREREJMCoCRARERERCTBqAkREREREAoyaABEJSHfeeec3Wn/79u088MADPqqmZe3atYvU1FTS0tKoqan5VvtYsWIFb7zxRpPvjxw5wogRI75rid/Yc889x0svvXTRjwvGf/vt27ezZ88eU+oQEWlJek+AiASklStXml2Cz7z55ptMmjTJ8y6Kb6qhoYG77rqrhau6ME6nk6CgIL/b11d27NjB5ZdfzvXXX9+i+xURudjUBIhIQLruuut4//332b59O88//zwdOnTg0KFD9OnTh8WLF2OxWCguLubpp5+mTZs2/OhHP/Jse+bMGRYuXMhHH31EQ0MDU6ZMYejQoTz55JOEhYUxZcoUtm7dyosvvsjy5ctp1cp70vW5557j888/58iRI3z++efce++9jB8/HoClS5fy+uuvA5CRkcGECRPOO4Z3332XRYsW4XQ6iY2NZf78+eTn5/O3v/2Nbdu2UVxczC9/+UvDNi+88AJvvvkm4eHhdO7cmT59+nD//fczbtw4evfuze7duxkxYgSnT5/m8ssv5/7772f//v08/vjjAAwcONCzr48++ojs7Gzq6+tpbGzkueeeo1u3buTn57N8+XLq6+vp168f8+bNIygoiHnz5rFv3z5qa2tJSUlh6tSpANjtdm677TbeeecdsrKyaNeuHc8++yxOp5MOHTrwyiuvAPDxxx8zbty4Jrn933/XO+64g3feeYecnByOHj3apBaAOXPmsH//fiwWC+np6UyYMIFx48Yxa9Ys4uLiqKqqIiMjg6KiIs++jxw5wsqVK2nVqhVvvvkmc+fO5YYbbjj/L5qIiJ9SEyAiAe/AgQMUFBTQqVMn7rrrLnbv3k1cXBxz587llVdeoWvXrkybNs2z/osvvkj//v3Jzc3l1KlTZGZmcvPNN/PII4+QkZHBDTfcwJNPPskf/vAHQwPwlU8//ZRly5ZRXV3Nbbfdxl133cXBgwdZs2YNq1atwuVycfvtt3PTTTdx7bXXnrPm2tpaZs+ezZ/+9Ce6d+/OrFmz+Mtf/sKECRPYvXs3t956K8OGDTNsU1JSwqZNm3jzzTepr69n7Nix9OnTx7O8vr6eNWvWAO5m5SvZ2dnk5ORw4403smjRIs/3K1euZPz48YwaNYq6ujoaGxs5fPgwGzZsYMWKFbRu3ZonnniCdevWMXr0aKZPn05YWBhOp5MJEybwj3/8g969ewMQFhbG2rVrqaqqYsyYMfz5z38mMjKSEydOnDe31q1bG8Z45swZ+vbty+zZszl8+DB//OMfm9QSFRWFw+Hgr3/9KwCnTp06/y/If3Tp0oU777zT0xyJiHyfqQkQkYDXt29fIiIiAOjduzdHjx7liiuuoEuXLnTr1g2AUaNGsWrVKgC2bdtGUVERL7/8MuCekH/xxRf07NmThQsXcs8995Cdnc0111xzzuMNGTIEq9VKeHg44eHhHDt2jN27dzN06FAuv/xyAJKSkti1a9fXNgGffvopXbp0oXv37gCMGTOGV1999bxnD/bs2UNiYiIhISGEhISQkJBgWD58+PAm25w6dYovv/ySG2+8EYC0tDS2bt0KQHx8PC+++CIVFRUkJyfTrVs33n33Xfbv309GRgYANTU1dOzYEYANGzawatUqGhoa+Ne//sXhw4c9TcBXx967dy833HADkZGRgLs5OF9uX/27fSUoKIiUlBSAr60lISGB8vJyFi5cyJAhQxg0aNDXZiYicqlSEyAiAc9qtXp+DgoKwul0NrvNb37zG3r06NHk+0OHDhEWFkZlZeUFH6+hoeEbVuwbbdq0+Ubrjxw5kn79+rFlyxYmTZrE/PnzcblcjBkzhkceecSwbnl5OS+//DKvvfYaoaGhzJ49m9ra2m907AvJLSQkxHMfwNfVApCfn8+2bdtYuXIlGzZsIDc3l6CgIFwuFwB1dXUXFoKIyPeUng4kInIOPXr04OjRo/zzn/8EoKCgwLNs0KBB/PnPf/ZMGA8cOADA0aNHWbp0KWvXrqW4uJgPPvjggo93ww03sHnzZs6ePcuZM2fYvHnzea837969O0ePHuWzzz4D3JPar/5a/3Wuv/563nrrLWprazl9+jRbtmxptq727dvTrl07du3aBcC6des8y8rLy4mMjGT8+PEkJiZy8OBBBgwYwMaNGzl27BgAJ06c4OjRo5w+fZo2bdrQrl07/v3vf1NcXHzO48XHx7Nr1y7Ky8s9239bX1dLVVUVLpeLlJQUpk2b5vn3u/rqq9m/fz8Af/vb3865zyuuuILTp09/65pERPyFzgSIiJxDSEgICxYsYNKkSZ4bg7+a/D300EM8/fTTjBo1isbGRrp06cKLL77InDlzmDVrFjabjaeeeors7Gxee+01QkJCmj1enz59GDt2LJmZmYD7xuCvLgVKS0sjPz+/SX25ubn8/Oc/99wY3NwTffr27YvdbmfUqFF07NiRXr160a5du2Zry83N5fHHH8disRhuDN6wYQP5+fkEBwdz5ZVX8sADDxAWFsa0adO47777aGxspHXr1uTk5BAfH8+1117LbbfdRkRExNc+XSc8PJwFCxbw8MMP09jYSMeOHVm6dGmzNZ5LVFTUOWu57LLLyM7OprGxEYAZM2YAcN999zFt2jRWrVrFkCFDzrnPhIQEpk6dSmFhoW4MFpHvNYvrqz9liYjIJe/06dNcccUVnD17lrvvvpuFCxcabg4WEZHAoDMBIiIBJCcnh48//pja2lrGjBmjBkBEJEDpTICIiIiISIDRjcEiIiIiIgFGTYCIiIiISIBREyAiIiIiEmDUBIiIiIiIBBg1ASIiIiIiAeb/Azqj6IAIKWu9AAAAAElFTkSuQmCC\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Observation:**\n",
        "\n",
        "Index no. 2 seems reasonable. getting hyperprmeter value below."
      ],
      "metadata": {
        "id": "i5L5SqzgsRal"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "cv_set_1.iloc[2,[4,5,10,16]]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "94XGrmmTsYqj",
        "outputId": "25479e60-4a50-4e39-d4e1-8ec9207e8ca3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "param_columntransformer__countvectorizer__max_features        2500\n",
              "param_multinomialnb__alpha                                       5\n",
              "mean_test_score                                           0.692022\n",
              "mean_train_score                                          0.719535\n",
              "Name: 2, dtype: object"
            ]
          },
          "metadata": {},
          "execution_count": 215
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Observation:**\n",
        "\n",
        "We get,\n",
        "\n",
        "max_features = 2500\n",
        "\n",
        "alpha = 5\n",
        "\n",
        "<h3>Training on whole X_train.</h3>"
      ],
      "metadata": {
        "id": "nJmJK1W7sfXj"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "param={'multinomialnb__alpha':5,\n",
        "   'multinomialnb__class_prior':None,\n",
        "   'columntransformer__tfidfvectorizer__max_features':2500}\n",
        "\n",
        "pipe_set2.set_params(**param)\n",
        "pipe_set2.fit(X_train,y_train)\n",
        "y_predict_test = pipe_set2.predict_proba(X_test)[:,1]\n",
        "y_predict__train = pipe_set2.predict_proba(X_train)[:,1]\n",
        "auc_score_set2_test = roc_auc_score(y_test, y_predict_test )\n",
        "auc_score_set2_train = roc_auc_score(y_train, y_predict_train )\n",
        "print('Test AUC Score:', auc_score_set2_test )\n",
        "print('Train AUC Score:', auc_score_set2_train)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "f249NAxbzWRi",
        "outputId": "4cf53e1a-e40e-4ed0-b94c-6f429dbaa3e9"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Test AUC Score: 0.6634126859462641\n",
            "Train AUC Score: 0.7128906807027795\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "confusion_matrix_train_set2 = confusion_matrix(y_train,pipe_set2.predict(X_train),normalize='true')\n",
        "confusion_matrix_train_set2"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pnb7CyrS6MUx",
        "outputId": "e0c1f340-ff4a-4987-d084-f42291bd674e"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.02905351, 0.97094649],\n",
              "       [0.00753465, 0.99246535]])"
            ]
          },
          "metadata": {},
          "execution_count": 167
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "confusion_matrix_test_set2 = confusion_matrix(y_test,pipe_set2.predict(X_test),normalize='true')\n",
        "confusion_matrix_test_set2"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ffQCmrWrHnSW",
        "outputId": "d32a7526-0a60-4d97-b1c1-3c138dfefa2b"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.01868474, 0.98131526],\n",
              "       [0.00921779, 0.99078221]])"
            ]
          },
          "metadata": {},
          "execution_count": 168
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "<h3>Trying class prior = [50,50]</h3>"
      ],
      "metadata": {
        "id": "ky2JBBBaslHa"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "param={'multinomialnb__alpha':5,\n",
        "   'multinomialnb__class_prior':[50,50],\n",
        "   'columntransformer__tfidfvectorizer__max_features':2500}\n",
        "\n",
        "pipe_set2.set_params(**param)\n",
        "pipe_set2.fit(X_train,y_train)\n",
        "y_predict_test = pipe_set2.predict_proba(X_test)[:,1]\n",
        "y_predict_train = pipe_set2.predict_proba(X_train)[:,1]\n",
        "auc_score_set2_test = roc_auc_score(y_test, y_predict_test )\n",
        "auc_score_set2_train = roc_auc_score(y_train, y_predict_train )\n",
        "print('Test AUC Score:', auc_score_set2_test )\n",
        "print('Train AUC Score:', auc_score_set2_train)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KtQ2G7bXsrir",
        "outputId": "96acc65f-5d81-48cd-f602-64c981701dc7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Test AUC Score: 0.6634126859462641\n",
            "Train AUC Score: 0.6738478258386486\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "confusion_matrix_train_set2 = confusion_matrix(y_train,pipe_set2.predict(X_train),normalize='true')\n",
        "confusion_matrix_train_set2 "
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aEOqqyQh6VVy",
        "outputId": "d4e9c41a-f080-478f-8d58-84f5230f7716"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.55950555, 0.44049445],\n",
              "       [0.31536071, 0.68463929]])"
            ]
          },
          "metadata": {},
          "execution_count": 218
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "confusion_matrix_test_set2 = confusion_matrix(y_test,pipe_set2.predict(X_test),normalize='true')\n",
        "confusion_matrix_test_set2 "
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LE9ya1lbIkxw",
        "outputId": "88a4bd04-0045-482c-d19c-f7df5ae12079"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.54680344, 0.45319656],\n",
              "       [0.31595463, 0.68404537]])"
            ]
          },
          "metadata": {},
          "execution_count": 219
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Observation:**\n",
        "\n",
        "- By using class prior of [50,50] we get better fit on data(close train & test auc score) and a significant improvement in overall confusion matrix statistics.\n",
        "\n",
        "- So, we continue using class prior of [50,50].\n",
        "\n",
        "- Now we find best threshold using max value of TPR(1-FPR).\n"
      ],
      "metadata": {
        "id": "js6KxiNis448"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "<h3>Finding optimal probability threshold</h3>"
      ],
      "metadata": {
        "id": "4kfjJX10koLQ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "fpr_train, tpr_train, thresholds = roc_curve( y_train,  y_predict_train )"
      ],
      "metadata": {
        "id": "YERlLqZCs_aC"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "fpr_test, tpr_test ,threshold_test = roc_curve( y_test ,  y_predict_test )"
      ],
      "metadata": {
        "id": "yVjpsGbsEHKn"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "sns.lineplot(fpr_train , tpr_train , label='Trian ROC Curve')\n",
        "sns.lineplot(fpr_test , tpr_test , label='Test ROC Curve')\n",
        "plt.title('Train and Test ROC Curve')\n",
        "plt.xlabel('False Positive Rate')\n",
        "plt.ylabel('True Positive Rate')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "d_h3kDnJtgAT",
        "outputId": "3b52d1ad-5673-4b8e-c07a-c2148b34b823"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Train AUC Score: 0.67\n",
        "\n",
        "Test AUC Score: 0.66"
      ],
      "metadata": {
        "id": "esd07XaOqNrV"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "threshold =  thresholds[np.argmax(tpr_train*(1-fpr_train))]\n",
        "\n",
        "y_train_after_thresholding = np.copy(y_predict_train)\n",
        "y_train_after_thresholding = np.where( y_train_after_thresholding > threshold , 1, 0)\n",
        "confusion_matrix_train_set2_thresh = confusion_matrix(y_train,y_train_after_thresholding,normalize='true')\n",
        "print('Train Confusion Matrix after thresholding probability:\\n',confusion_matrix_train_set2_thresh)\n",
        "\n",
        "y_test_after_thresholding = np.copy(y_predict_test)\n",
        "y_test_after_thresholding = np.where( y_test_after_thresholding > threshold , 1, 0)\n",
        "confusion_matrix_test_set2_thresh = confusion_matrix(y_test,y_test_after_thresholding,normalize='true')\n",
        "print('\\nTest Confusion Matrix after thresholding probability:\\n',confusion_matrix_test_set2_thresh)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qRzVeHHjtKHb",
        "outputId": "80a04319-4239-4014-bcb7-6c2d3adc4dfb"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Confusion Matrix after thresholding probability:\n",
            " [[0.62690607 0.37309393]\n",
            " [0.37243411 0.62756589]]\n",
            "\n",
            "Test Confusion Matrix after thresholding probability:\n",
            " [[0.61732918 0.38267082]\n",
            " [0.37387638 0.62612362]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Observation:**\n",
        "\n",
        "- We see some change in proportion of TP & FP after probability thresholding.\n",
        "- So, we continue with new threshold.\n"
      ],
      "metadata": {
        "id": "5oDisaditS0_"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "<h3>Confusion Matrix(normalized) after threshold</h3>"
      ],
      "metadata": {
        "id": "gqtPGNHJl8R0"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "sns.set_style('white')\n",
        "ConfusionMatrixDisplay.from_predictions(y_train, y_train_after_thresholding, normalize='true')\n",
        "plt.title('Confusion Matrix-Train set')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "hlgO0RNCtoQQ",
        "outputId": "5e0a772c-e99b-493a-ea5d-768c6d02b248"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "ConfusionMatrixDisplay.from_predictions(y_test, y_test_after_thresholding, normalize='true') \n",
        "plt.title('Confusion Matrix-Test set')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "Zx4q2HHQuMrk",
        "outputId": "87b80afc-654d-444b-c2e9-333ddd2b8cfe"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAATwAAAEWCAYAAAD7MitWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3de1hU1f4/8PfAMIiXYNAc8IioiZcURUXTEJFRUUFFEQ71TT3l8Y51Kk+WleQlLz/T1LQyDh5Ks8yTmiKWljfUviKmfjG7kAaKIqNySxAYmFm/PzhMEZeZUYaZYb9fz7Ofhz177bU/e5CPa6+1194yIYQAEZEEOFg7ACKixsKER0SSwYRHRJLBhEdEksGER0SSwYRHRJLBhNfISktLMXv2bPTv3x/PPffcfdezb98+TJs2rQEjs47p06djz5491g6DJELG+/Bql5iYiISEBGRkZKBFixbo3r07Zs+eDX9//weq94svvsDHH3+MHTt2QC6XN1C0DSclJQVTp07FiBEj8O677xo+/+mnnxAeHo6BAwdi27ZtRuvZuHEjrl69ijVr1jRofNnZ2QgLCzOs37t3Dy4uLpDJZACAf/3rX2b/jtRqNd588008/vjjDRqrpb4Dun+29xdnAxISEhAXF4clS5ZgyJAhcHJywokTJ3D48OEHTnjZ2dno2LGjTSa7Ku7u7rhw4QLy8/OhVCoBAHv27EHHjh0b7BhCCAgh4OBg3kVGu3btcP78ecN6t27dsHfvXnh7ezdYbNSECarmt99+E35+fuLAgQN1likrKxNvvvmmCAgIEAEBAeLNN98UZWVlQgghTp8+LQIDA8WWLVvEoEGDREBAgPj888+FEEJs2LBB9OzZUzz66KPCz89P7Ny5U7zzzjti/vz5hrqzsrJE165dRXl5uRBCiF27dgm1Wi38/PxEcHCw2Lt3r+HzJ554wrDfd999JyIiIkS/fv1ERESE+O677wzbJk+eLNatWyeio6OFn5+feOaZZ0Rubm6t51YV/6JFi8THH38shBCioqJCDBkyRGzcuFFMnjzZUHbZsmVi6NChom/fvmLixIkiNTVVCCHE8ePHq53nuHHjDHG8/fbbIjo6Wvj6+orMzEwxefJksXPnTiGEELGxsWLevHmG+levXi2mTp0q9Hp9nb+Lrl27iszMTMPvZdWqVSIoKEgMHjxYLFq0SJSUlAghhMjNzRUzZ84U/fv3FwMGDBBPPvmk0Ol04p///Kfo1q2b8PX1FX5+fiIuLq7GMeraVwghcnJyxLx588Rjjz0mgoODxUcffVTvd0DWxYT3J8ePHxc9evQwJJzarF+/XkRFRYk7d+6I3NxcER0dLdatWyeEqEwYPXr0EOvXrxdarVYcO3ZM9O7dWxQUFAghRI0EV1/CKy4uFn379hVXrlwRQgih0WhEenq6EKJ6wsvPzxf+/v5iz549ory8XCQmJgp/f3+Rl5cnhKhMNMOHDxe//vqrKCkpEZMnTxZvvfVWredWlfC+++47ERkZKYQQ4tixY2LatGli586d1RLeF198IfLy8kR5ebnYsmWLePzxx0VpaWmt51UVR1BQkEhPTxfl5eVCq9VWS3j37t0TISEhYteuXSI1NVUMHDhQ3Lx5s97f1x8T3vLly8WsWbNEfn6+uHv3rpg1a5ZYs2aNEEKINWvWiEWLFgmtViu0Wq1ITU01JNLg4GBx6tSpOo9R1746nU5MnDhRbNy4UZSVlYlr164JtVotkpOT6/wOyLo4aPEnBQUFUCqV9V5yJiYmIiYmBq1bt4a7uztiYmKwb98+w3a5XI6YmBg4OTkhKCgIzZs3R0ZGxn3F4+DggF9++QWlpaVo27YtfHx8apQ5duwYvL29MWHCBMjlcowdOxadO3fG0aNHDWUiIiLQqVMnNGvWDKNHj8aPP/5Y73H79euHwsJC/Prrr/jiiy8QHh5eo0x4eLjhu5o2bRq0Wq3R85w4cSJ8fHwgl8vh5ORUbZuLiwtWr16NVatW4aWXXsKiRYvg4eFRb31VhBDYuXMnXn31Vbi5uaFly5aYNWsWkpKSAFT+Tm7fvo3s7Gw4OTnB39/f0O9nTF37Xrx4EXl5eZg3bx4UCgW8vLzw17/+FQcOHDCpXmp8THh/4ubmhvz8fFRUVNRZ5tatW2jXrp1hvV27drh161a1Ov6YMF1cXHDv3j2zY2nevDnWrVuHHTt2YMiQIZg5cyauXLliNJ6qmDQajWH94YcfNjue8ePHY/v27UhJScHIkSNrbN+yZQvGjBmD/v37w9/fH3fv3kV+fn69dXp6eta7vU+fPmjfvj2EEBgzZozh8+nTp6Nv377o27dvtf9cquTl5aGkpAQRERHw9/eHv78/pk+fbojn73//O7y9vTFt2jQMHz4ccXFxRs+/Sl373rhxA7du3TIcz9/fH5s3b8adO3dMrpsal+32nFtJ3759oVAo8M0332D06NG1lmnbti2ys7MNra2bN2+ibdu293U8FxcXlJaWGtb//McSGBiIwMBAlJaWYv369Vi0aBE++eSTWuP5o5s3byIwMPC+YqoSHh6OkJAQTJgwAS4uLtW2nT17FvHx8fjwww/h4+MDBwcHDBgwAOK/g/51tZ6Mtaq2b9+O8vJytG3bFvHx8Zg1axYAID4+vt79lEolmjVrhqSkJKhUqhrbW7ZsiVdeeQWvvPIK0tPT8be//Q2+vr4YPHhwvfXWt6+npyfat2+PQ4cO3de5UuNjC+9PWrVqheeeew5Lly7FN998g5KSEpSXl+P48eNYvXo1ACAsLAzvv/8+8vLykJeXh3fffRfjxo27r+P16NEDqampyM7Oxt27d/HBBx8Ytt25cwfffPMN7t27B4VCgebNm9c6qhkUFITMzEwkJiaioqICBw4cwOXLlzFs2LD7iqmKl5cXtm3bhueff77GtuLiYjg6OsLd3R0VFRXYtGkTioqKDNtbt26NGzduQK/Xm3y8jIwMrF+/Hm+99RZWr16N+Ph4o5feVRwcHBAVFYUVK1YgNzcXAKDRaHDixAkAwNGjR3H16lUIIdCqVSs4OjoaElKbNm2QlZVVZ9117du7d2+0aNECcXFxKC0thU6nQ3p6OtLS0u77OyDLYsKrxbRp0/DKK6/gvffew+DBgzFs2DBs374dI0aMAADMnTsXvXr1wvjx4zF+/Hj07NkTc+fOva9jBQQEIDQ0FOPHj0dERASCg4MN2/R6PT788EMEBgZi4MCBSE1NxeLFi2vUoVQqsXnzZiQkJOCxxx5DfHw8Nm/eDHd39/uK6Y/8/f1rbTENGTIEgYGBGDVqFNRqNZydnatdrla1jh977DFMnDjR6HEqKirw0ksvYcaMGejevTs6duyIF154AQsWLIBWqzUp1pdeegne3t7461//in79+uHpp5829ClevXoVzzzzDPr27Yvo6Gg8+eSTGDRoEABg5syZeP/99+Hv748tW7bUqLeufR0dHbF582b89NNPGD58OAYNGoTXX3/dkPjN/Q7I8njjMRFJBlt4RCQZTHhEJBlMeEQkGUx4RCQZNnUf3sABPvDw4BiKPcn4/iFrh0BmqHAsw4+XLz1QHX+f0hX5haaVVT4cWOvIt7XYVMLz8BCI32z+jASynindh1g7BDJDdpu0B64jvxD43MSJKpFz659509hsKuERkX3Qwz5vpmbCIyKzCCFQLnTWDuO+MOERkVkE2MIjIgnR2ekELSY8IjKLgIAeTHhEJBE6JjwikoLKPjwmPCKSAAGgnH14RCQVvKQlIkkQAHT2me+Y8IjIfPZ5Fx4THhGZSQDQwT5fUMSER0RmqRy0YMIjIglgC4+IpEMAerbwiEgKBGRs4RGRdOjs9O0QTHhEZBYB+72ktc80TURWIyCDVjiatBiTnJyMUaNGYeTIkYiLq/258QcOHEBoaCjCwsIwf/58w+d79uxBSEgIQkJCsGfPHpNiZwuPiMymb4C2kk6nw9KlS5GQkACVSoXIyEio1Wp06dLFUCYzMxNxcXH49NNP4erqitzcXABAQUEBNm3ahF27dkEmkyEiIgJqtRqurq71HpMtPCIyS9WghSlLfdLS0uDt7Q0vLy8oFAqEhYXh8OHD1crs3LkTTz31lCGRtW7dGgBw8uRJBAQEwM3NDa6urggICMCJEyeMxs4WHhGZpXIurWltpfz8fERERBjWo6OjER0dDQDQaDTw8PAwbFOpVEhLq/5WtczMTADAE088Ab1ej3nz5mHo0KG17qvRaIzGw4RHRGbTm3hbilKpxO7du+/7ODqdDlevXsW2bduQk5ODyZMnIzEx8b7r4yUtEZmlctBCbtJSH5VKhZycHMO6RqOBSqWqUUatVsPJyQleXl7o2LEjMjMzTdq3Nkx4RGSWyiceO5i01MfX1xeZmZnIysqCVqtFUlIS1Gp1tTIjRozAmTNnAAB5eXnIzMyEl5cXhgwZgpMnT6KwsBCFhYU4efIkhgwx/lJ4XtISkXmEDLoGuA9PLpcjNjYW06dPh06nw6RJk+Dj44MNGzagV69eGD58OAIDA3Hq1CmEhobC0dERCxYsgFKpBADMnTsXkZGRAICYmBi4ubkZPaZMCNt5VvP4cV0Qv/metcMgM0zpHmLtEMgM2W3ScDHj3APVERzWA3PfaW5S2U9f8n6gPryGxhYeEZlNb+Iora1hwiMis1Teh8eER0QSUPkAUOPTxmwREx4RmUlm8o3HtoYJj4jMUnlbin0+LYUJj4jMxhYeEUkCBy2ISDr4TgsikgoBGcqNzJO1VfYZNRFZDV/TSESSwpkWRCQJfE0jEUkKW3hEJAmVgxacWkZEEiAEbzwmIgnhfXhEJAnmzbTQWzQWczHhEZFZBNjCIyIJMfaCnt/pLBqHuZjwiMgsAjKU6zloQURSIHgfHhFJBOfSStT/HXXDtsWdodcBw57UYHzMjRplTie2xq51HSCTAR16FGPepnRkXmqBhFc7o6RIDgcHgfBnr2Pw+DtWOAPp6R+Yj9mvZ8DBEfhqZ1v8J659te2hT+Zg7FM50OuB0mJHvLPoEVy73ByOcj2eX34Fj/QshqNc4PCeh7Hzg/Z1HKXp46BFLZKTk7F8+XLo9XpERUVh5syZljxco9LrgA9f74yFn1yCu6cWi8b2Qb+ReWjftcRQJiejGfa92x6Ld6ehhZsOhXecAADOLjrMWf8LPDqVIj9HgdfD+qB3UD5auNpWB29T4+AgELP4V7z6dE/cyVFgw640pBxxx7XLv79j9VhiGxz41AMA8Jg6DzMWZmLR3x9F4JhcOCn0mDvWD87NdPjgyws4tr8Nbt1oZq3TsRoBmd1e0losap1Oh6VLlyI+Ph5JSUnYv38/Ll++bKnDNborF1pB1bEUbb3LIFcIDBp/G98dcq9W5sgnKoz8Ww5auFUmMtc25QAAz86l8OhUCgBQemjxUOty3M1zatwTkKCuvYuQfdUFOVnNUFHugONJbTBoeF61MveKfm8DNGuuQ9Vr6oUAmjXXw8FRQNFMj/JyGe4V2ef0qoagh8ykxdZYrIWXlpYGb29veHl5AQDCwsJw+PBhdOnSxVKHbFR5OQq0bqc1rLt7anHlfKtqZXJ+dQEALJ7oC71OhkkvXEOf4IJqZa6cb4mKchnaepdaPmiJa+NRhts3FYb1OzkKdOtTVKPc2KduImJaNuROAq9M6QkAOPlVawwenodPvk2FczM94lZ0RFGhNP+Tqhyltc9kb7EWnkajgYeHh2FdpVJBo9FY6nA2SaeTQZPhgtd3fo95m35G/MtdUFz4+z+UfI0T3n++K2au/QUO9nmF0CTt3+6JacP7499veePJudcBAN16F0Gvl+GpAH88HdwPEdOy4eElzf+kxH8f8W7KYmv4Z3af3D20yM3+vbWQd1MBpUdZ9TKeWvQbmQe5k0DbDmXw7FyCnIzKVt+9u45Y8/SjiFpwFT79arYyqOHdyXHGw56/t8rbeGiRq1HUWf74/jYYPLLyknfYuDs4m+wGXYUDCvMU+OHcQ/DpJd3fm71e0los4alUKuTk5BjWNRoNVCqVpQ7X6Dr3uYucTBfcuuaMCq0Mp/c9jP4jq/cH+Yfk4sfTrgCAu3ly3PzVBW29S1GhlWH9jO4YMukWHgvLtUb4kpR+sSXadSyBqn0p5E56BIXdwenD1ftd23n/Pug0MDgfNzIrByVu31Sgz+BCAJWDTt397iLrv10WUmSvLTyL9eH5+voiMzMTWVlZUKlUSEpKwtq1ay11uEbnKAeeXvYr/t/kntDrgKDoW2jfrQSfr+mATr2L0D8kD72HFeBishteUveFg4PA/7yWiVbKCpzc/TB+SnkId/PlSP5PWwDArLcvo2PPYiufVdOm18nw/pLOePPfP8DRUeDQ5ypcu9wcU/5xDekXWyLliDvGTclB38cLUFEhQ1GhHGsX+AAAEj/2xIurLmPzgfOQyYBDu9oi8+cWVj4j67DnUVqZEFXjUA3v+PHjWLFiBXQ6HSZNmoQ5c+bUW378uC6I33zPUuGQBUzpHmLtEMgM2W3ScDHj3APV0WeUP7q8MdCksmJNDnbv3v1Ax2tIFr0PLygoCEFBQZY8BBE1NjPeS2trF7WcaUFEZjM14dnazStMeERklso+PCY8IpKIhhqBNTb9dPfu3Vi9erXhDo/JkycjKioKANCjRw907doVAODp6YnNmzcbPR4THhGZRQANco9d1fTThIQEqFQqREZGQq1W15iNFRoaitjY2Br7N2vWDHv37jXrmPY5tkxE1iOACr2DSUt9/jj9VKFQGKafWhJbeERkFnP68PLz8xEREWFYj46ORnR0NIDap5+mpaXVqOPQoUNITU1Fp06dsHDhQnh6egIAysrKEBERAblcjpkzZ2LEiBFG42HCIyKzmZrwlErlA92HFxwcjLFjx0KhUGDHjh14+eWXsXXrVgDA0aNHoVKpkJWVhb/97W/o2rUrOnToUG99vKQlIrMJITNpqY8p00+VSiUUisr5zlFRUbh06VK1/QHAy8sLAwcOxA8//GA0biY8IjJL1aDFgz484I/TT7VaLZKSkqBWq6uVuXXrluHnI0eO4JFHHgEAFBYWQqutfBBEXl4ezp07Z9Kj53hJS0Rma4jbUuRyOWJjYzF9+nTD9FMfHx9s2LABvXr1wvDhw7Ft2zYcOXIEjo6OcHV1xcqVKwEAV65cwRtvvAGZTAYhBGbMmGFSwrPoXFpzcS6t/eFcWvvSEHNpHx3xGFwXBJtU1nNzunTm0hJR02Ssf85WMeERkdls8Vl3pmDCIyKzCAC20xFmHiY8IjKTbT6+3RRMeERkFiEAnZFpY7aKCY+IzMZLWiKSDI7SEpFkMOERkSQIG30FoymY8IjIbOzDIyLJ0HOUloikwk4beEx4RGQeAQDswyMiSRCw2yYeEx4Rma3J3ZaybNkyyGR1n9Trr79ukYCIyNbJoNc3sYTXq1evxoyDiOxJU2vhTZw4sdp6SUkJXFxcLB4QEdk2Iez3PjyjN9OcP38eoaGhGDNmDADgp59+wuLFiy0dFxHZMmHiYmOMJrwVK1Zgy5YtcHNzAwB0794dZ8+etXhgRGS7GuI1jdZg0iht1Zu+qzg42Odd1kTUQGyw9WYKownP09MT586dg0wmQ3l5ObZu3Wp4NyQRSZAAhJ2O0hptqi1evBjbt2+HRqNBYGAgfvzxR8TGxjZGbERks2QmLrbFaAvP3d0da9eubYxYiMhe2OklrdEWXlZWFmbPno1BgwZh8ODBmDNnDrKyshojNiKyVU11lHb+/PkYPXo0Tp48iRMnTmD06NF48cUXGyM2IrJJssobj01ZbIzRhFdSUoIJEyZALpdDLpcjPDwcZWVljREbEdmoqpuPjS22ps4+vIKCAgDA0KFDERcXh9DQUMhkMhw4cABBQUGNFiAR2RgBwE5HaetMeBEREZDJZBD/TdM7duwwbJPJZJg/f77loyMimySzwdabKepMeEeOHGnMOIjInjS1hPdH6enpuHz5MrRareGzCRMmWCwoIrJxNjggYQqjCW/Tpk1ISUnBlStXEBQUhOTkZPTv358Jj0iqbPSWE1MYHaU9ePAgPvroI7Rp0wYrV67E3r17cffu3caIjYhsld7ExcYYbeE5OzvDwcEBcrkcRUVFaN26NW7evNkYsRGRrbLTS1qjLbxevXrht99+Q1RUFCIiIjBx4kT07du3MWIjIhslE6YtxiQnJ2PUqFEYOXIk4uLiamzfvXs3Bg0ahPDwcISHh+M///mPYduePXsQEhKCkJAQ7Nmzx6S4jbbwqh72+eSTTyIwMBBFRUXo3r27SZUTURPUQH14Op0OS5cuRUJCAlQqFSIjI6FWq9GlS5dq5UJDQ2s8sKSgoACbNm3Crl27IJPJEBERAbVaDVdX13qPWWfCu3TpUp07Xbp0CT179jTlnIiIapWWlgZvb294eXkBAMLCwnD48OEaCa82J0+eREBAgOHBxAEBAThx4gTGjh1b7351JrxVq1bVuZNMJsPWrVuNBmWujLSWeMqLl8v25GD2KWuHQGaInNkwfW+m3nicn5+PiIgIw3p0dDSio6MBABqNBh4eHoZtKpUKaWlpNeo4dOgQUlNT0alTJyxcuBCenp617qvRaIzGU2fC27Ztm2lnRETSY+LUMqVSid27d9/3YYKDgzF27FgoFArs2LEDL7/88gM1tvisdiIyXwM8HkqlUiEnJ8ewrtFooFKpqpVRKpVQKBQAgKioKENXmyn71oYJj4jMYuoIrbHLXl9fX2RmZiIrKwtarRZJSUlQq9XVyty6dcvw85EjRwyvlxgyZAhOnjyJwsJCFBYW4uTJkxgyZIjR2E2aWkZEVE0DjNLK5XLExsZi+vTp0Ol0mDRpEnx8fLBhwwb06tULw4cPx7Zt23DkyBE4OjrC1dUVK1euBAC4ublh7ty5iIyMBADExMQYBjDqIxOi/qdWCSGwb98+ZGVlYd68ecjOzsadO3fQu3fvBz/jP/H19oNHlk+D10uWczD7grVDIDNEzpRh9770B6qja0AAdBMnmVS2z7cnH6gPr6GZ9BKfCxcuICkpCQDQokULLFmyxOKBEZHtaqgbjxub0YSXlpaGN954A87OzgAAV1dXlJeXWzwwIrJRQlY5SmvKYmOM9uHJ5XLodDrIZJXB5+Xl8UXcRBJni603UxhNeFOmTEFMTAxyc3Oxbt06fPXVV3j++ecbIzYislVNNeGNHz8ePXv2xOnTpyGEwHvvvWcYGiYiaWqyLbzs7Gy4uLggODi42mft2rWzaGBEZMOaasKbNWuW4eeysjJcv34dnTp1MozaEpG0yARs8uGepjCa8BITE6utX7p0CZ988onFAiIishSzZ1r07Nmz1icaEJGENNVL2oSEBMPPer0eP/zwA9q2bWvRoIjItjXZQYvi4mLDz46OjggKCsKoUaMsGhQR2bimmPB0Oh2Ki4vx8ssvN1Y8RGTr7Pg1jXUmvIqKCsjlcpw7d64x4yEiOyBraqO0UVFR2LNnD7p3747Zs2dj9OjRaN68uWF7SEhIowRIRLanyfbhabVaKJVKpKSkVPucCY9IwppawsvNzUVCQgJ8fHwgk8nwx8fmVT1IgIgkqCn24en1+mojtEREVZrcJe3DDz+MefPmNWYsRGQvmlrCM/LkdyKSKJlogqO0H374YSOGQUR2xU7bQ3UmPFPeAERE0tTk+vCIiOrEhEdEksGER0SSYKOvYDQFEx4RmY0Jj4ikgwmPiCSDCY+IpIKXtEQkDU3x4QFERLWRoQlOLSMiqgsvaYlIOpjwiEgS2IdHRFJir5e0DtYOgIjsjYBMb9piTHJyMkaNGoWRI0ciLi6uznIHDx5Et27dcPHiRQDA9evX0bt3b4SHhyM8PByxsbEmRc4WHhGZrwFaeDqdDkuXLkVCQgJUKhUiIyOhVqvRpUuXauWKioqwdetW9OnTp9rnHTp0wN69e806Jlt4RGSe/z48wJSlPmlpafD29oaXlxcUCgXCwsJw+PDhGuU2bNiAGTNmwNnZ+YFDZ8IjIvMJ05b8/HxEREQYls8++8xQhUajgYeHh2FdpVJBo9FUO8ylS5eQk5ODYcOG1Qjh+vXrmDBhAiZPnoyzZ8+aFDYvaYnILDKYPmihVCqxe/fu+zqOXq/HqlWrsHLlyhrb2rZti6NHj0KpVOL7779HTEwMkpKS0LJly3rrZAuPiMxnYguvPiqVCjk5OYZ1jUYDlUplWC8uLkZ6ejqmTp0KtVqNCxcuYM6cObh48SIUCgWUSiUAoFevXujQoQMyMjKMhs0WHhGZp4HeWubr64vMzExkZWVBpVIhKSkJa9euNWxv1aoVUlJSDOtTpkzBggUL4Ovri7y8PLi6usLR0RFZWVnIzMyEl5eX0WMy4RGR2RriPjy5XI7Y2FhMnz4dOp0OkyZNgo+PDzZs2IBevXph+PDhde6bmpqKd955B3K5HA4ODliyZIlJLx6TCRt6Aa2vtx88snysHQaZ4WD2BWuHQGaInCnD7n3pD1THo30D8NCjESaVbVdy6r778CyBLTwiMpu9zrRgwnsA/sN+w+xl2XB0EPjyU3fs3KSqtj1syh2MezoXej1QUuyADS954dovzRA8MR9Rc28ZynXqUYqYUV3x6yWXxj4FyUk92gqbF/0FOr0MY57MRfSzt2qUOb7PDR+v9QBkAp0fLcXC965Cc90JS6d1gl4vQ0UFED7tDsZOzbXCGdgIJrzqFi5ciGPHjqF169bYv3+/pQ5jNQ4OAjErbmDhE51x56YTNh74BacPuuLaL80MZY7uUSJpWxsAwKCQQsxanI3XnuqMo3uUOLqncoSpY/cSvPHvTCa7RqDTAe++2h4rd1xBG89yPBvaFYNGFcK7a5mhzI1fFfhsY1u8vfcXtHLToeBO5Z+Ie9sKrEv8BQpngZJiB8wK7o7BIYVo7VFhrdOxngYatLAGi92WEhERgfj4eEtVb3Xd+t5DdqYCOdecUVHugGN73TB4VGG1MveKHA0/N2uuR229pcETCnB8r/HOVnpwP59vjnYdy+DprYWTQmBYeD7+96BrtTJfbm+NcU/fQSs3HQDArU1lQnNSCCicK3+B5WUy6O30D74hVD0A1JTF1lishTdgwABcv37dUpUvhAMAAAvVSURBVNVbXWuPctzOVhjW79x0Qvd+92qUG/f0HUTMvA0nhcCCqEdqbB86vgCLn+lowUipSm6OEx5uV25Yb+NZjp/ONa9W5vqvlS30F8Z3gV4vw+T5ORgQfBcAcOuGE2KndkZ2hjOmL8qWZuuuiu2MdZqFNx5bWOKHbfDM4z2wZbkn/ucf1afNdOtbjLISB1z9mZeztkKnA25kOOOtXZex8L2rWP9PLxQVVrbU2/6lHJsP/4yEb3/A1/9RIv+2RLvAG2gurTUw4d2nytaC1rDexrMcd2461Vn+2BdueHx09UveYeEFOPYFL2cbS2Wr/Pff0Z2bTmjjWV6tTBvPcgwK+Q1yJ8CjgxbtHynDjQzFn+qpQMdupfg+pUWjxG2TGmCmhTUw4d2nny80x186aaHyKoPcSY9h4QU4fah6f1C7Tr93hg8c8RtuZPz+tAeZTGDouAIcY/9do+nmdw83MpyRc02Bcq0Mx/YqMSjkt2plHh9diLT/rZyPWZjriOtXnOHZQYvb2U4oK5EBAO4WOOJSagu0f6SsxjGkwl5beBJtkz84vU6Gd1/7C1Z88iscHIFDO9xxNb0Zpr6Ug/T/c8HpQ64Y/8wd9Au8i4oKGYoKHLHmHx0M+/sOKsbt7MpBD2ocjnIgZvl1vPo/naHXyRDyRB46divFR6s90LXPPQwe9Rv8h93FueOtMCOoOxwcBWYsysZD7jp8d9wF/1raubLHXgCRs2+jU49Sa5+S1ZjycE8AgKPxIo3JYjMtXnzxRZw5cwb5+flo3bo1nn32WURFRdW7D2da2B/OtLAvDTHTomfvALTuOMGksm3k/yuNmRZvv/22paomIiuzxctVU/CSlojMZ+olrY1hwiMi89lnvmPCIyLz2OoIrCmY8IjITKa9gtEWMeERkfnsM98x4RGR+WR2OpeWCY+IzCMA2OCTUEzBhEdEZmMLj4ikwz7zHRMeEZlJmDGX1sYw4RGR+XhJS0RSUPWId3vEhEdEZhJs4RGRRNjo04xNwYRHRGaT2elr25jwiMh89pnvmPCIyEyCNx4TkZQw4RGRNHCUloikhH14RCQJgqO0RCQlvKQlIsmw04TnYO0AiMjOVD0A1JTFiOTkZIwaNQojR45EXFxcneUOHjyIbt264eLFi4bPPvjgA4wcORKjRo3CiRMnTAqdLTwiMpNokPvwdDodli5dioSEBKhUKkRGRkKtVqNLly7VyhUVFWHr1q3o06eP4bPLly8jKSkJSUlJ0Gg0eOaZZ3Dw4EE4OjrWe0y28IjIfEKYttQjLS0N3t7e8PLygkKhQFhYGA4fPlyj3IYNGzBjxgw4OzsbPjt8+DDCwsKgUCjg5eUFb29vpKWlGQ2bCY+IzCMA6PQmLfn5+YiIiDAsn332maEajUYDDw8Pw7pKpYJGo6l2qEuXLiEnJwfDhg2r9rkp+9aGl7REZCbTbzxWKpXYvXv3fR1Fr9dj1apVWLly5X3tXxsmPCIyXwP04alUKuTk5BjWNRoNVCqVYb24uBjp6emYOnUqAOD27duYM2cO3n//faP71oWXtERkHgFAL0xb6uHr64vMzExkZWVBq9UiKSkJarXasL1Vq1ZISUnBkSNHcOTIEfj5+eH999+Hr68v1Go1kpKSoNVqkZWVhczMTPTu3dto6GzhEZH5xIPPtJDL5YiNjcX06dOh0+kwadIk+Pj4YMOGDejVqxeGDx9e574+Pj4YM2YMQkND4ejoiNjYWKMjtAAgE8J27iD09faDR5aPtcMgMxzMvmDtEMgMkTNl2L0v/YHq8PUZgPZlASaVdfG/dt99eJbAFh4RmUfAbmdaMOERkfmY8IhIMpjwiEgaBMDHQxGRJLAPj4gkRccWHhFJgoBogPvwrIEJj4jMUzXTwg4x4RGR+diHR0TSwFFaIpIKjtISkZQInc7aIdwXJjwiMo8w/ugnW8WER0Tm420pRCQVgi08IpIGwRYeEUmE4KBFg2ju4YxW/e3zi5SqyLm+1g6BzHBDc+OB6xg4oj/y8/NNKqtUKh/4eA3Jph7xTkRkSXxrGRFJBhMeEUkGEx4RSQYTHhFJBhMeEUkGEx4RSQYTngUlJydj1KhRGDlyJOLi4qwdDplg4cKFGDx4MMaOHWvtUMgCmPAsRKfTYenSpYiPj0dSUhL279+Py5cvWzssMiIiIgLx8fHWDoMshAnPQtLS0uDt7Q0vLy8oFAqEhYXh8OHD1g6LjBgwYABcXV2tHQZZCBOehWg0Gnh4eBjWVSoVNBqNFSMiIiY8IpIMJjwLUalUyMnJMaxrNBqoVCorRkRETHgW4uvri8zMTGRlZUGr1SIpKQlqtdraYRFJGp+WYkHHjx/HihUroNPpMGnSJMyZM8faIZERL774Is6cOYP8/Hy0bt0azz77LKKioqwdFjUQJjwikgxe0hKRZDDhEZFkMOERkWQw4RGRZDDhEZFkMOHZkR49eiA8PBxjx47Fc889h5KSkvuu65VXXsFXX30FAHjttdfqfbBBSkoKzp07Z/Yx1Go18vLyTP78j/r27WvWsTZu3IgtW7aYtQ9JDxOeHWnWrBn27t2L/fv3w8nJCTt27Ki2vaKi4r7qXb58Obp06VLn9jNnzuD8+fP3VTeRLbGp99KS6fz9/fHzzz8jJSUFGzZswEMPPYSMjAwcOHAAa9aswZkzZ6DVavHUU0/hiSeegBACy5Ytw6lTp+Dp6QknJydDXVOmTMGCBQvg6+uL5ORkrFu3DjqdDkqlEsuXL8eOHTvg4OCAffv2YdGiRejcuTPeeOMNZGdnAwBeffVV9O9f+a7S+fPnQ6PRwM/PD6bc4jl37lzk5OSgrKwMU6dORXR0tGHbihUrcOrUKbRp0wbr1q2Du7s7rl27hiVLliA/Px/NmjXDsmXL8MgjjzT8F0xNkyC74efnJ4QQory8XMyePVts375dnD59WvTp00dcu3ZNCCHEjh07xLvvviuEEKKsrExMnDhRXLt2TRw8eFA8/fTToqKiQuTk5Ij+/fuLL7/8UgghxOTJk0VaWprIzc0VQ4cONdSVn58vhBDinXfeEfHx8YY4XnzxRZGamiqEEOLGjRti9OjRQgghli1bJjZu3CiEEOLo0aOia9euIjc3t8Z5BAcHGz6vOkZJSYkICwsTeXl5QgghunbtKvbu3SuEEGLjxo1iyZIlQgghpk6dKjIyMoQQQly4cEFMmTKl1hiJasMWnh0pLS1FeHg4gMoWXmRkJM6fPw9fX194eXkBAE6dOoWff/4ZBw8eBADcvXsXV69eRWpqKsLCwuDo6AiVSoVBgwbVqP/ChQvw9/c31OXm5lZrHN9++221Pr+ioiIUFxcjNTUVmzZtAgAMGzbMpOfKbdu2DV9//TUA4ObNm7h69SqUSiUcHBwQGhoKAAgPD8e8efNQXFyM8+fP4x//+Idhf61Wa/QYRFWY8OxIVR/enzVv3tzwsxACr7/+OgIDA6uVOX78eIPFodfrsXPnTjg7Oz9QPSkpKfj222/x2WefwcXFBVOmTEFZWVmtZWUyGYQQeOihh2r9DohMwUGLJmbIkCH49NNPUV5eDgDIyMjAvXv3MGDAAHz55ZfQ6XS4desWUlJSauzr5+eHs2fPIisrCwBQUFAAAGjRogWKi4urHWPbtm2G9R9//BFA5dOCExMTAVQm2MLCwnpjvXv3LlxdXeHi4oIrV67gwoULhm16vd7QSk1MTET//v3RsmVLtG/fHl9++SWAyuT+008/mfcFkaQx4TUxUVFR6NKlCyIiIjB27FjExsZCp9Nh5MiR8Pb2RmhoKF5++WX4+fnV2Nfd3R1Lly7Fs88+i/Hjx+OFF14AAAQHB+Prr79GeHg4zp49i9deew3ff/89xo0bh9DQUHz66acAgJiYGJw9exZhYWH4+uuv0a5du3pjHTp0KCoqKjBmzBisXbu2WkzNmzdHWloaxo4di9OnTyMmJgYA8NZbb+Hzzz/H+PHjERYWhm+++aahvjqSAD4thYgkgy08IpIMJjwikgwmPCKSDCY8IpIMJjwikgwmPCKSDCY8IpKM/w/DBI/ftpkuggAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "<h3>Getting top 50 features:</h3>"
      ],
      "metadata": {
        "id": "gDKpVluzEjGM"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "top_50_features_0 , top_50_features_1 = top_features(pipe_set2,50)\n",
        "print('top 50 features for predicting class 0:\\n', top_50_features_0,'\\n' )\n",
        "print('top 50 features for predicting class 1:\\n', top_50_features_1 )"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "TJRetBBTurK3",
        "outputId": "427308b6-b50d-403f-8671-af5fa7a44839"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "top 50 features for predicting class 0:\n",
            " ['onehotencoder__teacher_prefix_mrs'\n",
            " 'kbinsdiscretizer__teacher_number_of_previously_posted_projects_0.0'\n",
            " 'onehotencoder__project_grade_category_grades_prek_2'\n",
            " 'onehotencoder__teacher_prefix_ms'\n",
            " 'kbinsdiscretizer__teacher_number_of_previously_posted_projects_1.0'\n",
            " 'kbinsdiscretizer__price_3.0'\n",
            " 'onehotencoder__project_grade_category_grades_3_5'\n",
            " 'kbinsdiscretizer__price_2.0' 'kbinsdiscretizer__price_1.0'\n",
            " 'onehotencoder__clean_categories_literacy_language'\n",
            " 'onehotencoder__clean_categories_math_science'\n",
            " 'kbinsdiscretizer__teacher_number_of_previously_posted_projects_2.0'\n",
            " 'kbinsdiscretizer__price_0.0'\n",
            " 'onehotencoder__project_grade_category_grades_6_8'\n",
            " 'onehotencoder__school_state_ca'\n",
            " 'onehotencoder__clean_categories_literacy_language math_science'\n",
            " 'onehotencoder__project_grade_category_grades_9_12'\n",
            " 'onehotencoder__teacher_prefix_mr'\n",
            " 'onehotencoder__clean_categories_health_sports'\n",
            " 'onehotencoder__school_state_tx'\n",
            " 'onehotencoder__clean_subcategories_literacy'\n",
            " 'onehotencoder__clean_subcategories_literacy mathematics'\n",
            " 'onehotencoder__school_state_fl'\n",
            " 'onehotencoder__clean_subcategories_mathematics'\n",
            " 'onehotencoder__school_state_ny' 'tfidfvectorizer__learning'\n",
            " 'onehotencoder__clean_subcategories_literature_writing mathematics'\n",
            " 'onehotencoder__clean_subcategories_literacy literature_writing'\n",
            " 'tfidfvectorizer__classroom' 'onehotencoder__clean_categories_music_arts'\n",
            " 'onehotencoder__clean_categories_specialneeds'\n",
            " 'onehotencoder__clean_subcategories_specialneeds'\n",
            " 'onehotencoder__school_state_nc'\n",
            " 'onehotencoder__clean_categories_appliedlearning'\n",
            " 'onehotencoder__clean_subcategories_literature_writing'\n",
            " 'tfidfvectorizer__help' 'tfidfvectorizer__learn'\n",
            " 'onehotencoder__school_state_il' 'onehotencoder__school_state_ga'\n",
            " 'onehotencoder__clean_categories_literacy_language specialneeds'\n",
            " 'tfidfvectorizer__materials'\n",
            " 'onehotencoder__clean_subcategories_appliedsciences mathematics'\n",
            " 'tfidfvectorizer__need' 'tfidfvectorizer__work'\n",
            " 'onehotencoder__school_state_sc' 'tfidfvectorizer__reading'\n",
            " 'onehotencoder__teacher_prefix_teacher' 'onehotencoder__school_state_mi'\n",
            " 'tfidfvectorizer__skills' 'tfidfvectorizer__come'] \n",
            "\n",
            "top 50 features for predicting class 1:\n",
            " ['onehotencoder__teacher_prefix_mrs'\n",
            " 'kbinsdiscretizer__teacher_number_of_previously_posted_projects_0.0'\n",
            " 'onehotencoder__project_grade_category_grades_prek_2'\n",
            " 'onehotencoder__teacher_prefix_ms'\n",
            " 'onehotencoder__project_grade_category_grades_3_5'\n",
            " 'kbinsdiscretizer__teacher_number_of_previously_posted_projects_1.0'\n",
            " 'kbinsdiscretizer__teacher_number_of_previously_posted_projects_2.0'\n",
            " 'kbinsdiscretizer__price_0.0' 'kbinsdiscretizer__price_1.0'\n",
            " 'kbinsdiscretizer__price_2.0' 'kbinsdiscretizer__price_3.0'\n",
            " 'onehotencoder__clean_categories_literacy_language'\n",
            " 'onehotencoder__project_grade_category_grades_6_8'\n",
            " 'onehotencoder__clean_categories_math_science'\n",
            " 'onehotencoder__school_state_ca'\n",
            " 'onehotencoder__clean_categories_literacy_language math_science'\n",
            " 'onehotencoder__project_grade_category_grades_9_12'\n",
            " 'onehotencoder__teacher_prefix_mr'\n",
            " 'onehotencoder__clean_categories_health_sports'\n",
            " 'onehotencoder__clean_subcategories_literacy'\n",
            " 'onehotencoder__clean_subcategories_literacy mathematics'\n",
            " 'onehotencoder__school_state_ny' 'onehotencoder__school_state_tx'\n",
            " 'onehotencoder__school_state_fl'\n",
            " 'onehotencoder__clean_subcategories_literature_writing mathematics'\n",
            " 'onehotencoder__clean_subcategories_literacy literature_writing'\n",
            " 'onehotencoder__clean_categories_music_arts' 'tfidfvectorizer__learning'\n",
            " 'tfidfvectorizer__classroom' 'onehotencoder__school_state_nc'\n",
            " 'onehotencoder__clean_subcategories_mathematics'\n",
            " 'onehotencoder__clean_subcategories_literature_writing'\n",
            " 'onehotencoder__school_state_il' 'tfidfvectorizer__reading'\n",
            " 'tfidfvectorizer__help' 'tfidfvectorizer__learn'\n",
            " 'onehotencoder__clean_subcategories_specialneeds'\n",
            " 'onehotencoder__clean_categories_specialneeds'\n",
            " 'onehotencoder__school_state_sc'\n",
            " 'onehotencoder__clean_categories_literacy_language specialneeds'\n",
            " 'onehotencoder__school_state_ga'\n",
            " 'onehotencoder__clean_categories_appliedlearning'\n",
            " 'tfidfvectorizer__books'\n",
            " 'onehotencoder__clean_subcategories_health_wellness'\n",
            " 'tfidfvectorizer__technology' 'tfidfvectorizer__work'\n",
            " 'tfidfvectorizer__need'\n",
            " 'onehotencoder__clean_subcategories_appliedsciences mathematics'\n",
            " 'tfidfvectorizer__use' 'tfidfvectorizer__love']\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Observation:**\n",
        "\n",
        "- Most of the top features are categorical( one hot encoded, k-bin discritize).\n",
        "- Most of them are same for both classes.\n",
        "\n",
        "Possible reason for difference in top features betweeen both set:\n",
        "- The most frequent words in essay vocabulary feature may have got high weightage in BoW but got penalize by 'idf' value in tfidf.\n",
        "- In TfidfVectorizer the final values are l2 or l1 normalize which makes all values <1. so essay feature's value will be less(<1) compared to one hot encoded(=1) and multinomial probability would have been calculated to a low value."
      ],
      "metadata": {
        "id": "Iu_P2d_L9EuD"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# https://www.geeksforgeeks.org/creating-tables-with-prettytable-library-python/\n",
        "\n",
        "columns = [\"Model\", \"True Positive-Test\", \"False Positive-Test\", \"AUC Score-Test\"]\n",
        "\n",
        "myTable = PrettyTable()\n",
        "\n",
        "myTable.add_column(columns[0], [\"Set-1(BoW)\" , \"Set-1(tfidf)\"])\n",
        "myTable.add_column(columns[1], [np.around(confusion_matrix_test_set1_thresh[1,1],2) , np.around(confusion_matrix_test_set2_thresh[1,1],2)])\n",
        "myTable.add_column(columns[2], [np.around(confusion_matrix_test_set1_thresh[0,1],2) , np.around(confusion_matrix_test_set2_thresh[0,1],2)])\n",
        "myTable.add_column(columns[3], [np.around(auc_score_set1_test,2) , np.around(auc_score_set2_test,2)])\n",
        "  \n",
        "print(myTable)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GI4zLe8j_UKw",
        "outputId": "2991cd8f-a6d0-44da-d20e-2e7f31f666da"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "+--------------+--------------------+---------------------+----------------+\n",
            "|    Model     | True Positive-Test | False Positive-Test | AUC Score-Test |\n",
            "+--------------+--------------------+---------------------+----------------+\n",
            "|  Set-1(BoW)  |        0.64        |         0.33        |      0.7       |\n",
            "| Set-1(tfidf) |        0.63        |         0.38        |      0.66      |\n",
            "+--------------+--------------------+---------------------+----------------+\n"
          ]
        }
      ]
    }
  ]
}