spam_detection.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/PouyaRepos/Spam_detection/blob/master/spam_detection.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "!git clone https://github.com/PouyaRepos/Spam_detection.git"
      ],
      "metadata": {
        "id": "gfszUx5tuWsJ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": 50,
      "metadata": {
        "id": "rRLL9B8kNLsq"
      },
      "outputs": [],
      "source": [
        "# Libraries\n",
        "import os\n",
        "# os.getcwd()\n",
        "from scipy.io import loadmat\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import seaborn as sns\n",
        "import matplotlib.pylab as plt\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.decomposition import TruncatedSVD\n",
        "from sklearn.model_selection import GridSearchCV, KFold\n",
        "from sklearn.model_selection import cross_val_score\n",
        "from sklearn.pipeline import Pipeline\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        "from sklearn.tree import DecisionTreeClassifier\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.svm import SVC\n",
        "from sklearn.metrics import accuracy_score \n",
        "from sklearn.metrics import classification_report\n",
        "from sklearn import metrics\n",
        "from keras.models import Sequential\n",
        "from keras.layers.core import Dense\n",
        "import tensorflow as tf\n",
        "from scipy.stats import zscore\n",
        "from keras.wrappers.scikit_learn import KerasClassifier\n",
        "from scikeras.wrappers import KerasClassifier\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "id": "bUPQE3t8NLss"
      },
      "outputs": [],
      "source": [
        "# load data\n",
        "path = \"Spam_detection/emails.mat\"\n",
        "ref = loadmat(path)\n",
        "features = ref['X'].T\n",
        "target = ref['Y'].T"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bmNCWvSGNLst"
      },
      "source": [
        "# 1. Preprocessing"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_yisSoaaNLsu",
        "outputId": "86c1d20d-5b53-4a1f-ff8b-0ef1190eb332"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((10000, 57173), (10000, 1))"
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ],
      "source": [
        "# convert data to DataFrame for further process\n",
        "\n",
        "X= pd.DataFrame(features.todense())\n",
        "Y = pd.DataFrame(target)\n",
        "\n",
        "# Shape of data\n",
        "X.shape, Y.shape\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bxFyibK1NLsv",
        "outputId": "91faac08-ce02-44b7-daba-aea4b88a0c76"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "sparsity = 0.99589\n"
          ]
        }
      ],
      "source": [
        "# Sparsity\n",
        "#sparsity = 1 - (#nonzero/ # matrix elements)\n",
        "print(\"sparsity =\", round(1 - (np.count_nonzero(X) / X.size), 5))\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Tm3iFviYNLsv",
        "outputId": "7595145e-5207-46f4-8ee0-69ca82deddda"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(0, 0)"
            ]
          },
          "metadata": {},
          "execution_count": 26
        }
      ],
      "source": [
        "# Check for NULL values\n",
        "sum(X.isna().sum()), sum(Y.isna().sum())"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SgJ7a67hNLsw"
      },
      "source": [
        "#### Plot density and boxplot"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 27,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8y209hSXNLsw",
        "outputId": "22c0c8b3-abd1-4d4c-a22b-d758f7cd4c6b"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((10000, 2), Index(['length', 'target'], dtype='object'))"
            ]
          },
          "metadata": {},
          "execution_count": 27
        }
      ],
      "source": [
        "# Preparing for visualizatoin\n",
        "# token_length =  X.sum() # [column-wise] sum of frequency of each token per all emails \n",
        "email_length = X.sum(axis = 1) # [row-wise] sum of frequency of each token per single email\n",
        "\n",
        "# put labels and dfx_tr (lenght of each email) in same DataFrame \n",
        "data = email_length.to_frame(name = 'length').copy()\n",
        "data['target'] = Y.copy()\n",
        "data.shape ,data.columns\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 933
        },
        "id": "zrsYS1IXNLsw",
        "outputId": "7775eabd-ca78-4c1e-87fb-8d17078e0058"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1500x1000 with 6 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABUoAAAOtCAYAAABXLl+UAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOzdeVxV1f7/8fc5TIdBwBEcUEy9DmlqaoRZWlFods1GNcvhWpZlaWqDZlo2eNM0hware1NzuJn9ulZmDmnmNyXHNC3taqlYCmoKCMp49u8POFuPDAIeOAd4PR+P89Cz9tp7rX026OLDZ61lMQzDEAAAAAAAAABUYVZ3dwAAAAAAAAAA3I1AKQAAAAAAAIAqj0ApAAAAAAAAgCqPQCkAAAAAAACAKo9AKQAAAAAAAIAqj0ApAAAAAAAAgCqPQCkAAAAAAACAKo9AKQAAAAAAAIAqj0ApAAAAAAAAgCqPQCkAAAAAoMp48cUXZbFYyqWtbt26qVu3bub79evXy2Kx6NNPPy2X9gcNGqTIyMhyaau0UlNT9dBDDyk8PFwWi0UjR450d5cAVGEESgFUeLt379Y999yjRo0ayWazqX79+rrllls0e/Zsd3cNAAAAZWjevHmyWCzmy2azqV69eoqNjdWsWbN05swZl7Rz9OhRvfjii9q5c6dLrudKnty34njttdc0b948DRs2TAsWLNCDDz5YaN3MzEzNnDlT7du3V3BwsEJDQ3XllVdq6NCh2rdvXzn2GkBlZTEMw3B3JwCgtDZt2qQbb7xRDRs21MCBAxUeHq4jR47ohx9+0G+//aYDBw64u4sAAAAoI/PmzdPgwYM1adIkNW7cWFlZWUpISND69eu1Zs0aNWzYUF988YWuuuoq85zs7GxlZ2fLZrMVu51t27apU6dOmjt3rgYNGlTs8zIzMyVJvr6+knIzSm+88UYtXbpU99xzT7GvU9q+ZWVlyW63y8/PzyVtlYVrr71W3t7e+v777y9Z9+9//7u+/vpr9evXT9HR0crKytK+ffu0fPlyvfzyyyV6NgBQEG93dwAALserr76qkJAQbd26VaGhoU7Hjh8/7p5OAQAAoFz16NFDHTt2NN+PHTtW69at0+23365evXpp79698vf3lyR5e3vL27tsfxQ+e/asAgICzACpu/j4+Li1/eI4fvy4WrVqdcl6W7du1fLly/Xqq69q3LhxTsfeeustJSUllVEPAVQlTL0HUKH99ttvuvLKK/MFSSWpTp065t8tFouGDx+uRYsWqXnz5rLZbOrQoYM2bNjgdM7hw4f12GOPqXnz5vL391fNmjV177336tChQ071HNO8vv/+ez355JOqXbu2QkND9cgjjygzM1NJSUkaMGCAqlevrurVq+uZZ54RCfwAAADl56abbtILL7ygw4cPa+HChWZ5QWuUrlmzRl26dFFoaKiCgoLUvHlzMxi3fv16derUSZI0ePBgc5r/vHnzJOWuQ9q6dWtt375dN9xwgwICAsxzL16j1CEnJ0fjxo1TeHi4AgMD1atXLx05csSpTmRkZIEZkhde81J9K2iN0rS0NI0ePVoRERHy8/NT8+bN9cYbb+QbqzrGz8uWLVPr1q3l5+enK6+8UitXriz4A7/I8ePHNWTIEIWFhclms6lt27aaP3++edyxXuvBgwf11VdfmX2/eNzt8Ntvv0mSrrvuunzHvLy8VLNmTfO94xnv27dP9913n4KDg1WzZk2NGDFC6enpTufOnTtXN910k+rUqSM/Pz+1atVK7777br42IiMjdfvtt2v9+vXq2LGj/P391aZNG61fv16S9Nlnn6lNmzbmzxk//vhjsT4nAJ6FjFIAFVqjRo0UFxenPXv2qHXr1kXW/e6777RkyRI9+eST8vPz0zvvvKPu3btry5Yt5rlbt27Vpk2b1LdvXzVo0ECHDh3Su+++q27duumXX35RQECA0zWfeOIJhYeH66WXXtIPP/yg999/X6Ghodq0aZMaNmyo1157TStWrNDUqVPVunVrDRgwoMw+CwAAADh78MEHNW7cOK1evVoPP/xwgXV+/vln3X777brqqqs0adIk+fn56cCBA9q4caMkqWXLlpo0aZImTJigoUOH6vrrr5ckde7c2bzGX3/9pR49eqhv37564IEHFBYWVmS/Xn31VVksFj377LM6fvy4ZsyYoZiYGO3cudPMfC2O4vTtQoZhqFevXvr22281ZMgQtWvXTqtWrdLTTz+tP//8U2+++aZT/e+//16fffaZHnvsMVWrVk2zZs3S3Xffrfj4eKfA5MXOnTunbt266cCBAxo+fLgaN26spUuXatCgQUpKStKIESPUsmVLLViwQE899ZQaNGig0aNHS5Jq165d4DUbNWokSVq0aJGuu+66YmUF33fffYqMjNTkyZP1ww8/aNasWTp9+rQ++ugjs867776rK6+8Ur169ZK3t7e+/PJLPfbYY7Lb7Xr88cedrnfgwAHdf//9euSRR/TAAw/ojTfe0N///nfNmTNH48aN02OPPSZJmjx5su677z79+uuvslrJTwMqFAMAKrDVq1cbXl5ehpeXlxEdHW0888wzxqpVq4zMzEynepIMSca2bdvMssOHDxs2m8248847zbKzZ8/mayMuLs6QZHz00Udm2dy5cw1JRmxsrGG3283y6Ohow2KxGI8++qhZlp2dbTRo0MDo2rWrK24ZAAAAeRxjsq1btxZaJyQkxGjfvr35fuLEicaFPwq/+eabhiTjxIkThV5j69athiRj7ty5+Y517drVkGTMmTOnwGMXjgG//fZbQ5JRv359IyUlxSz/5JNPDEnGzJkzzbJGjRoZAwcOvOQ1i+rbwIEDjUaNGpnvly1bZkgyXnnlFad699xzj2GxWIwDBw6YZZIMX19fp7Jdu3YZkozZs2fna+tCM2bMMCQZCxcuNMsyMzON6OhoIygoyOneGzVqZPTs2bPI6xmGYdjtdvOzDgsLM/r162e8/fbbxuHDh/PVdTzjXr16OZU/9thjhiRj165dZllB4//Y2FjjiiuucCpr1KiRIcnYtGmTWbZq1SpDkuHv7+/Uj/fee8+QZHz77beXvC8AnoVfbQCo0G655RbFxcWpV69e2rVrl6ZMmaLY2FjVr19fX3zxhVPd6OhodejQwXzfsGFD3XHHHVq1apVycnIkyek3+FlZWfrrr7/UtGlThYaGaseOHfnaHzJkiNPUraioKBmGoSFDhphlXl5e6tixo37//XeX3TcAAACKJygoSGfOnCn0uGMJp88//1x2u71Ubfj5+Wnw4MHFrj9gwABVq1bNfH/PPfeobt26WrFiRanaL64VK1bIy8tLTz75pFP56NGjZRiGvv76a6fymJgYNWnSxHx/1VVXKTg4+JLj2hUrVig8PFz9+vUzy3x8fPTkk08qNTVV3333XYn7brFYtGrVKr3yyiuqXr26/vOf/+jxxx9Xo0aN1KdPnwLXKL04I/SJJ54w++dw4fg/OTlZJ0+eVNeuXfX7778rOTnZ6fxWrVopOjrafB8VFSUpd5mHhg0b5itn/A9UPARKAVR4nTp10meffabTp09ry5YtGjt2rM6cOaN77rlHv/zyi1mvWbNm+c7929/+prNnz+rEiROScqcJTZgwwVyzqVatWqpdu7aSkpLyDZQkOQ2IJCkkJESSFBERka/89OnTl32vAAAAKJnU1FSnoOTF+vTpo+uuu04PPfSQwsLC1LdvX33yySclCprWr1+/RBs3XTwutVgsatq0aaHrc7rK4cOHVa9evXyfR8uWLc3jF7p4rCtJ1atXv+S49vDhw2rWrFm+aeeFtVNcfn5+ev7557V3714dPXpU//nPf3Tttdfqk08+0fDhw/PVv/hzbtKkiaxWq9PnvHHjRsXExCgwMFChoaGqXbu2ucbsxeP/koz9JTH+ByogAqUAKg1fX1916tRJr732mt59911lZWVp6dKlJbrGE088oVdffVX33XefPvnkE61evVpr1qxRzZo1Cxwse3l5FXidgsoNNnMCAAAoV3/88YeSk5PVtGnTQuv4+/trw4YN+uabb/Tggw/qp59+Up8+fXTLLbeYs44upSTrihbXxRtOORS3T65Q2FjXE8a1devWVd++fbVhwwY1a9ZMn3zyibKzs4s85+LP9LffftPNN9+skydPavr06frqq6+0Zs0aPfXUU5KUb/xfkrG/5BmfE4CSYTMnAJVSx44dJUnHjh0zy/bv35+v3v/+9z8FBASYi8Z/+umnGjhwoKZNm2bWSU9PL3AqDwAAADzbggULJEmxsbFF1rNarbr55pt18803a/r06Xrttdf0/PPP69tvv1VMTEyhQcvSunhcahiGDhw4oKuuusosq169eoFj0MOHD+uKK64w35ekb40aNdI333yjM2fOOGWV7tu3zzzuCo0aNdJPP/0ku93ulFXq6nak3Cn9V111lfbv36+TJ08qPDzcPLZ//341btzYfH/gwAHZ7XZFRkZKkr788ktlZGToiy++cMoW/fbbb13WPwAVCxmlACq0b7/9tsDf1DrWHWrevLlZFhcX57TO6JEjR/T555/r1ltvNX8L7OXlle96s2fPLtff3AMAAODyrVu3Ti+//LIaN26s/v37F1rv1KlT+cratWsnScrIyJAkBQYGSpLLfnn+0UcfOa2b+umnn+rYsWPq0aOHWdakSRP98MMPyszMNMuWL1+uI0eOOF2rJH277bbblJOTo7feesup/M0335TFYnFq/3LcdtttSkhI0JIlS8yy7OxszZ49W0FBQeratWuJr7l//37Fx8fnK09KSlJcXJyqV69uJj84vP32207vZ8+eLUnmfTp+Brhw/J+cnKy5c+eWuH8AKgcySgFUaE888YTOnj2rO++8Uy1atFBmZqY2bdqkJUuWKDIy0mlR/datWys2NlZPPvmk/Pz89M4770iSXnrpJbPO7bffrgULFigkJEStWrVSXFycvvnmG9WsWbPc7w0AAADF8/XXX2vfvn3Kzs5WYmKi1q1bpzVr1qhRo0b64osvZLPZCj130qRJ2rBhg3r27KlGjRrp+PHjeuedd9SgQQN16dJFUm7QMjQ0VHPmzFG1atUUGBioqKgop2zFkqhRo4a6dOmiwYMHKzExUTNmzFDTpk318MMPm3Ueeughffrpp+revbvuu+8+/fbbb1q4cKHT5kol7dvf//533XjjjXr++ed16NAhtW3bVqtXr9bnn3+ukSNH5rt2aQ0dOlTvvfeeBg0apO3btysyMlKffvqpNm7cqBkzZhS5Zmxhdu3apfvvv189evTQ9ddfrxo1aujPP//U/PnzdfToUc2YMSPfFPiDBw+qV69e6t69u+Li4rRw4ULdf//9atu2rSTp1ltvla+vr/7+97/rkUceUWpqqj744APVqVPHaWYagKqDQCmACu2NN97Q0qVLtWLFCr3//vvKzMxUw4YN9dhjj2n8+PHmLqaS1LVrV0VHR+ull15SfHy8WrVqpXnz5jlNcZo5c6a8vLy0aNEipaen67rrrtM333xzyelaAAAAcJ8JEyZIyl2zvkaNGmrTpo1mzJihwYMHXzIo16tXLx06dEgffvihTp48qVq1aqlr16566aWXzE15fHx8NH/+fI0dO1aPPvqosrOzNXfu3FIHSseNG6effvpJkydP1pkzZ3TzzTfrnXfeUUBAgFknNjZW06ZN0/Tp0zVy5Eh17NhRy5cv1+jRo52uVZK+Wa1WffHFF5owYYKWLFmiuXPnKjIyUlOnTs133cvh7++v9evX67nnntP8+fOVkpKi5s2ba+7cuRo0aFCprnnDDTfo5Zdf1tdff63p06frxIkTqlatmtq3b6/XX39dd999d75zlixZogkTJui5556Tt7e3hg8frqlTp5rHmzdvrk8//VTjx4/XmDFjFB4ermHDhql27dr6xz/+UdrbB1CBWQxWFwZQBVgsFj3++OP5phkBAAAAqFxefPFFvfTSSzpx4oRq1arl7u4AqEBYoxQAAAAAAABAlUegFAAAAAAAAECVR6AUAAAAAAAAQJVHoBRAlWAYBuuTAkA5evvttxUZGSmbzaaoqCht2bKlyPpLly5VixYtZLPZ1KZNG61YscLpuGEYmjBhgurWrSt/f3/FxMRo//795vFDhw5pyJAhaty4sfz9/dWkSRNNnDhRmZmZTtf56aefdP3118tmsykiIkJTpkxx3U0DADzCiy++KMMwWJ8UQIkRKAUAAIBLLVmyRKNGjdLEiRO1Y8cOtW3bVrGxsTp+/HiB9Tdt2qR+/fppyJAh+vHHH9W7d2/17t1be/bsMetMmTJFs2bN0pw5c7R582YFBgYqNjZW6enpkqR9+/bJbrfrvffe088//6w333xTc+bM0bhx48xrpKSk6NZbb1WjRo20fft2TZ06VS+++KLef//9sv1AAAAAUCGw6z0AAABcKioqSp06dTIz+e12uyIiIvTEE0/oueeey1e/T58+SktL0/Lly82ya6+9Vu3atdOcOXNkGIbq1aun0aNHa8yYMZKk5ORkhYWFad68eerbt2+B/Zg6dareffdd/f7775Kkd999V88//7wSEhLk6+srSXruuee0bNky7du3z6WfAQAAACoe79KeaLfbdfToUVWrVk0Wi8WVfQIAAEAZMAxDZ86cUb169WS1ls3EoszMTG3fvl1jx441y6xWq2JiYhQXF1fgOXFxcRo1apRTWWxsrJYtWyZJOnjwoBISEhQTE2MeDwkJUVRUlOLi4goNlCYnJ6tGjRpO7dxwww1mkNTRzuuvv67Tp0+revXq+a6RkZGhjIwM873dbtepU6dUs2ZNxsAAAAAVQEnGwKUOlB49elQRERGlPR0AAABucuTIETVo0KBMrn3y5Enl5OQoLCzMqTwsLKzQrM2EhIQC6yckJJjHHWWF1bnYgQMHNHv2bL3xxhtO7TRu3DjfNRzHCgqUTp48WS+99FKBbQAAAKDiKM4YuNSB0mrVqpmNBAcHl/YyAAAAKCcpKSmKiIgwx3GV1Z9//qnu3bvr3nvv1cMPP3xZ1xo7dqxTtmtycrIaNmzIGBgAAKCCKMkYuNSBUsdUo+DgYAaJAAAAFUhZThmvVauWvLy8lJiY6FSemJio8PDwAs8JDw8vsr7jz8TERNWtW9epTrt27ZzOO3r0qG688UZ17tw53yZNhbVzYRsX8/Pzk5+fX75yxsAAAAAVS3HGwOx6DwAAAJfx9fVVhw4dtHbtWrPMbrdr7dq1io6OLvCc6Ohop/qStGbNGrN+48aNFR4e7lQnJSVFmzdvdrrmn3/+qW7duqlDhw6aO3duvjWooqOjtWHDBmVlZTm107x58wKn3QMAAKBqIVAKAAAAlxo1apQ++OADzZ8/X3v37tWwYcOUlpamwYMHS5IGDBjgtNnTiBEjtHLlSk2bNk379u3Tiy++qG3btmn48OGScn/7P3LkSL3yyiv64osvtHv3bg0YMED16tVT7969JZ0PkjZs2FBvvPGGTpw4oYSEBKc1TO+//375+vpqyJAh+vnnn7VkyRLNnDkz30ZSAAAAqJpKPfUeAAAAKEifPn104sQJTZgwQQkJCWrXrp1WrlxpbpwUHx/vlO3ZuXNnLV68WOPHj9e4cePUrFkzLVu2TK1btzbrPPPMM0pLS9PQoUOVlJSkLl26aOXKlbLZbJJyM0MPHDigAwcO5Fuk3zAMSVJISIhWr16txx9/XB06dFCtWrU0YcIEDR06tKw/EgAAAFQAFsMxciyhlJQUhYSEKDk5mfWZAAAAKgDGb5ePzxAAAKBiKcn4jan3AAAAAAAAAKo8AqUAAAAAAAAAqjwCpQAAAADKXHaOXTuPJCk7x+7urgAAABSIQCkAAACAMrfgh8Pq/fZGfbjxoLu7AgAAUCACpQAAAADK3M9HUyRJ8afOurknAAAABSNQCgAAAKDMJSSnS5LOZTL1HgAAeCYCpQAAAADK3LHkc5Kkc1nZbu4JAABAwQiUAgAAAChziSkZkqRzmTlu7gkAAEDBCJQCAAAAKFNn0rOUmpGbSXoui0ApAADwTARKAQAAAJQpx/qkEhmlAADAcxEoBQAAAFCmjl0YKCWjFAAAeCgCpQAAAADKVEIKgVIAAOD5CJQCAAAAKFNMvQcAABWBt7s74GqGYSg9Pb1Y9TIycnfe9PPzk8VicVkfbDabS68HAAAAVGTHCJQCAIAKoNIFStPT09WjRw+39uHrr7+Wv7+/W/sAAAAAeIqE5HPm389l5cgwDBILAACAx2HqPQAAAIAylZCSYf7dbkgZ2XY39gYAAKBglS6j9EKp7frJsBZyizlZqrbrY0nSmbZ9JS+fy2rLYs9W0M7/XNY1AAAAgMrowoxSSUrPypHNx8tNvQEAAChYpQ6UGlbv4gVAvXwuO1BqXNbZAAAAQOWUnpWj02eznMrOZeUo1D3dAQAAKBRT7wEAAACUGceO9/4+Xqpmy83TOMuGTgAAwAMRKAUAAABQZhJScgOl4SE2BfjmTrdn53sAAOCJCJQCAAAAKDOOjNLwYJsCfHMzSs9lESgFAACeh0ApAAAAgDJzLC9QWjfEZm7gREYpAADwRARKAQAAAJQZx4734SE2+fvk/vhBRikAAPBEBEoBAAAAlBnHGqV1Qy6Yek9GKQAA8EAESgEAAACUGXON0hD/81PvySgFAAAeiEApAAAAgDJz7ILNnPzZ9R4AAHgwAqUAAAAAykRWjl0nUjMk5a5RGkBGKQAA8GAESgEAAACUiRNnMmQYko+XRTUDfckoBQAAHo1AKQAAAIAy4Zh2HxZsk9VqYY1SAADg0QiUAgAAACgTjo2c6obYJEkBeRmlZ8koBQAAHohAKQAAAIAykXwuS5IUGuArSfLPyyhNJ6MUAAB4IAKlAAAAAMpERnZuQNQx5d7GGqUAAMCDESgFAAAAUCYysu2SJD/v3B87HLvenyWjFAAAeCACpQAAAADKhGOKvSNQ6tj1Pp2MUgAA4IEIlAIAAAAoE+czSnMDpP7seg8AADwYgVIAAAAAZSIjKzdQavNxzig9m5nttj4BAAAUhkApAAAAgDLh2Mzp4ozS9LwAKgAAgCchUAoAAACgTJhT7y/KKGXqPQAA8EQESgEAAACUiYt3vXdklDL1HgAAeCICpQAAAADKxPld7/Om3vuen3pvtxtu6xcAAEBBCJQCAAAAKBOFZZReeAwAAMBTECgFAAAAUCYy8jJKbT7OmzlJTL8HAACeh0ApAAAAgDJxcUap1Wox/86GTgAAwNMQKAUAAABQJi7e9V66cJ1SAqUAAMCzECgFAAAAUCYyLtrMSZIC8qbfn80kUAoAADwLgVIAAAC43Ntvv63IyEjZbDZFRUVpy5YtRdZfunSpWrRoIZvNpjZt2mjFihVOxw3D0IQJE1S3bl35+/srJiZG+/fvd6rz6quvqnPnzgoICFBoaGiB7Vgslnyvjz/++LLuFYW7eOq9JNnyMkrPESgFAAAehkApAAAAXGrJkiUaNWqUJk6cqB07dqht27aKjY3V8ePHC6y/adMm9evXT0OGDNGPP/6o3r17q3fv3tqzZ49ZZ8qUKZo1a5bmzJmjzZs3KzAwULGxsUpPTzfrZGZm6t5779WwYcOK7N/cuXN17Ngx89W7d2+X3Dfyy8jOyyi9cOp9XkYpa5QCAABPQ6AUAAAALjV9+nQ9/PDDGjx4sFq1aqU5c+YoICBAH374YYH1Z86cqe7du+vpp59Wy5Yt9fLLL+vqq6/WW2+9JSk3m3TGjBkaP3687rjjDl111VX66KOPdPToUS1btsy8zksvvaSnnnpKbdq0KbJ/oaGhCg8PN182m81l9w5nGVm5GaW2C6fek1EKAAA8FIFSAAAAuExmZqa2b9+umJgYs8xqtSomJkZxcXEFnhMXF+dUX5JiY2PN+gcPHlRCQoJTnZCQEEVFRRV6zaI8/vjjqlWrlq655hp9+OGHMgyj0LoZGRlKSUlxeqH4CtrMyUZGKQAA8FDe7u4AAAAAKo+TJ08qJydHYWFhTuVhYWHat29fgeckJCQUWD8hIcE87igrrE5xTZo0STfddJMCAgK0evVqPfbYY0pNTdWTTz5ZYP3JkyfrpZdeKlEbyGW3G8rMcaxRej6jlKn3AADAUxEoBQAAQJXxwgsvmH9v37690tLSNHXq1EIDpWPHjtWoUaPM9ykpKYqIiCjzflYGjmxSyXkzJ6beAwAAT8XUewAAALhMrVq15OXlpcTERKfyxMREhYeHF3hOeHh4kfUdf5bkmsUVFRWlP/74QxkZGQUe9/PzU3BwsNOrMrDbDe35M1lHk86VWRuOjZwk50CpP4FSAADgoQiUAgAAwGV8fX3VoUMHrV271iyz2+1au3atoqOjCzwnOjraqb4krVmzxqzfuHFjhYeHO9VJSUnR5s2bC71mce3cuVPVq1eXn5/fZV2novjpjySNWbpL17z2jW6f/b16vbVRmRdkfrqSI6PUy2qRtxdrlAIAAM/H1HsAAAC41KhRozRw4EB17NhR11xzjWbMmKG0tDQNHjxYkjRgwADVr19fkydPliSNGDFCXbt21bRp09SzZ099/PHH2rZtm95//31JksVi0ciRI/XKK6+oWbNmaty4sV544QXVq1dPvXv3NtuNj4/XqVOnFB8fr5ycHO3cuVOS1LRpUwUFBenLL79UYmKirr32WtlsNq1Zs0avvfaaxowZU66fjzs9umC7jianm+9PpmYo/lSamtap5vK2zu9475yb4Zh6f5aMUgAA4GEIlAIAAMCl+vTpoxMnTmjChAlKSEhQu3bttHLlSnMzpvj4eFmt54NnnTt31uLFizV+/HiNGzdOzZo107Jly9S6dWuzzjPPPKO0tDQNHTpUSUlJ6tKli1auXCmbzWbWmTBhgubPn2++b9++vSTp22+/Vbdu3eTj46O3335bTz31lAzDUNOmTTV9+nQ9/PDDZf2ReISU9CwzSDr/H9do6qp92vNnivYnppZNoDRv6r2fj5dTuWMzp3QySgEAgIchUAoAAACXGz58uIYPH17gsfXr1+cru/fee3XvvfcWej2LxaJJkyZp0qRJhdaZN2+e5s2bV+jx7t27q3v37oUer+wOnUyTJNWu5qeuf6utz3f+qT1/pujA8dQyac8x9d7vooxSx9R7MkoBAICnYY1SAAAAoAo4mBcobVwzUJLUtE6QJOnAibIJlDoyRi8OlAb45uZqsEYpAADwNARKAQAAgCrADJTWyguU1s4LlJZ5RulFU+99c38EYeo9AADwNBVi6r1hGEpPz11PyWazyWKxuLlHlQefLQAAQNXgCJRG1nLOKP3tRKrsdkNWq2vHgY41Sm0+zrkZ/ky9BwAAHqpCZJSmp6erR48e6tGjhxnUg2vw2QIAAFQNF2eUNqwRIF8vq9Kz7Poz6ZzL23Psep8/ozRv6j2BUgAA4GEqRKAUAAAAQOkZhmEGSq+onRso9fayKrJWgKSyWafUnHpfSEYpU+8BAICnIVAKAAAAVHJ/pWXqTHq2LJbcTFIHc/p9GaxTWthmTky9BwAAnopAKQAAAFDJHcrLJq0X4i+bz/mp8E3rVJMk7U8sw4zSfFPvc9+z6z0AAPA0BEoBAACASu73i9YndXBklJbN1PtCMkoJlAIAAA9FoBQAAACo5A4VFiitnRcoPZ4qwzBc2qa5mZPPRRmlee8zs+3Ksbu2TQAAgMtBoBQAAACo5BwbOUVeFCi9onagLBYp+VyWTqZmurTN81PvnX/kCPA9HzglqxQAAHgSAqUAAABAJWfueH9RoNTm46WI6rmbOx1w8YZO5tT7i3a9vzBweo4NnQAAgAchUAoAAABUYna7oUN/FZxRKpXdOqXpWQVv5mSxWMzp9wRKAQCAJyFQCgAAAFRiiWfSlZ5ll7fVogbV/fMddwRKfyurjFLv/D9yBLChEwAA8EAESgEAAIBK7OCJ3GzSiBoB8vHKP/y/cEMnVypsjVIpd8q/RKAUAAB4FgKlAAAAQCV28K+Cd7x3aFInt/x3F0+9d+x6b7to13tJ8s/LKD2bme3SNgEAAC4HgVIAAACgEnNklEbWLDhQWj80dzOnxDMZyrEbLmu3OFPv08koBQAAHoRAKQAAAFCJOTZyaly74EBp7Wp+8rJalGM3dDI1w2Xtmps5FZBRak69z7S7rD0AAIDLRaAUAAAAqMSOJqVLUoEbOUmSl9Wi2kF+kqSE5HSXtVtURqlj13um3gMAAE9CoBQAAACoxBxZoo5gaEHCQ2ySpGMuDZQWvpmTI1DK1HsAAOBJCJQCAAAAlZTdbuivtExJUq2iAqXBuYHSxJSyCJQWNPU+98cQx/R8AAAAT0CgFAAAAKikks5lmRs01Qj0LbRe2WSU5maLOoKiF/JnMycAAOCBCJQCAAAAlZRj2n2Iv498C5gC71A3L1CakHzOZW1nZBWeUeooS88mUAoAADwHgVIAAACgknIESmsFFZ5NKp3PKE1w4dR7R7aoXwEZpTZzjVKm3gMAAM9BoBQAAACopE6mXnp9Uun8GqWu3fW+8M2czq9RSkYpAADwHARKAQAAgErq5Jm8jNJqlwiUXpBRahjGZbdrGMYlNnMioxQAAHgeAqUAAACQJP3+++/u7gJczJx6X8RGTpIUlpdRmp5lV/K5rMtuNzPnfAC0oM2cbN5klAIAAM9DoBQAAACSpKZNm+rGG2/UwoULlZ7uuinYcJ/za5QWnVFq8/FSjbxgqit2vndkk0qXyiglUAoAADwHgVIAAABIknbs2KGrrrpKo0aNUnh4uB555BFt2bLF3d3CZfjLsUbpJabeS+ezSl2xoZNjx3uLRfLxsuQ7bgZK2fUeAAB4EAKlAAAAkCS1a9dOM2fO1NGjR/Xhhx/q2LFj6tKli1q3bq3p06frxIkT7u4iSqi4GaWSVDfEdRs6mTvee1tlsRQUKHVMvWeNUgAA4DkIlAIAAMCJt7e37rrrLi1dulSvv/66Dhw4oDFjxigiIkIDBgzQsWPH3N1FFNP5Xe+LXqNUuiCj1IVT7wuadi9Jfky9BwAAHohAKQAAAJxs27ZNjz32mOrWravp06drzJgx+u2337RmzRodPXpUd9xxh7u7iGIwDEMn3JRRmpF9PqO0IDZvAqUAAMDzeLu7AwAAAPAM06dP19y5c/Xrr7/qtttu00cffaTbbrtNVmtusKtx48aaN2+eIiMj3dtRFMuZjGxl5mV2FidQGp6XUXrMFWuU5rXrWIv0Yky9BwAAnohAKQAAACRJ7777rv7xj39o0KBBqlu3boF16tSpo3//+9/l3DOUhmMjp0BfL/n7FhywvFB4XkZpoisySrMcU+8LySjNC6BmsJkTAADwIARKAQAAIElas2aNGjZsaGaQOhiGoSNHjqhhw4by9fXVwIED3dRDlIS5kVMxdryXzk+9P5Z87rLbduxm7+dTcKDU31yjlIxSAADgOVijFAAAAJKkJk2a6OTJk/nKT506pcaNG7uhR7gcJ88Uf31SSQrLC5SmpGfrbGb2ZbV9PqO0sKn3rFEKAAA8D4FSAAAASMrNHC1IamqqbDZbOfcGl8vMKC3GjveSVM3PW4F5U/Qvd0OnS27mlJdpmm03lJ1DVikAAPAMTL0HAACo4kaNGiVJslgsmjBhggICAsxjOTk52rx5s9q1a+em3qG0TuStUVqzmBmlFotFYSE2/X4iTQnJ6bqidlCp23Zs5nSpNUolKT3briAv8jcAAID7ESgFAACo4n788UdJuRmlu3fvlq/v+QxEX19ftW3bVmPGjHFX91BK5zNKixcolXLXKf39RJoSLnPn+0vten9hADU9K0dBfvxYAgAA3I8RCQAAQBX37bffSpIGDx6smTNnKjg42M09giv8lRcorV3MqfeSFB7sL0k6drlT77OKnnpvsVjk521VRraddUoBAIDHIFAKAAAASdLcuXPd3QW40Mm8qfclySgND8mtm+iijNLCNnOScrNNcwOlrFEKAAA8A4FSAACAKuyuu+7SvHnzFBwcrLvuuqvIup999lk59QquYE69r1aSQGluRunRJBdllPoUvvaozceq5HMioxQAAHgMAqUAAABVWEhIiCwWi/l3VB4nz+QGSmsGFn/qfVheUPXEGVdllBYVKM3NNiVQCgAAPAWBUgAAgCrswun2TL2vPM5l5igtMzcAWZKM0tpmoDTjstov1tR7b0eglKn3AADAMxT+K14AAABUKefOndPZs2fN94cPH9aMGTO0evVqN/YKpeGYdu/rbVW1EuwoXyfYJkk6kZohwzBK3X5Gdm6Q1naJqfcSGaUAAMBzkFEKU48ePUp1nsViUUBAgHx8fBQcHKz4+HhZrVb5+/vLx8dHzzzzjDp37mzWHz58uPbs2SNJ6tatmyIiIrRo0SK1atVKv/zyi2644Qbt3btXI0aM0N69e7Vo0SKnstWrV2v9+vVO5/bv318tW7bUlClTlJWV5dS/gvqwadMmzZw5U7feeqtWr16tESNGOB0vC442C2qrqGOFueeee3Ty5EnVqlVLn376aaH1XnzxRfPzevHFFy/nFgCUQmm+v8vjWp6mMt9bYTZt2qQpU6ZIUr7/p9zljjvu0F133aVHH31USUlJuuaaa+Tr66uTJ09q+vTpGjZsmLu7iGI6ae5472curVActYJyp+ln5RhKOpul6iWYtn8hR5ZoURmlfo6p99kESgEAgGcgoxSXzTAMpaWlKSkpSfHx8ZIku91ulk2bNk3p6bnrXMXHx5tBUklav369Fi5cKLvdrj179shut2v9+vVKTEzUtGnTzGOOsjfeeEPr16/Pd+7ChQs1depUJSUlKS0tzel1cR/S09M1ffp0JSYmatGiRUpMTNT06dPN42XhwjYvbquoY4XZs2ePTp48KUk6efKk02d6ocTERKfPKzEx8fJvBkCxleb7uzyu5Wkq870VJj09XdOmTVNSUlK+/6fcaceOHbr++uslSZ9++qnCw8N1+PBhffTRR5o1a1aJrvX2228rMjJSNptNUVFR2rJlS5H1ly5dqhYtWshms6lNmzZasWKF03HDMDRhwgTVrVtX/v7+iomJ0f79+53qvPrqq+rcubMCAgIUGhpaYDvx8fHq2bOnAgICVKdOHT399NPKzs4u0b1VBOd3vC9ZoNPP20uhAT6ScrNKS8uRUVr0Zk5MvQcAAJ6FQCnK3F9//aXFixdLkh599NF8xwub1vXXX3/lO3bq1KkCzzUMQ6dPny5WHxYtWqS//vpLUm5A9+LjZeHCNi9uq6hjhXniiSeKfO8wfPjwYtUDUDZK8/1dHtfyNJX53gpz4T1LnnPfZ8+eVbVq1SRJq1ev1l133SWr1aprr71Whw8fLvZ1lixZolGjRmnixInasWOH2rZtq9jYWB0/frzA+ps2bVK/fv00ZMgQ/fjjj+rdu7d69+7t9IvAKVOmaNasWZozZ442b96swMBAxcbGOgWYMzMzde+99xaa+ZqTk6OePXsqMzNTmzZt0vz58zVv3jxNmDCh2PdWUTgySmsGFX99UofaeeccT7mMQGnWpTdz8mfqPQAA8DAVIlB6YbAsPT1d586dK/TllI1xGesqlaKTxe6jJ702b95cLh/PokWLNGfOHKd1z8rbokWLtG3bNi1evDhfANYwDC1evFh//PGHy9v9448/nNq8sK2ijhVm6tSpBfZ/6tSpTmUrV67UiRMnnMqOHz+ulStXuuK2AFxCab6/y+NanqYy31thHPd8sUWLFrn9vps2baply5bpyJEjWrVqlW699VZJuf9/BAcHF/s606dP18MPP6zBgwerVatWmjNnjgICAvThhx8WWH/mzJnq3r27nn76abVs2VIvv/yyrr76ar311luScr8uZsyYofHjx+uOO+7QVVddpY8++khHjx7VsmXLzOu89NJLeuqpp9SmTZsC21m9erV++eUXLVy4UO3atVOPHj308ssv6+2331ZmZmax768icOx4X9KMUkmqE5y3oVNq6bOci7WZE7veAwAAD1PsNUozMjKUkXH+t8opKSll0qHC2na48847i3+iPVtS6dZVKjH7+SlbJepjFZGTk6OPP/7Y7X2YNGmSmUV6McMwNHPmTE2ZMqVEa3kVxXHNgspnzJhRYDtF9SMzM1NfffVVgW199dVXGjFihHx9fZWTk5MvcOowdepU3XLLLfLyKvwHFwCXp6jv/ZL+O+PKa3maynxvhXHcW05O/sBQTk6OZsyYoalTp7rtvidMmKD7779fTz31lG6++WZFR0dLyg0wtm/fvljXyMzM1Pbt2zV27FizzGq1KiYmRnFxcQWeExcXp1GjRjmVxcbGmkHQgwcPKiEhQTExMebxkJAQRUVFKS4uTn379i1W3+Li4tSmTRuFhYU5tTNs2DD9/PPPBd6jO8fAl+OvNMfUezdllDqm3heRUerY9d4RVAUAAHC3YmeUTp48WSEhIeYrIiKiLPsFlImUlJRCp/rn5ORo69at5jqrrhAfH6+tW7fm+4E4JydH27ZtK/RYYf2YPHlyke05ji9fvrzAH8Id11++fHlJbgNACRX1vV/Sf2dceS1PU5nvrTCOey7Mtm3b3Hrf99xzj+Lj47Vt2zanGQg333yz3nzzzWJd4+TJk8rJyXEKRkpSWFiYEhISCjwnISGhyPqOP0tyzZK0c2EbF6uoY+BTeYHSGqXYjMnc+f7M5QRKc4OfjqzRgrDrPQAA8DTFzigdO3as02/6U1JSym2g6Od3/jfh//3vf2Wz2Qqtm56efj6j01rs27t8F7R1qT56Crvdrj59+ig1NdXdXSk3wcHBOnPmTIHBUi8vL3Xo0EENGzZ0WXsNGzZUp06dtGPHDqdAgJeXl66++mpJKvBYYf0YO3asvv3220Lbc2Tv3H777Zo1a1aBwVJvb2/dfvvtpb4nAJdW1Pd+Sf+dceW1PE1lvrfCOO65sGBpp06d3H7f4eHhCg8Pdyq75ppr3NQb93PnGPhyJJ3LkiSF+PuU+Fwzo/QyAqWO4GeRGaVMvQcAAB6m2Bmlfn5+Cg4OdnqVlwunn9lsNvn7+xf6cgpQlue0tRL00VNegYGBGjduXLl8PF5eXsWeFleWfZg4caKs1oK/7C0Wi0aMGOHS6Y6OaxZUPnLkyEKPFdYPX19f9ezZs8C2/v73v8vXNzdrxMvLS08//XSB9Z555hmm3QNlrKjv/ZL+O+PKa3maynxvhXHcW0H/Dnt5eWnkyJFuve+0tDS98MIL6ty5s5o2baorrrjC6VUctWrVkpeXlxITE53KExMT8wVgHcLDw4us7/izJNcsSTsXtnExd46BL0fyZQRKzTVKXZBRWtSu937seg8AADxMhdjMCWWnuOuNXa7+/fvr0UcfVUBAQLm0V1gfOnTooPvvvz/fD6EWi0X333+/6tev7/J2GzRo4NTmhW0VdawwTz/9dIH9Hz16tFNZ9+7dVbt2baeyOnXqmBtzAChbpfn+Lo9reZrKfG+Fcdzzxfr37+/2+37ooYf073//W9dff72GDx+uESNGOL2Kw9fXVx06dNDatWvNMrvdrrVr15prnl4sOjraqb4krVmzxqzfuHFjhYeHO9VJSUnR5s2bC71mYe3s3r1bx48fd2onODhYrVq1KvZ1KoIUF2SUnkh1xa73TL0HAAAVB4FSlLlatWqZPxDOmTMn3/HCMmdq1qyZ71iNGjUKPNdqtap69erF6kP//v1Vs2ZN87yLj5eFC9u8uK2ijhVm9uzZRb53cOwWfKl6AMpGab6/y+NanqYy31thLrxnyXPu++uvv9bSpUv1+uuvmzMfShoolaRRo0bpgw8+0Pz587V3714NGzZMaWlpGjx4sCRpwIABTps9jRgxQitXrtS0adO0b98+vfjii9q2bZuGDx8u6fxMjFdeeUVffPGFdu/erQEDBqhevXrq3bu3eZ34+Hjt3LlT8fHxysnJ0c6dO7Vz505zmaFbb71VrVq10oMPPqhdu3Zp1apVGj9+vB5//HGnpZ4qAzOjNKD0GaXHUy5n1/vib+aUzmZOAADAQxAoxWWzWCwKDAxUaGioua6a1Wo1y0aNGmUuidCwYUO1bt3aPLdbt2564IEHZLVa1bp1a1mtVnXr1k1hYWEaPXq0ecxRNmbMGHXr1i3fuf3799fTTz+t0NBQBQYGOr0u7oPNZtOoUaMUFham/v37KywsTE899VSZrit7YZsXt1XUscK0bt1atWrVkpT7w/WFn+mFwsLCnD6vizewAFC2SvP9XR7X8jSV+d4KY7PZNHr0aIWGhub7f8qdqlevnu+XkqXRp08fvfHGG5owYYLatWunnTt3auXKleb/Q/Hx8Tp27JhZv3Pnzlq8eLHef/99tW3bVp9++qmWLVvm9P/bM888oyeeeEJDhw5Vp06dlJqaqpUrVzp9bhMmTFD79u01ceJEpaamqn379mrfvr22bdsmKXd5g+XLl8vLy0vR0dF64IEHNGDAAE2aNOmy79mTGIZhBkpD/Uu+mVPtoNzPNCU9u9TZnsXbzCn32LlMMkoBAIBnsBiFbQF+CSkpKQoJCVFycnKZr9V07tw59ejRQ1JupoO/v3+x6p65+kHJq5DfoudkqdqOBZeuV1wXXO9SffQkJflsAQBAxXap8dvChQv1+eefa/78+W5dLseTlecYuLRSM7LVeuIqSdLeSd3l71uytckNw1DzF1YqM9uu/3vmRkXUKNnXgmEYajJuheyGtGXczaoTXPAvAT7b8YdGfbJL1zerpQVDokrUBgAAQHGVZPxWjtvCAwAAwJNNmzZNv/32m8LCwhQZGSkfH+dfJO/YscNNPUNJJJ3NlCT5elnNdUBLwmKxqHaQn/5MOqcTqRklDpRm2w3Z81Ixil6jlF3vAQCAZyFQCgAAAElyWu8TFZdj2n2wv0+ha8FfSu1quYHS4ykl39Ap44I1R4va9f78Zk6sUQoAADwDgVIAAABIkiZOnOjuLsAFzI2c/Es/1K9TLXdDpxOppQiUXpAh6utVjM2cyCgFAAAegs2cAAAAYEpKStK//vUvjR07VqdOnZKUO+X+zz//dHPPUFwpZqC09Gvw13YESkux870jo9TXyyqrtfCMVpuvY9d7AqUAAMAzkFEKAAAASdJPP/2kmJgYhYSE6NChQ3r44YdVo0YNffbZZ4qPj9dHH33k7i6iGMwd7wNKvuO9Q51quRswlSqjNC9QWtS0e+nCjFKm3gMAAM9ARikAAAAkSaNGjdKgQYO0f/9+2Wzndyq/7bbbtGHDBjf2DCWR7MqM0jMlD5Q6ptIXtZGTdOEapWSUAgAAz0CgFAAAAJKkrVu36pFHHslXXr9+fSUkJLihRygNVwRKHWuUHi9FoNTMKPW+REZp3q73GWSUAgAAD0GgFAAAAJIkPz8/paSk5Cv/3//+p9q1a7uhRyiNpLPnd70vLZdklF5q6n1eoDQzx64cu1HidgAAAFyNQCkAAAAkSb169dKkSZOUlZUbaLNYLIqPj9ezzz6ru+++2829Q3G5JKM0+Hyg1F7CIOa5zNxAaaBv0dsh2C4IpGawoRMAAPAABEoBAAAgSZo2bZpSU1NVu3ZtnTt3Tl27dlXTpk1VrVo1vfrqq+7uHorJFYHSmoG5gdJsu6GkvOsVV1pmtiQpwPcSa5ResIYpGzoBAABPwK73AAAAkCSFhIRozZo12rhxo3bt2qXU1FRdffXViomJcXfXUAIpjl3vLyNQ6uttVY1AX51Ky9TxM+mqEehb7HPPZuRllPoV/aOG1WqRr5dVmTl2NnQCAAAegUApAAAAZLfbNW/ePH322Wc6dOiQLBaLGjdurPDwcBmGIYvF4u4uopjMjNKA0gdKJal2kJ9OpWXqxJkMtQgv/nnFzSiVctcxJVAKAAA8BVPvAQAAqjjDMNSrVy899NBD+vPPP9WmTRtdeeWVOnz4sAYNGqQ777zT3V1ECbhi6r10fkOn4ykl29DpbDHXKJXOb+jE1HsAAOAJyCgFAACo4ubNm6cNGzZo7dq1uvHGG52OrVu3Tr1799ZHH32kAQMGuKmHKC673XBZoLROXqD0RGrJAqVpGXkZpX6Xzih1bOh0joxSAADgAcgoBQAAqOL+85//aNy4cfmCpJJ000036bnnntOiRYvc0DOUVGpmthyb1FeIjNK8DZ0yCJQCAAAPQKAUAACgivvpp5/UvXv3Qo/36NFDu3btKsceobSSz+Zmk/p6W81p7aXlCJSeLNOM0ryp99kESgEAgPsRKAUAAKjiTp06pbCwsEKPh4WF6fTp0+XYI5RWsgt2vHeoFVS6QGlJMkr9WaMUAAB4EAKlAAAAVVxOTo68vQsPanl5eSk7O7sce4TSSnHR+qRS6QOlJd31XhK73gMAAI/AZk4AAABVnGEYGjRokPz8/Ao8npFRskAZ3MdVGzlJUq1qvpKkk6mZJTrvbEZeRqkfu94DAICKhUApAABAFTdw4MBL1mHH+4rBpYHSvIzS02czlZ1jl7dX8SajlSSj9HyglIxSAADgfgRKAQAAqri5c+e6uwtwkSQXBkqrB/jKapHshnQqLVN1gm3FOs9co7Q4GaXeeVPv2cwJAAB4ANYoBQAAACoJR0ZpsAsCpV5Wi2oE5maVnijBOqXmrvclyihl6j0AAHA/AqUAAABAJeHKqfeSVCuo5OuUlmTXe1veZk4ZTL0HAAAegEApAAAAUEk4AqWhAa4JlNaulptRevJM8TJKDcM4v0apH2uUAgCAioVAKQAAAFBJpLg8ozQvUFrMqffpWXYZRu7fi5dRytR7AADgOQiUAgAAAJVEWU29P1HMjFJHNqkk+ftcOqPUj82cAACAByFQCgAAAFQSrg+Uliyj9GxGbsAzwNdLVqvlkvWZeg8AADwJgVIAAACgkkg6W1aB0uJt5mSuT1qMaffS+UDpOabeAwAAD0CgFAAAAKgE7HZDKekuDpRWK2FGaV6gNLAYGzlJ53e9J6MUAAB4AgKlAAAAQCVwJiPb3Egp2MVrlBY3UJpmTr0vXkapYx3TDAKlAADAAxAoBQAAACoBx473Nh+rOaX9ctXOm3p/Ki1TOXbjkvXPmlPvi5tRyq73AADAcxAoBQAAACoBV2/kJEk1An1lsUh2IzdYeilnM89v5lQc5tR7dr0HAAAegEApAAAAUAmURaDU28uq6gHFn36flhcoDSzm1Hs/b3a9BwAAnoNAKQAAAFAJuHrHe4eSrFN6NiNv6n2xN3Ni6j0AAPAcBEoBAACASqAsMkolqVbeOqVlkVHKrvcAAMCTECgFAAAAKgFHoNRVO947mIHSM8VYo7SUGaUZ2XYZxqU3iwIAAChLBEoBAACASiDpXG4g07GmqKuUbUbp+YBqRjbT7wEAgHsRKAUAAAAqgaS03IzSUFdnlFbLDbyeKM4apZl5GaXF3fXe+/yPI0y/BwAA7la8X/W6mc1m09dff23+Ha7DZwsAAFA5ODJKQwPLKqP00lPv0zLyMkr9ivdjhreXVd5Wi7LtBhs6AQAAt6sQgVKLxSJ/f393d6NS4rMFAACoHE7n7XpfPcC1GaW1zTVKXZ9RKuVOv0/NyCajFAAAuB1T7wEAAIBKIPmsY+p9xVmjVJLMne+zCZQCAAD3IlAKAAAAVAKnz+ZNvXdxRqljjdK/0jJltxe9M31Jd72XJD/v3LpMvQcAAO5GoBQAAACo4AzDUNK5vIxSFwdKawbmZpTm2M+3UZizpcgo9c+bpn8uk4xSAADgXgRKAQAAgAruXFaOMrNzMzKrB7h26r2vt1Uh/rnB10tNv0/LW6M0sAQZpcG23KBq8iWCsAAAAGWNQCkAAABc7u2331ZkZKRsNpuioqK0ZcuWIusvXbpULVq0kM1mU5s2bbRixQqn44ZhaMKECapbt678/f0VExOj/fv3O9U5deqU+vfvr+DgYIWGhmrIkCFKTU01jx86dEgWiyXf64cffnDdjbtJUt76pD5elhJtpFRctYJyg6+X2tDpbN6u9wElyCitkZexeiots5S9AwAAcA0CpQAAAHCpJUuWaNSoUZo4caJ27Nihtm3bKjY2VsePHy+w/qZNm9SvXz8NGTJEP/74o3r37q3evXtrz549Zp0pU6Zo1qxZmjNnjjZv3qzAwEDFxsYqPT3drNO/f3/9/PPPWrNmjZYvX64NGzZo6NCh+dr75ptvdOzYMfPVoUMH138I5ez8+qS+slgsLr++Y0OnE0VklGZm25WZk5vVWpKp9zUCc7NVT6VderMoAACAskSgFAAAAC41ffp0Pfzwwxo8eLBatWqlOXPmKCAgQB9++GGB9WfOnKnu3bvr6aefVsuWLfXyyy/r6quv1ltvvSUpN5t0xowZGj9+vO644w5dddVV+uijj3T06FEtW7ZMkrR3716tXLlS//rXvxQVFaUuXbpo9uzZ+vjjj3X06FGn9mrWrKnw8HDz5ePj2jU93eH8jvdlcy+1q+UFSovIKL1wjVH/EmS1ns8oZeo9AABwLwKlAAAAcJnMzExt375dMTExZpnValVMTIzi4uIKPCcuLs6pviTFxsaa9Q8ePKiEhASnOiEhIYqKijLrxMXFKTQ0VB07djTrxMTEyGq1avPmzU7X7tWrl+rUqaMuXbroiy++KPJ+MjIylJKS4vTyRKfzAqWuXp/UITzYJklKTEkvtI5jfVJfL6t8vYv/Y0bNwNw+k1EKAADcjUApAAAAXObkyZPKyclRWFiYU3lYWJgSEhIKPCchIaHI+o4/L1WnTp06Tse9vb1Vo0YNs05QUJCmTZumpUuX6quvvlKXLl3Uu3fvIoOlkydPVkhIiPmKiIi41EfgFo6p9yEu3vHeITwkN1B6LLnwQOnZvEBpQAk2cpKk6nmB0r9YoxQAALhZ8RcPAgAAACqwWrVqadSoUeb7Tp066ejRo5o6dap69epV4Dljx451OiclJcUjg6WOHeOrl1GgtG6IvyQpoYhAaVreRk4lWZ9UujCjlEApAABwLzJKAQAA4DK1atWSl5eXEhMTncoTExMVHh5e4Dnh4eFF1nf8eak6F28WlZ2drVOnThXariRFRUXpwIEDhR738/NTcHCw08sTnU47v5lTWagbeumMUsfU+4ASrE8qSTXyAqWnCZQCAAA3I1AKAAAAl/H19VWHDh20du1as8xut2vt2rWKjo4u8Jzo6Gin+pK0Zs0as37jxo0VHh7uVCclJUWbN28260RHRyspKUnbt28366xbt052u11RUVGF9nfnzp2qW7duyW/UwyTlZZSGlllG6fk1Su12o8A6Z/MySgP8SpZRWuOCqfeGUfC1AQAAygNT7wEAAOBSo0aN0sCBA9WxY0ddc801mjFjhtLS0jR48GBJ0oABA1S/fn1NnjxZkjRixAh17dpV06ZNU8+ePfXxxx9r27Ztev/99yVJFotFI0eO1CuvvKJmzZqpcePGeuGFF1SvXj317t1bktSyZUt1795dDz/8sObMmaOsrCwNHz5cffv2Vb169SRJ8+fPl6+vr9q3by9J+uyzz/Thhx/qX//6Vzl/Qq6XlLdGaah/2WSU1g7yk9UiZdsNnUzNUJ28zZ0u5MgoDSxlRmlGtl1nM3MUWMJAKwAAgKswCgEAAIBL9enTRydOnNCECROUkJCgdu3aaeXKleZmTPHx8bJaz09s6ty5sxYvXqzx48dr3LhxatasmZYtW6bWrVubdZ555hmlpaVp6NChSkpKUpcuXbRy5UrZbOcDdosWLdLw4cN18803y2q16u6779asWbOc+vbyyy/r8OHD8vb2VosWLbRkyRLdc889ZfyJlL2ks2W7Rqm3l1V1qtmUkJKuY8npBQZKz2bmZZSWMFAa4OslP2+rMrLtOpWWSaAUAAC4DaMQAAAAuNzw4cM1fPjwAo+tX78+X9m9996re++9t9DrWSwWTZo0SZMmTSq0To0aNbR48eJCjw8cOFADBw4svNMVWFnvei9J4SHnA6VtC9jP6nygtGQ/YlgsFtUI9NWx5HSdSstURI0AV3QXAACgxFijFAAAAKjgzu96XzZT76Xz65QmJJ8r8PjZjLyp934lyyiVzk+/P8WGTgAAwI0IlAIAAAAVmGEY5tT7strMScrNKJWkYynpBR5PK2VGqeS8oRMAAIC7ECgFAAAAKrDUjGxl5+1EXz4ZpQUHSs+WcjMnSaqZFyg9TaAUAAC4EYFSAAAAoAJzZJPafKyy+ZQ8SFlc4SH+kqRjhQRK0zLyMkpLsRlTdTJKAQCAByBQCgAAAFRg5rR7/7LLJpXKJ6P0VFpGKXsHAABw+QiUAgAAABWYY8f7slyfVJLCg88HSg3DyHf88tYo9ZPEZk4AAMC9CJQCAAAAFVjSubLfyEmSwoJtslikzBx7gQHNy9v1PrfvBEoBAIA7ESgFAAAAKrCkvIzSstzISZJ8va2qFZSb+VnQOqVklAIAgIqOQCkAAABQgZ1OK5+MUqnodUrNNUpLlVHKZk4AAMD9CJQCAAAAFVjSOccapWWbUSqdX6f0WPK5fMfMXe9LlVGa2/cz6dnKzLZfRg8BAABKj0ApAAAAUIGd3/W+/DJKC5p6f37X+5IHSkP9fWS15P7dsZQAAABAeSNQCgAAAFRg5bVGqSSFh/hLyj/13m43dNaxRmkppt5brRaz/0y/BwAA7kKgFAAAAKjATudllIaU4xqlF2eUnsvKMf9emoxS6fz0ezZ0AgAA7kKgFAAAAKjAks/lBkrLJ6M0bzOnFOdAqSO46eNlkc2ndD9iVCdQCgAA3IxAKQAAAFCBnT7r2MypPDNKz8kwDLN877EUSVKzOtVksVhKde2aBEoBAICbESgFAAAAKii73TAzSssjUBqWt+t9epbdbFeSfj6aGyhtVS+41Nd2TL1njVIAAOAuBEoBAACACiolPUuOxM5Q/7Kfem/z8TIzPy9cp/SXvIzSVnUvP1B6Ki3jMnoIAABQegRKAQAAgAoqKW8jp0BfL/l6l8/Qvm5oblbp4b/SzLJf8jJKr3RBRunptKxL1AQAACgbBEoBAACACur8+qRln03q0C4iVJL0w++nJEnJZ7P0Z9I5SVJLl0y9J6MUAAC4B4FSAAAAoIJKKsf1SR2ua1JLkrTpt5OSzk+7j6jhr2Bb6ftRg82cAACAmxEoBQAAACqoxLx1QmtX8yu3Nq+9oqYk6X+JqTpxJkM/H02WdHnrk0oESgEAgPsRKAUAAAAqqMOnzkqSGtYIKLc2qwf6mkHRuN//umAjp5DLum7NwNxg7+mzWbLbjcvrJAAAQCkQKAUAAAAqqHg3BEolqXOT3KzSTQdOmhs5tbqM9UklqXpg7rT9HLuh5HNs6AQAAMofgVIAAACggjqSFyiNKOdA6XVNc9cp3fC/EzpwPFXS5e14L0l+3l6qH+ovSdp++PTldRAAAKAUCJQCAAAAFZQjo7RRzfINlHZqXENeVouOJqcr224oNMBHdUNsl33dmJZ1JEmrfk647GsBAACUVKUOlFrs2VJOVuEvh6LqFPNlsWe770YBAABQ5SSfy1LS2dwxbUT18g2UBvl5q22D82uStqobLIvFctnXjW0dLkn6Zm+isnPsl309AACAkvB2dwfKUtDO/xSrXrVdH5dxTwAAAADXcky7rxXkq0C/8h/WX9e0lnbEJ0m6/B3vHa6JrKHqAT46fTZLWw6dUucmtVxyXQAAgOKo1BmlAAAAQGXlrvVJHaLzNnSSLn8jJwdvL6tiWoZJklbtYfo9AAAoX5Uuo9Rms+nrr7++ZD3DMJSRkSFJ8vPzc8lUoQv7AAAAAJQld+1473B1w+ry9/HSuawcXXXBNPzL1b11uJZu/0Orfk7UxL9fKavVdeN0AACAolS6QKnFYpG/v3+x6gYEuGdQCQAAAFyuw24OlNp8vPRO/6t1/Ey6mtap5rLrXte0lgJ9vZSQkq6f/kxWu4hQl10bAACgKEy9BwAAACogd0+9l6QbW9RRn04NXXpNm4+XurWoI0la9TPT7wEAQPkhUAoAAABUQO6eel+WYq8MlyQt/+moMrJz3NwbAABQVRAoBQAAACqY7By7/jx9TpLUqGblC5Te1KKOagT66sipc5qy8ld3dwcAAFQRBEoBAACACuZYcrqy7YZ8vawKq1b5NhIN8vPW1HuukiT9+/uD+nbfcTf3CAAAVAUESgEAAIAKxrE+aYMa/pV2V/ibW4ZpUOdISdLopbt0PCXdvR0CAACVHoFSAAAAoIKpzOuTXmjsbS3Uqm6wTqVlavjiH5WexXqlAACg7BAoBQAAACqYw1UkUOrn7aXZ97dXNT9vbTl0SiM/3qkcu+HubgEAgEqKQCkAAABQwVSVjFJJalI7SO8P6ChfL6tW/pygiV/skWEQLAUAAK5HoBQAAACoYBxrlEZUgUCpJEU3qakZfdvJYpEW/hCvl5fvVXaO3d3dAgAAlQyBUgAAAKCCcWSUNqpZNQKlknRbm7p6qdeVkqQPNx7UA//erBNnMtzcKwAAUJkQKAUAAAAqkORzWUo6myVJiqhedQKlkjQgOlLv9r9aQX7e+uH3U+o56//09e5jTMUHAAAuQaAUAAAAqEC2HjwlSapTzU+Bft5u7k3569Gmrj4ffp2a1QnS8TMZGrZoh3q9tVHrfz1OwBQAAFwWAqUAAABABTJ300FJ0p3t67u5J+7TpHaQPh9+nZ68uZkCfb20+89kDZq7VTdN+07vrD+g4ynp7u4iAACogAiUAgAAABXEvoQUbTzwl7ysFg3oHOnu7rhVgK+3Rt3yN2145kYN6dJYAb5eOngyTVNW/qrof67TQ/O3ac0viWz6BAAAiq3qzdUBAAAAKqi53x+SJHW/Mlz1Q/3d2xkPUTPITy/c3kqjbvmbvvrpmD7eGq8d8Un6Zm+ivtmbqFpBvurRuq56XlVXnSJryMtqcXeXAQCAhyKjFAAAAC739ttvKzIyUjabTVFRUdqyZUuR9ZcuXaoWLVrIZrOpTZs2WrFihdNxwzA0YcIE1a1bV/7+/oqJidH+/fud6pw6dUr9+/dXcHCwQkNDNWTIEKWmpjrV+emnn3T99dfLZrMpIiJCU6ZMcc0Nl4O/UjP0351/SpL+0SXSvZ3xQIF+3rqvU4Q+e+w6rXnqBj18fWPVCPTVydRMLfjhsPq+/4PavbRa/f/1g6as3Kf//viHth46pWPJ55SelSPDMGQYhpLPZmnPn8n6fOefmvTlL7rn3U267p/r1OX1dbphyrfq/fZGPf/f3Vq8OV67jiQpPSvH3bcOAABchIxSAAAAuNSSJUs0atQozZkzR1FRUZoxY4ZiY2P166+/qk6dOvnqb9q0Sf369dPkyZN1++23a/Hixerdu7d27Nih1q1bS5KmTJmiWbNmaf78+WrcuLFeeOEFxcbG6pdffpHNZpMk9e/fX8eOHdOaNWuUlZWlwYMHa+jQoVq8eLEkKSUlRbfeeqtiYmI0Z84c7d69W//4xz8UGhqqoUOHlt8HVEr/2RKvzGy72jYI0dUNq7u7Ox6tWVg1Pd+zlZ7p3kIbD5zUVz8d06qfE5SSnq2NB/7SxgN/5TvH22qRr7dVZzOLDnzGnzqrnUeSzPdeVoua1g5S41qBql/dXw2q+6t+qH/u30MDFOzvLYuFLFYAACoCi1HKrSFTUlIUEhKi5ORkBQcHu7pfAAAAcLHyGr9FRUWpU6dOeuuttyRJdrtdEREReuKJJ/Tcc8/lq9+nTx+lpaVp+fLlZtm1116rdu3aac6cOTIMQ/Xq1dPo0aM1ZswYSVJycrLCwsI0b9489e3bV3v37lWrVq20detWdezYUZK0cuVK3Xbbbfrjjz9Ur149vfvuu3r++eeVkJAgX19fSdJzzz2nZcuWad++fcW6N3eMgf+XeEYfbzmiJVvjlZaZo5l92+mOdlV3I6fSys6x63+Jqdr1R5J++iNJh06e1ZHTZ3UsOV05ducfiWoF+alhDX+1qR+ithGhiqwVKIsku2HoaFK6fj6aop+PJuvnoyk6lZZZZLtBft6qH+qvuqE2VQ/wVYi/jwJ8veRo0fHTmCFDuqAbPl5W+Xlb5edjlc3HS37eVvl6W2VRbtC1oNirIyBrcSqTLMoNAtt8rPL38ZLNx0u2vOs6Xn7eVnlbLR4b1LXbDaVmZutMerbOpGcpNT1bZzKyZbcb8rJa5ONllZfVIm+rJe9Pa175Be+9zh/3uei9J987AODylGT8VuqMUkd8NSUlpbSXAAAAQDlyjNtK+XvyYsnMzNT27ds1duxYs8xqtSomJkZxcXEFnhMXF6dRo0Y5lcXGxmrZsmWSpIMHDyohIUExMTHm8ZCQEEVFRSkuLk59+/ZVXFycQkNDzSCpJMXExMhqtWrz5s268847FRcXpxtuuMEMkjraef3113X69GlVr54/SzMjI0MZGRnm++TkZEnlMwY+k56lRxds164/ks2yNvVDdF3DQMbgpdQgSGrQIlQ9W4SaZXa7obTMbKVlZCsj26461Wzy9/Uq9BrNqgepa+MgSfVkGIYSU9K1L+GM/jx9VkeT03U06ZyOJqXrWNI5nTqbpZSM3K+XvfFlf3+u4GW1yGq1yNua+3dvi0VeeUFHL6tF3l4WeV0QUDSU+2/KhYFfQ0bunxf8U3NxHTnqFFDmuK7dMJRtN5SZYygru+w35bJaJC8va4H37p0XVLW6Y43bsvsn2+OaLcv/nwpts9xbdP7eKNd23XC37rhXd32+7uCu75kcu6HsHLuyDUPZOUbue7shL4tFFsf/IZbc/zccf1qtFlkt5/8fqR7gqwUPRZVLn0syBi51oPTMmTOSpIiIiNJeAgAAAG5w5swZhYSElMm1T548qZycHIWFhTmVh4WFFZq1mZCQUGD9hIQE87ijrKg6F0/r9/b2Vo0aNZzqNG7cON81HMcKCpROnjxZL730Ur5yd42Bj0iq/bRbmgYAAHCpkNHl215xxsClDpTWq1dPR44cUbVq1cplikJKSooiIiJ05MgRpvp7OJ5VxcLzqjh4VhUHz6piqUrPyzAMnTlzRvXq1XN3VyqMsWPHOmW72u12nTp1SjVr1vS4abpV6Wu5ouHZeC6ejWfj+Xguno3n4tnkV5IxcKkDpVarVQ0aNCjt6aUWHBzMg64geFYVC8+r4uBZVRw8q4qlqjyvssokdahVq5a8vLyUmJjoVJ6YmKjw8PACzwkPDy+yvuPPxMRE1a1b16lOu3btzDrHjx93ukZ2drZOnTrldJ2C2rmwjYv5+fnJz8/PqSw0NLTAup6iqnwtV0Q8G8/Fs/FsPB/PxbPxXDwbZ8UdA1vLuB8AAACoQnx9fdWhQwetXbvWLLPb7Vq7dq2io6MLPCc6OtqpviStWbPGrN+4cWOFh4c71UlJSdHmzZvNOtHR0UpKStL27dvNOuvWrZPdbldUVJRZZ8OGDcrKynJqp3nz5gVOuwcAAEDVQqAUAAAALjVq1Ch98MEHmj9/vvbu3athw4YpLS1NgwcPliQNGDDAabOnESNGaOXKlZo2bZr27dunF198Udu2bdPw4cMl5e7kPXLkSL3yyiv64osvtHv3bg0YMED16tVT7969JUktW7ZU9+7d9fDDD2vLli3auHGjhg8frr59+5rTrO6//375+vpqyJAh+vnnn7VkyRLNnDkz30ZSAAAAqJpKPfW+vPn5+WnixIn5pj7B8/CsKhaeV8XBs6o4eFYVC8/L9fr06aMTJ05owoQJSkhIULt27bRy5Upz46T4+HhZred/X9+5c2ctXrxY48eP17hx49SsWTMtW7ZMrVu3Nus888wzSktL09ChQ5WUlKQuXbpo5cqVstlsZp1FixZp+PDhuvnmm2W1WnX33Xdr1qxZ5vGQkBCtXr1ajz/+uDp06KBatWppwoQJGjp0aDl8KmWPr2XPxbPxXDwbz8bz8Vw8G8/Fs7k8FsMwDHd3AgAAAAAAAADcian3AAAAAAAAAKo8AqUAAAAAAAAAqjwCpQAAAAAAAACqPAKlAAAAAAAAAKq8ChMoffvttxUZGSmbzaaoqCht2bLF3V2q1CZPnqxOnTqpWrVqqlOnjnr37q1ff/3VqU56eroef/xx1axZU0FBQbr77ruVmJjoVCc+Pl49e/ZUQECA6tSpo6efflrZ2dlOddavX6+rr75afn5+atq0qebNm1fWt1ep/fOf/5TFYtHIkSPNMp6VZ/nzzz/1wAMPqGbNmvL391ebNm20bds287hhGJowYYLq1q0rf39/xcTEaP/+/U7XOHXqlPr376/g4GCFhoZqyJAhSk1Ndarz008/6frrr5fNZlNERISmTJlSLvdXWeTk5OiFF15Q48aN5e/vryZNmujll1/WhXsg8qzcY8OGDfr73/+uevXqyWKxaNmyZU7Hy/O5LF26VC1atJDNZlObNm20YsUKl98vUFyMl8sW4+OKg/GwZ2Hs65kY63oWxrcexKgAPv74Y8PX19f48MMPjZ9//tl4+OGHjdDQUCMxMdHdXau0YmNjjblz5xp79uwxdu7cadx2221Gw4YNjdTUVLPOo48+akRERBhr1641tm3bZlx77bVG586dzePZ2dlG69atjZiYGOPHH380VqxYYdSqVcsYO3asWef33383AgICjFGjRhm//PKLMXv2bMPLy8tYuXJlud5vZbFlyxYjMjLSuOqqq4wRI0aY5Twrz3Hq1CmjUaNGxqBBg4zNmzcbv//+u7Fq1SrjwIEDZp1//vOfRkhIiLFs2TJj165dRq9evYzGjRsb586dM+t0797daNu2rfHDDz8Y//d//2c0bdrU6Nevn3k8OTnZCAsLM/r372/s2bPH+M9//mP4+/sb7733Xrneb0X26quvGjVr1jSWL19uHDx40Fi6dKkRFBRkzJw506zDs3KPFStWGM8//7zx2WefGZKM//73v07Hy+u5bNy40fDy8jKmTJli/PLLL8b48eMNHx8fY/fu3WX+GQAXY7xc9hgfVwyMhz0LY1/PxVjXszC+9RwVIlB6zTXXGI8//rj5Picnx6hXr54xefJkN/aqajl+/Lghyfjuu+8MwzCMpKQkw8fHx1i6dKlZZ+/evYYkIy4uzjCM3G90q9VqJCQkmHXeffddIzg42MjIyDAMwzCeeeYZ48orr3Rqq0+fPkZsbGxZ31Klc+bMGaNZs2bGmjVrjK5du5oDQ56VZ3n22WeNLl26FHrcbrcb4eHhxtSpU82ypKQkw8/Pz/jPf/5jGIZh/PLLL4YkY+vWrWadr7/+2rBYLMaff/5pGIZhvPPOO0b16tXN5+dou3nz5q6+pUqrZ8+exj/+8Q+nsrvuusvo37+/YRg8K09x8UCyPJ/LfffdZ/Ts2dOpP1FRUcYjjzzi0nsEioPxcvljfOx5GA97Hsa+nouxrudifOteHj/1PjMzU9u3b1dMTIxZZrVaFRMTo7i4ODf2rGpJTk6WJNWoUUOStH37dmVlZTk9lxYtWqhhw4bmc4mLi1ObNm0UFhZm1omNjVVKSop+/vlns86F13DU4dmW3OOPP66ePXvm+zx5Vp7liy++UMeOHXXvvfeqTp06at++vT744APz+MGDB5WQkOD0WYeEhCgqKsrpeYWGhqpjx45mnZiYGFmtVm3evNmsc8MNN8jX19esExsbq19//VWnT58u69usFDp37qy1a9fqf//7nyRp165d+v7779WjRw9JPCtPVZ7PhX8X4SkYL7sH42PPw3jY8zD29VyMdSsOxrfly+MDpSdPnlROTo7Tf1iSFBYWpoSEBDf1qmqx2+0aOXKkrrvuOrVu3VqSlJCQIF9fX4WGhjrVvfC5JCQkFPjcHMeKqpOSkqJz586Vxe1USh9//LF27NihyZMn5zvGs/Isv//+u9599101a9ZMq1at0rBhw/Tkk09q/vz5ks5/3kX9m5eQkKA6deo4Hff29laNGjVK9ExRtOeee059+/ZVixYt5OPjo/bt22vkyJHq37+/JJ6VpyrP51JYHZ4byhvj5fLH+NjzMB72TIx9PRdj3YqD8W358nZ3B+D5Hn/8ce3Zs0fff/+9u7uCAhw5ckQjRozQmjVrZLPZ3N0dXILdblfHjh312muvSZLat2+vPXv2aM6cORo4cKCbe4cLffLJJ1q0aJEWL16sK6+8Ujt37tTIkSNVr149nhUAVHGMjz0L42HPxdjXczHWBQrm8RmltWrVkpeXV74dCRMTExUeHu6mXlUdw4cP1/Lly/Xtt9+qQYMGZnl4eLgyMzOVlJTkVP/C5xIeHl7gc3McK6pOcHCw/P39XX07ldL27dt1/PhxXX311fL29pa3t7e+++47zZo1S97e3goLC+NZeZC6deuqVatWTmUtW7ZUfHy8pPOfd1H/5oWHh+v48eNOx7Ozs3Xq1KkSPVMU7emnnzZ/096mTRs9+OCDeuqpp8xMFZ6VZyrP51JYHZ4byhvj5fLF+NjzMB72XIx9PRdj3YqD8W358vhAqa+vrzp06KC1a9eaZXa7XWvXrlV0dLQbe1a5GYah4cOH67///a/WrVunxo0bOx3v0KGDfHx8nJ7Lr7/+qvj4ePO5REdHa/fu3U7frGvWrFFwcLD5n2V0dLTTNRx1eLbFd/PNN2v37t3auXOn+erYsaP69+9v/p1n5Tmuu+46/frrr05l//vf/9SoUSNJUuPGjRUeHu70WaekpGjz5s1OzyspKUnbt28366xbt052u11RUVFmnQ0bNigrK8uss2bNGjVv3lzVq1cvs/urTM6ePSur1fm/SS8vL9ntdkk8K09Vns+FfxfhKRgvlw/Gx56L8bDnYuzruRjrVhyMb8uZu3eTKo6PP/7Y8PPzM+bNm2f88ssvxtChQ43Q0FCnHQnhWsOGDTNCQkKM9evXG8eOHTNfZ8+eNes8+uijRsOGDY1169YZ27ZtM6Kjo43o6GjzeHZ2ttG6dWvj1ltvNXbu3GmsXLnSqF27tjF27Fizzu+//24EBAQYTz/9tLF3717j7bffNry8vIyVK1eW6/1WNhfu8mkYPCtPsmXLFsPb29t49dVXjf379xuLFi0yAgICjIULF5p1/vnPfxqhoaHG559/bvz000/GHXfcYTRu3Ng4d+6cWad79+5G+/btjc2bNxvff/+90axZM6Nfv37m8aSkJCMsLMx48MEHjT179hgff/yxERAQYLz33nvler8V2cCBA4369esby5cvNw4ePGh89tlnRq1atYxnnnnGrMOzco8zZ84YP/74o/Hjjz8akozp06cbP/74o3H48GHDMMrvuWzcuNHw9vY23njjDWPv3r3GxIkTDR8fH2P37t3l92EAeRgvlz3GxxUL42HPwNjXczHW9SyMbz1HhQiUGoZhzJ4922jYsKHh6+trXHPNNcYPP/zg7i5VapIKfM2dO9esc+7cOeOxxx4zqlevbgQEBBh33nmncezYMafrHDp0yOjRo4fh7+9v1KpVyxg9erSRlZXlVOfbb7812rVrZ/j6+hpXXHGFUxsonYsHhjwrz/Lll18arVu3Nvz8/IwWLVoY77//vtNxu91uvPDCC0ZYWJjh5+dn3Hzzzcavv/7qVOevv/4y+vXrZwQFBRnBwcHG4MGDjTNnzjjV2bVrl9GlSxfDz8/PqF+/vvHPf/6zzO+tMklJSTFGjBhhNGzY0LDZbMYVV1xhPP/880ZGRoZZh2flHt9++22B/0cNHDjQMIzyfS6ffPKJ8be//c3w9fU1rrzySuOrr74qs/sGLoXxctlifFyxMB72HIx9PRNjXc/C+NZzWAzDMMovfxUAAAAAAAAAPI/Hr1EKAAAAAAAAAGWNQCkAAAAAAACAKo9AKQAAAAAAAIAqj0ApAAAAAAAAgCqPQCkAAAAAAACAKo9AKQAAAAAAAIAqj0ApAAAAAAAAgCqPQCkAAAAAAACAKo9AKYAKr1u3bho5cqS7u6H169fLYrEoKSnJ3V0BAABAJccYGABcj0ApAJSCpwxMAQAAgPLCGBhAZUegFAAAAAAAAECVR6AUQKWSkZGhMWPGqH79+goMDFRUVJTWr19vHp83b55CQ0O1atUqtWzZUkFBQerevbuOHTtm1snOztaTTz6p0NBQ1axZU88++6wGDhyo3r17S5IGDRqk7777TjNnzpTFYpHFYtGhQ4fM87dv366OHTsqICBAnTt31q+//lpOdw8AAICqiDEwALgGgVIAlcrw4cMVFxenjz/+WD/99JPuvfdede/eXfv37zfrnD17Vm+88YYWLFigDRs2KD4+XmPGjDGPv/7661q0aJHmzp2rjRs3KiUlRcuWLTOPz5w5U9HR0Xr44Yd17NgxHTt2TBEREebx559/XtOmTdO2bdvk7e2tf/zjH+Vy7wAAAKiaGAMDgGt4u7sDAOAq8fHxmjt3ruLj41WvXj1J0pgxY7Ry5UrNnTtXr732miQpKytLc+bMUZMmTSTlDiwnTZpkXmf27NkaO3as7rzzTknSW2+9pRUrVpjHQ0JC5Ovrq4CAAIWHh+frx6uvvqquXbtKkp577jn17NlT6enpstlsZXPjAAAAqLIYAwOA6xAoBVBp7N69Wzk5Ofrb3/7mVJ6RkaGaNWua7wMCAswBoiTVrVtXx48flyQlJycrMTFR11xzjXncy8tLHTp0kN1uL1Y/rrrqKqdrS9Lx48fVsGHDkt8UAAAAUATGwADgOgRKAVQaqamp8vLy0vbt2+Xl5eV0LCgoyPy7j4+P0zGLxSLDMFzWjwuvb7FYJKnYA0wAAACgJBgDA4DrsEYpgEqjffv2ysnJ0fHjx9W0aVOnV0HTgwoSEhKisLAwbd261SzLycnRjh07nOr5+voqJyfHpf0HAAAASooxMAC4DhmlACqNv/3tb+rfv78GDBigadOmqX379jpx4oTWrl2rq666Sj179izWdZ544glNnjxZTZs2VYsWLTR79mydPn3a/M24JEVGRmrz5s06dOiQgoKCVKNGjbK6LQAAAKBQjIEBwHXIKAVQqcydO1cDBgzQ6NGj1bx5c/Xu3Vtbt24t0dpIzz77rPr166cBAwYoOjpaQUFBio2NdVqIfsyYMfLy8lKrVq1Uu3ZtxcfHl8XtAAAAAJfEGBgAXMNiuHJREgCohOx2u1q2bKn77rtPL7/8sru7AwAAAJQ5xsAAqiKm3gPARQ4fPqzVq1era9euysjI0FtvvaWDBw/q/vvvd3fXAAAAgDLBGBgAmHoPAPlYrVbNmzdPnTp10nXXXafdu3frm2++UcuWLd3dNQAAAKBMMAYGAKbeAwAAAAAAAAAZpQAAAAAAAABAoBQAAAAAAABAlUegFAAAAAAAAECVR6AUAAAAAAAAQJVHoBQAAAAAAABAlUegFAAAAAAAAECVR6AUAAAAAAAAQJVHoBQAAAAAAABAlUegFAAAAAAAAECVR6AUAAAAAAAAQJVHoBQAAAAAAABAlUegFAAAAAAAAECVR6AUAAAAAAAAQJVHoBQAAAAAAABAlUegFAAAAAAAAECVR6AUAAAAAAAAQJVHoBQAAAAAAABAlUegFAAAAAAAAECVR6AUAAAAAAAAQJVHoBQAAAAAAABAlUegFAAAAAAAAECVR6AUAAAAAAAAQJVHoBQAAAAAAABAlUegFAAAAAAAAECVR6AUAAAAAAAAQJVHoBQAAAAAAABAlUegFAAAAADg0V588UVZLJZyaatbt27q1q2b+X79+vWyWCz69NNPy6X9QYMGKTIyslzaKq3U1FQ99NBDCg8Pl8Vi0ciRI93dJQBwCQKlAMrdvHnzZLFYZLPZ9Oeff+Y73q1bN7Vu3doNPQMAAEBZc4wFHS+bzaZ69eopNjZWs2bN0pkzZ1zSztGjR/Xiiy9q586dLrmeK3ly34rjtdde07x58zRs2DAtWLBADz74YKF1IyMjZbFY9MQTT+Q7Vt5BaAC4FAKlANwmIyND//znP93dDQAAALjBpEmTtGDBAr377rtmEG3kyJFq06aNfvrpJ6e648eP17lz50p0/aNHj+qll14qcTBy9erVWr16dYnOKami+vbBBx/o119/LdP2L9e6det07bXXauLEiXrggQfUoUOHS57zwQcf6OjRo+XQOwAoPQKlANymXbt2DJgAAACqqB49euiBBx7Q4MGDNXbsWK1atUrffPONjh8/rl69ejkFRr29vWWz2cq0P2fPnpUk+fr6ytfXt0zbKoqPj4/8/Pzc1n5xHD9+XKGhocWuf+WVVyonJ4ckCQAej0ApALcZN25csQZM2dnZevnll9WkSRP5+fkpMjJS48aNU0ZGhlO9yMhI3X777fr+++91zTXXyGaz6YorrtBHH31UrP6cOXNGI0eOVGRkpPz8/FSnTh3dcsst2rFjh1nHsSzA9u3b1blzZ/n7+6tx48aaM2eO07UyMzM1YcIEdejQQSEhIQoMDNT111+vb7/91qneoUOHZLFY9MYbb+jtt9/WFVdcoYCAAN166606cuSIDMPQyy+/rAYNGsjf31933HGHTp06Vaz7AQAAqGhuuukmvfDCCzp8+LAWLlxolhe0RumaNWvUpUsXhYaGKigoSM2bN9e4ceMk5U7p7tSpkyRp8ODB5jT/efPmSXIe091www0KCAgwz714jVKHnJwcjRs3TuHh4QoMDFSvXr105MgRpzqRkZEaNGhQvnMvvOal+lbQGqVpaWkaPXq0IiIi5Ofnp+bNm+uNN96QYRhO9SwWi4YPH65ly5apdevW8vPz05VXXqmVK1cW/IFf5Pjx4xoyZIjCwsJks9nUtm1bzZ8/3zzumCp/8OBBffXVV2bfDx06VOR1IyMjNWDAgGInSfz444/q0aOHgoODFRQUpJtvvlk//PCDUx3HEg4bN27UqFGjVLt2bQUGBurOO+/UiRMninW/RX0NXXi/S5YsueSz/7//+z/de++9atiwofz8/BQREaGnnnoqXyb0oEGDFBQUpPj4eN1+++0KCgpS/fr19fbbb0uSdu/erZtuukmBgYFq1KiRFi9eXKx7AeAaBEoBuE3jxo2LNWB66KGHNGHCBF199dV688031bVrV02ePFl9+/bNV/fAgQO65557dMstt2jatGmqXr26Bg0apJ9//vmS/Xn00Uf17rvv6u6779Y777yjMWPGyN/fX3v37nWqd/r0ad12223q0KGDpkyZogYNGmjYsGH68MMPzTopKSn617/+pW7duun111/Xiy++qBMnTig2NrbAKVaLFi3SO++8oyeeeEKjR4/Wd999p/vuu0/jx4/XypUr9eyzz2ro0KH68ssvNWbMmEveCwAAQEXlWO+yqOnvP//8s26//XZlZGRo0qRJmjZtmnr16qWNGzdKklq2bKlJkyZJkoYOHaoFCxZowYIFuuGGG8xr/PXXX+rRo4fatWunGTNm6MYbbyyyX6+++qq++uorPfvss3ryySe1Zs0axcTElHhJgOL07UKGYahXr15688031b17d02fPl3NmzfX008/rVGjRuWr//333+uxxx5T3759NWXKFKWnp+vuu+/WX3/9VWS/zp07p27dumnBggXq37+/pk6dqpCQEA0aNEgzZ840+75gwQLVqlVL7dq1M/teu3btS973888/r+zs7EsmSfz888+6/vrrtWvXLj3zzDN64YUXdPDgQXXr1k2bN2/OV/+JJ57Qrl27NHHiRA0bNkxffvmlhg8ffsn+XOpr6ELFefZLly7V2bNnNWzYMM2ePVuxsbGaPXu2BgwYkO96OTk56tGjhyIiIjRlyhRFRkZq+PDhmjdvnrp3766OHTvq9ddfV7Vq1TRgwAAdPHjwkvcDwEUMAChnc+fONSQZW7duNX777TfD29vbePLJJ83jXbt2Na688krDMAxj586dhiTjoYcecrrGmDFjDEnGunXrzLJGjRoZkowNGzaYZcePHzf8/PyM0aNHX7JfISEhxuOPP15kna5duxqSjGnTppllGRkZRrt27Yw6deoYmZmZhmEYRnZ2tpGRkeF07unTp42wsDDjH//4h1l28OBBQ5JRu3ZtIykpySwfO3asIclo27atkZWVZZb369fP8PX1NdLT0y95PwAAAJ7owrFgYUJCQoz27dub7ydOnGhc+OPrm2++aUgyTpw4Ueg1tm7dakgy5s6dm++YY0w3Z86cAo917drVfP/tt98akoz69esbKSkpZvknn3xiSDJmzpxpljVq1MgYOHDgJa9ZVN8GDhxoNGrUyHy/bNkyQ5LxyiuvONW75557DIvFYhw4cMAsk2T4+vo6le3atcuQZMyePTtfWxeaMWOGIclYuHChWZaZmWlER0cbQUFBTvfeqFEjo2fPnkVer6C6gwcPNmw2m3H06FHDMM5/tkuXLjXr9+7d2/D19TV+++03s+zo0aNGtWrVjBtuuMEsc3wdxcTEGHa73Sx/6qmnDC8vL6exdUGK8zVUkmd/9uzZfOdPnjzZsFgsxuHDh82ygQMHGpKM1157zSw7ffq04e/vb1gsFuPjjz82y/ft22dIMiZOnFjkvQBwHTJKAbjVFVdcoQcffFDvv/++jh07lu/4ihUrJCnfb8tHjx4tSfrqq6+cylu1aqXrr7/efF+7dm01b95cv//++yX7Ehoaqs2bN19yOpC3t7ceeeQR872vr68eeeQRHT9+XNu3b5ckeXl5mWtb2e12nTp1StnZ2erYsaPTVH6He++9VyEhIeb7qKgoSdIDDzwgb29vp/LMzEz9+eefl7wfAACAiiooKEhnzpwp9LhjfczPP/9cdru9VG34+flp8ODBxa4/YMAAVatWzXx/zz33qG7duuZ4taysWLFCXl5eevLJJ53KR48eLcMw9PXXXzuVx8TEqEmTJub7q666SsHBwZccD69YsULh4eHq16+fWebj46Mnn3xSqamp+u677y77XsaPH19kVmlOTo5Wr16t3r1764orrjDL69atq/vvv1/ff/+9UlJSnM4ZOnSo07IM119/vXJycnT48OEi+1KSr6HiPHt/f3/z72lpaTp58qQ6d+4swzD0448/5rvmQw895NSX5s2bKzAwUPfdd59Z3rx5c4WGhhbrZxkArkGgFIDbFTVgOnz4sKxWq5o2bepUHh4ertDQ0HwDoIYNG+a7RvXq1XX69GlJuYOvhIQEp1dmZqYkacqUKdqzZ48iIiJ0zTXX6MUXXyxwUFKvXj0FBgY6lf3tb3+TJKf1mebPn6+rrrpKNptNNWvWVO3atfXVV18pOTk53zUv7rcjaBoREVFgueN+AAAAKqPU1FSnwNTF+vTpo+uuu04PPfSQwsLC1LdvX33yySclCprWr1+/RJs2NWvWzOm9xWJR06ZNL7k+5+U6fPiw6tWrl+/zaNmypXn8QpcaDxfVTrNmzWS1OocJCmunNC6VJHHixAmdPXtWzZs3z3esZcuWstvt+dYGvfh+q1evLun8eDk5Odlp7O9Y778kX0PFefbx8fEaNGiQatSooaCgINWuXVtdu3Y1+3Ahm82Wb7mCkJAQNWjQIN9avCEhIYz9gXJEoBSA211xxRV64IEHCh0wSco3YCiMl5dXgeVG3kL3R44cUd26dZ1emzZtkiTdd999+v333zV79mzVq1dPU6dO1ZVXXpnvt/TFsXDhQg0aNEhNmjTRv//9b61cuVJr1qzRTTfdVODgq7B+X+p+AAAAKps//vhDycnJ+X5RfiF/f39t2LBB33zzjR588EH99NNP6tOnj2655Rbl5OQUq50LMwBdpbAxa3H75AqePn50rFX6+uuvu+R6l7rfESNGOI3977rrLkmu+RpyyMnJ0S233GKuY7ps2TKtWbPG3KDr4vE/Y3/AcxEoBeARHFmlFw+YGjVqJLvdrv379zuVJyYmKikpSY0aNSpRO+Hh4VqzZo3Tq23btubxunXr6rHHHtOyZct08OBB1axZU6+++qrTNY4ePaq0tDSnsv/973+SZO5Q+umnn+qKK67QZ599pgcffFCxsbGKiYlRenp6ifoLAABQ1SxYsECSFBsbW2Q9q9Wqm2++WdOnT9cvv/yiV199VevWrdO3334rqfi/aC+ui8ejhmHowIEDTjvUV69eXUlJSfnOvTgbsyR9a9SokY4ePZpvKYJ9+/aZx12hUaNG2r9/f76gnqvbadKkiR544AG99957+ZIkateurYCAAP3666/5ztu3b5+sVmu+GVeX8swzzziN/adNm2Yeu9TXkMOlnv3u3bv1v//9T9OmTdOzzz6rO+64QzExMapXr16J+grA/QiUAvAIFw6YEhISzPLbbrtNkjRjxgyn+tOnT5ck9ezZs0Tt2Gw2xcTEOL2qV6+unJycfFNi6tSpo3r16ikjI8OpPDs7W++99575PjMzU++9955q166tDh06SDr/2+ALf/u7efNmxcXFlai/AAAAVcm6dev08ssvq3Hjxurfv3+h9RzTpy/Url07STLHbo6lkgoKXJbGRx995BSs/PTTT3Xs2DH16NHDLGvSpIl++OEHc2knSVq+fHm+6eIl6dttt92mnJwcvfXWW07lb775piwWi1P7l+O2225TQkKClixZYpZlZ2dr9uzZCgoKMqeRu8L48eOVlZWlKVOmOJV7eXnp1ltv1eeff+40rT0xMVGLFy9Wly5dFBwcXKK2WrVq5TT2d4zXi/M15HCpZ1/Q2N8wDM2cObNEfQXgft6XrgIA5eP555/XggUL9Ouvv+rKK6+UJLVt21YDBw7U+++/r6SkJHXt2lVbtmzR/Pnz1bt3b914440uafvMmTNq0KCB7rnnHrVt21ZBQUH65ptvtHXrVqffOku5a5S+/vrrOnTokP72t79pyZIl2rlzp95//335+PhIkm6//XZ99tlnuvPOO9WzZ08dPHhQc+bMUatWrZSamuqSPgMAAFRkX3/9tfbt26fs7GwlJiZq3bp1WrNmjRo1aqQvvvhCNput0HMnTZqkDRs2qGfPnmrUqJGOHz+ud955Rw0aNFCXLl0k5QYtQ0NDNWfOHFWrVk2BgYGKiopS48aNS9XfGjVqqEuXLho8eLASExM1Y8YMNW3aVA8//LBZ56GHHtKnn36q7t2767777tNvv/2mhQsXOm2uVNK+/f3vf9eNN96o559/XocOHVLbtm21evVqff755xo5cmS+a5fW0KFD9d5772nQoEHavn27IiMj9emnn2rjxo2aMWNGkWvGlpQjSWL+/Pn5jr3yyitas2aNunTposcee0ze3t567733lJGRkS+wejmK8zXkcKln36JFCzVp0kRjxozRn3/+qeDgYP2///f/WFsUqIAIlALwGE2bNi1wwPSvf/1LV1xxhebNm6f//ve/Cg8P19ixYzVx4kSXtR0QEKDHHntMq1ev1meffSa73a6mTZvqnXfe0bBhw5zqVq9eXfPnz9cTTzyhDz74QGFhYXrrrbecBsmDBg1SQkKC3nvvPa1atUqtWrXSwoULtXTpUq1fv95l/QYAAKioJkyYIEny9fVVjRo11KZNG82YMUODBw++ZFCuV69eOnTokD788EOdPHlStWrVUteuXfXSSy+Zm1/6+Pho/vz5Gjt2rB599FFlZ2dr7ty5pQ6Ujhs3Tj/99JMmT56sM2fO6Oabb9Y777yjgIAAs05sbKymTZum6dOna+TIkerYsaOWL1+u0aNHO12rJH2zWq364osvNGHCBC1ZskRz585VZGSkpk6dmu+6l8Pf31/r16/Xc889p/nz5yslJUXNmzfX3LlzNWjQIJe14zB+/HgtXLgw33qgV155pf7v//5PY8eO1eTJk2W32xUVFaWFCxcqKirKZe0X52vI4VLP3sfHR19++aWefPJJTZ48WTabTXfeeaeGDx/utMwXAM9nMVgVGACKrVu3bjp58qT27Nnj7q4AAAAAKEPr16/XjTfeqKVLl+qee+5xd3cAlAPWKAUAAAAAAABQ5REoBQAAAAAAAFDlESgFAAAAAAAAUOWxRikAAAAAAACAKo+MUgAAAAAAAABVHoFSAAAAAAAAAFWed2lPtNvtOnr0qKpVqyaLxeLKPgEAAKAMGIahM2fOqF69erJa+X15aTAGBgAAqFhKMgYudaD06NGjioiIKO3pAAAAcJMjR46oQYMG7u5GhcQYGAAAoGIqzhi41IHSatWqmY0EBweX9jIAAAAoJykpKYqIiDDHcSg5xsAAAAAVS0nGwKUOlDqmGgUHBzNIBAAAqECYMl56jIEBAAAqpuKMgVmcCgAAAAAAAECVR6AUAAAAAAAAQJVHoBQAAAAAAABAlUegFAAAAAAAAECVR6AUAAAAAAAAQJVHoBQAAAAAAABAlUegFAAAAC739ttvKzIyUjabTVFRUdqyZUuR9ZcuXaoWLVrIZrOpTZs2WrFihdNxwzA0YcIE1a1bV/7+/oqJidH+/fud6pw6dUr9+/dXcHCwQkNDNWTIEKWmpjrVWbVqla699lpVq1ZNtWvX1t13361Dhw655J5xnt1uuLsLAAAAJUagFAAAAC61ZMkSjRo1ShMnTtSOHTvUtm1bxcbG6vjx4wXW37Rpk/r166chQ4boxx9/VO/evdW7d2/t2bPHrDNlyhTNmjVLc+bM0ebNmxUYGKjY2Filp6ebdfr376+ff/5Za9as0fLly7VhwwYNHTrUPH7w4EHdcccduummm7Rz506tWrVKJ0+e1F133VV2H0YVtO3QKXV69Rs98Z8f3d0VAACAErEYhlGqX/empKQoJCREycnJCg4OdnW/AAAA4GLlNX6LiopSp06d9NZbb0mS7Ha7IiIi9MQTT+i5557LV79Pnz5KS0vT8uXLzbJrr71W7dq105w5c2QYhurVq6fRo0drzJgxkqTk5GSFhYVp3rx56tu3r/bu3atWrVpp69at6tixoyRp5cqVuu222/THH3+oXr16+vTTT9WvXz9lZGTIas3NF/jyyy91xx13KCMjQz4+Ppe8N8bARdty8JQGzd2is5k5kqRPH41Wx8gabu4VAACoykoyfiOjFAAAAC6TmZmp7du3KyYmxiyzWq2KiYlRXFxcgefExcU51Zek2NhYs/7BgweVkJDgVCckJERRUVFmnbi4OIWGhppBUkmKiYmR1WrV5s2bJUkdOnSQ1WrV3LlzlZOTo+TkZC1YsEAxMTGFBkkzMjKUkpLi9ELBNv/+lxkkDfD1kiS9+c3/3NwrAACA4iNQCgAAAJc5efKkcnJyFBYW5lQeFhamhISEAs9JSEgosr7jz0vVqVOnjtNxb29v1ahRw6zTuHFjrV69WuPGjZOfn59CQ0P1xx9/6JNPPin0fiZPnqyQkBDzFRERcamPoEpKzcjWQx9t09nMHF3frJa+GN5FPl4WbTzwl7YcPOXu7gEAABQLgVIAAABUCQkJCXr44Yc1cOBAbd26Vd999518fX11zz33qLDVqMaOHavk5GTzdeTIkXLudcWw9eApnUnPVv1Qf30woKOa1gnSvR1zg8pvriGrFAAAVAze7u4AAAAAKo9atWrJy8tLiYmJTuWJiYkKDw8v8Jzw8PAi6zv+TExMVN26dZ3qtGvXzqxz8WZR2dnZOnXqlHn+22+/rZCQEE2ZMsWss3DhQkVERGjz5s269tpr8/XNz89Pfn5+xbn1Ku2H3/+SJHVpWks2n9xp94/f2FRLtx1R3O9/6Yff/9K1V9R0ZxcBAAAuiYxSAAAAuIyvr686dOigtWvXmmV2u11r165VdHR0gedER0c71ZekNWvWmPUbN26s8PBwpzopKSnavHmzWSc6OlpJSUnavn27WWfdunWy2+2KioqSJJ09e9bcxMnBy8vL7CNKzxEovbbJ+Y2b6of6q2+nhpKkuRsPuqVfAAAAJUGgFAAAAC41atQoffDBB5o/f7727t2rYcOGKS0tTYMHD5YkDRgwQGPHjjXrjxgxQitXrtS0adO0b98+vfjii9q2bZuGDx8uSbJYLBo5cqReeeUVffHFF9q9e7cGDBigevXqqXfv3pKkli1bqnv37nr44Ye1ZcsWbdy4UcOHD1ffvn1Vr149SVLPnj21detWTZo0Sfv379eOHTs0ePBgNWrUSO3bty/fD6kSOZOepd1/JkuSoho7Z43eeXV9SdL2w6cLXd4AAADAUzD1HgAAAC7Vp08fnThxQhMmTFBCQoLatWunlStXmpsxxcfHO2V2du7cWYsXL9b48eM1btw4NWvWTMuWLVPr1q3NOs8884zS0tI0dOhQJSUlqUuXLlq5cqVsNptZZ9GiRRo+fLhuvvlmWa1W3X333Zo1a5Z5/KabbtLixYs1ZcoUTZkyRQEBAYqOjtbKlSvl7+9fDp9M5bTt0GnZDalRzQDVC3X+HK+sFyxfL6tOpmbqyKlzalgzwE29BAAAuDSLUcpf7aakpCgkJETJyckKDg52db8AAADgYozfLh+fYX6TV+zVext+V5+OEXr9nqvyHe/99kbtPJKkmX3b6Y529d3QQwAAUJWVZPzG1HsAAAAApfbDwVOSpKgrahR4/OqG1SVJOw6fLrc+AQAAlEaFnnpvGIbS09MLLM/IyJCUu1OpxWIpdRs2m+2yzgcAAAAqqzPpWdrjWJ+0kF3t2zcMlTZKPx5JKr+OAQAAlEKFDpSmp6erR48eZdrG119/zZpVAAAAQAG2HT6tHLuhhjUCVD+04DFz+4ahkqRfjqYoPStHNh+vcuwhAABA8TH1HgAAAECp/PD7X5KkawuZdi9J9UP9Vaean7LthnbnZZ8CAAB4ogqdUXqh1Hb9ZFjzbicnS9V2fSxJOtO2r+TlU6JrWezZCtr5H1d3EQAAAKhUfvg9d33SawuZdi9JFotF7RuGatXPifox/rQ6RRYeVAUAAHCnSpNRali9cwOijpfDhWXFfJkBVwAAAAAFys6x65ejuRmiHRsVHfw8v6FTUll3CwAAoNQqTaAUAAAAQPn54/Q5ZeUYsvlY1aB60Wv6t3cESuNPyzCM8ugeAABAiREoBQAAAFBiv59MlSRF1gyU1Wopsm6b+iHytlp0/EyGjiWnl0f3AAAASoxAKQAAAIAS+/1EmiTpitqBl6zr7+ullnWDJeVmlQIAAHgiAqUAAAAASuz3k3mB0lpBxarfvmGoJOn/s3fncVHV+x/H3zPDMiwCIgKiqLiUmmtaXM3SkhuaLbbntVwybbM0U8tySyuvlaatttzUSq9e+5W3vGaaS93Sq6mZS2VaKqYCKrLvzPn9gTMysggKDDCv5+MxD+Cc7znnc74zA18+811+OpJcRREBAABcHBKlAAAAACrs4JkepVEh5+9RKkmXhteTJP1+5jgAAICahkQpAAAAgAqzz1FanqH3ktSyYWHP099PpFdZTAAAABeDRCkAAACACsnIyVdCao6k8vcotSdKjyRlKjuvoMpiAwAAuFAkSgEAAABUyMEz85MG+3kpyNerXMeE+HspwOohmyEdPpVZleEBAABcEBKlAAAAACrk7EJO5etNKkkmk0ktQxl+DwAAai4SpQAAAAAq5I8zic7yDru3c8xTmkiiFAAA1DwkSgEAAABUiH3ofYszic/yYkEnAABQk5EoBQAAAFAhf5woTJRWvEdpYfnfzxwPAABQk5AoBQAAAFBuhmE4epTaE5/lVXSOUsMwKj02AACAi0GiFAAAAEC5nUjLUXpOvswmqWkD3wod2zTYVx5mkzJzCxSfml1FEQIAAFwYEqUAAAAAys2+4n2T+r7y9rBU6FhPi1nNziRXf09k+D0AAKhZSJQCAAAAKDf7sPuKzk9qx4JOAACgpiJRCgAAAKDc/jiT4GxRwflJ7YrOUwoAAFCTkCgFAAAAUG72HqUt6FEKAADqGBKlAAAAAMrt8KlMSVLzC06UFh7HHKUAAKCmIVEKAAAAoFwMw9Cx5CxJUuMgnws6R4szPUrjU7OVnpNfabEBAABcLBKlAAAAAMolNStfGbkFkqSIC0yUBvp4KsTfW9LZ+U4BAABqAhKlAAAAAMrl6JnepMF+XrJ6Wi74PPbh9wcSSZQCAICag0QpAAAAgHKxD7uPCLJe1HlanEmU2uc7BQAAqAlIlAIAAAAol2MpZxKlgRc27N4uMthXknQkiUQpAACoOUiUAgAAACiXY8nZki58flK7pmcSpXEkSgEAQA1CohQAAABAuVzsivd29kTpYRKlAACgBiFRCgAAAKBczs5RWjmJ0hNpOcrKLbjouAAAACoDiVIAAAAA5VJZizkF+niqntVDknTkNL1KAQBAzUCiFAAAAMB55RfYFJ9aOXOUmkwmNWtwZp5SVr4HAAA1BIlSAAAAAOeVkJYjmyF5Wkxq6O990edjQScAAFDTkCgFAAAAcF7Hzwy7Dw+0ymw2XfT5IkmUAgCAGoZEKQAAAIDzOmqfnzTw4obd29GjFAAA1DQkSgEAAACc17HkwvlJG1/k/KR2JEoBAEBNQ6IUAAAAwHmdXfG+chOlR5IyZbMZlXJOAACAi0GiFAAAAMB5VXaiNCLIRxazSTn5Np1Iz6mUcwIAAFwMEqUAAAAAzutYSuHQ+0ZB1ko5n6fFrIgz52L4PQAAqAlIlAIAAKDSvfnmm2revLmsVquio6O1devWMssvX75cbdq0kdVqVYcOHbRq1Sqn/YZhaMqUKWrUqJF8fHwUExOj/fv3O5VJSkrSoEGDFBAQoKCgIA0fPlzp6enFzvPKK6/okksukbe3txo3bqwXXnihcm66jrP3KK2sOUqlIvOUniJRCgAAXI9EKQAAACrVsmXLNHbsWE2dOlU7duxQp06dFBsbq8TExBLLb9q0SQMHDtTw4cP1448/asCAARowYID27NnjKPPSSy/ptdde0/z587Vlyxb5+fkpNjZW2dnZjjKDBg3S3r17tXbtWq1cuVLffvutRo4c6XSt0aNH6/3339crr7yiX3/9VZ9//rmuvPLKqqmIOiQ9J18pWXmSpEaBldOjVDqbKD1Mj1IAAFADkCgFAABApZozZ45GjBihYcOGqV27dpo/f758fX31wQcflFh+3rx56tu3r8aPH6+2bdtqxowZuvzyy/XGG29IKuwFOnfuXE2aNEm33HKLOnbsqA8//FDHjh3TihUrJEm//PKLVq9erffff1/R0dHq2bOnXn/9dS1dulTHjh1zlHn77bf173//WzfffLOioqLUtWtX/fWvf62WeqnNjp/pTRpg9VA9q2elnTeyyIJOAAAArlYrEqWGYSgrK0tZWVkyjLq/Iqa73S8AAKg7cnNztX37dsXExDi2mc1mxcTEaPPmzSUes3nzZqfykhQbG+sof/DgQcXHxzuVCQwMVHR0tKPM5s2bFRQUpG7dujnKxMTEyGw2a8uWLZKkL774Qi1atNDKlSsVFRWl5s2b64EHHlBSUlKp95OTk6PU1FSnhzs6WskLOdk1C/aTxBylAACgZqgVidLs7Gz169dP/fr1cxpeVVe52/0CAIC64+TJkyooKFBYWJjT9rCwMMXHx5d4THx8fJnl7V/PVyY0NNRpv4eHh4KDgx1l/vjjDx0+fFjLly/Xhx9+qIULF2r79u264447Sr2fmTNnKjAw0PGIjIw8XxXUSceSC9uklTk/qVRkjlISpQAAoAaoFYlSAAAA4GLZbDbl5OToww8/1NVXX63evXvrH//4hzZs2KB9+/aVeMzEiROVkpLieBw5cqSao64ZjqdUTY9Se6L0RFqOsnILKvXcAAAAFUWiFAAAAJUmJCREFotFCQkJTtsTEhIUHh5e4jHh4eFllrd/PV+ZcxeLys/PV1JSkqNMo0aN5OHhoUsuucRRpm3btpKkuLi4EmPz9vZWQECA08Md2YfeNwqqvIWcJCnQ11MBVg9J0pHT9CoFAACuRaIUAAAAlcbLy0tdu3bVunXrHNtsNpvWrVun7t27l3hM9+7dncpL0tq1ax3lo6KiFB4e7lQmNTVVW7ZscZTp3r27kpOTtX37dkeZ9evXy2azKTo6WpJ01VVXKT8/X7///rujzG+//SZJatas2cXcdp0Xn1I49L4yV7y3a9qgsFfp4VMkSgEAgGt5uDoAAAAA1C1jx47VkCFD1K1bN1155ZWaO3euMjIyNGzYMEnS4MGD1bhxY82cOVOSNHr0aPXq1UuzZ89W//79tXTpUm3btk3vvvuuJMlkMmnMmDF6/vnn1bp1a0VFRWny5MmKiIjQgAEDJBX2DO3bt69GjBih+fPnKy8vT6NGjdI999yjiIgISYWLO11++eW6//77NXfuXNlsNj366KP661//6tTLFMUlpBYmSsMCKj9RGlnfV3uOpupPepQCAAAXI1EKAACASnX33XfrxIkTmjJliuLj49W5c2etXr3asRhTXFyczOazA5t69OihJUuWaNKkSXrmmWfUunVrrVixQu3bt3eUmTBhgjIyMjRy5EglJyerZ8+eWr16tazWs4m7xYsXa9SoUerTp4/MZrNuv/12vfbaa479ZrNZX3zxhR577DFdc8018vPzU79+/TR79uxqqJXaLSE1R1LVJEqb1C+c9/RIUlalnxsAAKAiSJQCAACg0o0aNUqjRo0qcd/GjRuLbbvzzjt15513lno+k8mk6dOna/r06aWWCQ4O1pIlS8qMKyIiQv/3f/9XZhk4S8/JV3pOvqQq6lF6ZkEn5igFAACuxhylAAAAAEplH3bv7+0hf+/K72cRWf9MojSJRCkAAHAtEqUAAAAASpWQYp+f1LtKzh8ZXDj0/s/TWTIMo0quAQAAUB4kSgEAAACUKiGtMFEaXgUr3ktSkzM9StNz8pWcmVcl1wAAACgPEqUAAAAAShWfcmYhp3pVkyi1elrUsF5hb1XmKQUAAK5EohQAAABAqexzlIZVUY9SSYqsf3b4PQAAgKuQKAUAAABQKnuiNLwKVry3c6x8z4JOAADAhUiUAgAAAChVfGrVLuYkSU3O9Chl6D0AAHAlEqUAAAAASnV21fuqHHpv71HK0HsAAOA6JEoBAAAAlMhmM5SYVriYU1Wtei8VGXpPj1IAAOBCJEoBAAAAlOhURq7ybYZMJinEv+qG3tt7lP55Oks2m1Fl1wEAACgLiVIAAAAAJbIv5BTi7y1PS9X969AoyCqzScrNt+lEek6VXQcAAKAsJEoBAAAAlCihGhZykiRPi1mNAs8s6MTK9wAAwEVIlAIAAAAokX3F+/AqXMjJLjK4MFH652kWdAIAAK5BohQAAABAiRJSC4fBV+WK93b2eUrpUQoAAFyFRCkAAACAEiWk2IfeV32itIk9UcrK9wAAwEVIlAIAAAAokSuG3h9JYug9AABwDRKlAAAAkCT98ccfrg4BNYxjMafA6kiU0qMUAAC4FolSAAAASJJatWqla6+9Vh9//LGys7NdHQ5qgOpa9V46O0fp8ZRs5RfYqvx6AAAA5yJRCgAAAEnSjh071LFjR40dO1bh4eF68MEHtXXrVleHBRfJzivQ6cw8SdUz9D60nre8PMwqsBk6nkKiHgAAVD8SpQAAAJAkde7cWfPmzdOxY8f0wQcf6Pjx4+rZs6fat2+vOXPm6MSJE64OEdXoRFrhivfeHmYF+nhW+fXMZpOa1C+cpzSOle8BAIALkCgFAACAEw8PD912221avny5Zs2apQMHDmjcuHGKjIzU4MGDdfz4cVeHiGoQn3p2xXuTyVQt12x2Zp5SEqUAAMAVSJQCAADAybZt2/TII4+oUaNGmjNnjsaNG6fff/9da9eu1bFjx3TLLbe4OkRUg/iU6lvx3q4piVIAAOBCHq4OAAAAADXDnDlztGDBAu3bt0833HCDPvzwQ91www0ymws/W4+KitLChQvVvHlz1waKalGdK97bRZIoBQAALkSiFAAAAJKkt99+W/fff7+GDh2qRo0alVgmNDRU//jHP6o5MriCI1Far+pXvLdz9Cg9RaIUAABUPxKlAAAAkCStXbtWTZs2dfQgtTMMQ0eOHFHTpk3l5eWlIUOGuChCVKeE1MLFnMKrsUdp0wb0KAUAAK7DHKUAAACQJLVs2VInT54stj0pKUlRUVEuiAiuZF/MKbQa5yiNrF+YKE3JylNKZl61XRcAAEAiUQoAAIAzDMMocXt6erqs1upLlqFmsA+9r87FnPy8PRTiXzjU/8hpepUCAIDqxdB7AAAANzd27FhJkslk0pQpU+Tr6+vYV1BQoC1btqhz584uig6uYBiGSxKlktQ02Ecn03N0+FSm2jcOrNZrAwAA90aiFAAAwM39+OOPkgqTY7t375aXl5djn5eXlzp16qRx48a5Kjy4QGpWvrLzbJKk0IDqW8xJKlzQaUdcMvOUAgCAakeiFAAAwM1t2LBBkjRs2DDNmzdPAQEBLo4IrmafnzTI11NWT0u1XrtpMAs6AQAA1yBRCgAAAEnSggULXB0CaghXDbuXpMgzidIjJEoBAEA1I1EKAADgxm677TYtXLhQAQEBuu2228os++mnn1ZTVHA1V6x4b9esgZ8kepQCAIDqR6IUAADAjQUGBspkMjm+ByQpIcXeo7R65yeVzg69P5qcpbwCmzwt5mqPAQAAuCcSpQAAAG6s6HB7ht7DLiHNdUPvQ+t5y8vDrNx8m44nZ6tpA99qjwEAALgnEqU1XL9+/SRJHh4eys/PL7FMWftKYzKZZBiG07amTZvqoYce0ksvvaS8vDzl5+crJydHkmSxWFRQUFDiuXr37q1ffvlF3t7eiouLk8VikdVqlaenp2666SZ98cUXysrKUm5uru69915J0uLFi9WuXTvt3btXXl5e8vDwkKenpyZMmKA1a9Zo48aNTnHl5ORo9OjR6tGjh2P7pk2b9NJLL0mSJkyY4Ni3adMmPffcc8rJyVHv3r01bdo0xzHTpk1znPu+++7T8OHDnc43b968Ytcp7/5zy0gqMb6yynXu3FnffPONfH19ddttt2nNmjW6/vrr9a9//cvxXNhjt9fjNddco507dyovL8+x3173a9ascTwvRY8dPny4U13Yt7dt21YvvPCCMjMz5evrq2effVY9evRwlG3fvr1OnDjhiPt89XG+Ojr3uH/84x9avHixBg0a5PTcVJbyPIcXeh77tuuvv15r1qxx7CvPNe33bX9PGIYhb29vx3s7JydH3t7e8vHx0YQJEyTJ8Tzde++9Gj58eJnXKe29Upl1Ut1qS9y1Jc6aorT3kavjKU8cJf1ez8vLc/xtK+l9WZPu1S4rK0uGYcjXtzAxdfjwYX322Wdq166drr/+ehdHh+oUn1LY7nDF0Huz2aTI+j76/USG4pIySZQCAIBqYzLOzZaVU2pqqgIDA5WSklLlK6NmZWU5EoZffvmlfHx8im1Pu/w+yeJZeEBBnurt+Kj49vIqcnzR61WXVatWOZIa1c3+nFaWcxOyJSVoiwoKClJycnKJ+0JCQvTxxx/LarUqOztbgwYN0qlTpyRJDRo00OLFiyVJAwcO1OnTpx3HLVu2TGFhYUpISNDdd9/tdM4VK1YoKChI2dnZuvfee3Xy5Emn69idb/+5ZRo0aCDDMJSUlOQUnz320sqV1/nqsawyJpNJ7777rkaMGFFs37n1HxwcrLlz52rw4MFO5Ro0aCBJOnXqVKn1UZKy6jE5OVm33XabbDabzGazPv30UwUFBZ33nOVVnufwQs8jybHNbDbLZrMpJCRE77//vh544IEyr1n0vsujQYMGstlsjte4yWTS0qVL9eijj5Z4ndLeK+e+Fi+mTqpbbYm7tsRZUxStr6LvI1fVW0Wev/P9Xi/6vju3fHXf6/nab9dff71uu+02PfTQQ0pOTtall14qLy8vnTx5UnPmzNHDDz9cpfHVBtXZBnalm17/TruPpuj9wd0U0y6s2q8/bMFWbdh3Qi/e2kF/i25a7dcHAAB1R0Xab0z4UwO5KkkqqVKTpJKKJerOl9wrLUkqFSbllixZIqmwJ6U98VN03+LFi52SpJL02GOPSZJGjRpV7JxTpkwpdr6i17E73/6SyhT9J7m02M8tV17l+XyjtDKGYZT6j+659Z+UlKSHHnqoWLlTp06dtz5KUlY9Tp482ZEstNlsjuemspTnObzQ8xTdZr+HU6dOafLkyee9ZtH7Lo9Tp045vcYNw9Bjjz1W6nVKe6+Udi+1QW2Ju7bEWVOU9j5yVb1V5Pk73+/1st6XNeFei9qxY4euvvpqSdInn3yi8PBwHT58WB9++KFee+01F0eH6mRf9T7MBT1KJRZ0AgAArlErEqVFEz7Z2dnKyspSVlaWsrOzixaqzAuWeL3qeJTUyw+FDMPQkiVLtG3bNkfv0aIWL16sjz76qNj2xMREvfzyyzpx4kSxfbt27dLq1au1ZMkSx+vMfp0///xTkvTnn3+Wub+kMiVZvHixtm3bdt5y1aEiUzVkZpb9D0pJ9VGSsupx27Zt2r17t1P5Xbt2adu2beWO80KvfbHnWbx4cYnPqWEY2r17d5nXLOm+L0RiYmKJ17HHe66SXosXWifVrbKey6pWW+KsKUr7HeqqeqvI81ee3/9S4fuu6PuyptzruTIzM1WvXj1J0po1a3TbbbfJbDbrL3/5iw4fPuzS2FB98gtsOpleOPQ+LLD6F3OSpMgzCzrFJWW45PoAAMA9lXuO0pycHKc5ElNTU6skoNKubXfrrbeWXMiWL8mrci5oO5tEKvV6cImCggI999xzJfbAK20OVUn6z3/+U+q+v//9747Vfu0Mw9C8efM0a9YszZs3r9gx9v323r8llSkpvunTp7s8SVoVitbHuXVZdH9J21999VXt37+/xPNOnz5dK1askNl84Z/plHXtsmIu73nKet2VdU3DMDR9+vRyH1sR9usYhlFifPbXYlnxladOqltlPZdVrbbEWVMUfb2Wtb+66q0iz19pZUtSUFBQ5n1Khb1LXf0aadWqlVasWKFbb71VX331lZ544glJhR/I1OVh5nB2Mj1XNkOymE0K8XNNorSpI1FKj1IAAFB9yp19mDlzpgIDAx2PyMjIqowLKJFhGEpLS6uS8xZVUFCgH374QVu2bNEPP/xQLNlk3x8XF6e4uLgSy5QkNTW1QsOsa4ui9VGS0uqooKBA27dvL/WDl9TUVG3ZsuWiYivr2mXFXN7zVETRa27ZsqXKPnCyX6esHrmpqakXXSfVrbKey6pWW+KsKez1Vdrvxuqut4o8fxX9vWB/X5ZW3mazufw1MmXKFI0bN07NmzdXdHS0unfvLqmwd2mXLl1cFheqV/yZYfeh9bxlNrsmaW9PlB4+lVknP2QGAAA1U7l7lE6cOFFjx451/JyamlptyVJv77OfZH/22WdOiyE4enyay30r51fkXEWvV9UyMzN1++23V8u1aiuTySR/f/9KT5aeu/CRxWJR165dFR0drSuuuEI7duxw+sfWvr9p08LFBUoqU5KAgAClp6fXuWTpufVxrqZNm5Zaj126dNFvv/1WYtIwMDBQ0dHRFxVbWdcuK+bynqciil4zMjJSAQEBVZIstV/HZrOVmiwNCAhQRkbGRdVJdaus57Kq1ZY4awp7fW3fvr3E343VXW8Vef4q+nvhiiuukGEY+vHHH0ssbzab1a1bN5e+Ru644w717NlTx48fV6dOnRzb+/TpwygbNxKf4tr5SaWzidK07HydzsxTsF8ljRwDAAAoQ7l7lHp7eysgIMDpUV2KDj+zWq3y8fGRj4+PcwKzMoeolXK9qn40aNBAl156aeXdRx1ksVg0bdq0EodiWyyWUo+78cYbS903ceLEYuczmUwaPXq0zGazRo8eXewY+36TyeT4vjyxT506tU4OuS1aH2XtL2n7E088UerCTVOnTr2oYffnu3ZZMZf3PBaLpczXXmnXNJvNlb5g1bnXGTNmTImx2V+LZcVXE1XWc1nVakucNUV5f39UV71V5Pkr7+9/qfB9N2bMGI0ZM6bUMva/Oa5+jYSHh6tLly5Ov3+vvPJKtWnTpkLnefPNN9W8eXNZrVZFR0dr69atZZZfvny52rRpI6vVqg4dOmjVqlVO+w3D0JQpU9SoUSP5+PgoJiam2NQtSUlJGjRokAICAhQUFKThw4crPT29xOsdOHBA9erVU1BQUIXuyx0kphUmSsNdmCj18bIoIrDw+gdPlvwcAgAAVLZasZiTO3nnnXdcHUKNZTKZ9Le//U1du3bVoEGDiu0fNGiQ7rvvvmLbQ0NDNW7cODVs2LDYvo4dOyo2NlZ/+9vfHP+Y2q/TuHFjSVKTJk3K3F9SmZIMGjRIXbt2PW+56uDhUf4e2L6+vmXuL6k+SlJWPXbr1k0dOnRwKt+xY0ddfvnl5Y7zQq99secZNGhQic+pyWRShw4dyrxmSfd9IUJDQ0u8jj3ec5X0WrzQOqlulfVcVrXaEmdNUdrvUFfVW0Wev/L8/pcK33dF35c15V7PlZGRocmTJ6tHjx5q1aqVWrRo4fQor2XLlmns2LGaOnWqduzYoU6dOik2NlaJiYkllt+0aZMGDhyo4cOH68cff9SAAQM0YMAA7dmzx1HmpZde0muvvab58+dry5Yt8vPzU2xsrNPinoMGDdLevXu1du1arVy5Ut9++61GjhxZ7Hp5eXkaOHCgrr766grUjvs426PUNfOT2rVo6C9J+uMECzoBAIDqQaK0BpowYYLLrh0YGFip5yvpH8Gy1K9fv9R9ISEhjqTPoEGD1KBBg2L7Bg0aVOwcr7/+uiTpjTfeKHZO+4I2Rc9X9Dp259tfUpng4ODzxn5uuaLKqqvyJFpL64lpMpn09ttvl7jv3F41DRo00Pz584uVCwkJOW99lKSsepwxY4YjZrPZXOkLHZXnObzQ8xTdZr+HkJAQzZgx47zXLHrf5dGgQQOn17jZbNbrr79e6nVKe6+Udi+1QW2Ju7bEWVOU9j5yVb1V5Pk73+/1st6XNeFei3rggQf0j3/8Q1dffbVGjRql0aNHOz3Ka86cORoxYoSGDRumdu3aaf78+fL19dUHH3xQYvl58+apb9++Gj9+vNq2basZM2bo8ssvd/ztNgxDc+fO1aRJk3TLLbeoY8eO+vDDD3Xs2DGtWLFCkvTLL79o9erVev/99xUdHa2ePXvq9ddf19KlS3Xs2DGn602aNElt2rTRXXfddWEVVcfZ5ygNC3Rdj1JJigrxkyQdPEmiFAAAVA8SpTXQtddeW2xbWT0AK9I70K6kRFvTpk311FNPKSgoSH5+fk5zw5Y1tLh3794KCwtzzKlmsVjk5+enoKAg3XvvvQoKCpK3t7dMJpPuvfde3XfffTKbzWrfvr1MJpO8vb0d5cePH6/evXsXiyssLExPPPGEY7oFq9WqJ598UkFBQQoKCtLYsWNltVpltVo1fvx4R+z22CQpLCzM6dz33XefIzFotVo1duzYYtexO9/+c8uMHTtW48aNKxbf+cr17t1bJpNJfn5+uvfeexUWFqb77rvP6bmQ5FSPvXv3djxn9kdQUJAGDRrk9LwUPbZ169bF6vm+++7ThAkT5Ofn54jhySefVNOmTR1l27dv74j7ySefLLM+SlJWPdpjNpvNGjRoUKUPhSzPc3ih5ym6zV7vTzzxhOO5L+uaRe/b/p6Q5Hhf2J97b29vBQUF6cknn9T48eMdz5P9eqVdp7T3SmXWSXWrLXHXljhritLeR66qt4o8f6X9Xrf/Pi76vju3fE2416K+/PJLLV++XLNmzdKYMWMuKFGam5ur7du3KyYmxrHNbDYrJiZGmzdvLvGYzZs3O5WXpNjYWEf5gwcPKj4+3qmMfR5re5nNmzcrKChI3bp1c5SJiYmR2Wx2Whhw/fr1Wr58ud58881y3U9OTo5SU1OdHnVdYmqOJNcOvZdIlAIAgOpnMi5wGcnU1FQFBgYqJSWlyucrzcrKUr9+/SQVNuB9fHyKbU+7/D7J4ll4QEGe6u34qPj28ipyfNHrVZfS7hcAAOBinK/9FhUVpVWrVqlt27YXfI1jx46pcePG2rRpk7p37+7YPmHCBH3zzTdOSUs7Ly8vLVq0SAMHDnRse+utt/Tcc88pISFBmzZt0lVXXaVjx46pUaNGjjJ33XWXTCaTli1bphdffFGLFi3Svn37nM4dGhqq5557Tg8//LBOnTqlLl266OOPP9Y111yjhQsXasyYMUpOTi71fqZNm6bnnnuu2PbqaAO7Ssycb3QgMV2LH4jWVa1CXBbHhn2JGrbgB7UJr6fVY65xWRwAAKB2q0gOkx6lAAAAkFQ4HciUKVOUmZnp6lCqxIgRI/S3v/1N11xT/qTbxIkTlZKS4ngcOXKkCiOsGRJqwKr3ktSiSI9Sm+2C+nYAAABUSMXHbAMAAKBOmj17tn7//XeFhYWpefPm8vR0HpWzY8eO854jJCREFotFCQkJTtsTEhIUHh5e4jHh4eFllrd/TUhIcOpRmpCQoM6dOzvKnLtYVH5+vpKSkhzHr1+/Xp9//rleeeUVSYVzn9psNnl4eOjdd9/V/fffXyw2b2/vYlPg1GUZOflKy8mXJIW7eI7SxkE+8rSYlJNv0/HUbDUOYpQVAACoWiRKAQAAIEkaMGDARZ/Dy8tLXbt21bp16xzns9lsWrdunUaNGlXiMd27d9e6des0ZswYx7a1a9c6hu5HRUUpPDxc69atcyRGU1NTtWXLFj388MOOcyQnJ2v79u3q2rWrpMLEqM1mU3R0tKTCeUwLCgoc1/j3v/+tWbNmadOmTWrcuPFF33tdkHBmISc/L4v8vV37r4KHxaymwb76/USGDp7IIFEKAACqHIlSAAAASJKmTp1aKecZO3ashgwZom7duunKK6/U3LlzlZGRoWHDhkmSBg8erMaNG2vmzJmSpNGjR6tXr16aPXu2+vfvr6VLl2rbtm169913JRUuQjlmzBg9//zzat26taKiojR58mRFREQ4krFt27ZV3759NWLECM2fP195eXkaNWqU7rnnHkVERDjKFLVt2zbHYnoolHBmISdXr3hvFxXiX5goPZmunq1dN18qAABwDyRKAQAA4JCcnKxPPvlEv//+u8aPH6/g4GDt2LFDYWFh5e51effdd+vEiROaMmWK4uPj1blzZ61evVphYWGSpLi4OJnNZ6fK79Gjh5YsWaJJkybpmWeeUevWrbVixQqnBOaECROUkZGhkSNHKjk5WT179tTq1atltZ5N6C1evFijRo1Snz59ZDabdfvtt+u1116rpJpxD/YepWH1akaitEVDP+kX6Y+TGa4OBQAAuAESpQAAAJAk7dq1SzExMQoMDNShQ4c0YsQIBQcH69NPP1VcXJw+/PDDcp9r1KhRpQ6137hxY7Ftd955p+68885Sz2cymTR9+nRNnz691DLBwcFasmRJuWMcOnSohg4dWu7y7uBYSpYkqVFQzUiURhVZ0AkAAKCqseo9AAAAJBUOmR86dKj279/v1FPzhhtu0LfffuvCyFBdjicX9iiNCKwZ84GSKAUAANWJRCkAAAAkST/88IMefPDBYtsbN26s+Ph4F0SE6na8hvUobXEmUXokKVO5+TYXRwMAAOo6EqUAAACQJHl7eys1NbXY9t9++00NGzZ0QUSobsfO9ChtVEMWc2pYz1t+XhbZDCkuKdPV4QAAgDqORCkAAAAkSTfffLOmT5+uvLw8SYXzgsbFxempp57S7bff7uLoUB0cPUpryNB7k8mkqIYMvwcAANWDRCkAAAAkSbNnz1Z6eroaNmyorKws9erVS61atVK9evX0wgsvuDo8VLHsvAKdzixMkteUOUolKSrEX5J08GS6iyMBAAB1HaveAwAAQJIUGBiotWvX6vvvv9dPP/2k9PR0XX755YqJiXF1aKgGx1MKh937elkU4FNz/k1gQScAAFBdak4LCAAAAC5js9m0cOFCffrppzp06FDhkOeoKIWHh8swDJlMJleHiCp2PNk+7N5ao55v+4JOv58gUQoAAKoWQ+8BAADcnGEYuvnmm/XAAw/o6NGj6tChgy677DIdPnxYQ4cO1a233urqEFENjp3pURoRVHOG3UtSy4aFQ+9/T2ToPQAAqFr0KAUAAHBzCxcu1Lfffqt169bp2muvddq3fv16DRgwQB9++KEGDx7soghRHew9SsMDasaK93atQv1lMkmnMnJ1Ii1HDet5uzokAABQR9GjFAAAwM3985//1DPPPFMsSSpJ1113nZ5++mktXrzYBZGhOtl7lDaqYT1Kfbwsat6gcPj9bwlpLo4GAADUZSRKAQAA3NyuXbvUt2/fUvf369dPP/30UzVGBFeITynsURoRWLN6lErSJWGFw+9/jSdRCgAAqg6JUgAAADeXlJSksLCwUveHhYXp9OnT1RgRXOF4De1RKkmXhgdIkn4jUQoAAKoQiVIAAAA3V1BQIA+P0qeut1gsys/Pr8aI4ArHkmtuj9JLw+pJkn5l6D0AAKhCLOYEAADg5gzD0NChQ+XtXfIiOTk5OdUcEapbRk6+UrMLk+E1s0dpYaJ0f0KabDZDZrPJxREBAIC6iEQpAACAmxsyZMh5y7Difd12/Mz8pPWsHvL3rnn/IjRv4CsvD7Mycwv05+ksNW3g6+qQAABAHVTzWkEAAACoVgsWLHB1CHCxY8ln5ietgcPuJcnDYlarhv76+Xiq9iWkkSgFAABVgjlKAQAAADcXb1/IKbDmDbu3sw+/3xef6uJIAABAXUWiFAAAAHBzx84MvY8Iqpk9SqUiidKEdBdHAgAA6ioSpQAAAICbO55cC3qUhtGjFAAAVC0SpQAAAICbs/coralzlEpne5T+cSJDufk2F0cDAADqIhKlAAAAgJs7fmaO0oigmtujtFGgVfWsHsq3GfrjJMPvAQBA5SNRCgAAALgxwzB0PLmwR2l4De5RajKZigy/T3NxNAAAoC4iUQoAAAC4sbScfGXkFkiSImrwHKWSdEk4iVIAAFB1SJQCAAAAbsy+kFOQr6d8vCwujqZsbc4kSn8+zoJOAACg8pEoBQAAANzYkaRMSVKT+jW7N6kkdWwSJEn66UiyDMNwbTAAAKDOIVEKAAAAuLEjpwsTpU2DfV0cyfm1bVRPXhazTmfmKe5MghcAAKCykCgFAAAA3Jg94RhZv+YnSr09LGobESBJ2nkk2bXBAACAOodEKQAAAODG7EPvI2tBj1JJ6hIZJIlEKQAAqHwkSgEAAAA3Zu9RWhuG3ktSp8hASYXzlAIAAFQmEqUAAACAmzIMQ0eSsiTVnkRp58j6kqQ9x1KVm29zcTQAAKAuIVEKAAAAuKmT6bnKyiuQySRFBNX8Ve8lqXkDXwX6eCo336Zf41NdHQ4AAKhDSJQCAAAAbso+7D4i0EdeHrXjXwOTyaROZ+YpZfg9AACoTLWjNQQAAACg0p1dyKl29Ca163wmUfojiVIAAFCJPFwdQHlYrVZ9+eWXju/rOne7XwAAALhGbVvIya7zmQWdWPkeAABUplqRKDWZTPLxqV2fcl8Md7tfAAAAuIajR2n92pUo7dQkSJL0x4kMpWTlKdDH07UBAQCAOoGh9wAAAICbcvQobVC7EqUN/L0d0wXs+jPZtcEAAIA6g0QpAAAA4KbOzlFauxKlktQ5sr4k6ce4ZNcGAgAA6gwSpQAAAIAbyskv0PHUbEm1b45SSbqyeWGi9PsDJ10cCQAAqCtIlAIAAABu6OjpLBmG5ONpUQM/L1eHU2FXt24oSdoRd1oZOfkujgYAANQFJEoBAAAAN3TkdJakwt6kJpPJxdFUXLMGvooM9lFegaEtB0+5OhwAAFAHkCgFAAAA3FBcLZ6fVJJMJpOjV+m3vzH8HgAAXDwSpQAAAIAbsi/kVBvnJ7W7ulWIJOm/+0+4OBIAAFAXkCgFAABApXvzzTfVvHlzWa1WRUdHa+vWrWWWX758udq0aSOr1aoOHTpo1apVTvsNw9CUKVPUqFEj+fj4KCYmRvv373cqk5SUpEGDBikgIEBBQUEaPny40tPTHfs3btyoW265RY0aNZKfn586d+6sxYsXV95N1zJxp+w9Sn1cHMmF69EyRGaT9PuJDB1LznJ1OAAAoJYjUQoAAIBKtWzZMo0dO1ZTp07Vjh071KlTJ8XGxioxMbHE8ps2bdLAgQM1fPhw/fjjjxowYIAGDBigPXv2OMq89NJLeu211zR//nxt2bJFfn5+io2NVXZ2tqPMoEGDtHfvXq1du1YrV67Ut99+q5EjRzpdp2PHjvq///s/7dq1S8OGDdPgwYO1cuXKqquMGuzI6drfozTQ11OdIoMkSd/tZ/g9AAC4OCbDMIwLOTA1NVWBgYFKSUlRQEBAZcdVLllZWerXr58kKe3y+ySLZ+GOgjzV2/FR8e3lVeT4L7/8Uj4+tfdTdgAAALvqar9FR0friiuu0BtvvCFJstlsioyM1GOPPaann366WPm7775bGRkZTgnLv/zlL+rcubPmz58vwzAUERGhJ598UuPGjZMkpaSkKCwsTAsXLtQ999yjX375Re3atdMPP/ygbt26SZJWr16tG264QX/++aciIiJKjLV///4KCwvTBx98UK57qwlt4MpgGIY6TlujtJx8rX3iGrUOq+fqkC7YnDX79Nr6A7qpU4ReH9jF1eEAAIAapiLtN3qUAgAAoNLk5uZq+/btiomJcWwzm82KiYnR5s2bSzxm8+bNTuUlKTY21lH+4MGDio+PdyoTGBio6OhoR5nNmzcrKCjIkSSVpJiYGJnNZm3ZsqXUeFNSUhQcHFzq/pycHKWmpjo96oLTmXlKy8mXJDWpX3t7lErS1ZcULuj03f4TstkuqA8IAACAJBKlAAAAqEQnT55UQUGBwsLCnLaHhYUpPj6+xGPi4+PLLG//er4yoaGhTvs9PDwUHBxc6nX/9a9/6YcfftCwYcNKvZ+ZM2cqMDDQ8YiMjCy1bG2yPyFNktSkvo98vCwujubidI4Mkr+3h05n5mnvsbqRyAYAAK5BohQAAABuZ8OGDRo2bJjee+89XXbZZaWWmzhxolJSUhyPI0eOVGOUVWd/YuEiV61D/V0cycXztJjVo2UDSdJXe0tOigMAAJQHiVIAAABUmpCQEFksFiUkJDhtT0hIUHh4eInHhIeHl1ne/vV8Zc5dLCo/P19JSUnFrvvNN9/opptu0quvvqrBgweXeT/e3t4KCAhwetQF9h6ltXlu0qJu7lw4B+2KnUcZfg8AAC4YiVIAAABUGi8vL3Xt2lXr1q1zbLPZbFq3bp26d+9e4jHdu3d3Ki9Ja9eudZSPiopSeHi4U5nU1FRt2bLFUaZ79+5KTk7W9u3bHWXWr18vm82m6Ohox7aNGzeqf//+mjVrlkaOHHnxN1xL1aUepZIU0zZM/t4e+vN0lrYdPu3qcAAAQC1FohQAAACVauzYsXrvvfe0aNEi/fLLL3r44YeVkZHhmAt08ODBmjhxoqP86NGjtXr1as2ePVu//vqrpk2bpm3btmnUqFGSJJPJpDFjxuj555/X559/rt27d2vw4MGKiIjQgAEDJElt27ZV3759NWLECG3dulXff/+9Ro0apXvuucex4v2GDRvUv39/Pf7447r99tsVHx+v+Ph4JSUlVW8F1QCORGkd6VFq9bSoX/vCnsOf/XjUxdEAAIDaikQpAAAAKtXdd9+tV155RVOmTFHnzp21c+dOrV692rEYU1xcnI4fP+4o36NHDy1ZskTvvvuuOnXqpE8++UQrVqxQ+/btHWUmTJigxx57TCNHjtQVV1yh9PR0rV69Wlar1VFm8eLFatOmjfr06aMbbrhBPXv21LvvvuvYv2jRImVmZmrmzJlq1KiR43HbbbdVQ63UHMmZuTqRliNJalVHepRK0q1dGkuS/rPrmHLyC1wcDQAAqI1MhmFc0CQ+qampCgwMVEpKisvmasrKylK/fv0kSWmX3ydZPAt3FOSp3o6Pim8vryLHf/nll/Lx8am0mAEAAFylJrTfaru6UIc/HErSnfM3q3GQj75/+jpXh1NpCmyGrvr7esWnZmv+vV3Vt33Jc+ICAAD3UpH2Gz1KAQAAADfy25mFnOpSb1JJsphNusW+qBPD7wEAwAUgUQoAAAC4kf0JhfOTXhJWtxKlknTr5YXD79f/mqjkzFwXRwMAAGobEqUAAACAGzngWPG+bizkVFSb8AC1axSg3AKbPtx82NXhAACAWoZEKQAAAOBGHEPv62CPUkl6sFcLSdIH3x9Uek6+i6MBAAC1CYlSAAAAwE2kZOYp8cyK963r2Byldjd2jFCLED8lZ+bpI3qVAgCACiBRCgAAALiJAycKe5M2CrSqntXTxdFUDYvZpEevbSVJev+/fygzl16lAACgfEiUAgAAAG7itzMLObUOq3vzkxZ1S+cINQ321amMXC3ZEufqcAAAQC1BohQAAABwE/YV7+vqsHs7D4tZj/RuKUl659s/lJad5+KIAABAbUCiFAAAAHAT+xMLh97X9USpJN12eRM1DfbVibQcvbjqV1eHAwAAagESpQAAAICb2Bd/JlFaR1e8L8rLw6xZt3eUJP1za5y+/e2EiyMCAAA1HYlSAAAAwA0cS85SYlqOLGaT2jYKcHU41aJ7ywYa2qO5JOnp/9ulVIbgAwCAMpAoBQAAANzAjrjTkqQ24fXk6+Xh4miqz4S+l6pZA18dS8nWtM/3yjAMV4cEAABqKBKlAAAAgBvYcThZknR50/quDaSa+Xp56OU7Oslkkj7dcVSvrz/g6pAAAEANRaIUAAAAcAP2HqWXNwtybSAucGVUsKbddJkkac7a37R4y2EXRwQAAGoiEqUAAABAHZedV6C9x1IkuV+PUrshPZrr8etaSZImrdijL3465uKIAABATUOiFAAAAKjj9h5LUV6BoQZ+Xmoa7OvqcFzmib9eooFXNpVhSI8v/VHzv/mdOUsBAIADiVIAAACgjvsxLlmS1KVpfZlMJtcG40Imk0nPD2iv+/7STIYh/f3LXzXhk13Kzbe5OjQAAFADkCgFAAAA6jh3np/0XBazSTMGtNe0m9rJbJKWb/9Td8zfpN9PpLs6NAAA4GIkSgEAAIA6zl1XvC/L0Kui9I+hVyjQx1O7/kxR/9f+q482H2IoPgAAboxEKQAAAFCHHUvOUnxqtixmkzo2CXR1ODXKtZeG6qsx16hnqxBl59k0+d97dc+7/9P+hDRXhwYAAFyARCkAAABQh9mH3bdtVE++Xh4ujqbmCQ+06sP7r9SUG9vJ6mnWloNJ6jfvv5q1+ldl5Ra4OjwAAFCN6kyi1GTLlwryzj7sim4r58Nky3fdjQAAAACViGH352c2m3R/zyitfaKXYtqGKt9m6O2Nvytmzjf6+ucEV4cHAACqSZ35SNl/5z9L3F7vp6XVHAkAAABQc2w5eEoSidLyiAz21ftDrtCavfF67oufdTQ5Sw98uE0xbcM0+ca2atbAz9UhAgCAKlRnepQCAAAAcHYkKVN7j6XKbJKubh3i6nBqjesvC9fasdfooV4t5WE26etfEvTXOd9q5pe/KC077/wnAAAAtVKt7lFqtVr15ZdfFttuGIZycnIkSd7e3jKZTBd1DQAAAKA2WnNm2PgVzYPVwN/bxdHULr5eHnq6XxvdfnljTV/5s/67/6Te+eYP/d/2P/Xk9Zfqrm6Rspgv/P8MAABQ89TqRKnJZJKPj0+J+3x9fas5GgAAAKBm+WpPvCSpb/twF0dSe7UOq6cP779SG/Yl6vmVv+iPkxma+Olufbj5sCb1b6urWtFTFwCAuoKh9wAAAEAddCItRz8cTpJUOJQcF85kMum6NmFaPeYaTb6xnQKsHvrleKoGvb9Ff3vvf9p++LSrQwQAAJWARCkAAABQB339S4IMQ+rYJFCNg0oehYWK8fIwa3jPKG0cf62G9mguT4tJm34/pdvf3qRhC7Zqz9EUV4cIAAAuAolSAAAAoA76am/hsPtYepNWumA/L027+TJtGNdbd5+Zq3TDvhO68fXv9PDH20mYAgBQS5EoBQAAAOqY1Ow8fX/gpCQSpVWpSX1fzbqjo74e20sDOkfIZJK+3BOvG1//Tne9s1mr98SrwGa4OkwAAFBOtXoxJwAAAADFbfg1UXkFhlo29FOrUH9Xh1PnRYX4ae49XfTIta30xvoDWrX7uLYeTNLWg0lqUt9HQ3s0111XRCrA6unqUAEAQBnoUQoAAADUMcu3/SmJ3qTV7ZKwenptYBd999R1evTalqrv66k/T2fp+f/8or+8uE5j/7VT3x84SS9TAABqKHqUAgAAAHXIjrjT+u7ASXmYTRp4ZVNXh+OWwgOtGh/bRo9d11orfjyqD74/qN8S0vXpjqP6dMdRNQq0akCXxrr98sZqFVrP1eECAIAzSJQCAAAAdcib6w9Ikm7t0liRwb4ujsa9WT0tuufKprr7ikjtiEvWpzv+1Bc/HdPxlGy9vfF3vb3xd10WEaC+l4Wrb/twtQr1l8lkcnXYAAC4LRKlAAAAQB2x52iK1v2aKLNJerh3S1eHgzNMJpO6Nquvrs3qa8pN7bT+l0T9346j2rgvUXuPpWrvsVTNXvubWjT0U+xl4ep7Wbg6NgkkaQoAQDUjUQoAAADUEW9tLOxNemPHCLVoyCJONZG3h0X9OjRSvw6NlJSRq69/TtDqvfH6bv9J/XEiw9HTtGE9b13dOkS9Lmmoq1s3VLCfl6tDBwCgziNRCgAAANQB+xPS9OWeeEnSo9e2cnE0KI9gPy/ddUWk7roiUmnZedq474RW743Xhl8TdSItxzGnqckkdWwSpF6tQxTdooG6NA2Srxf/ygEAUNn46woAAADUctl5BXriXztlGFLsZWG6NJwFgmqbelZP3dQpQjd1ilBOfoG2Hzqtb347oW9+O6Ff49P005Fk/XQkWVp/QB5mky5rHKgrm9dXt+bB6tasvhr4e7v6FgAAqPVMhmEYF3JgamqqAgMDlZKSooCAgMqOCwAAAJWM9tvFq4l1aBiGJnyyS8u3/6lgPy998VhPNQ7ycXVYqETxKdn6dv8Jfbf/pH44lKTjKdnFykQEWnVZ40C1jwhU+8YBat84UKH1vN16nlPDMJSSlaeT6bk6nZmr5Mw8ZeUVKDu3QLkFNkmSyVQ4HYK/t4fqWT0U7OelsACr6vt6unXdAUBdUpH2Gz1KAQAAgFrsn1uPaPn2P2U2Sa8P7EKStA4KD7Tqrm6RuqtbpAzD0J+ns7TtcJK2HjytHw4l6UBiuo6lZOtYSrbW/pzgOC7I11OtGvqrZUN/tQr1V8tQP7VqWE+N6/vIYq47ScDsvAIdPJmhP05k6I8T6Tp4MkO/nyz8Pi07/4LO6WUxKzLYR61C/c8+GtZTy1A/pj0AgDqM3/AAAABALbXix6Oa9vleSdL42Da6qlWIiyNCVTOZTIoM9lVksK9u7dJEkpSWnaefj6Vqz7FU7T2aoj3HUnQgMV3JmXnadvi0th0+7XQObw+zokL81LKhvyKDfdWkvk/hOev7qHF9H3l7WFxxa2Wy2QwdT83WHyfSHQnRP84kR4+lZKmscZL1vD3UwN9Lgb5e8vW0yNfLIg+LSYYh2Qwpt8Cm9Ow8pefk61R6rk5l5Cq3wKbfT2To9xMZ+mrv2eSzySRFhfipfUSgLoso7Ll7WUSAgnxZbAsA6gISpQAAAKh0b775pl5++WXFx8erU6dOev3113XllVeWWn758uWaPHmyDh06pNatW2vWrFm64YYbHPsNw9DUqVP13nvvKTk5WVdddZXefvtttW7d2lEmKSlJjz32mL744guZzWbdfvvtmjdvnvz9z67+vmvXLj366KP64Ycf1LBhQz322GOaMGFC1VRCFcrKLdDUz/foX9v+lCTd0CFcD/Vq4eKo4Cr1rJ6KbtFA0S0aOLZl5xXojxMZOnAiXb8npju+/nEyQzn5Nv0an6Zf49NKPF9YgLci6xcmYxsFWhVaz1uhAVaFBXgrtJ5VDet5y+pZ+cnU/AKbjqdk68/TWfrzdKbikjL1x4kM/X4iXYdOZSg7z1bqsQFWD7Vo6K8WDQsTwFEhfmrR0E/NG/hVONbcfJsSUrN18GSGDpypuwOJhfV3KiP3TKI2Q5//dMxxTJP6PoWJ04hAR/I0NMB6wXVRVxiGodTsfJ1Mz9HJtBylZOUpJ9+mnHybcvNtys0vnAbBbDLJeiaJ7eNpkdXLIl9Pi4L9vNSwnrcCfZgKAUD1YI5SAAAAN1Fd7bdly5Zp8ODBmj9/vqKjozV37lwtX75c+/btU2hoaLHymzZt0jXXXKOZM2fqxhtv1JIlSzRr1izt2LFD7du3lyTNmjVLM2fO1KJFixQVFaXJkydr9+7d+vnnn2W1FiYj+vXrp+PHj+udd95RXl6ehg0bpiuuuEJLlixx3P8ll1yimJgYTZw4Ubt379b999+vuXPnauTIkeW6N1e3gdNz8vXZj0f1wXcHdfBkhkwm6fHrWuvxPq3r1FBqVJ0Cm6E/T2fq9zM9M/88naUjSZk6cjpTR5KylJVXUK7zBFg9FBpQmEQNsHqqntVD/lYP1bN6KsDqIaunRSaTZFLh67Lweykrr0CpWflKycpTanaeUrPylJyZp6PJWYpPzVaBrfR/Tz0tJjUN9nUkRFuE+BV+H+KnYD+vakmknUzP0d5jqdpzNEV7j6Vo77FUHT6VWWLZhvW81TrUX80a+KlZA181b+CrZg381DTYV37etbvPUkZOvhLTcnQiLUeJadlKTM3RiTPJ0JPpOTqZnquT6Tk6lZ7rmA/2YnhaTArx91aIv7ca1vNWWIC3wgKsCg+wKuzMIzyQuWUBlKwi7TcSpQAAAG6iutpv0dHRuuKKK/TGG29Ikmw2myIjI/XYY4/p6aefLlb+7rvvVkZGhlauXOnY9pe//EWdO3fW/PnzZRiGIiIi9OSTT2rcuHGSpJSUFIWFhWnhwoW655579Msvv6hdu3b64Ycf1K1bN0nS6tWrdcMNN+jPP/9URESE3n77bT377LOKj4+Xl1fhMNmnn35aK1as0K+//lque6vuNnBOfoF+PZ6mn/5M1o7Dp7X25wRl5BYmshrW89a8uzurB8PtUUkMw1BSRq6OnOnReSQpS/EpWUpMyznzyFZCao5y8y8+8VUaLw+zmgQVTgEQGex7JhnqpxYh/mpS30ceFnOVXftCpWQVTn1gT5zuOZqi30+kq4ycr/y8LGrg760Qfy/H1yBfL/l7e8jH0yI/b4t8vDzk52WRr5eHfL0Kt3lazPKwmOVpNsnDYpbFbJKnxSQPs1keZpPMZz4wMQxDhiEZkmyO7wu/5hbYlJ1XoJy8wq/ZebbCRa7OPAqT2XlKOeeRlJF7JjGao8zc8iXU7ep5eyiknreCfD3l7WGWl4flzFezvC1m2QxDWXkFysw9G0NmToFOZeQqJSuv3NfxspgVGuDtlEBt4O+lIF9P1ff1UpCPp4J8vRTo6yk/L4usnoVxkFxFdTIMQ/k2o7B3dV6Bo4d1YW/rAqfvc/JsRfYXOPXINgxDhuT0/jabTPI6897yspgd33sX+dnqaTnzMBf24Pa0OL56e5gdv0fqkmpZzMmeX01NTb3QUwAAAKAa2dttF/g5ebnk5uZq+/btmjhxomOb2WxWTEyMNm/eXOIxmzdv1tixY522xcbGasWKFZKkgwcPKj4+XjExMY79gYGBio6O1ubNm3XPPfdo8+bNCgoKciRJJSkmJkZms1lbtmzRrbfeqs2bN+uaa65xJEnt15k1a5ZOnz6t+vXrF4stJydHOTk5jp9TUlIkVU8bOL/App6z1is9xzkh0TzEV3d1i9QtnRsr0MeT9jgqlaekFoFmtQj0l5r7F9tvGIZSs/J1Mj1bJ9IKew2m5+QpNTtf6TkFysjJU1p2vrLzCmQ4jpF05p94q6elsNepj6fqWS0K8PFUgLenwgO91STIVw38vUv5J71AmRnpVXfjF8Ek6bKGnrqsYYjUqfCDi6zcAu1LSNXhk5k6kpSpuNOZikvK0pGkDKVk5SstR0pLS9Ohyo7FpDLna61Mvl7mwh6e/laF+HsppJ63Gvh7qYFfYfLX/jXYz+uipmrIyS9Qkr2H6plk7QlH4j5biWm5OpGaraTMPGVListMV1x8+c9vMqkwYeRhkbenRT5eFlk9LPK0mGQxm2QymWQxFX4vk2QxFyajzCaTKpJfrUjqqWLnLX/hKotXhXVjP8ZkOhNVqduce5oX631uLyhDNlthst9mFP7+sX9f+AHA2e/PfrWXPVPeJhUUKVtgK/L9me3Gme9thiGbrci5znwvXfybymYUTuuRU2BPcl70KauMt6dZVqeEamEC1dNy9gMZs0mO7z3Mhe8Hi7nwvWI+8/XscylJJsfzLRVu9/f21BN/vaRa7qkibeALTpSmpRXOZxMZGXmhpwAAAIALpKWlKTAwsErOffLkSRUUFCgsLMxpe1hYWKm9NuPj40ssHx8f79hv31ZWmXOH9Xt4eCg4ONipTFRUVLFz2PeVlCidOXOmnnvuuWLbXdkGPiLpv5Iec1kEAFBon6sDAFCrTavm65WnDXzBidKIiAgdOXJE9erVq5Zu6qmpqYqMjNSRI0cY6n+BqMPKQT1WDurx4lGHlYN6vHjUYeWojno0DENpaWmKiIiokvPXRRMnTnTq7Wqz2ZSUlKQGDRq45VBN3u+lo25KR92UjfopHXVTOuqmdNRN6dy1birSBr7gRKnZbFaTJk0u9PALFhAQ4FZPZlWgDisH9Vg5qMeLRx1WDurx4lGHlaOq67GqepLahYSEyGKxKCEhwWl7QkKCwsPDSzwmPDy8zPL2rwkJCWrUqJFTmc6dOzvKJCYmOp0jPz9fSUlJTucp6TpFr3Eub29veXt7O20LCgoqsaw74f1eOuqmdNRN2aif0lE3paNuSkfdlM4d66a8beCaNxM2AAAAai0vLy917dpV69atc2yz2Wxat26dunfvXuIx3bt3dyovSWvXrnWUj4qKUnh4uFOZ1NRUbdmyxVGme/fuSk5O1vbt2x1l1q9fL5vNpujoaEeZb7/9Vnl5eU7XufTSS0scdg8AAAD3QqIUAAAAlWrs2LF67733tGjRIv3yyy96+OGHlZGRoWHDhkmSBg8e7LTY0+jRo7V69WrNnj1bv/76q6ZNm6Zt27Zp1KhRkgoXgBgzZoyef/55ff7559q9e7cGDx6siIgIDRgwQJLUtm1b9e3bVyNGjNDWrVv1/fffa9SoUbrnnnscw6z+9re/ycvLS8OHD9fevXu1bNkyzZs3r9hCUgAAAHBPFzz0vrp5e3tr6tSpxYY+ofyow8pBPVYO6vHiUYeVg3q8eNRh5ahL9Xj33XfrxIkTmjJliuLj49W5c2etXr3asXBSXFyczOazn9f36NFDS5Ys0aRJk/TMM8+odevWWrFihdq3b+8oM2HCBGVkZGjkyJFKTk5Wz549tXr1almtVkeZxYsXa9SoUerTp4/MZrNuv/12vfbaa479gYGBWrNmjR599FF17dpVISEhmjJlikaOHFkNtVI31KXXaWWjbkpH3ZSN+ikddVM66qZ01E3pqJvzMxmGYbg6CAAAAAAAAABwJYbeAwAAAAAAAHB7JEoBAAAAAAAAuD0SpQAAAAAAAADcHolSAAAAAAAAAG6vViRK33zzTTVv3lxWq1XR0dHaunWrq0NymZkzZ+qKK65QvXr1FBoaqgEDBmjfvn1OZbKzs/Xoo4+qQYMG8vf31+23366EhASnMnFxcerfv798fX0VGhqq8ePHKz8/36nMxo0bdfnll8vb21utWrXSwoULq/r2XOLvf/+7TCaTxowZ49hGHZbP0aNHde+996pBgwby8fFRhw4dtG3bNsd+wzA0ZcoUNWrUSD4+PoqJidH+/fudzpGUlKRBgwYpICBAQUFBGj58uNLT053K7Nq1S1dffbWsVqsiIyP10ksvVcv9VbWCggJNnjxZUVFR8vHxUcuWLTVjxgwVXWOPOizu22+/1U033aSIiAiZTCatWLHCaX911tny5cvVpk0bWa1WdejQQatWrar0+60qZdVjXl6ennrqKXXo0EF+fn6KiIjQ4MGDdezYMadzuHs9nu+1WNRDDz0kk8mkuXPnOm139zpE7VPX2+W0tcuPNrQz2sUlo717Fm3Y0tEuLR3tTRcwarilS5caXl5exgcffGDs3bvXGDFihBEUFGQkJCS4OjSXiI2NNRYsWGDs2bPH2Llzp3HDDTcYTZs2NdLT0x1lHnroISMyMtJYt26dsW3bNuMvf/mL0aNHD8f+/Px8o3379kZMTIzx448/GqtWrTJCQkKMiRMnOsr88ccfhq+vrzF27Fjj559/Nl5//XXDYrEYq1evrtb7rWpbt241mjdvbnTs2NEYPXq0Yzt1eH5JSUlGs2bNjKFDhxpbtmwx/vjjD+Orr74yDhw44Cjz97//3QgMDDRWrFhh/PTTT8bNN99sREVFGVlZWY4yffv2NTp16mT873//M/773/8arVq1MgYOHOjYn5KSYoSFhRmDBg0y9uzZY/zzn/80fHx8jHfeeada77cqvPDCC0aDBg2MlStXGgcPHjSWL19u+Pv7G/PmzXOUoQ6LW7VqlfHss88an376qSHJ+Oyzz5z2V1edff/994bFYjFeeukl4+effzYmTZpkeHp6Grt3767yOqgMZdVjcnKyERMTYyxbtsz49ddfjc2bNxtXXnml0bVrV6dzuHs9nu+1aPfpp58anTp1MiIiIoxXX33VaZ+71yFqF3dol9PWLh/a0M5oF5eO9u5ZtGFLR7u0dLQ3q1+NT5ReeeWVxqOPPur4uaCgwIiIiDBmzpzpwqhqjsTEREOS8c033xiGUfhLxNPT01i+fLmjzC+//GJIMjZv3mwYRuEbzWw2G/Hx8Y4yb7/9thEQEGDk5OQYhmEYEyZMMC677DKna919991GbGxsVd9StUlLSzNat25trF271ujVq5ejkUcdls9TTz1l9OzZs9T9NpvNCA8PN15++WXHtuTkZMPb29v45z//aRiGYfz888+GJOOHH35wlPnyyy8Nk8lkHD161DAMw3jrrbeM+vXrO+rVfu1LL720sm+p2vXv39+4//77nbbddtttxqBBgwzDoA7L49zGQnXW2V133WX079/fKZ7o6GjjwQcfrNR7rA5lNbrstm7dakgyDh8+bBgG9Xiu0urwzz//NBo3bmzs2bPHaNasmVPDlTpEbeOO7XLa2sXRhi6OdnHpaO+WjDZs6WiXlo72ZvWo0UPvc3NztX37dsXExDi2mc1mxcTEaPPmzS6MrOZISUmRJAUHB0uStm/frry8PKc6a9OmjZo2beqos82bN6tDhw4KCwtzlImNjVVqaqr27t3rKFP0HPYydaneH330UfXv37/YfVKH5fP555+rW7duuvPOOxUaGqouXbrovffec+w/ePCg4uPjneogMDBQ0dHRTvUYFBSkbt26OcrExMTIbDZry5YtjjLXXHONvLy8HGViY2O1b98+nT59uqpvs0r16NFD69at02+//SZJ+umnn/Tdd9+pX79+kqjDC1GddVbX3+PnSklJkclkUlBQkCTqsTxsNpvuu+8+jR8/Xpdddlmx/dQhahN3bZfT1i6ONnRxtItLR3u3fGjDVgzt0rNob1a+Gp0oPXnypAoKCpz+kEpSWFiY4uPjXRRVzWGz2TRmzBhdddVVat++vSQpPj5eXl5ejl8YdkXrLD4+vsQ6te8rq0xqaqqysrKq4naq1dKlS7Vjxw7NnDmz2D7qsHz++OMPvf3222rdurW++uorPfzww3r88ce1aNEiSWfroaz3b3x8vEJDQ532e3h4KDg4uEJ1XVs9/fTTuueee9SmTRt5enqqS5cuGjNmjAYNGiSJOrwQ1VlnpZWpa3UqFc4599RTT2ngwIEKCAiQRD2Wx6xZs+Th4aHHH3+8xP3UIWoTd2yX09YujjZ0yWgXl472bvnQhi0/2qXOaG9WPg9XB4AL9+ijj2rPnj367rvvXB1KrXLkyBGNHj1aa9euldVqdXU4tZbNZlO3bt304osvSpK6dOmiPXv2aP78+RoyZIiLo6sd/vWvf2nx4sVasmSJLrvsMu3cuVNjxoxRREQEdYgaIy8vT3fddZcMw9Dbb7/t6nBqje3bt2vevHnasWOHTCaTq8MBcAFoazujDV062sWlo72LykS71BntzapRo3uUhoSEyGKxFFspMSEhQeHh4S6KqmYYNWqUVq5cqQ0bNqhJkyaO7eHh4crNzVVycrJT+aJ1Fh4eXmKd2veVVSYgIEA+Pj6VfTvVavv27UpMTNTll18uDw8PeXh46JtvvtFrr70mDw8PhYWFUYfl0KhRI7Vr185pW9u2bRUXFyfpbD2U9f4NDw9XYmKi0/78/HwlJSVVqK5rq/Hjxzs+Ze/QoYPuu+8+PfHEE45eGtRhxVVnnZVWpi7Vqb0xevjwYa1du9bxqb1EPZ7Pf//7XyUmJqpp06aOvzWHDx/Wk08+qebNm0uiDlG7uFu7nLZ2cbShS0e7uHS0d8uHNuz50S4tjvZm1ajRiVIvLy917dpV69atc2yz2Wxat26dunfv7sLIXMcwDI0aNUqfffaZ1q9fr6ioKKf9Xbt2laenp1Od7du3T3FxcY466969u3bv3u30ZrH/orH/ge/evbvTOexl6kK99+nTR7t379bOnTsdj27dumnQoEGO76nD87vqqqu0b98+p22//fabmjVrJkmKiopSeHi4Ux2kpqZqy5YtTvWYnJys7du3O8qsX79eNptN0dHRjjLffvut8vLyHGXWrl2rSy+9VPXr16+y+6sOmZmZMpudfw1bLBbZbDZJ1OGFqM46q+vvcXtjdP/+/fr666/VoEEDp/3UY9nuu+8+7dq1y+lvTUREhMaPH6+vvvpKEnWI2sVd2uW0tUtHG7p0tItLR3u3fGjDlo12aclob1YR164ldX5Lly41vL29jYULFxo///yzMXLkSCMoKMhppUR38vDDDxuBgYHGxo0bjePHjzsemZmZjjIPPfSQ0bRpU2P9+vXGtm3bjO7duxvdu3d37M/Pzzfat29vXH/99cbOnTuN1atXGw0bNjQmTpzoKPPHH38Yvr6+xvjx441ffvnFePPNNw2LxWKsXr26Wu+3uhRdsdMwqMPy2Lp1q+Hh4WG88MILxv79+43Fixcbvr6+xscff+wo8/e//90ICgoy/v3vfxu7du0ybrnlFiMqKsrIyspylOnbt6/RpUsXY8uWLcZ3331ntG7d2hg4cKBjf3JyshEWFmbcd999xp49e4ylS5cavr6+xjvvvFOt91sVhgwZYjRu3NhYuXKlcfDgQePTTz81QkJCjAkTJjjKUIfFpaWlGT/++KPx448/GpKMOXPmGD/++KNj1cvqqrPvv//e8PDwMF555RXjl19+MaZOnWp4enoau3fvrr7KuAhl1WNubq5x8803G02aNDF27tzp9Pem6GqY7l6P53stnuvcVUgNgzpE7eIO7XLa2hVDG7oQ7eLS0d49izZs6WiXlo72ZvWr8YlSwzCM119/3WjatKnh5eVlXHnllcb//vc/V4fkMpJKfCxYsMBRJisry3jkkUeM+vXrG76+vsatt95qHD9+3Ok8hw4dMvr162f4+PgYISEhxpNPPmnk5eU5ldmwYYPRuXNnw8vLy2jRooXTNeqacxt51GH5fPHFF0b79u0Nb29vo02bNsa7777rtN9msxmTJ082wsLCDG9vb6NPnz7Gvn37nMqcOnXKGDhwoOHv728EBAQYw4YNM9LS0pzK/PTTT0bPnj0Nb29vo3Hjxsbf//73Kr+36pCammqMHj3aaNq0qWG1Wo0WLVoYzz77rNMffOqwuA0bNpT4e3DIkCGGYVRvnf3rX/8yLrnkEsPLy8u47LLLjP/85z9Vdt+Vrax6PHjwYKl/bzZs2OA4h7vX4/lei+cqqeHq7nWI2qeut8tpa1cMbeizaBeXjPbuWbRhS0e7tHS0N6ufyTAMo3L6pgIAAAAAAABA7VSj5ygFAAAAAAAAgOpAohQAAAAAAACA2yNRCgAAAAAAAMDtkSgFAAAAAAAA4PZIlAIAAAAAAABweyRKAQAAAAAAALg9EqUAAAAAAAAA3B6JUgAAAAAAAABuj0QpgFqvd+/eGjNmjKvD0MaNG2UymZScnOzqUAAAAFDH0QYGgMpHohQALkBNaZgCAAAA1YU2MIC6jkQpAAAAAAAAALdHohRAnZKTk6Nx48apcePG8vPzU3R0tDZu3OjYv3DhQgUFBemrr75S27Zt5e/vr759++r48eOOMvn5+Xr88ccVFBSkBg0a6KmnntKQIUM0YMAASdLQoUP1zTffaN68eTKZTDKZTDp06JDj+O3bt6tbt27y9fVVjx49tG/fvmq6ewAAALgj2sAAUDlIlAKoU0aNGqXNmzdr6dKl2rVrl+6880717dtX+/fvd5TJzMzUK6+8oo8++kjffvut4uLiNG7cOMf+WbNmafHixVqwYIG+//57paamasWKFY798+bNU/fu3TVixAgdP35cx48fV2RkpGP/s88+q9mzZ2vbtm3y8PDQ/fffXy33DgAAAPdEGxgAKoeHqwMAgMoSFxenBQsWKC4uThEREZKkcePGafXq1VqwYIFefPFFSVJeXp7mz5+vli1bSipsWE6fPt1xntdff10TJ07UrbfeKkl64403tGrVKsf+wMBAeXl5ydfXV+Hh4cXieOGFF9SrVy9J0tNPP63+/fsrOztbVqu1am4cAAAAbos2MABUHhKlAOqM3bt3q6CgQJdcconT9pycHDVo0MDxs6+vr6OBKEmNGjVSYmKiJCklJUUJCQm68sorHfstFou6du0qm81Wrjg6duzodG5JSkxMVNOmTSt+UwAAAEAZaAMDQOUhUQqgzkhPT5fFYtH27dtlsVic9vn7+zu+9/T0dNpnMplkGEalxVH0/CaTSZLK3cAEAAAAKoI2MABUHuYoBVBndOnSRQUFBUpMTFSrVq2cHiUNDypJYGCgwsLC9MMPPzi2FRQUaMeOHU7lvLy8VFBQUKnxAwAAABVFGxgAKg89SgHUGZdccokGDRqkwYMHa/bs2erSpYtOnDihdevWqWPHjurfv3+5zvPYY49p5syZatWqldq0aaPXX39dp0+fdnwyLknNmzfXli1bdOjQIfn7+ys4OLiqbgsAAAAoFW1gAKg89CgFUKcsWLBAgwcP1pNPPqlLL71UAwYM0A8//FChuZGeeuopDRw4UIMHD1b37t3l7++v2NhYp4nox40bJ4vFonbt2qlhw4aKi4uritsBAAAAzos2MABUDpNRmZOSAEAdZLPZ1LZtW911112aMWOGq8MBAAAAqhxtYADuiKH3AHCOw4cPa82aNerVq5dycnL0xhtv6ODBg/rb3/7m6tAAAACAKkEbGAAYeg8AxZjNZi1cuFBXXHGFrrrqKu3evVtff/212rZt6+rQAAAAgCpBGxgAGHoPAAAAAAAAAPQoBQAAAAAAAAASpQAAAAAAAADcHolSAAAAAAAAAG6PRCkAAAAAAAAAt0eiFAAAAAAAAIDbI1EKAAAAAAAAwO2RKAUAAAAAAADg9kiUAgAAAAAAAHB7JEoBAAAAAAAAuD0SpQAAAAAAAADcHolSAAAAAAAAAG6PRCkAAAAAAAAAt0eiFAAAAAAAAIDbI1EKAAAAAAAAwO2RKAUAAAAAAADg9kiUAgAAAAAAAHB7JEoBAAAAAAAAuD0SpQAAAAAAAADcHolSAAAAAAAAAG6PRCkAAAAAAAAAt0eiFAAAAAAAAIDbI1EKAAAAAAAAwO2RKAUAAAAAAADg9kiUAgAAAAAAAHB7JEoBAAAAAAAAuD0SpQAAAAAAAADcHolSAAAAAAAAAG6PRCmAWsNkMmnatGmOnxcuXCiTyaRDhw65LKYL1bx5cw0dOtTx88aNG2UymbRx40aXxQQAAGA3bdo0mUymarlW79691bt3b8fP9nbRJ598Ui3XHzp0qJo3b14t17pQ6enpeuCBBxQeHi6TyaQxY8Zc9DnPrfdDhw7JZDJp4cKFF33u6lbSc3ju/w7uYujQofL393d1GECtRaIUQI3w1ltvyWQyKTo62tWhAAAA1Cn2D5ftD6vVqoiICMXGxuq1115TWlpapVzn2LFjmjZtmnbu3Fkp56tMNTm28njxxRe1cOFCPfzww/roo4903333uTokt7Fp0yZNmzZNycnJrg7FITMzU9OmTaOTBVAFPFwdAABI0uLFi9W8eXNt3bpVBw4cUKtWrVwdUpXat2+fzGY+qwIAANVn+vTpioqKUl5enuLj47Vx40aNGTNGc+bM0eeff66OHTs6yk6aNElPP/10hc5/7NgxPffcc2revLk6d+5c7uPWrFlToetciLJie++992Sz2ao8houxfv16/eUvf9HUqVNdHUqNVJXP4aZNm/Tcc89p6NChCgoKqpJrVFRmZqaee+45SXLqFQzg4vFfOgCXO3jwoDZt2qQ5c+aoYcOGWrx4satDqnLe3t7y9PR0dRgAAMCN9OvXT/fee6+GDRumiRMn6quvvtLXX3+txMRE3XzzzcrKynKU9fDwkNVqrdJ4MjMzJUleXl7y8vKq0muVxdPTU97e3i67fnkkJibWmCRdTVQbnsO6xDAMp98XQF1CohSAyy1evFj169dX//79dccdd1RJovTXX3/VHXfcoeDgYFmtVnXr1k2ff/65Uxn7sLTvvvtOjz/+uBo2bKigoCA9+OCDys3NVXJysgYPHqz69eurfv36mjBhggzDcDrHK6+8oh49eqhBgwby8fFR165dS5xf69w5Skuyf/9+3X777QoPD5fValWTJk10zz33KCUl5aLrAwAAQJKuu+46TZ48WYcPH9bHH3/s2F7SHKVr165Vz549FRQUJH9/f1166aV65plnJBXOK3rFFVdIkoYNG+YY5m+f77J3795q3769tm/frmuuuUa+vr6OY8+dK9OuoKBAzzzzjMLDw+Xn56ebb75ZR44ccSpTWpuq6DnPF1tJ81tmZGToySefVGRkpLy9vXXppZfqlVdeKdb2M5lMGjVqlFasWKH27dvL29tbl112mVavXl1yhZ8jMTFRw4cPV1hYmKxWqzp16qRFixY59tvnaz148KD+85//OGIva47+BQsW6LrrrlNoaKi8vb3Vrl07vf322+WKp7ySk5M1ZswYR/20atVKs2bNcurVaZ/z9JVXXtGbb76pFi1ayNfXV9dff72OHDkiwzA0Y8YMNWnSRD4+PrrllluUlJTkdJ1///vf6t+/vyIiIuTt7a2WLVtqxowZKigocCpXnnlm09LSNGbMGDVv3lze3t4KDQ3VX//6V+3YsaPUY6ZNm6bx48dLkqKioorVf35+vmbMmKGWLVvK29tbzZs31zPPPKOcnByn89hsNk2bNk0RERHy9fXVtddeq59//rnE1+/56vbQoUNq2LChJOm5555zxHTufKxHjx7VgAED5O/vr4YNG2rcuHHF6s1ms2nu3Lm67LLLZLVaFRYWpgcffFCnT592Kte8eXPdeOON+uqrr9StWzf5+PjonXfeKbO+gdqKofcAXG7x4sW67bbb5OXlpYEDB+rtt9/WDz/84GjQXqy9e/fqqquuUuPGjfX000/Lz89P//rXvzRgwAD93//9n2699Van8o899pjCw8P13HPP6X//+5/effddBQUFadOmTWratKlefPFFrVq1Si+//LLat2+vwYMHO46dN2+ebr75Zg0aNEi5ublaunSp7rzzTq1cuVL9+/cvd8y5ubmKjY1VTk6OI56jR49q5cqVSk5OVmBgYKXUDQAAwH333adnnnlGa9as0YgRI0oss3fvXt14443q2LGjpk+fLm9vbx04cEDff/+9JKlt27aaPn26pkyZopEjR+rqq6+WJPXo0cNxjlOnTqlfv3665557dO+99yosLKzMuF544QWZTCY99dRTSkxM1Ny5cxUTE6OdO3fKx8en3PdXntiKMgxDN998szZs2KDhw4erc+fO+uqrrzR+/HgdPXpUr776qlP57777Tp9++qkeeeQR1atXT6+99ppuv/12xcXFqUGDBqXGlZWVpd69e+vAgQMaNWqUoqKitHz5cg0dOlTJyckaPXq02rZtq48++khPPPGEmjRpoieffFKSHImykrz99tu67LLLdPPNN8vDw0NffPGFHnnkEdlsNj366KPlrrfSZGZmqlevXjp69KgefPBBNW3aVJs2bdLEiRN1/PhxzZ0716n84sWLlZubq8cee0xJSUl66aWXdNddd+m6667Txo0b9dRTT+nAgQN6/fXXNW7cOH3wwQeOYxcuXCh/f3+NHTtW/v7+Wr9+vaZMmaLU1FS9/PLLFYr7oYce0ieffKJRo0apXbt2OnXqlL777jv98ssvuvzyy0s85rbbbtNvv/2mf/7zn3r11VcVEhIi6Wz9P/DAA1q0aJHuuOMOPfnkk9qyZYtmzpypX375RZ999pnjPBMnTtRLL72km266SbGxsfrpp58UGxur7OzsCtdtw4YN9fbbb+vhhx/Wrbfeqttuu02SnKbOKCgoUGxsrKKjo/XKK6/o66+/1uzZs9WyZUs9/PDDjnIPPvigFi5cqGHDhunxxx/XwYMH9cYbb+jHH3/U999/7zQCbt++fRo4cKAefPBBjRgxQpdeemmF6h+oNQwAcKFt27YZkoy1a9cahmEYNpvNaNKkiTF69OhiZSUZU6dOdfy8YMECQ5Jx8ODBMq/Rp08fo0OHDkZ2drZjm81mM3r06GG0bt262PliY2MNm83m2N69e3fDZDIZDz30kGNbfn6+0aRJE6NXr15O18rMzHT6OTc312jfvr1x3XXXOW1v1qyZMWTIEMfPGzZsMCQZGzZsMAzDMH788UdDkrF8+fIy7w0AAOB87G2cH374odQygYGBRpcuXRw/T5061Sj67+Krr75qSDJOnDhR6jl++OEHQ5KxYMGCYvt69eplSDLmz59f4r6ibSp7u6hx48ZGamqqY/u//vUvQ5Ixb948x7Zz21SlnbOs2IYMGWI0a9bM8fOKFSsMScbzzz/vVO6OO+4wTCaTceDAAcc2SYaXl5fTtp9++smQZLz++uvFrlXU3LlzDUnGxx9/7NiWm5trdO/e3fD393e692bNmhn9+/cv83x257ZHDcMwYmNjjRYtWjhtO7eODh48WGodFTVjxgzDz8/P+O2335y2P/3004bFYjHi4uKcztewYUMjOTnZUW7ixImGJKNTp05GXl6eY/vAgQMNLy8vpzZ7Sffy4IMPGr6+vk7lzn0ODaP4/w6BgYHGo48+Wua9leTll18u8X+OnTt3GpKMBx54wGn7uHHjDEnG+vXrDcMwjPj4eMPDw8MYMGCAU7lp06YZkpxev+Wt2xMnThS7P7shQ4YYkozp06c7be/SpYvRtWtXx8///e9/DUnG4sWLncqtXr262PZmzZoZkozVq1eXUENA3cLQewAutXjxYoWFhenaa6+VVDh86e6779bSpUuLDQ25EElJSVq/fr3uuusupaWl6eTJkzp58qROnTql2NhY7d+/X0ePHnU6Zvjw4U5DzaKjo2UYhoYPH+7YZrFY1K1bN/3xxx9Oxxbt3XD69GmlpKTo6quvLnNIT0nsPUa/+uorx/xdAAAAVcXf319paWml7rfPj/nvf//7ghfN8fb21rBhw8pdfvDgwapXr57j5zvuuEONGjXSqlWrLuj65bVq1SpZLBY9/vjjTtuffPJJGYahL7/80ml7TEyMWrZs6fi5Y8eOCggIKNZOLOk64eHhGjhwoGObp6enHn/8caWnp+ubb765oPiLtkdTUlJ08uRJ9erVS3/88UelTOG0fPlyXX311apfv76jbX3y5EnFxMSooKBA3377rVP5O++802k0VHR0tCTp3nvvlYeHh9P23Nxcp7Z50Xuxt+WvvvpqZWZm6tdff61Q3EFBQdqyZYuOHTtWoeNKY38djh071mm7vdfvf/7zH0nSunXrlJ+fr0ceecSp3GOPPVbsnBWt27I89NBDTj9fffXVTq/J5cuXKzAwUH/961+drtW1a1f5+/trw4YNTsdHRUUpNja23NcHaiuG3gNwmYKCAi1dulTXXnutDh486NgeHR2t2bNna926dbr++usv6hoHDhyQYRiaPHmyJk+eXGKZxMRENW7c2PFz06ZNnfbbG3aRkZHFtp87f8/KlSv1/PPPa+fOnU5zE507x9f5REVFaezYsZozZ44WL16sq6++WjfffLPuvfdeht0DAIBKl56ertDQ0FL333333Xr//ff1wAMP6Omnn1afPn1022236Y477pDZXL7+N40bN67Qok2tW7d2+tlkMqlVq1Zlzs9ZGQ4fPqyIiAinJK1UOITfvr+oc9uOklS/fv1i7cSSrtO6deti9Vfadcrr+++/19SpU7V58+ZiH7inpKRcdFty//792rVrV6nD/xMTE51+rkjbWpJTve3du1eTJk3S+vXrlZqa6lS+oknfl156SUOGDFFkZKS6du2qG264QYMHD1aLFi0qdB67w4cPy2w2q1WrVk7bw8PDFRQU5Hj+7F/PLRccHKz69es7bato3ZbGarUWO8e5r8n9+/crJSWl1Pf9udeKiooq17WB2o5EKQCXWb9+vY4fP66lS5dq6dKlxfYvXrz4ohOl9h4P48aNK/UT0HMbLRaLpcRyJW03ikzo/9///lc333yzrrnmGr311ltq1KiRPD09tWDBAi1ZsqTCsc+ePVtDhw7Vv//9b61Zs0aPP/64Zs6cqf/9739q0qRJhc8HAABQkj///FMpKSnF2kRF+fj46Ntvv9WGDRv0n//8R6tXr9ayZct03XXXac2aNaW2n849R2Ur7cPogoKCcsVUGUq7jnHOwk/V4ffff1efPn3Upk0bzZkzR5GRkfLy8tKqVav06quvXnBv4KJsNpv++te/asKECSXuv+SSS5x+rkjbWjpbb8nJyerVq5cCAgI0ffp0tWzZUlarVTt27NBTTz1V4Xu56667dPXVV+uzzz7TmjVr9PLLL2vWrFn69NNP1a9fvwqdq6iKdogoS0XrtjTlee3bbDaFhoaWupDuuYnWqnj/AjURiVIALrN48WKFhobqzTffLLbv008/1Weffab58+df1B9l+yfEnp6eiomJueDzlMf//d//yWq16quvvpK3t7dj+4IFCy74nB06dFCHDh00adIkbdq0SVdddZXmz5+v559/vjJCBgAA0EcffSRJ5x1Wazab1adPH/Xp00dz5szRiy++qGeffVYbNmxQTExMpSaMpMIeb0UZhqEDBw44LVpTv359JScnFzv28OHDTj0FKxJbs2bN9PXXXystLc2pV6l9qHezZs3Kfa7zXWfXrl2y2WxOvUov5jpffPGFcnJy9Pnnnzv15Dx3GPXFaNmypdLT06u8bb1x40adOnVKn376qa655hrH9qIj0SqqUaNGeuSRR/TII48oMTFRl19+uV544YUyE6WlvXaaNWsmm82m/fv3O3oBS1JCQoKSk5Mdz5/964EDB5x6ZZ46dapYr+Py1m1lvNdatmypr7/+WldddRVJUKAI5igF4BJZWVn69NNPdeONN+qOO+4o9hg1apTS0tL0+eefX9R1QkND1bt3b73zzjs6fvx4sf0nTpy4qPMXZbFYZDKZnOZWPXTokFasWFHhc6Wmpio/P99pW4cOHWQ2m52G9AMAAFyM9evXa8aMGYqKitKgQYNKLZeUlFRsW+fOnSXJ0Tbx8/OTpBITlxfiww8/dJo39ZNPPtHx48edklotW7bU//73P+Xm5jq2rVy5UkeOHHE6V0Viu+GGG1RQUKA33njDafurr74qk8l0Ub0Pz71OfHy8li1b5tiWn5+v119/Xf7+/urVq1eFz2nvSVi0N2tKSspFfXB/rrvuukubN2/WV199VWxfcnJysTbshSrpXnJzc/XWW29V+FwFBQXFhuqHhoYqIiLivG3r0l47N9xwgyRp7ty5TtvnzJkjSerfv78kqU+fPvLw8NDbb7/tVO7c15dU/rr19fUtMaaKuOuuu1RQUKAZM2YU25efn19p72OgtqFHKQCX+Pzzz5WWlqabb765xP1/+ctf1LBhQy1evFh33333RV3rzTffVM+ePdWhQweNGDFCLVq0UEJCgjZv3qw///xTP/3000Wd365///6aM2eO+vbtq7/97W9KTEzUm2++qVatWmnXrl0VOtf69es1atQo3XnnnbrkkkuUn5+vjz76SBaLRbfffnulxAsAANzLl19+qV9//VX5+flKSEjQ+vXrtXbtWjVr1kyff/65rFZrqcdOnz5d3377rfr3769mzZopMTFRb731lpo0aaKePXtKKkxaBgUFaf78+apXr578/PwUHR19wXMbBgcHq2fPnho2bJgSEhI0d+5ctWrVSiNGjHCUeeCBB/TJJ5+ob9++uuuuu/T777/r448/dlpcqaKx3XTTTbr22mv17LPP6tChQ+rUqZPWrFmjf//73xozZkyxc1+okSNH6p133tHQoUO1fft2NW/eXJ988om+//57zZ07t9gcqeVx/fXXy8vLSzfddJMefPBBpaen67333lNoaGiJnQYuxPjx4/X555/rxhtv1NChQ9W1a1dlZGRo9+7d+uSTT3To0CGFhIRc9HV69Oih+vXra8iQIXr88cdlMpn00UcfXdCUBmlpaWrSpInuuOMOderUSf7+/vr666/1ww8/aPbs2WUe27VrV0nSs88+q3vuuUeenp666aab1KlTJw0ZMkTvvvuuY5qArVu3atGiRRowYIBjsdqwsDCNHj1as2fP1s0336y+ffvqp59+0pdffqmQkBCn3qHlrVsfHx+1a9dOy5Yt0yWXXKLg4GC1b99e7du3L3ed9OrVSw8++KBmzpypnTt36vrrr5enp6f279+v5cuXa968ebrjjjsqXNdAbUeiFIBLLF68WFarVX/9619L3G82m9W/f38tXrxYp06dUoMGDS74Wu3atdO2bdv03HPPaeHChTp16pRCQ0PVpUsXTZky5YLPe67rrrtO//jHP/T3v/9dY8aMUVRUlGbNmqVDhw5VOFHaqVMnxcbG6osvvtDRo0fl6+urTp066csvv9Rf/vKXSosZAAC4D3u7x8vLS8HBwerQoYPmzp2rYcOGnTcpd/PNN+vQoUP64IMPdPLkSYWEhKhXr1567rnnHIvweHp6atGiRZo4caIeeugh5efna8GCBRecKH3mmWe0a9cuzZw5U2lpaerTp4/eeustR286qXC6gNmzZ2vOnDkaM2aMunXrppUrVzpWHrerSGxms1mff/65pkyZomXLlmnBggVq3ry5Xn755WLnvRg+Pj7auHGjnn76aS1atEipqam69NJLtWDBAg0dOvSCznnppZfqk08+0aRJkzRu3DiFh4fr4YcfVsOGDXX//fdXSty+vr765ptv9OKLL2r58uX68MMPFRAQoEsuucTp9XCxGjRo4HguJ02apPr16+vee+9Vnz59Krz6uq+vrx555BGtWbNGn376qWw2m1q1aqW33npLDz/8cJnHXnHFFZoxY4bmz5+v1atXy2az6eDBg/Lz89P777+vFi1aaOHChfrss88UHh6uiRMnaurUqU7nmDVrlnx9ffXee+/p66+/Vvfu3bVmzRr17NnT6QOKitTt+++/r8cee0xPPPGEcnNzNXXq1AolSiVp/vz56tq1q9555x0988wz8vDwUPPmzXXvvffqqquuqtC5gLrCZLhihmkAAAAAAAA3lZycrPr16+v555/Xs88+6+pwAJzBHKUAAAAAAABVJCsrq9g2+9ymvXv3rt5gAJSJofcAAAAAAABVZNmyZVq4cKFuuOEG+fv767vvvtM///lPXX/99QxxB2oYEqUAAAAAAABVpGPHjvLw8NBLL72k1NRUxwJPzz//vKtDA3AO5igFAAAAAAAA4PaYoxQAAAAAAACA2yNRCgAAAAAAAMDtXfAcpTabTceOHVO9evVkMpkqMyYAAABUAcMwlJaWpoiICJnNfF5+IWgDAwAA1C4VaQNfcKL02LFjioyMvNDDAQAA4CJHjhxRkyZNXB1GrUQbGAAAoHYqTxv4ghOl9erVc1wkICDgQk8DAACAapKamqrIyEhHOw4VRxsYAACgdqlIG/iCE6X2oUYBAQE0EgEAAGoRhoxfONrAAAAAtVN52sBMTgUAAAAAAADA7ZEoBQAAAAAAAOD2SJQCAAAAAAAAcHskSgEAAAAAAAC4PRKlAAAAAAAAANweiVIAAAAAAAAAbo9EKQAAAOBmDMPQnDX79Mb6/a4OBQAAoMbwcHUAAAAAAKrX4VOZem39AUnSoOhmqu/n5eKIAAAAXI8epQAAAICb2X74tOP7P06muzASAACAmoNEKQAAAOBmdsSdTZT+fiLDhZEAAADUHCRKAQAAADfj1KOURCkAAIAkEqUAAACAW0nLztNvCWmOnw8y9B4AAEASiVIAAADArfx0JEU24+zP9CgFAAAoRKIUAAAAcCP2+UkvbxokSTp8KlMFRTOnAAAAbopEKQAAAOBG7POT3tgxQl4eZuUW2PTn6UwXRwUAAOB6JEoBAAAAN2GzGfrxTI/SK5oHK6qBnyTpj5MMvwcAACBRCgAAALiJ30+kKzU7Xz6eFrVpVE8tGp5JlDJPKQAAgDxcHUBlMAxD2dnZpe7LycmRJHl7e8tkMl3wdaxW60UdDwAAALiSfdh9xyaB8rSYFRViT5SmuzIsAACAGqFOJEqzs7PVr1+/Kr/Ol19+KR8fnyq/DgAAAFAV7As5dW1WX5LUoqG/JOkgQ+8BAAAYeg8AAIDK9+abb6p58+ayWq2Kjo7W1q1byyy/fPlytWnTRlarVR06dNCqVauc9huGoSlTpqhRo0by8fFRTEyM9u/f71TmhRdeUI8ePeTr66ugoKBi1/jpp580cOBARUZGysfHR23bttW8efMu+l5rkx1xyZKky5vaE6UMvQcAALCrEz1Ki0rvPFCGuchtFeSp3k9LJUlpne6RLJ4VOp/Jli//nf+szBABAADqtGXLlmns2LGaP3++oqOjNXfuXMXGxmrfvn0KDQ0tVn7Tpk0aOHCgZs6cqRtvvFFLlizRgAEDtGPHDrVv316S9NJLL+m1117TokWLFBUVpcmTJys2NlY///yzrFarJCk3N1d33nmnunfvrn/84x/FrrN9+3aFhobq448/VmRkpDZt2qSRI0fKYrFo1KhRVVspNcSx5CxJUqvQwp6kLc4MvY9PzVZGTr78vOvcvwcAAADlVud6lBpmj8JkaNGH3bnby/FwSroCAADgvObMmaMRI0Zo2LBhateunebPny9fX1998MEHJZafN2+e+vbtq/Hjx6tt27aaMWOGLr/8cr3xxhuSCnuTzp07V5MmTdItt9yijh076sMPP9SxY8e0YsUKx3mee+45PfHEE+rQoUOJ17n//vs1b9489erVSy1atNC9996rYcOG6dNPP630OqiJ8gpsyswtkCQF+hS2kYN8vRTs5yWJ4fcAAAB1LlEKAAAA18nNzdX27dsVExPj2GY2mxUTE6PNmzeXeMzmzZudyktSbGyso/zBgwcVHx/vVCYwMFDR0dGlnrO8UlJSFBwcXOr+nJwcpaamOj1qq5SsPMf3AT5nOxPYe5X+QaIUAAC4ORKlAAAAqDQnT55UQUGBwsLCnLaHhYUpPj6+xGPi4+PLLG//WpFzlsemTZu0bNkyjRw5stQyM2fOVGBgoOMRGRl5wddzNXuitJ63hyxmk2M7K98DAAAUIlEKAAAAt7Nnzx7dcsstmjp1qq6//vpSy02cOFEpKSmOx5EjR6oxysplT5QW7U0qSfaV71nQCQAAuDsSpQAAAKg0ISEhslgsSkhIcNqekJCg8PDwEo8JDw8vs7z9a0XOWZaff/5Zffr00ciRIzVp0qQyy3p7eysgIMDpUVvZE6WBxRKl9qH39CgFAADujUQpAAAAKo2Xl5e6du2qdevWObbZbDatW7dO3bt3L/GY7t27O5WXpLVr1zrKR0VFKTw83KlMamqqtmzZUuo5S7N3715de+21GjJkiF544YUKHVvbpZaSKI2s7ytJOp6cXe0xAQAA1CQs6Q4AAIBKNXbsWA0ZMkTdunXTlVdeqblz5yojI0PDhg2TJA0ePFiNGzfWzJkzJUmjR49Wr169NHv2bPXv319Lly7Vtm3b9O6770qSTCaTxowZo+eff16tW7dWVFSUJk+erIiICA0YMMBx3bi4OCUlJSkuLk4FBQXauXOnJKlVq1by9/fXnj17dN111yk2NlZjx451zG9qsVjUsGHD6qsgFymtR2l9v8Kfk7PyZBiGTCZTsWMBAADcAYlSAAAAVKq7775bJ06c0JQpUxQfH6/OnTtr9erVjsWY4uLiZDafHdjUo0cPLVmyRJMmTdIzzzyj1q1ba8WKFWrfvr2jzIQJE5SRkaGRI0cqOTlZPXv21OrVq2W1Wh1lpkyZokWLFjl+7tKliyRpw4YN6t27tz755BOdOHFCH3/8sT7++GNHuWbNmunQoUNVVR01RkpmKYlSXy9JUoHNUGp2frH9AAAA7sJkGIZxIQempqYqMDBQKSkpLp+rKSsrS/369ZMkpV1+n2Qp0rgryFO9HR+VvK88ihz/5ZdfysfHp1JiBgAAqG41qf1WW9XmOnx+5c96/7uDGnlNCz1zQ1unfW0mf6nsPJu+Gd9bzRr4uShCAACAyleR9htzlAIAAABuoLSh99LZXqWnz/Q6BQAAcEckSgEAAAA3YE+UBpSQKA1yJEpzqzUmAACAmoREKQAAAOAGyu5RemZBJxKlAADAjZEoBQAAANxAuYbeZzD0HgAAuC8SpQAAAIAbSC0jURpEj1IAAAASpQAAAIA7YDEnAACAspEoBQAAAOq4vAKbMnILJJXdo5TFnAAAgDsjUQoAAADUcfZh95IUYPUott/eozSZHqUAAMCNkSgFAAAA6jj7sHt/bw95WIr/C1Dfjx6lAAAAJEoBAACAOq6s+UklKYgepQAAACRKAQAAgLrOnigNKCVRenYxJ3qUAgAA90WiFAAAAKjjzvYoLT4/qSTVP7OYU2ZugXLyC6otLgAAgJqERCkAAABQx6WeZ+h9gNVTZlPh9wy/BwAA7opEKQAAAFDHnW+OUrPZ5NjH8HsAAOCuSJQCAAAAddz5EqVSkXlKM+hRCgAA3BOJUgAAAKCOK0+iNOjMPKXJ9CgFAABuikQpAAAAUMdVqEcpc5QCAAA3RaIUAAAAqOPsidKAMnuU2hOl9CgFAADuiUQpAAAAUMelZOVLOl+PUobeAwAA90aiFAAAAKjjUssz9N6PofcAAMC9kSgFAAAA6jgWcwIAADg/EqUAAABAHZZfYFN6TnmG3tOjFAAAuDcSpQAAAKh0b775ppo3by6r1aro6Ght3bq1zPLLly9XmzZtZLVa1aFDB61atcppv2EYmjJliho1aiQfHx/FxMRo//79TmVeeOEF9ejRQ76+vgoKCirxOnFxcerfv798fX0VGhqq8ePHKz8//6LutaZLzT57f2Ut5lSfxZwAAICbI1EKAACASrVs2TKNHTtWU6dO1Y4dO9SpUyfFxsYqMTGxxPKbNm3SwIEDNXz4cP34448aMGCABgwYoD179jjKvPTSS3rttdc0f/58bdmyRX5+foqNjVV2drajTG5uru688049/PDDJV6noKBA/fv3V25urjZt2qRFixZp4cKFmjJlSuVWQA1jH3bv52WRp6X05n99P/vQe3qUAgAA90SiFAAAAJVqzpw5GjFihIYNG6Z27dpp/vz58vX11QcffFBi+Xnz5qlv374aP3682rZtqxkzZujyyy/XG2+8IamwN+ncuXM1adIk3XLLLerYsaM+/PBDHTt2TCtWrHCc57nnntMTTzyhDh06lHidNWvW6Oeff9bHH3+szp07q1+/fpoxY4befPNN5ebW3V6U5ZmfVDrbozQ5M1c2m1HlcQEAANQ0JEoBAABQaXJzc7V9+3bFpl/JMAAAWbtJREFUxMQ4tpnNZsXExGjz5s0lHrN582an8pIUGxvrKH/w4EHFx8c7lQkMDFR0dHSp5yztOh06dFBYWJjTdVJTU7V3794Sj8nJyVFqaqrTo7axJ0rLGnYvnV3MyWZIadl1ezoCAACAkpAoBQAAQKU5efKkCgoKnJKRkhQWFqb4+PgSj4mPjy+zvP1rRc5ZkesUvca5Zs6cqcDAQMcjMjKy3NerKcrbo9TbwyJfL4sk5ikFAADuiUQpAAAAUIqJEycqJSXF8Thy5IirQ6qw8iZKJRZ0AgAA7o1EKQAAACpNSEiILBaLEhISnLYnJCQoPDy8xGPCw8PLLG//WpFzVuQ6Ra9xLm9vbwUEBDg9apvUCiRK7cPvWdAJAAC4IxKlAAAAqDReXl7q2rWr1q1b59hms9m0bt06de/evcRjunfv7lRektauXesoHxUVpfDwcKcyqamp2rJlS6nnLO06u3fvVmJiotN1AgIC1K5du3Kfp7ahRykAAED5eLg6AAAAANQtY8eO1ZAhQ9StWzddeeWVmjt3rjIyMjRs2DBJ0uDBg9W4cWPNnDlTkjR69Gj16tVLs2fPVv/+/bV06VJt27ZN7777riTJZDJpzJgxev7559W6dWtFRUVp8uTJioiI0IABAxzXjYuLU1JSkuLi4lRQUKCdO3dKklq1aiV/f39df/31ateune677z699NJLio+P16RJk/Too4/K29u7WuuoOqVkVrxH6Wl6lAIAADdEohQAAACV6u6779aJEyc0ZcoUxcfHq3Pnzlq9erVj4aS4uDiZzWcHNvXo0UNLlizRpEmT9Mwzz6h169ZasWKF2rdv7ygzYcIEZWRkaOTIkUpOTlbPnj21evVqWa1WR5kpU6Zo0aJFjp+7dOkiSdqwYYN69+4ti8WilStX6uGHH1b37t3l5+enIUOGaPr06VVdJS7l6FHqW/4epckX0KN00aZDysor0N3dIlXfz6vCxwMAALharUiUGoah7OxsSZLVapXJZHJxRK5DXQAAgNpg1KhRGjVqVIn7Nm7cWGzbnXfeqTvvvLPU85lMJk2fPr3MpObChQu1cOHCMuNq1qyZVq1aVWaZuiY1uyJD7+09SiuWKD2ZnqOpn++VJM39+jfd0bWJHundShFBPhWMFgAAwHVqxRyl2dnZ6tevn/r16+dIEror6gIAAAAVkZadL0ny9z5/H4kgxxylFRt6H59ytl2anWfTx/+L0yOLd1ToHAAAAK5WKxKlAAAAAC5MRk5hotSvHInS+n72Ve8r3qNUkto2CtDrAwunPDiSlFmhcwAAALgaiVIAAACgDkvPqUCPUp/CHqX2eU3L60RaYaK0YT1vXRkVLElKzsqTzWZU6DwAAACuRKIUAAAAqMMycwskla9HqX3Bp+QKDr0/mV7YAzXE30tBZ85RYDMcw/4BAABqAxKlAAAAQB1lGIYycu1D7y3nLR90ZsGnlAonSs/0KPX3lreHRX5ehdeq6KJQAAAArkSiFAAAAKijMnMLZJwZ/V6RxZzScvKVV2Ar93XsidIQf2+n8ySRKAUAALUIiVIAAACgjrIv5GQyST6e5+9RGmA9m0xNrcA8pY5Eab3CBGmwX+HXii4KBQAA4EokSgEAAIA6yr6Qk5+Xh0wm03nLe1jMqncmWZpckURpWmFCtKG/VZIc85QmZVRsCD8AAIArkSgFAAAA6qiMHPtCTufvTWoXdAELOp2gRykAAKgDSJQCAAAAddTZhZzOPz+pXZBPYZIzJat8Sc78Aptj0Sb7HKX17XOUZpAoBQAAtQeJUgAAAKCOss9RWp6FnOwq2qM0KSNXhiGZTWcTpPavpyvQKxUAAMDVSJQCAAAAdVTROUrLK9CnYolS+7D7YD9vWf6/vXuPi6rM/wD+mQszw30EFMQ0KS0xMBWTMEt3ZcOyC2mmRqLmarXRaqSWpthSreVmqekvft00N0jzt+VurlEsau4Wi4Ka18zdTDQFQeR+n3l+fwxzYGAGZmBgZuDzfr3mBZzznHOe850LD1+ei9wwD2ofT8M5rrFHKREREbkQJkqJiIiIiHqopjlKO9Cj1MrFnIoqjMPuVdK2ph6lTJQSERGR62CilIiIiIiohzIOvbdpMSfjHKVWJjmLyg09Svt6q6VtTJQSERGRK2KilIiIiIioh5KG3ndpj9LGRKlXU6LUeA7OUUpERESuhIlSIiIiIqIeqqrO9sWcbJ6jtLFHaUCzHqV+nqrGc9RBCGH1tYmIiIgciYlSIiIiIqIeqsI4R6kNizlpG4fN29qj1NwcpfU6IfVqJSIiInJ2TJQSEREREfVQHZqjtHHYvNVzlEqLOTX1KHVXKaBxM/ypYW3PVCIiIiJHY6KUiIiIiKiHMiZKbRl6r3Xv2BylzROlQFOv0uJKLuhEREREroGJUiIiIiKiHso47N3DljlKjT1Kq+uh17c/v2h7iVKufE9ERESugolSIiIiIqIeqlJazMn6offGxZyEAMpr2p5fVKcXUo/Rvt4tEqWehvMwUUpERESugolSIiIiIrK7zZs3Y/DgwdBoNIiMjMTBgwfbLL9z504MGzYMGo0G4eHh2LNnj8l+IQSSkpLQv39/uLu7Izo6GmfPnjUpU1xcjLi4OPj4+ECr1WL+/PmoqKgwKfPVV1/h9ttvh7e3N/r27Ytp06bh559/tss9O6OqDizmpFYq4KEyJFZLqttOcl6trIVeAHJZ00r3RlKP0krOUUpERESugYlSIiIiIgIA/PTTT3Y5z44dO5CYmIjVq1fj8OHDuPXWWxETE4MrV66YLf/dd99h1qxZmD9/Po4cOYLY2FjExsbixIkTUpm1a9di48aNSElJQXZ2Njw9PRETE4OamhqpTFxcHE6ePImMjAzs3r0bBw4cwMKFC6X9586dw4MPPohf//rXOHr0KL766isUFRVh6tSpdrlvZ1QhLeZkfaIUaDZPaTsLMRWVGxKpfp4qKOQyk30cek9ERESuholSIiIiIgIADBkyBL/61a/w8ccfmyQgbfXmm29iwYIFmDdvHoYPH46UlBR4eHjgww8/NFt+w4YNmDx5MpYuXYrQ0FC8/PLLGD16NDZt2gTA0Jt0/fr1WLlyJR588EGMGDEC27Ztw6VLl7Br1y4AwOnTp5Geno73338fkZGRGD9+PN5++21s374dly5dAgDk5uZCp9PhlVdewY033ojRo0djyZIlOHr0KOrre2avx44s5gQAvo1JzvYWdLI0PykA9PFkopSIiIhcCxOlRERERAQAOHz4MEaMGIHExEQEBQXhiSeeaHfIfEt1dXXIzc1FdHS0tE0ulyM6OhpZWVlmj8nKyjIpDwAxMTFS+XPnziE/P9+kjK+vLyIjI6UyWVlZ0Gq1GDNmjFQmOjoacrkc2dnZAICIiAjI5XJs2bIFOp0OpaWl+POf/4zo6Gi4ubnZdJ+uQK8XqKxrHHrf4R6lbSc520yUNi4KxaH3RERE5CqYKCUiIiIiAMDIkSOxYcMGXLp0CR9++CEuX76M8ePHIywsDG+++SYKCwvbPUdRURF0Oh0CAwNNtgcGBiI/P9/sMfn5+W2WN35tr0y/fv1M9iuVSvj5+UllQkJC8PXXX2PFihVQq9XQarW4ePEiPv30U4v3U1tbi7KyMpOHq6iq10nfe9qwmBMAaBuTnKVW9yhVtdrnxx6lRERE5GKYKCUiIiIiE0qlElOnTsXOnTvx+uuv4z//+Q+WLFmCgQMHIj4+HpcvX3Z0FTskPz8fCxYswJw5c3Do0CF88803UKlUePjhhyGEMHvMmjVr4OvrKz0GDhzYzbXuOOOwe7kMcHfrWKK03TlKK8yveG84hyFRWlzJRCkRERG5BiZKiYiIiMhETk4Ofve736F///548803sWTJEvz3v/9FRkYGLl26hAcffNDisQEBAVAoFCgoKDDZXlBQgKCgILPHBAUFtVne+LW9Mi0Xi2poaEBxcbFUZvPmzfD19cXatWsxatQo3HXXXfj444+RmZkpDc9vafny5SgtLZUeFy5csHjvzsaYKPVUKSGTydopbcrXvXGO0nYSpYXllofe+3lYdw4iIiIiZ8FEKREREREBMCzCFB4ejnHjxuHSpUvYtm0bzp8/j1deeQUhISG48847sXXrVhw+fNjiOVQqFSIiIpCZmSlt0+v1yMzMRFRUlNljoqKiTMoDQEZGhlQ+JCQEQUFBJmXKysqQnZ0tlYmKikJJSQlyc3OlMnv37oVer0dkZCQAoKqqCnK5afNXoVBIdTRHrVbDx8fH5OEqKms7Nj8p0KxHaXXH5yg1nuNaVZ3FHrtEREREzsT2VhMRERER9UjvvPMOHn/8ccydOxf9+/c3W6Zfv3744IMP2jxPYmIi5syZgzFjxmDs2LFYv349KisrMW/ePABAfHw8BgwYgDVr1gAAFi1ahAkTJmDdunWYMmUKtm/fjpycHLz77rsAAJlMhsWLF+OVV17B0KFDERISglWrViE4OBixsbEAgNDQUEyePBkLFixASkoK6uvrkZCQgJkzZyI4OBgAMGXKFLz11ltITk7GrFmzUF5ejhUrVuD666/HqFGj7BFCp1Jh7FFq4/ykQNNiTqXW9ig1M/TeOEdpbYMe1fU6eKj4pwcRERE5N7ZWiIiIiAiAoRfnoEGDWvW6FELgwoULGDRoEFQqFebMmdPmeWbMmIHCwkIkJSUhPz8fI0eORHp6urQYU15ensk1xo0bh7S0NKxcuRIrVqzA0KFDsWvXLoSFhUllli1bhsrKSixcuBAlJSUYP3480tPTodFopDKpqalISEjApEmTIJfLMW3aNGzcuFHa/+tf/xppaWlYu3Yt1q5dCw8PD0RFRSE9PR3u7u6dip0zMg699+pUj1Lr5ig1t5iTh0oBlUKOOp0e16rqmSglIiIip8fWChEREREBAG688UZcvny51erxxcXFCAkJgU6ns3BkawkJCUhISDC7b//+/a22TZ8+HdOnT7d4PplMhuTkZCQnJ1ss4+fnh7S0tDbrNXPmTMycObPNMj1FZZ0hUdqRBGXTHKWWh97r9ALFlYYepX3NDL2XyWTo4+mGgrJaXKuswwBtz0tGExERUc/COUqJiIiICAAsziNZUVFh0nOTXEPT0PuO9ygtbaNH6bWqOugFIJM1DbNvqU/jgk7X2ki4EhERETkL9iglIiIi6uUSExMBGHoAJiUlwcPDQ9qn0+mQnZ2NkSNHOqh21FFVjYs5eXVkjlLj0PuqegghIJPJWpUxLuTk56GCUmG+/4XxPMWVTJQSERGR82OilIiIiKiXO3LkCABDj9Ljx49DpWrqHahSqXDrrbdiyZIljqoedVCnepQ2Dr1v0AtU1unMznMqLeRkZti9kbGnaUk7i0IREREROQMmSomIiIh6uX379gEA5s2bhw0bNsDHx8fBNSJ76MxiTho3OVRKOeoa9CipqjN7DmOP0gBv88PuAUDbOPSePUqJiIjIFXCOUiIiIiICAGzZsoVJ0h7EuJhTR3qUymQyaN2bht+bU1RuXPG+jR6lHu0vCkVERETkLNijlIiIiKgXmzp1KrZu3QofHx9MnTq1zbKfffZZN9WK7KGicY7SjiRKAcP8olfKay0u6CT1KG0jUSrNUcqh90REROQCmCglIiIi6sV8fX2lhXp8fX0dXBuyJ+PQe0+V7Ys5AU3zlFrqUVpoRaK0D3uUEhERkQthopSIiIioF9uyZYvZ78n1VXZiMScA8G3sDVpSbT7JWVRh2N7X23Ki1FtjuHZ5TUOH6kBERETUnZgodWH33HOP3c+pUqlQV1cHuVwOvV5v9XFKpRJ6vb7VMRMnTsTRo0dRUVGBhoYGhIWF4dy5c6iqqsItt9yCkydPQqVSQalUws3NDcuWLcPp06fx5z//2eTcXl5eWLZsGcaNGydt/+677/Dqq6+iqqoKjz32GObPny9t/8Mf/oDa2lqo1WqsXr1aOu6ll17C/v37AQBhYWE4ceKE9P2jjz6KDRs2YNGiRRg3bpxUduLEiXjppZfw3Xffmew3p2WZDz74AKmpqYiLi5Pq17IcAKxduxYAcP/99+Ozzz5DVVUVJkyYgNOnT2PRokVIS0uT6goAs2fPBgCkpqbirrvuwunTpxEaGopvvvkGKpUK7u7uuP/++/H111+jb9++Jsca7+eDDz4wifPs2bMRGhoqxdTDwwMvvviiSSzCwsJQWFgo3Z81MWkvRs1Zipe9dKS+1p7HuO3uu+/G119/bVOMjPc9fPhwnDx5EgqFAg0NDVCrDX94Gl/L7u7uWLZsGQDDa6a+vh5ubm64//778cUXXwBAq/eJsW7G11jL/V0dc3tpGUd7PZddreXnCLXN0vvI0fWxph7mPteN71FL70tnulej6upqCCHg4eEBADh//jw+//xzDB8+HHfffbeDa0e2Ms5R2pHFnAC0O0dp06r3lhdz8mk8R1kNh94TERGR85MJIURHDiwrK4Ovry9KS0u7fNL/6upqKSn45Zdfwt3d3eL+8tGzAYVb005dPbwP/9n8Pms0O97ctbtb83vtifr06YOSkhKYe1n6+/sjNTUVGo0GNTU1ePTRR1FcXAzAsODA559/Do1Gg1mzZuHatWsm5/zkk09QWlqKGTNmWLy2VqtFSUkJAgIC8OabbyI+Pl7at23bNiQmJqKoqAgBAQH4+OOPodFoTI6vqanBY489JpXZvHkzZs2aBb1eD7lcjs8++wxardaknL+/P4QQJvfR8t6N9WrOXLmW2irz3nvvYcGCBRZjYOTn54f169ebxAIAAgIC8P777+O3v/1tmzFpqWWMmh9TUlKCqVOntoqXvbR17c6eB4C0zfhPBmtj1Py+rdHyNQOYPtfN3yfG+sbFxeHq1aut9nd1zO2lZcw78tpzhIKCApPPnB07diAwMNCBNXJuzZ/n5u8jRz2/tnxmtPW5Dph/XzrqXttrv919992YOnUqnnzySZSUlODmm2+GSqVCUVER3nzzTTz11FNdWj9X0J1t4M761Rv7ca6oEp8+EYWxIX42H//q30/hvX+ew8K7bsCKe0Nb7b/t1X+gsLwWu58Zj7AB5qdtOHWpDPdu/Cf6eqtx6MVom+tARERE1Fm2tN+46j05lWvXrllM7l29ehVpaWkADD0pm/8RKoRAUlISUlNTTZKkxnOmpaUhISGhzWsbE4RXr17Fk08+abLvySeflBJNzevRXGpqqkmZhIQEKfml1+uRlJRktlzL+7BUr+as+f9GW2Us/aHb8lrFxcWtYgEY6r1q1ap2Y9JSy3tvfsyqVavMxste2rp2Z8/TfJvxHqyNUfP7tkbL1wxg+ly3vE7zurXc39Uxt5eWMe/Ia88RWn7mPPPMMw6qiWuw9D5y1PNry2dGW5/r5o53tntt7vDhw7jzzjsBAP/3f/+HoKAgnD9/Htu2bcPGjRsdXDuyVYU09L5jc5T6eRpGNhQ19hxtTq8XKK60fuh9mYUFoYiIiIiciUskSpsnAWpqalBdXW3yqKmpaV7Y3hdv89rd/Zg5c6Z978/FpKamIicnR+rF19yxY8dMhpI3t23bNhQWFlp1DSEEqqqqTLZVVVVJr0MhBNLS0nDx4kVp/8WLF5GWlmZSpuX1jh07hvT0dJNyjtLQYP08YS1jARju7/jx423GpCVzMTIek5OTg+PHj5uUP3bsGHJycqyuZ1vaunZnz5Oammr2ObUmRubu2x5SU1Nx8eJFqb7m9qenp3dpzO3FXMxtfe05Qnp6eqvPgCtXriA9Pd1BNXJuLZ9nI0c9v7Z8Zliqe0st35fOcq8tVVVVwdvbGwDw9ddfY+rUqZDL5bj99ttx/vx5h9aNbGeco7SjQ++DtYYezpdKq1vtu1ZVB53e8Dr282x/6H1tgx61DboO1YOIiIiou1jdaqqtrUVtbdN/k8vKyrqkQpaubfTQQw+1XVjfAMByY81m+qaEUrvXpi6n0+nw0ksvOTzRKITAhg0bpHkfN2zYYNVxf/rTn2zqPehKmsfEuHpyy33mjnnrrbdw9uxZs+dMTk7Grl27IJd3/H86bV3bUn1tOY9OZ/0ffc2vKYRAcnKy1cfaQqfTYf369Rbrp9Pp8Kc//cnssfaIub1YirmlctY+l12trfj+6U9/wm9+8xsoFB3r3dUTGZ8/S5/r3f382vKZYe1rFDC8Ltq6T8DQu9TRr+UhQ4Zg165deOihh/DVV1/h2WefBWBI9Dv7MHMypdcLVNUZfgd4qDqWKO3va5hy6lJJTat9xoWc+ni4wU1h+XdG8yRteU0D1F78/CMiIiLnZfVfwmvWrIGvr6/0GDhwYFfWi8iiiooKR1cBOp0Ohw4dQl5eHvLy8nDo0CGrEmY6nc7hSd6u0jwmLVmKkU6nQ25ursV/vJSVlSE7O7tT9Wrr2pbqa8t5bNH8mtnZ2V36D6ecnJw2e4daug97xNxerI25rc9lV9u9e7fFOut0Ouzevbuba+TcjM+zpX8idffza8tnhq2fC4cOHUJOTo7F8nq93uGv5aSkJCxZsgSDBw9GZGQkoqKiABh6l44aNcph9SLbGRdyAjrfozS/tAZ6vWn7pajCuJCT5WH3AKCQy+DdeP3yGutHtBARERE5gtWtpuXLlyMxMVH6uaysrNuSpcaVnwFIC/Y0V1NT09TbU96xhqBFzc5n7trdqbS0tNcPvQcALy8vhydLFQoFIiIiMGjQIADAbbfdhsOHD7f7x7JCoYBer++RydKWMWlu0KBBZmOkUCgwatQo/Pjjj2aThr6+voiMjOxUvdq6tqX62nIeWzS/5sCBA+Hj49NlydLbbrsNQgiLyVKFQmH2PuwRc3uxNua2Ppdd7b777sPGjRvN1lmpVOK+++5zQK2cl/F5zs3NNZss7e7n15bPDFs/F4zvyyNHjpgtL5fLMWbMGIe+lh9++GGMHz8ely9fxq233iptnzRpEkfWuBhjb1K5DNC4dWyUQKCPBjIZUKfT42plnclcpMYV79uan9TIW6NEeW0D5yklIiIip2d1q0mtVsPHx8fk0V2aDz/TaDRwd3c3eZgkL+09VK2da3fnIygoCL6+5lcU7S0UCgX+8Ic/OHx4rUwmw6JFiyCTyaTvrbFs2TKnGNLcFZrHxNI+c9ufffZZi4sIrV69utPxauvalupry3kUCoXVw6ibX1Mul3fZ4kkKhQKLFy/G4sWLzdZNoVBg2bJlZo+1R8ztxdr3lq3PZVdTKBRYunSp2X3Lli3jsPsW2nv+uvv5teUzw5bP/+bvS0vkcrlTvJaDgoIwatQok8+CsWPHYtiwYQ6sFdmqaSEnZYdfU24KOfo1JkIvlZjOU2ptj1KgaZ5S9iglIiIiZ+ccfw2T1bZv3+7oKjhUXFwcIiIi8Nhjj7XaN2LECMyePdvscfHx8ejbt69V15DJZPDw8DDZ5uHhIf2RIZPJ8Oijj2LAgAHS/uuuuw6PPvqoSZmW1xsxYgRiYmJMyjmKUml9z+uWsQAM9xceHt5mTFoyFyPjMWPGjEF4eLhJ+REjRmD06NFW17MtbV27s+eJi4sz+5xaEyNz920PcXFxGDBggFRfc/tjYmK6NOb2Yi7mtr72HGHy5MmtPgP69euHu+++20E1cm4tn2cjRz2/tnxmWKp7Sy3fl85yry1VVlZi1apVGDduHIYMGYIbbrjB5EGuo7MLORkZ5ym93GJBp0IbEqXeGkMdymrYo5SIiIicGxOl5FT8/Pws/rEZEBAgJX3i4uLg5+cn7ZPL5UhOTkZcXBz69OnT6pyPPvooNm3a1Oa1tVqtdJ2UlBSTfSkpKfD3929Vj+bi4uJMymzatEnqjWOsn7lyze/D3L23vB9L5Vpqq1fgO++8Y3a7MQZG/v7+rWIBGOr98ssvtxuTllree/NjXn75ZbPxspe2rt3Z8zTfZrwHa2PU/L6t4e/vb/KaAUxfDy2v07xuLfd3dcztpWXMO/Lac4SWnzlvv/22g2riGiy9jxz1/NrymdHW57q5453tXpv77W9/iw8++AB33nknEhISsGjRIpMHuY7mPUo7Y4DW/IJOReWGxZwCvNtfRNVHY+hRyqH3RERE5OyYKCUTKpWhsWvr0FulUmn2mIkTJ0Kr1Uo9GMPCwuDp6QmZTIawsDDIZDKo1Wp4enpCq9ViyZIlrXqLKpVKaLVaJCYmStMsaDQaLFmyRDpXXFwctFotNBoNli5dKs1rq1arsWTJEmg0GgQGBmLixInSecPCwky+X7ZsGQIDA/Hss89i0KBBUtmJEydi0KBBSExMlPabm6tWo9GYlAkMDERcXBzkcrlUv5blEhMTsWTJEmi1Wmi1Wjz22GPSPU2cOBGBgYFYunSpSV0B4LHHHsPs2bMhl8ulchMnTpTiqdVqERcXh8DAwFbHTpw4EUOHDm3V+3b27NlYtmyZdH1PT08899xzJrEICwuT7s/4nLQVk/Zi1PwYY51bxste2rp2Z8/TfJsx7tbGqPl9G98TxveLWq02eS1rtVo899xz0mvG+L557LHHpNdQ8/eJsb7PPfec2f1dHXN7aRnzjrz2HKH5Z47xfUqWWXofOer5teUzw9LnuvE9au596Uz32tyXX36JnTt34vXXX8fixYuZKHVhlbWGOUo9VZ2b7qO/r+F12Zmh98YepRx6T0RERM5OJjq4qkxZWRl8fX1RWlra5fOVVldX45577gFgaMC7u7tb3F8+ejagcGvaqauH9+E/m99njWbHm7t2d2svFkRERESWtNd+CwkJwZ49exAaGuqA2rmG7mwDd8auI79g8Y6jGHejP9IW3N7h83zwr3N4efcpTAnvj81xTdOzTNn4T5y8VIYtc2/Dr4b1a/McSX89gW1Z5/HMr4fgubtv7nBdiIiIiDrClvYbe5QSEREREQDDlBxJSUmoqqpydFWokyrr7DX0vrFHaan5HqXWrnoPsEcpEREROb/OtZyIiIiIqMdYt24d/vvf/yIwMBCDBw+Gm5vpSJzDhw87qGZkK3sv5tR86L1eL1BU0ThHqTWr3nOOUiIiInIR7FFKRERERACA2NhYaS7khx9+GA8++KDJwxabN2/G4MGDodFoEBkZiYMHD7ZZfufOnRg2bBg0Gg3Cw8OxZ88ek/1CCCQlJaF///5wd3dHdHQ0zp49a1KmuLgYcXFx8PHxgVarxfz581FRUdHqPG+88QZuuukmqNVqDBgwAK+++qpN9+YKKoxzlKo7OUdpY4/SK+W1qNfpAQAl1fXQ6Q2zd/l7WbGYk3tjopQ9SomIiMjJsUcpEREREQEAVq9ebZfz7NixA4mJiUhJSUFkZCTWr1+PmJgYnDlzBv36tZ7P8rvvvsOsWbOwZs0a3HfffUhLS0NsbCwOHz4sLUq4du1abNy4ER999BFCQkKwatUqxMTE4NSpU9JCWHFxcbh8+TIyMjJQX1+PefPmYeHChUhLS5OutWjRInz99dd44403EB4ejuLiYhQXF9vlvp1JpZ1WvQ/wVEOlkKNOp0dBWQ2u6+MhDbvXerjBTdF+vwvj0PuyGvYoJSIiIufGHqVEREREJCkpKcH777+P5cuXSwnEw4cP45dffrH6HG+++SYWLFiAefPmYfjw4UhJSYGHhwc+/PBDs+U3bNiAyZMnY+nSpQgNDcXLL7+M0aNHY9OmTQAMvUDXr1+PlStX4sEHH8SIESOwbds2XLp0Cbt27QIAnD59Gunp6Xj//fcRGRmJ8ePH4+2338b27dtx6dIlqcw777yDv/71r3jggQcQEhKCiIgI/OY3v+lExJyTlChVdS5RKpfLECStfF8DACgqt37Fe6Bp6D3nKCUiIiJnx0QpEREREQEAjh07hptuugmvv/463njjDZSUlAAAPvvsMyxfvtyqc9TV1SE3NxfR0dHSNrlcjujoaGRlZZk9Jisry6Q8AMTExEjlz507h/z8fJMyvr6+iIyMlMpkZWVBq9VizJgxUpno6GjI5XJkZ2cDAL744gvccMMN2L17N0JCQjB48GD89re/bbNHaW1tLcrKykwerqC8MVFq7M3ZGf0bE6WXGxd0KqwwJkrbH3bfvA6co5SIiIicHROlRERERAQASExMxNy5c3H27FlpODsA3HvvvThw4IBV5ygqKoJOp0NgYKDJ9sDAQOTn55s9Jj8/v83yxq/tlWk5rF+pVMLPz08q89NPP+H8+fPYuXMntm3bhq1btyI3NxcPP/ywxftZs2YNfH19pcfAgQPbC4FTMPbe7OxiTgAwQGtc0KmxR2njQk59vTUWj2nOOEdpOYfeExERkZNjopSIiIiIAACHDh3CE0880Wr7gAEDLCY5XYler0dtbS22bduGO++8ExMnTsQHH3yAffv24cyZM2aPWb58OUpLS6XHhQsXurnWHVPRmJT0bhz23hnGBZ2MK98XlnesR2l5bQP0jYtAERERETkjJkqJiIiICACgVqvNDi3/8ccf0bdvX6vOERAQAIVCgYKCApPtBQUFCAoKMntMUFBQm+WNX9src+XKFZP9DQ0NKC4ulsr0798fSqUSN910k1QmNDQUAJCXl2e2bmq1Gj4+PiYPV2DsUepjl6H3hh6lxqH3RRUdm6NUCKCyjvOUEhERkfNiopSIiIiIAAAPPPAAkpOTUV9v6I0ok8mQl5eH559/HtOmTbPqHCqVChEREcjMzJS26fV6ZGZmIioqyuwxUVFRJuUBICMjQyofEhKCoKAgkzJlZWXIzs6WykRFRaGkpAS5ublSmb1790Kv1yMyMhIAcMcdd6ChoQH//e9/pTI//vgjAOD666+36v5chTFRao8epa2H3hsSpX2tTJRq3BRQKQx/dpRxQSciIiJyYkyUEhEREREAYN26daioqEDfvn1RXV2NCRMmYMiQIfD29sarr75q9XkSExPx3nvv4aOPPsLp06fx1FNPobKyEvPmzQMAxMfHmywOtWjRIqSnp2PdunX44Ycf8NJLLyEnJwcJCQkADAnbxYsX45VXXsHf/vY3HD9+HPHx8QgODkZsbCwAQ8/QyZMnY8GCBTh48CC+/fZbJCQkYObMmQgODgZgWNxp9OjRePzxx3HkyBHk5ubiiSeewG9+8xuTXqY9gXE+UC979Cg1Dr0vrca1yjp8f6EEABDka90cpQDg484FnYiIiMj5db7lREREREQ9gq+vLzIyMvDtt9/i+++/R0VFBUaPHt1qRfr2zJgxA4WFhUhKSkJ+fj5GjhyJ9PR0aTGmvLw8yOVN/68fN24c0tLSsHLlSqxYsQJDhw7Frl27EBYWJpVZtmwZKisrsXDhQpSUlGD8+PFIT083WXQqNTUVCQkJmDRpEuRyOaZNm4aNGzdK++VyOb744gs888wzuOuuu+Dp6Yl77rkH69at62jInJJOL1BZpwNgr1XvDT1KS6rqseqvJ3Ctqh43BXrh9hv8rT6Ht8YNRRV1Uk9XIiIiImfERCkRERERQa/XY+vWrfjss8/w888/QyaTSUPehRCQyWQ2nS8hIUHqEdrS/v37W22bPn06pk+fbvF8MpkMycnJSE5OtljGz88PaWlpbdYrODgYf/nLX9os4+oqapuSkfZIlPpolPBSK1FR24Ddxy4DANZMDYdKaf3gNONcqexRSkRERM6MQ++JiIiIejkhBB544AH89re/xS+//ILw8HDccsstOH/+PObOnYuHHnrI0VUkGxiH3asUcqiVik6fTyaToX+zYfZxkYMQcb2fTecwzpVaXstEKRERETkv9iglIiIi6uW2bt2KAwcOIDMzE7/61a9M9u3duxexsbHYtm0b4uPjHVRDskXTQk72a+r317rj7JUK9PVWY9nkYTYf3zRHKYfeExERkfNij1IiIiKiXu6TTz7BihUrWiVJAeDXv/41XnjhBaSmpjqgZtQRxqH39kyU/ia0HzxVCrw2NRy+7m42H+9j7FFawx6lRERE5LyYKCUiIiLq5Y4dO4bJkydb3H/PPffg+++/78YaUWcYk5HG4e72MDtqMI69FINJoYEdOt6YtC3jYk5ERETkxJgoJSIiIurliouLpRXpzQkMDMS1a9e6sUbUGcah915q+86ypZDbtqBXc+xRSkRERK6AiVIiIiKiXk6n00GptJxUUygUaGhgT0BX0RVzlHaW1KOUc5QSERGRE3Oe1hMREREROYQQAnPnzoVarTa7v7a2tptrRJ3RlCi139D7zvJpnNe0jD1KiYiIyIkxUUpERETUy82ZM6fdMlzx3nU0zVHqPE19Y9KWc5QSERGRM3Oe1hMREREROcSWLVscXQWyI2cceu/TWJfyavYoJSIiIufFOUqJiIiIiHqQilrnS5SyRykRERG5AiZKiYiIiIh6kKah9840R2njYk6co5SIiIicGBOlREREREQ9SJkTDr03Jm3rGvSoqdc5uDZERERE5jFRSkRERETUgxjnKPVSO1GiVK2ETGb4vpzD74mIiMhJMVFKRERERNSDVNQ639B7uVwGL1Xjgk4cfk9EREROiolSIiIiIqIexNhj08eJht4DgI87F3QiIiIi58ZEKRERERFRDyGEkBKlztSjFGiaM5U9SomIiMhZMVFKRERERNRDVNfroNMLAICXs/UobUzcllWzRykRERE5JyZKiYiIiIh6iIrG3qRyGeCpUji4NqZ83NmjlIiIiJwbE6VERERERD1EWbMV72XGZeadhHEqgDImSomIiMhJMVFKRERERNRDGHtrOtv8pEDT4lIcek9ERETOiolSIiIiIqIeoqLWuJCTc81PCgC+javel1TXObgmREREROYxUUpERERE1EM0rXjvfIlSfy81AOBqBROlRERE5JyYKCUiIiIiu9u8eTMGDx4MjUaDyMhIHDx4sM3yO3fuxLBhw6DRaBAeHo49e/aY7BdCICkpCf3794e7uzuio6Nx9uxZkzLFxcWIi4uDj48PtFot5s+fj4qKCrPX+89//gNvb29otdpO3aezceah9/5eKgBMlBIREZHzYqKUiIiIiOxqx44dSExMxOrVq3H48GHceuutiImJwZUrV8yW/+677zBr1izMnz8fR44cQWxsLGJjY3HixAmpzNq1a7Fx40akpKQgOzsbnp6eiImJQU1NjVQmLi4OJ0+eREZGBnbv3o0DBw5g4cKFra5XX1+PWbNm4c4777T/zTuYU/co9TT0KC2qrHVwTYiIiIjMY6KUiIiIiOzqzTffxIIFCzBv3jwMHz4cKSkp8PDwwIcffmi2/IYNGzB58mQsXboUoaGhePnllzF69Ghs2rQJgKE36fr167Fy5Uo8+OCDGDFiBLZt24ZLly5h165dAIDTp08jPT0d77//PiIjIzF+/Hi8/fbb2L59Oy5dumRyvZUrV2LYsGF45JFHujQOjlDebNV7ZxPAHqVERETk5JgoJSIiIiK7qaurQ25uLqKjo6Vtcrkc0dHRyMrKMntMVlaWSXkAiImJkcqfO3cO+fn5JmV8fX0RGRkplcnKyoJWq8WYMWOkMtHR0ZDL5cjOzpa27d27Fzt37sTmzZutup/a2lqUlZWZPJxZU49SZxx6b+hRWlpdj7oGvYNrQ0RERNSa8/2r2QyNRoMvv/xS+r43YyyIiIjImRUVFUGn0yEwMNBke2BgIH744Qezx+Tn55stn5+fL+03bmurTL9+/Uz2K5VK+Pn5SWWuXr2KuXPn4uOPP4aPj49V97NmzRr84Q9/sKqsM2iao9T5mvladzco5DLo9ALXquoQ6MO2LBERETkXl+hRKpPJ4O7uDnd3d8hkMkdXx6EYCyIiIqKOWbBgAR599FHcddddVh+zfPlylJaWSo8LFy50YQ07z5nnKJXLZfDzNAy/L6rgPKVERETkfFwiUUpEREREriEgIAAKhQIFBQUm2wsKChAUFGT2mKCgoDbLG7+2V6blYlENDQ0oLi6WyuzduxdvvPEGlEollEol5s+fj9LSUiiVSovzp6rVavj4+Jg8nFlFrfMmSgHA35PzlBIREZHzYqKUiIiIiOxGpVIhIiICmZmZ0ja9Xo/MzExERUWZPSYqKsqkPABkZGRI5UNCQhAUFGRSpqysDNnZ2VKZqKgolJSUIDc3Vyqzd+9e6PV6REZGAjDMY3r06FHpkZycDG9vbxw9ehQPPfSQfQLgYNLQe7XzzVEKAAGN85Re5cr3RERE5ISc81/NREREROSyEhMTMWfOHIwZMwZjx47F+vXrUVlZiXnz5gEA4uPjMWDAAKxZswYAsGjRIkyYMAHr1q3DlClTsH37duTk5ODdd98FYJh6aPHixXjllVcwdOhQhISEYNWqVQgODkZsbCwAIDQ0FJMnT8aCBQuQkpKC+vp6JCQkYObMmQgODpbKNJeTkwO5XI6wsLBuikzXc+ah9wDgz5XviYiIyIk5ZwuKiIiIiFzWjBkzUFhYiKSkJOTn52PkyJFIT0+XFmPKy8uDXN40sGncuHFIS0vDypUrsWLFCgwdOhS7du0ySWAuW7YMlZWVWLhwIUpKSjB+/Hikp6ebLG6ZmpqKhIQETJo0CXK5HNOmTcPGjRu778adQFljotTLWROlnoYepUVMlBIREZETcs4WFBERERG5tISEBCQkJJjdt3///lbbpk+fjunTp1s8n0wmQ3JyMpKTky2W8fPzQ1pamtV1nDt3LubOnWt1eVdQUWsYeu+jcc6h9009Sjn0noiIiJwP5yglIiIiIuoB6nV61NTrATjv0PsAY6K0kj1KiYiIyPkwUUpERERE1AMY5ycFAC+1cyZKjUPv2aOUiIiInBETpUREREREPUBFY6LU3U0BpcI5m/nGofeco5SIiIickXO2oIiIiIiIyCZlNYb5SZ112D0ABHg19iitrIUQwsG1ISIiIjLFRCkRERERUQ9gHHrvzIlSY4/Smno9Kut0Dq4NERERkSkmSomIiIiIeoByqUepc654DwAeKiXc3RQAOE8pEREROR8mSomIiIiIeoCKWufvUQpwnlIiIiJyXkyUEhERERH1AK4w9B5oNk8pe5QSERGRk2GilIiIiIioByitNgy993HiofcAENDYo/RqJXuUEhERkXNhopSIiIiIqAcoauyhaeyx6az8PdmjlIiIiJwTE6VERERERD1AYbkh8djX28kTpZyjlIiIiJwUE6VERERERD2A6yRKG3uUcug9ERERORkmSomIiIiIegBXGXovzVHKofdERETkZJgoJSIiIiLqAVymR6k0Ryl7lBIREZFzYaKUiIiIiMjFVdY2oLJOB8AFEqXSqvfsUUpERETOhYlSIiIiIiIXZxx27+6mgKdK4eDatM2YKC2urINOLxxcGyIiIqImTJQSEREREbk447D7AG8VZDKZg2vTNj8PQ6JUL4CSKg6/JyIiIufBRCkRERERkYsz9ijt6+QLOQGAUiFHHw83AABXviciIiJnwkQpEREREZGLc5WFnIz8GxO6xgQvERERkTNgopSIiIiIyMW5XKLUs3FBJ658T0RERE6EiVIiIiIiIhdXKA291zi4JtYJ9DHU85eSagfXhIiIiKgJE6VERERERC6usNzQMzPAW+XgmljnpkAvAMCPBeUOrgkRERFREyZKiYiIiMjuNm/ejMGDB0Oj0SAyMhIHDx5ss/zOnTsxbNgwaDQahIeHY8+ePSb7hRBISkpC//794e7ujujoaJw9e9akTHFxMeLi4uDj4wOtVov58+ejoqJC2r9//348+OCD6N+/Pzw9PTFy5Eikpqba76YdqNCFFnMCgKGB3gCYKCUiIiLnwkQpEREREdnVjh07kJiYiNWrV+Pw4cO49dZbERMTgytXrpgt/91332HWrFmYP38+jhw5gtjYWMTGxuLEiRNSmbVr12Ljxo1ISUlBdnY2PD09ERMTg5qaGqlMXFwcTp48iYyMDOzevRsHDhzAwoULTa4zYsQI/OUvf8GxY8cwb948xMfHY/fu3V0XjG5S5GJzlN7cmCg9W1ABnV44uDZEREREBjIhRIdaJmVlZfD19UVpaSl8fHzsXS+bVFdX45577gEAlI+eDSjcmnbq6uF9+M/m91mj2fFffvkl3N3d7VJnIiIiou7WXe23yMhI3Hbbbdi0aRMAQK/XY+DAgXjmmWfwwgsvtCo/Y8YMVFZWmiQsb7/9dowcORIpKSkQQiA4OBjPPfcclixZAgAoLS1FYGAgtm7dipkzZ+L06dMYPnw4Dh06hDFjxgAA0tPTce+99+LixYsIDg42W9cpU6YgMDAQH374oVX35kxtYCMhBG5emY46nR7/ev5XuK6Ph6Or1C6dXuCW1emoqddj35KJCAnwdHSViIiIqIeypf3GHqVEREREZDd1dXXIzc1FdHS0tE0ulyM6OhpZWVlmj8nKyjIpDwAxMTFS+XPnziE/P9+kjK+vLyIjI6UyWVlZ0Gq1UpIUAKKjoyGXy5GdnW2xvqWlpfDz87O4v7a2FmVlZSYPZ1NW04A6nR4AEOAiQ+8VchmG9jP0Kj2Tz+H3RERE5ByYKCUiIiIiuykqKoJOp0NgYKDJ9sDAQOTn55s9Jj8/v83yxq/tlenXr5/JfqVSCT8/P4vX/fTTT3Ho0CHMmzfP4v2sWbMGvr6+0mPgwIEWyzpKYeOwe2+NEho3hYNrY72bOE8pERERORkmSomIiIio19m3bx/mzZuH9957D7fccovFcsuXL0dpaan0uHDhQjfW0jqFLjY/qdHNQYaV788wUUpEREROgolSIiIiIrKbgIAAKBQKFBQUmGwvKChAUFCQ2WOCgoLaLG/82l6ZlotFNTQ0oLi4uNV1v/nmG9x///146623EB8f3+b9qNVq+Pj4mDycjauteG8k9Sjl0HsiIiJyEkyUEhEREZHdqFQqREREIDMzU9qm1+uRmZmJqKgos8dERUWZlAeAjIwMqXxISAiCgoJMypSVlSE7O1sqExUVhZKSEuTm5kpl9u7dC71ej8jISGnb/v37MWXKFLz++utYuHBh52/YCRhXvA9wuR6lhkTpuaJK1DboHFwbIiIiIkDp6AoQERERUc+SmJiIOXPmYMyYMRg7dizWr1+PyspKaS7Q+Ph4DBgwAGvWrAEALFq0CBMmTMC6deswZcoUbN++HTk5OXj33XcBADKZDIsXL8Yrr7yCoUOHIiQkBKtWrUJwcDBiY2MBAKGhoZg8eTIWLFiAlJQU1NfXIyEhATNnzpRWvN+3bx/uu+8+LFq0CNOmTZPmLlWpVG0u6OTsXLVHaZCPBt4aJcprGnCuqBLDgpyvty4RERH1LkyUEhEREZFdzZgxA4WFhUhKSkJ+fj5GjhyJ9PR0aTGmvLw8yOVNA5vGjRuHtLQ0rFy5EitWrMDQoUOxa9cuhIWFSWWWLVuGyspKLFy4ECUlJRg/fjzS09Oh0WikMqmpqUhISMCkSZMgl8sxbdo0bNy4Udr/0UcfoaqqCmvWrJGStAAwYcIE7N+/vwsj0rVcdY5SmUyGmwO9kXP+Gs7klzNRSkRERA7HRCkRERER2V1CQgISEhLM7jOXlJw+fTqmT59u8XwymQzJyclITk62WMbPzw9paWkW92/duhVbt261uN9VuWqiFABuCjIkSrnyPRERETkDzlFKREREROTCilx06D0A3Ny4oNOZ/AoH14SIiIioB/YolekbIJpv0NWb/96G8xEREREROSuX7lHamCg9e4U9SomIiMjxelyi1OvoJxb3eX+/vRtrQkRERETUtXR6gauVdQBcNVHqBQDIK65CVV0DPFQ97s8TIiIiciEcek9ERERE5KKuVdVBpxeQyQA/T5Wjq2Mzfy81ArxUEAI4fZm9SomIiMixesS/bDUaDb788kuz+4QQqK01DEdSq9WQyWSdug4RERERkbMwzk/ax0MFN4Vr9oGIDPHH349fxu5jlxBxfR9HV4eIiIh6sR6RKJXJZHB3d7e438PDoxtrQ0RERETUPaT5SV1wISej6WOuw9+PX8auI7/ghXuGQa1UOLpKRERE1Eu55r+diYiIiIgIPxVWAgCCta478unOoX0R5KPBtap6/OPUFUdXh4iIiHoxJkqJiIiIiFzU9xdLAADh12kdWo/OUMhleDjiOgDApzkXHFwbIiIi6s2YKCUiIiIiclHHLpYCAG69ztfBNekcY6L0wNlCXCqpdnBtiIiIqLdiopSIiIiIyAVV1Dbgv4UVAIARLtyjFAAGB3giMsQPQgCfHb7o6OoQERFRL8VEKRERERGRCzp+sRRCAMG+GvT1dt3FnIweGTMQAPBpzkXo9MLBtSEiIqLeiIlSIiIiIiIXdPyXEgCu35vU6J7wIHhrlMgrrsJbGT86ujpERETUCzFRSkRERETkgr5vnJ90xEDXnp/UyEOlxCuxYQCATfv+g4xTBQ6uEREREfU2TJQSEREREbmgY40r3o8YoHVoPezpwZEDMHfcYABA4o6jOFdU6dgKERERUa/CRCkRERERkYsprqzDhWLD6vDhLr7ifUsr7g1FxPV9UF7bgDkfHkT2T1cdXSUiIiLqJZgoJSIiIiJyMcbepCEBnvB1d3NsZexMpZTjf+JGI9hXg7ziKsx499948fPjuFZZ5+iqERERUQ/HRCkRERERkYs5bpyftIf1JjUK9NHgy0V3YdbYgQCA1Ow8RLySgekp32Hzvv/g4LliVNQ2OLiWRERE1NMoHV0BIiIiIiKyjbSQUw9Z8d4cXw83rJk6Ag/cOgAv7z6FU5fLcOjnazj08zUAgExm6FEbMagPbhvshzGD+yAkwBMymczBNSciIiJXxUQpEREREZGLMQ69v7WH9ihtLupGf+xZdCcuFFdh/4+FOPBjIY5fLEV+WQ1+KqzET4WV2Jl7EQDg76lCxPV9MPr6Prg50BtD+nlhgNYdcjmTp0RERNQ+JkqJiIiIiFzIuaJKXCmvhUIuwy3BPT9RajTQzwOzb78es2+/HgBQWF6LYxdLkHP+GnJ/voajF0twtbIOX58qwNenCqTjPFQKDO3nhaGB3ggL9sGYwX4I7e8DBZOnRERE1AITpURERERELkIIgZf+dhIAMO5Gf7irFA6ukeP09VZjUmggJoUGAgBqG3Q48UspDv18Dcd/KcXZgnKcK6pEVZ0O318sxfcXS/F/uYZjvdRKjL6+D267vg9uC/HDyIFaaNx6byyJiIjIgIs5EREREZHdbd68GYMHD4ZGo0FkZCQOHjzYZvmdO3di2LBh0Gg0CA8Px549e0z2CyGQlJSE/v37w93dHdHR0Th79qxJmeLiYsTFxcHHxwdarRbz589HRUWFSZljx47hzjvvhEajwcCBA7F27Vr73HA3+dv3l/DNj4VQKeR46YFbHF0dp6JWKhBxvR+enHAjNj86Gl8/OwGnkifjH4l34Z240VgcPRQTbuoLb7USFbUNOPBjIdZl/IiZ7/4b4S99han/8y3+uOc0/i/3Io5eKEFpdb2jb4mIiIi6GXuUEhEREZFd7dixA4mJiUhJSUFkZCTWr1+PmJgYnDlzBv369WtV/rvvvsOsWbOwZs0a3HfffUhLS0NsbCwOHz6MsLAwAMDatWuxceNGfPTRRwgJCcGqVasQExODU6dOQaPRAADi4uJw+fJlZGRkoL6+HvPmzcPChQuRlpYGACgrK8Pdd9+N6OhopKSk4Pjx43j88ceh1WqxcOHC7gtQB5VU1SH5i1MAgIRfD8GNfb0cXCPn56aQY0g/bwzp5417wvsDAHR6gR/yy5Dz8zUc/LkYh84V40p5LQ7nleBwXonJ8V5qJYJ8NQj0UUProYLW3Q0eKgUUcjmUchkUcpnhq6Lxq1wON4UMfp4q9PVSo5+PBn291fBS888uIiIiVyATQoiOHFhWVgZfX1+UlpbCx8fH3vUiIiIiIjvrrvZbZGQkbrvtNmzatAkAoNfrMXDgQDzzzDN44YUXWpWfMWMGKisrsXv3bmnb7bffjpEjRyIlJQVCCAQHB+O5557DkiVLAAClpaUIDAzE1q1bMXPmTJw+fRrDhw/HoUOHMGbMGABAeno67r33Xly8eBHBwcF455138OKLLyI/Px8qlQoA8MILL2DXrl344YcfrLo3R7WB80tr8NLfTiL9ZD6G9vPC339/J1RKDg6zByEELhRX49DPxThy4Rr+e6US/y2swJXyWrtdw0OlQF9vNfp5qxHgpUZfbzX6Nn7VerhB7aaAu5sCGumrHEqFHAqZIRmrkMsM3ytk0jalXGa3RaqMfxLKZK4xb6sQAnU6PWob9BB6ADJAJgPkMhlkMHxfrxNo0OnRoBeo1+nRoDN8rdcJNOj1JvsVchm81Ep4qBTwUivh2fi9q8SDiIjaZkv7jf/aJCIiIiK7qaurQ25uLpYvXy5tk8vliI6ORlZWltljsrKykJiYaLItJiYGu3btAgCcO3cO+fn5iI6Olvb7+voiMjISWVlZmDlzJrKysqDVaqUkKQBER0dDLpcjOzsbDz30ELKysnDXXXdJSVLjdV5//XVcu3YNffr0sUcI7KauQY8vT1zG/+VexL/+UwQhDAmg16aFM0lqRzKZDIP8PTDI3wPTIq6TtlfVNeBSSQ0ul1ajsLwWpdX1KKmqR029Djq9QINeNPuql36ua9DjamUdCstrcaWsBpV1OlTV6XD+ahXOX62ya901bnJ4qJRwd1PAU62Au0oJpVwGIQQEACHQ+FWgXidQ16Br/KpHnU6P+gY9anV61Ov0EAJQymVQKeVwUxgeaqXhoXFTwEOlgLvKkMht/lXjpoBSLoNMZkjiymWQErj6ZnHSicavUtz00Okhxa6uwZD4rG3Qo7Zeh9oGfeM2XavttQ16u8bRHJkM8FQp4alWSIlTDzcl3FVNsfCQ4qA0JLeb9So2JrObJ1tlJue3tL319zK0TtganuHG75t1fTL5vnn5xtcERNOxTa+PpnMatxkLNN+vF82fP9Pn0/j8NugE9ML4/DaVNb4W6nV6VNXpUFnbYPha14CqWh0a9MLw2pE1vpbkhu9VSjlUCjnUbnKolQrpNalSNvu5cZ9SIYMMhtegaeJcBpkMJvuM24xl5MYyzZLtLbfJpWMM5zLuk87Zcpvx51bPXtcyPn9CtH5Ojf8UES1fB80+K6RzNJbRC8NzrxeG93TL7wUgPXfyxn/kyOWGGCgat0k/t/issFR36Wcz3fpEq1KWyrUsY10fwZbFrL5eq+PMlTFzLrOV6GAdrCrT+JoQps+t4eembUJ6jo0/Gxoh0vMqN7y+LT3HbT3/xn0aNwV+Naz1SCNH63Ci1PgEl5WV2a0yRERERNR1jO22Dg4oskpRURF0Oh0CAwNNtgcGBlrstZmfn2+2fH5+vrTfuK2tMi2H9SuVSvj5+ZmUCQkJaXUO4z5zidLa2lrU1jb1LCwtLQXQPW3g2gYdXtxxEGU1DQCAiEF9MOeOwRjaR8k2eDfppwH6adRAoLrD56isbUBRRS0Ky2tRXFmHoopaFJU3fq2sRUVNA6obE4DVdTrU1utQ06BDvR4QjcklS6pqAXumXusaHz2JsUeuW+P0CEq5HEqFDEqFzJAQlstRr9ejqrZBSmobPyLLagC+04iIuoa/pxu+WfbrbrmWLW3gDidKy8vLAQADBw7s6CmIiIiIyAHKy8vh6+vr6Gq4hDVr1uAPf/hDq+2OaANfALCr269KREREZH8XAPi+2r3XtKYN3OFEaXBwMC5cuABvb+9umbulrKwMAwcOxIULFzgnaicwjp3HGNoH49h5jKF9MI6dxxjaR3fEUQiB8vJyBAcHd8n5ASAgIAAKhQIFBQUm2wsKChAUFGT2mKCgoDbLG78WFBSgf//+JmVGjhwplbly5YrJORoaGlBcXGxyHnPXaX6NlpYvX24yLYBer0dxcTH8/f2dfv5Cvjdtx5jZjjHrGMbNdoyZ7Rgz2zFmHePscbOlDdzhRKlcLsd1113XfkE78/HxccqguxrGsfMYQ/tgHDuPMbQPxrHzGEP76Oo4dnVPUpVKhYiICGRmZiI2NhaAIbmYmZmJhIQEs8dERUUhMzMTixcvlrZlZGQgKioKABASEoKgoCBkZmZKidGysjJkZ2fjqaeeks5RUlKC3NxcREREAAD27t0LvV6PyMhIqcyLL76I+vp6uLm5Sde5+eabLc5PqlaroVabDrvWarU2x8WR+N60HWNmO8asYxg32zFmtmPMbMeYdYwzx83aNjBngSciIiIiu0pMTMR7772Hjz76CKdPn8ZTTz2FyspKzJs3DwAQHx9vstjTokWLkJ6ejnXr1uGHH37ASy+9hJycHCmxKpPJsHjxYrzyyiv429/+huPHjyM+Ph7BwcFSMjY0NBSTJ0/GggULcPDgQXz77bdISEjAzJkzpd4Djz76KFQqFebPn4+TJ09ix44d2LBhQ6uFpIiIiIiod+Kq90RERERkVzNmzEBhYSGSkpKQn5+PkSNHIj09XVo4KS8vD3J50//rx40bh7S0NKxcuRIrVqzA0KFDsWvXLoSFhUllli1bhsrKSixcuBAlJSUYP3480tPTodFopDKpqalISEjApEmTIJfLMW3aNGzcuFHa7+vri6+//hpPP/00IiIiEBAQgKSkJCxcuLAbokJEREREzs5lEqVqtRqrV69uNfSJbMM4dh5jaB+MY+cxhvbBOHYeY2gfPS2OCQkJFofa79+/v9W26dOnY/r06RbPJ5PJkJycjOTkZItl/Pz8kJaW1ma9RowYgX/+859tlukpetprqjswZrZjzDqGcbMdY2Y7xsx2jFnH9KS4yYQQwtGVICIiIiIiIiIiInIkzlFKREREREREREREvR4TpURERERERERERNTrMVFKREREREREREREvR4TpURERERERERERNTruUyidPPmzRg8eDA0Gg0iIyNx8OBBR1fJIdasWYPbbrsN3t7e6NevH2JjY3HmzBmTMjU1NXj66afh7+8PLy8vTJs2DQUFBSZl8vLyMGXKFHh4eKBfv35YunQpGhoaTMrs378fo0ePhlqtxpAhQ7B169auvj2HeO211yCTybB48WJpG2NonV9++QWPPfYY/P394e7ujvDwcOTk5Ej7hRBISkpC//794e7ujujoaJw9e9bkHMXFxYiLi4OPjw+0Wi3mz5+PiooKkzLHjh3DnXfeCY1Gg4EDB2Lt2rXdcn/dQafTYdWqVQgJCYG7uztuvPFGvPzyy2i+zh7jaOrAgQO4//77ERwcDJlMhl27dpns78547dy5E8OGDYNGo0F4eDj27Nlj9/vtKm3Fsb6+Hs8//zzCw8Ph6emJ4OBgxMfH49KlSybn6O1xbO+12NyTTz4JmUyG9evXm2zv7TGkrtOb285sL3cO28bWY1vYNmz3WodtXduxXWs7tmPbIFzA9u3bhUqlEh9++KE4efKkWLBggdBqtaKgoMDRVet2MTExYsuWLeLEiRPi6NGj4t577xWDBg0SFRUVUpknn3xSDBw4UGRmZoqcnBxx++23i3Hjxkn7GxoaRFhYmIiOjhZHjhwRe/bsEQEBAWL58uVSmZ9++kl4eHiIxMREcerUKfH2228LhUIh0tPTu/V+u9rBgwfF4MGDxYgRI8SiRYuk7Yxh+4qLi8X1118v5s6dK7Kzs8VPP/0kvvrqK/Gf//xHKvPaa68JX19fsWvXLvH999+LBx54QISEhIjq6mqpzOTJk8Wtt94q/v3vf4t//vOfYsiQIWLWrFnS/tLSUhEYGCji4uLEiRMnxCeffCLc3d3F//7v/3br/XaVV199Vfj7+4vdu3eLc+fOiZ07dwovLy+xYcMGqQzjaGrPnj3ixRdfFJ999pkAID7//HOT/d0Vr2+//VYoFAqxdu1acerUKbFy5Urh5uYmjh8/3uUxsIe24lhSUiKio6PFjh07xA8//CCysrLE2LFjRUREhMk5ensc23stGn322Wfi1ltvFcHBweKtt94y2dfbY0hdo7e3ndle7ji2ja3HtrDt2O61Dtu6tmO71nZsx1rmEonSsWPHiqefflr6WafTieDgYLFmzRoH1so5XLlyRQAQ33zzjRDC8CHg5uYmdu7cKZU5ffq0ACCysrKEEIY3hFwuF/n5+VKZd955R/j4+Ija2lohhBDLli0Tt9xyi8m1ZsyYIWJiYrr6lrpNeXm5GDp0qMjIyBATJkyQGoOMoXWef/55MX78eIv79Xq9CAoKEn/605+kbSUlJUKtVotPPvlECCHEqVOnBABx6NAhqcyXX34pZDKZ+OWXX4QQQvzP//yP6NOnjxRX47Vvvvlme9+SQ0yZMkU8/vjjJtumTp0q4uLihBCMY3ta/lLvzng98sgjYsqUKSb1iYyMFE888YRd77E7tNU4Mjp48KAAIM6fPy+EYBxbshTDixcvigEDBogTJ06I66+/3qSByRhSV2Hb2RTby9Zh29g2bAvbju1e27Gtazu2a23Hdqwppx96X1dXh9zcXERHR0vb5HI5oqOjkZWV5cCaOYfS0lIAgJ+fHwAgNzcX9fX1JvEaNmwYBg0aJMUrKysL4eHhCAwMlMrExMSgrKwMJ0+elMo0P4exTE+K+dNPP40pU6a0uk/G0Dp/+9vfMGbMGEyfPh39+vXDqFGj8N5770n7z507h/z8fJMY+Pr6IjIy0iSOWq0WY8aMkcpER0dDLpcjOztbKnPXXXdBpVJJZWJiYnDmzBlcu3atq2+zy40bNw6ZmZn48ccfAQDff/89/vWvf+Gee+4BwDjaqjvj1dPf4y2VlpZCJpNBq9UCYBytodfrMXv2bCxduhS33HJLq/2MIXUFtp1bY3vZOmwb24ZtYdux3dt5bOvaB9u17evN7VinT5QWFRVBp9OZ/NIFgMDAQOTn5zuoVs5Br9dj8eLFuOOOOxAWFgYAyM/Ph0qlkt7wRs3jlZ+fbzaexn1tlSkrK0N1dXVX3E632r59Ow4fPow1a9a02scYWuenn37CO++8g6FDh+Krr77CU089hd///vf46KOPADTFoa33bn5+Pvr162eyX6lUws/Pz6ZYu7IXXngBM2fOxLBhw+Dm5oZRo0Zh8eLFiIuLA8A42qo742WpTE+Kp1FNTQ2ef/55zJo1Cz4+PgAYR2u8/vrrUCqV+P3vf292P2NIXYFtZ1NsL1uHbWPbsS1sO7Z7O49t3c5ju9Y6vbkdq3TYlanTnn76aZw4cQL/+te/HF0Vl3LhwgUsWrQIGRkZ0Gg0jq6Oy9Lr9RgzZgz++Mc/AgBGjRqFEydOICUlBXPmzHFw7VzHp59+itTUVKSlpeGWW27B0aNHsXjxYgQHBzOO5BTq6+vxyCOPQAiBd955x9HVcRm5ubnYsGEDDh8+DJlM5ujqEPVabC+3j23jjmFb2HZs95KjsV1rnd7ejnX6HqUBAQFQKBStVlUsKChAUFCQg2rleAkJCdi9ezf27duH6667TtoeFBSEuro6lJSUmJRvHq+goCCz8TTua6uMj48P3N3d7X073So3NxdXrlzB6NGjoVQqoVQq8c0332Djxo1QKpUIDAxkDK3Qv39/DB8+3GRbaGgo8vLyADTFoa33blBQEK5cuWKyv6GhAcXFxTbF2pUtXbpU+u96eHg4Zs+ejWeffVbq0cE42qY742WpTE+Kp7Exef78eWRkZEj/dQcYx/b885//xJUrVzBo0CDpd8358+fx3HPPYfDgwQAYQ+oabDs3YXvZOmwbdwzbwrZju7fz2NbtOLZrrdfb27FOnyhVqVSIiIhAZmamtE2v1yMzMxNRUVEOrJljCCGQkJCAzz//HHv37kVISIjJ/oiICLi5uZnE68yZM8jLy5PiFRUVhePHj5u8qI0fFMZf9lFRUSbnMJbpCTGfNGkSjh8/jqNHj0qPMWPGIC4uTvqeMWzfHXfcgTNnzphs+/HHH3H99dcDAEJCQhAUFGQSg7KyMmRnZ5vEsaSkBLm5uVKZvXv3Qq/XIzIyUipz4MAB1NfXS2UyMjJw8803o0+fPl12f92lqqoKcrnpR7FCoYBerwfAONqqO+PV09/jxsbk2bNn8Y9//AP+/v4m+xnHts2ePRvHjh0z+V0THByMpUuX4quvvgLAGFLXYNuZ7WVbsW3cMWwL247t3s5jW7dj2K61Ta9vxzpsGSkbbN++XajVarF161Zx6tQpsXDhQqHVak1WVewtnnrqKeHr6yv2798vLl++LD2qqqqkMk8++aQYNGiQ2Lt3r8jJyRFRUVEiKipK2t/Q0CDCwsLE3XffLY4ePSrS09NF3759xfLly6UyP/30k/Dw8BBLly4Vp0+fFps3bxYKhUKkp6d36/12l+YrewrBGFrj4MGDQqlUildffVWcPXtWpKamCg8PD/Hxxx9LZV577TWh1WrFX//6V3Hs2DHx4IMPipCQEFFdXS2VmTx5shg1apTIzs4W//rXv8TQoUPFrFmzpP0lJSUiMDBQzJ49W5w4cUJs375deHh4iP/93//t1vvtKnPmzBEDBgwQu3fvFufOnROfffaZCAgIEMuWLZPKMI6mysvLxZEjR8SRI0cEAPHmm2+KI0eOSKtWdle8vv32W6FUKsUbb7whTp8+LVavXi3c3NzE8ePHuy8YndBWHOvq6sQDDzwgrrvuOnH06FGT3zfNV63s7XFs77XYUsvVQoVgDKlr9Pa2M9vLnce2cfvYFrYd273WYVvXdmzX2o7tWMtcIlEqhBBvv/22GDRokFCpVGLs2LHi3//+t6Or5BAAzD62bNkilamurha/+93vRJ8+fYSHh4d46KGHxOXLl03O8/PPP4t77rlHuLu7i4CAAPHcc8+J+vp6kzL79u0TI0eOFCqVStxwww0m1+hpWjYGGUPrfPHFFyIsLEyo1WoxbNgw8e6775rs1+v1YtWqVSIwMFCo1WoxadIkcebMGZMyV69eFbNmzRJeXl7Cx8dHzJs3T5SXl5uU+f7778X48eOFWq0WAwYMEK+99lqX31t3KSsrE4sWLRKDBg0SGo1G3HDDDeLFF180+aXNOJrat2+f2c/BOXPmCCG6N16ffvqpuOmmm4RKpRK33HKL+Pvf/95l921vbcXx3LlzFn/f7Nu3TzpHb49je6/Flsw1MHt7DKnr9Oa2M9vLnce2sXXYFrYN273WYVvXdmzX2o7tWMtkQghhn76pRERERERERERERK7J6ecoJSIiIiIiIiIiIupqTJQSERERERERERFRr8dEKREREREREREREfV6TJQSERERERERERFRr8dEKREREREREREREfV6TJQSERERERERERFRr8dEKREREREREREREfV6TJQSkcubOHEiFi9e7OhqYP/+/ZDJZCgpKXF0VYiIiIioh2MbmIjI/pgoJSLqAGdpmBIRERERdRe2gYmop2OilIiIiIiIiIiIiHo9JkqJqEepra3FkiVLMGDAAHh6eiIyMhL79++X9m/duhVarRZfffUVQkND4eXlhcmTJ+Py5ctSmYaGBvz+97+HVquFv78/nn/+ecyZMwexsbEAgLlz5+Kbb77Bhg0bIJPJIJPJ8PPPP0vH5+bmYsyYMfDw8MC4ceNw5syZbrp7IiIiIuqN2AYmIrIPJkqJqEdJSEhAVlYWtm/fjmPHjmH69OmYPHkyzp49K5WpqqrCG2+8gT//+c84cOAA8vLysGTJEmn/66+/jtTUVGzZsgXffvstysrKsGvXLmn/hg0bEBUVhQULFuDy5cu4fPkyBg4cKO1/8cUXsW7dOuTk5ECpVOLxxx/vlnsnIiIiot6JbWAiIvtQOroCRET2kpeXhy1btiAvLw/BwcEAgCVLliA9PR1btmzBH//4RwBAfX09UlJScOONNwIwNCyTk5Ol87z99ttYvnw5HnroIQDApk2bsGfPHmm/r68vVCoVPDw8EBQU1Koer776KiZMmAAAeOGFFzBlyhTU1NRAo9F0zY0TERERUa/FNjARkf0wUUpEPcbx48eh0+lw0003mWyvra2Fv7+/9LOHh4fUQASA/v3748qVKwCA0tJSFBQUYOzYsdJ+hUKBiIgI6PV6q+oxYsQIk3MDwJUrVzBo0CDbb4qIiIiIqA1sAxMR2Q8TpUTUY1RUVEChUCA3NxcKhcJkn5eXl/S9m5ubyT6ZTAYhhN3q0fz8MpkMAKxuYBIRERER2YJtYCIi++EcpUTUY4waNQo6nQ5XrlzBkCFDTB7mhgeZ4+vri8DAQBw6dEjaptPpcPjwYZNyKpUKOp3OrvUnIiIiIrIV28BERPbDHqVE1GPcdNNNiIuLQ3x8PNatW4dRo0ahsLAQmZmZGDFiBKZMmWLVeZ555hmsWbMGQ4YMwbBhw/D222/j2rVr0n/GAWDw4MHIzs7Gzz//DC8vL/j5+XXVbRERERERWcQ2MBGR/bBHKRH1KFu2bEF8fDyee+453HzzzYiNjcWhQ4dsmhvp+eefx6xZsxAfH4+oqCh4eXkhJibGZCL6JUuWQKFQYPjw4ejbty/y8vK64naIiIiIiNrFNjARkX3IhD0nJSEi6oH0ej1CQ0PxyCOP4OWXX3Z0dYiIiIiIuhzbwETUG3HoPRFRC+fPn8fXX3+NCRMmoLa2Fps2bcK5c+fw6KOPOrpqRERERERdgm1gIiIOvSciakUul2Pr1q247bbbcMcdd+D48eP4xz/+gdDQUEdXjYiIiIioS7ANTETEofdERERERERERERE7FFKRERERERERERExEQpERERERERERER9XpMlBIREREREREREVGvx0QpERERERERERER9XpMlBIREREREREREVGvx0QpERERERERERER9XpMlBIREREREREREVGvx0QpERERERERERER9XpMlBIREREREREREVGv9//41nYPfRJcNQAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# density deistribution of spam and non-spam emails\n",
        "\n",
        "plt.figure(figsize=(15,10))\n",
        "\n",
        "plt.subplot(3,2,1)\n",
        "plt.tight_layout(pad = 5)\n",
        "\n",
        "plt.title(\"Spam\")\n",
        "# data[data.target == 1].length.plot(kind = 'bar')\n",
        "sns.boxplot(x = data[data.target == 1].length)\n",
        "\n",
        "plt.subplot(3,2,3)\n",
        "plt.title(\"Non-spam\")\n",
        "sns.boxplot(x = data[data.target == -1].length)\n",
        "\n",
        "plt.subplot(3,2,5)\n",
        "plt.title(\"All emails\")\n",
        "sns.boxplot(x = data.length)\n",
        "\n",
        "\n",
        "plt.subplot(3,2,2)\n",
        "\n",
        "plt.title(\"Distribution of Spam\")\n",
        "# data[data.target == 1].length.plot(kind = 'bar')\n",
        "sns.kdeplot(data[data.target == 1].length)\n",
        "# plt.tick_params(\n",
        "#     axis='x',          # changes apply to the x-axis\n",
        "#     which='both',      # both major and minor ticks are affected\n",
        "#     bottom=False,      # ticks along the bottom edge are off\n",
        "#     top=False,         # ticks along the top edge are off\n",
        "#     labelbottom=False) \n",
        "\n",
        "plt.subplot(3,2,4)\n",
        "plt.title(\"Distribution of Non-spam\")\n",
        "sns.kdeplot(data[data.target == -1].length)\n",
        "\n",
        "# plt.tick_params(\n",
        "#     axis='x',          # changes apply to the x-axis\n",
        "#     which='both',      # both major and minor ticks are affected\n",
        "#     bottom=False,      # ticks along the bottom edge are off\n",
        "#     top=False,         # ticks along the top edge are off\n",
        "#     labelbottom=False) \n",
        "\n",
        "plt.subplot(3,2,6)\n",
        "plt.title(\"Distribution of all emails together\")\n",
        "sns.kdeplot(data.length)\n",
        "\n",
        "plt.show()\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {
        "id": "GRhh3DJXNLsx"
      },
      "outputs": [],
      "source": [
        "# Function for explained variance by different number of component in TruncatedSVD \n",
        "def explained_variance(data,n_components):\n",
        "    svd = TruncatedSVD(n_components =n_components ,n_iter=7, random_state=42)\n",
        "    svd.fit(data)\n",
        "    return (svd.explained_variance_ratio_).sum()\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3u6Y1tMmNLsx"
      },
      "source": [
        "### Model fitting and hyperparameter tuning by applying nested cross-validation\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "id": "j99zT0afNLsy"
      },
      "outputs": [],
      "source": [
        "# # load data\n",
        "# path = \"Spam_detection/emails.mat\"\n",
        "# ref = loadmat(path)\n",
        "# X = ref['X'].T\n",
        "# Y = ref['Y'].T"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 29,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ra6fj4uuNLsy",
        "outputId": "c32e328f-72f7-47d4-b33c-3ee1374f291c"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((200, 400), (200,))"
            ]
          },
          "metadata": {},
          "execution_count": 29
        }
      ],
      "source": [
        "# For decreasing cost and time of computation and for the sake of only showing the procedure a small slice\n",
        "# of main data will be taken. Thus, performance and evaluation doesn't reflects the actual result. It is just an example.\n",
        "X= X.iloc[:200, :400]\n",
        "y = Y.iloc[:200].values.ravel()\n",
        "# Shape of data\n",
        "X.shape, y.shape\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RGqbgpvYNLsy",
        "outputId": "a7faf6ab-6b8d-41ac-f6a0-325ce76ce244"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((160, 400), (40, 400), (160,), (40,))"
            ]
          },
          "metadata": {},
          "execution_count": 30
        }
      ],
      "source": [
        "X_train, X_test, y_train, y_test = train_test_split(X,y, test_size=.2)\n",
        "X_train.shape, X_test.shape, y_train.shape, y_test.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 31,
      "metadata": {
        "id": "q4vr9f6MNLsz"
      },
      "outputs": [],
      "source": [
        "# Initializing Classifiers\n",
        "# Several algorithms such as logistic regression, K-nearest neighbors, decision tree, SVC, and random forest will be tried \n",
        "clf1 = LogisticRegression(max_iter= 7000)\n",
        "clf2 = KNeighborsClassifier(algorithm='ball_tree', leaf_size=50)\n",
        "clf3 = DecisionTreeClassifier(random_state=1)\n",
        "clf4 = SVC(random_state=1, probability=True)\n",
        "clf5 = RandomForestClassifier(random_state=1)\n",
        "\n",
        "\n",
        "# Building the pipelines\n",
        "pipe1 = Pipeline(steps=[\n",
        "                        (\"svd\", TruncatedSVD(n_iter=7, random_state=42)),\n",
        "                        (\"scaler\", StandardScaler()),\n",
        "                        (\"clf1\",clf1 )])\n",
        "\n",
        "pipe2 = Pipeline(steps=[\n",
        "                        (\"svd\", TruncatedSVD(n_iter=7, random_state=42)),\n",
        "                        (\"scaler\", StandardScaler()),\n",
        "                        (\"clf2\", clf2)])\n",
        "pipe3 = Pipeline(steps=[\n",
        "                        (\"svd\", TruncatedSVD(n_iter=7, random_state=42)),\n",
        "                        (\"clf3\", clf3)])\n",
        "pipe4 = Pipeline(steps=[\n",
        "                        (\"svd\", TruncatedSVD(n_iter=7, random_state=42)),\n",
        "                        (\"scaler\", StandardScaler()),\n",
        "                        (\"clf4\", clf4)])\n",
        "pipe5 = Pipeline(steps=[\n",
        "                        (\"svd\", TruncatedSVD(n_iter=7, random_state=42)),\n",
        "                        (\"clf5\", clf5)])\n",
        "\n",
        "\n",
        "\n",
        "# Setting up the parameter grids\n",
        "param_grid1 = [{\n",
        "                'svd__n_components':[100,200,300],\n",
        "                'clf1__penalty': ['l2'],\n",
        "                'clf1__C':[.1, 1, 10],\n",
        "                'clf1__solver':['lbfgs','saga','newton-cg']},\n",
        "                {\n",
        "                'svd__n_components':[100,200,300],\n",
        "                'clf1__penalty': ['l1'],                \n",
        "                'clf1__C':[.1, 1, 10],\n",
        "                'clf1__solver':['liblinear','saga']}\n",
        "                ]\n",
        "param_grid2 = [{\n",
        "                'svd__n_components':[100,200,300],                \n",
        "                'clf2__n_neighbors': list(range(1, 10)),\n",
        "                'clf2__p': [1, 2]}]\n",
        "\n",
        "param_grid3 = {\n",
        "                'svd__n_components':[100,200,300],                \n",
        "                'clf3__max_depth': list(range(1, 10)) + [None],\n",
        "                'clf3__criterion': ['gini', 'entropy']}\n",
        "\n",
        "param_grid4 = [{\n",
        "                'svd__n_components':[100,200,300],\n",
        "                'clf4__kernel': ['rbf'],\n",
        "                'clf4__C': np.power(10., np.arange(-4, 4)),\n",
        "                'clf4__gamma': np.power(10., np.arange(-5, 0))},\n",
        "               {\n",
        "                'svd__n_components':[100,200,300],\n",
        "                'clf4__kernel': ['linear'],\n",
        "                'clf4__C': np.power(10., np.arange(-4, 4))}]\n",
        "\n",
        "param_grid5 = [{\n",
        "                'svd__n_components':[100,200,300],\n",
        "                'clf5__n_estimators': [10, 100, 500, 1000, 10000]}]\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 32,
      "metadata": {
        "id": "f4ucAe9jNLsz"
      },
      "outputs": [],
      "source": [
        "# Setting up multiple GridSearchCV objects, 1 for each algorithm\n",
        "gridcvs = {}\n",
        "inner_cv = KFold(n_splits=2, shuffle=True, random_state=1)\n",
        "\n",
        "for pgrid, est, name in zip((param_grid1, param_grid2, param_grid3, param_grid4, param_grid5),\n",
        "                            (pipe1, pipe2, pipe3, pipe4, pipe5 ),\n",
        "                            ('LogisticReg', 'KNN', 'DTree', 'SVM', 'RForest' )):\n",
        "    gcv = GridSearchCV(estimator=est,\n",
        "                       param_grid=pgrid,\n",
        "                       scoring='accuracy',\n",
        "                       n_jobs=-1,\n",
        "                       cv=inner_cv,\n",
        "                       verbose=0,\n",
        "                       refit=True)\n",
        "    gridcvs[name] = gcv"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 33,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5CqlPP1FNLsz",
        "outputId": "d76e5d5f-f53c-4ffb-c918-bf6f76adbd8c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "-------------------------------------------------- \n",
            "\n",
            "Algorithm: DTree\n",
            "    Inner loop:\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 81.25%\n",
            "        Best parameters: {'clf3__criterion': 'gini', 'clf3__max_depth': 1, 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 90.62%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 89.06%\n",
            "        Best parameters: {'clf3__criterion': 'gini', 'clf3__max_depth': 1, 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 65.62%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 84.38%\n",
            "        Best parameters: {'clf3__criterion': 'gini', 'clf3__max_depth': 3, 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 84.38%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 82.81%\n",
            "        Best parameters: {'clf3__criterion': 'gini', 'clf3__max_depth': 4, 'svd__n_components': 300}\n",
            "        ACC (on outer test fold) 90.62%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 82.81%\n",
            "        Best parameters: {'clf3__criterion': 'gini', 'clf3__max_depth': 1, 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 84.38%\n",
            "\n",
            "    Outer Loop:\n",
            "        ACC 83.12% +/- 9.19\n",
            "-------------------------------------------------- \n",
            "\n",
            "Algorithm: KNN\n",
            "    Inner loop:\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 82.03%\n",
            "        Best parameters: {'clf2__n_neighbors': 2, 'clf2__p': 1, 'svd__n_components': 200}\n",
            "        ACC (on outer test fold) 87.50%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 90.62%\n",
            "        Best parameters: {'clf2__n_neighbors': 1, 'clf2__p': 1, 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 65.62%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 83.59%\n",
            "        Best parameters: {'clf2__n_neighbors': 1, 'clf2__p': 1, 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 87.50%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 83.59%\n",
            "        Best parameters: {'clf2__n_neighbors': 1, 'clf2__p': 1, 'svd__n_components': 300}\n",
            "        ACC (on outer test fold) 84.38%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 82.81%\n",
            "        Best parameters: {'clf2__n_neighbors': 3, 'clf2__p': 1, 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 84.38%\n",
            "\n",
            "    Outer Loop:\n",
            "        ACC 81.88% +/- 8.24\n",
            "-------------------------------------------------- \n",
            "\n",
            "Algorithm: LogisticReg\n",
            "    Inner loop:\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 82.81%\n",
            "        Best parameters: {'clf1__C': 1, 'clf1__penalty': 'l1', 'clf1__solver': 'saga', 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 90.62%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 90.62%\n",
            "        Best parameters: {'clf1__C': 1, 'clf1__penalty': 'l1', 'clf1__solver': 'liblinear', 'svd__n_components': 300}\n",
            "        ACC (on outer test fold) 46.88%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 84.38%\n",
            "        Best parameters: {'clf1__C': 0.1, 'clf1__penalty': 'l2', 'clf1__solver': 'lbfgs', 'svd__n_components': 300}\n",
            "        ACC (on outer test fold) 78.12%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 82.81%\n",
            "        Best parameters: {'clf1__C': 0.1, 'clf1__penalty': 'l2', 'clf1__solver': 'lbfgs', 'svd__n_components': 200}\n",
            "        ACC (on outer test fold) 84.38%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 82.81%\n",
            "        Best parameters: {'clf1__C': 0.1, 'clf1__penalty': 'l1', 'clf1__solver': 'liblinear', 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 84.38%\n",
            "\n",
            "    Outer Loop:\n",
            "        ACC 76.88% +/- 15.51\n",
            "-------------------------------------------------- \n",
            "\n",
            "Algorithm: RForest\n",
            "    Inner loop:\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 82.03%\n",
            "        Best parameters: {'clf5__n_estimators': 1000, 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 90.62%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 89.06%\n",
            "        Best parameters: {'clf5__n_estimators': 10, 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 65.62%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 85.16%\n",
            "        Best parameters: {'clf5__n_estimators': 100, 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 84.38%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 84.38%\n",
            "        Best parameters: {'clf5__n_estimators': 10, 'svd__n_components': 300}\n",
            "        ACC (on outer test fold) 90.62%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 82.81%\n",
            "        Best parameters: {'clf5__n_estimators': 10, 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 90.62%\n",
            "\n",
            "    Outer Loop:\n",
            "        ACC 84.38% +/- 9.68\n",
            "-------------------------------------------------- \n",
            "\n",
            "Algorithm: SVM\n",
            "    Inner loop:\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 82.03%\n",
            "        Best parameters: {'clf4__C': 0.0001, 'clf4__gamma': 1e-05, 'clf4__kernel': 'rbf', 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 87.50%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 90.62%\n",
            "        Best parameters: {'clf4__C': 0.01, 'clf4__kernel': 'linear', 'svd__n_components': 300}\n",
            "        ACC (on outer test fold) 53.12%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 84.38%\n",
            "        Best parameters: {'clf4__C': 1.0, 'clf4__gamma': 0.01, 'clf4__kernel': 'rbf', 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 87.50%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 83.59%\n",
            "        Best parameters: {'clf4__C': 1.0, 'clf4__gamma': 0.1, 'clf4__kernel': 'rbf', 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 90.62%\n",
            "\n",
            "        Best ACC (avg. of inner test folds) 82.81%\n",
            "        Best parameters: {'clf4__C': 0.0001, 'clf4__gamma': 1e-05, 'clf4__kernel': 'rbf', 'svd__n_components': 100}\n",
            "        ACC (on outer test fold) 84.38%\n",
            "\n",
            "    Outer Loop:\n",
            "        ACC 80.62% +/- 13.89\n"
          ]
        }
      ],
      "source": [
        "# Implementing nested cross-validation\n",
        "# Fitting model on inner loop and outer loop of training set\n",
        "for name, gs_est in sorted(gridcvs.items()):\n",
        "\n",
        "    print(50 * '-', '\\n')\n",
        "    print('Algorithm:', name)\n",
        "    print('    Inner loop:')\n",
        "    \n",
        "    outer_scores = []\n",
        "    outer_cv = KFold(n_splits=5, shuffle=True, random_state=1)\n",
        "    \n",
        "    \n",
        "    for train_idx, valid_idx in outer_cv.split(X_train, y_train):\n",
        "        \n",
        "        gridcvs[name].fit(X_train.iloc[train_idx,:], y_train[train_idx]) # run inner loop hyperparam tuning\n",
        "        print('\\n        Best ACC (avg. of inner test folds) %.2f%%' % (gridcvs[name].best_score_ * 100))\n",
        "        print('        Best parameters:', gridcvs[name].best_params_)\n",
        "        \n",
        "        # perf on test fold (valid_idx)\n",
        "        outer_scores.append(gridcvs[name].best_estimator_.score(X_train.iloc[valid_idx,:], y_train[valid_idx]))\n",
        "        print('        ACC (on outer test fold) %.2f%%' % (outer_scores[-1]*100))\n",
        "    \n",
        "    print('\\n    Outer Loop:')\n",
        "    print('        ACC %.2f%% +/- %.2f' % \n",
        "              (np.mean(outer_scores) * 100, np.std(outer_scores) * 100))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 34,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "67zEodE3NLs0",
        "outputId": "481922b3-1dad-47b8-f496-67c5ed0480f4"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Fitting 2 folds for each of 15 candidates, totalling 30 fits\n",
            "Best CV accuracy: 84.38%\n",
            "Best parameters: {'clf5__n_estimators': 10, 'svd__n_components': 100}\n"
          ]
        }
      ],
      "source": [
        "# Fitting best estiamtor of test set\n",
        "# The best estiamtor can be defferent time to time because of randomness of K-fold.\n",
        "gcv_model_select = GridSearchCV(estimator=pipe5,\n",
        "                                param_grid=param_grid5,\n",
        "                                scoring='accuracy',\n",
        "                                n_jobs=-1,\n",
        "                                cv=inner_cv,\n",
        "                                verbose=1,\n",
        "                                refit=True)\n",
        "\n",
        "gcv_model_select.fit(X_train, y_train)\n",
        "print('Best CV accuracy: %.2f%%' % (gcv_model_select.best_score_*100))\n",
        "print('Best parameters:', gcv_model_select.best_params_)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ti6XR5OiNLs0"
      },
      "source": [
        "### Model evaluation"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 35,
      "metadata": {
        "id": "Yy6VP97VNLs0"
      },
      "outputs": [],
      "source": [
        "# Function for visualizing performance of best estiamtor.\n",
        "\n",
        "def model_accuracy(X_train, y_train, X_test, y_test, model):\n",
        "    y_train_pred_=model.predict(X_train)\n",
        "    y_test_pred_=model.predict(X_test)\n",
        "    train_acc = accuracy_score(y_true=y_train, y_pred=y_train_pred_)\n",
        "    test_acc = accuracy_score(y_true=y_test, y_pred=y_test_pred_)\n",
        "\n",
        "    print('Training Accuracy: %.2f%%' % (100 * train_acc))\n",
        "    print('Test Accuracy: %.2f%%' % (100 * test_acc))\n",
        "\n",
        "    tn, fp, fn, tp = metrics.confusion_matrix(y_test, y_test_pred_).ravel()\n",
        "    print('legitimat_prediction_rate: %.2f%%' % ((fp + fn) /len(y_test)))\n",
        "    target_names = ['Spam', 'Non-spam']\n",
        "    print(classification_report(y_test, y_test_pred_, target_names=target_names, zero_division= 1))\n",
        "    y_pred_proba = model.predict_proba(X_test)[::,1]\n",
        "    fpr, tpr, _ = metrics.roc_curve(y_test,  y_pred_proba)\n",
        "    auc = metrics.roc_auc_score(y_test, y_pred_proba)\n",
        "    plt.plot(fpr,tpr,label=\"auc=\"+str(round(auc, 3)))\n",
        "    plt.title(\"ROC Curve\")\n",
        "    plt.legend(loc=4)\n",
        "    \n",
        "    return plt.show()\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 39,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 666
        },
        "id": "8ey8cfW2NLs1",
        "outputId": "323baba8-d917-4d55-ce0f-43c73e0277dc"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Training Accuracy: 89.38%\n",
            "Test Accuracy: 85.00%\n",
            "legitimat_prediction_rate: 0.15%\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "        Spam       0.60      0.43      0.50         7\n",
            "    Non-spam       0.89      0.94      0.91        33\n",
            "\n",
            "    accuracy                           0.85        40\n",
            "   macro avg       0.74      0.68      0.71        40\n",
            "weighted avg       0.84      0.85      0.84        40\n",
            "\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Applying model_accuracy function\n",
        "model_accuracy(X_train, y_train, X_test, y_test, gcv_model_select)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zCunT_FkNLs1"
      },
      "source": [
        "### Fitting Neural Network and hyperparameter tuning by applying nested cross-validation\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 55,
      "metadata": {
        "id": "01kKDSKDNLs1"
      },
      "outputs": [],
      "source": [
        "# Hyperparameters such as different unites of dense layers, optimizers, batch_size and epochs will be tuned.\n",
        "# Function to create model, required for KerasClassifier\n",
        "def create_model(unit1, unit2, optimizer):\n",
        "\t# create model\n",
        "\tmodel = Sequential()\n",
        "\tmodel.add(Dense(unit1, input_shape=(X_train.shape[1],), activation='relu'))\n",
        "\tmodel.add(Dense(unit2, activation='relu'))\n",
        "\tmodel.add(Dense(20, activation='relu'))\n",
        "\tmodel.add(Dense(1, activation='sigmoid'))\n",
        "\t# Compile model\n",
        "\tmodel.compile(loss='binary_crossentropy', optimizer =optimizer , metrics=['accuracy'], run_eagerly=True)\n",
        "\treturn model\n",
        "\n",
        "# create model\n",
        "model = KerasClassifier(model=create_model, verbose=0)\n",
        "\n",
        "# define the grid search parameters\n",
        "unit1 = [100,80]\n",
        "unit2 = [80, 60]\n",
        "optimizer = ['SGD','Adam', 'Adamax'] # 'RMSprop', 'Adagrad', 'Adadelta', 'Adam', 'Adamax', 'Nadam'\n",
        "batch_size = [30, 60]\n",
        "epochs = [20, 30]\n",
        "\n",
        "param_grid = dict(model__unit1=unit1, model__unit2=unit2 ,model__optimizer=optimizer,batch_size=batch_size, epochs=epochs)\n",
        "# param_grid = dict(model__unit1=unit1, model__unit2=unit2,model__optimizer=optimizer)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 56,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "23vE42zBNLs2",
        "outputId": "cd58e4d0-ca1d-4de3-a993-c92f534d2f88"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best: 0.862683 using {'batch_size': 30, 'epochs': 30, 'model__optimizer': 'Adam', 'model__unit1': 80, 'model__unit2': 80}\n",
            "0.837526 (0.023298) with: {'batch_size': 30, 'epochs': 20, 'model__optimizer': 'SGD', 'model__unit1': 100, 'model__unit2': 80}\n",
            "0.837526 (0.023298) with: {'batch_size': 30, 'epochs': 20, 'model__optimizer': 'SGD', 'model__unit1': 100, 'model__unit2': 60}\n",
            "0.831237 (0.030847) with: {'batch_size': 30, 'epochs': 20, 'model__optimizer': 'SGD', 'model__unit1': 80, 'model__unit2': 80}\n",
            "0.831237 (0.030847) with: {'batch_size': 30, 'epochs': 20, 'model__optimizer': 'SGD', 'model__unit1': 80, 'model__unit2': 60}\n",
            "0.837526 (0.023298) with: {'batch_size': 30, 'epochs': 20, 'model__optimizer': 'Adam', 'model__unit1': 100, 'model__unit2': 80}\n",
            "0.843932 (0.021983) with: {'batch_size': 30, 'epochs': 20, 'model__optimizer': 'Adam', 'model__unit1': 100, 'model__unit2': 60}\n",
            "0.831353 (0.025899) with: {'batch_size': 30, 'epochs': 20, 'model__optimizer': 'Adam', 'model__unit1': 80, 'model__unit2': 80}\n",
            "0.825064 (0.031652) with: {'batch_size': 30, 'epochs': 20, 'model__optimizer': 'Adam', 'model__unit1': 80, 'model__unit2': 60}\n",
            "0.843699 (0.023817) with: {'batch_size': 30, 'epochs': 20, 'model__optimizer': 'Adamax', 'model__unit1': 100, 'model__unit2': 80}\n",
            "0.831237 (0.030847) with: {'batch_size': 30, 'epochs': 20, 'model__optimizer': 'Adamax', 'model__unit1': 100, 'model__unit2': 60}\n",
            "0.831237 (0.030847) with: {'batch_size': 30, 'epochs': 20, 'model__optimizer': 'Adamax', 'model__unit1': 80, 'model__unit2': 80}\n",
            "0.831237 (0.030847) with: {'batch_size': 30, 'epochs': 20, 'model__optimizer': 'Adamax', 'model__unit1': 80, 'model__unit2': 60}\n",
            "0.825064 (0.031652) with: {'batch_size': 30, 'epochs': 30, 'model__optimizer': 'SGD', 'model__unit1': 100, 'model__unit2': 80}\n",
            "0.831237 (0.030847) with: {'batch_size': 30, 'epochs': 30, 'model__optimizer': 'SGD', 'model__unit1': 100, 'model__unit2': 60}\n",
            "0.831237 (0.030847) with: {'batch_size': 30, 'epochs': 30, 'model__optimizer': 'SGD', 'model__unit1': 80, 'model__unit2': 80}\n",
            "0.831237 (0.030847) with: {'batch_size': 30, 'epochs': 30, 'model__optimizer': 'SGD', 'model__unit1': 80, 'model__unit2': 60}\n",
            "0.850221 (0.039701) with: {'batch_size': 30, 'epochs': 30, 'model__optimizer': 'Adam', 'model__unit1': 100, 'model__unit2': 80}\n",
            "0.850105 (0.040298) with: {'batch_size': 30, 'epochs': 30, 'model__optimizer': 'Adam', 'model__unit1': 100, 'model__unit2': 60}\n",
            "0.862683 (0.031060) with: {'batch_size': 30, 'epochs': 30, 'model__optimizer': 'Adam', 'model__unit1': 80, 'model__unit2': 80}\n",
            "0.856511 (0.037460) with: {'batch_size': 30, 'epochs': 30, 'model__optimizer': 'Adam', 'model__unit1': 80, 'model__unit2': 60}\n",
            "0.837410 (0.032459) with: {'batch_size': 30, 'epochs': 30, 'model__optimizer': 'Adamax', 'model__unit1': 100, 'model__unit2': 80}\n",
            "0.837410 (0.032459) with: {'batch_size': 30, 'epochs': 30, 'model__optimizer': 'Adamax', 'model__unit1': 100, 'model__unit2': 60}\n",
            "0.837410 (0.032459) with: {'batch_size': 30, 'epochs': 30, 'model__optimizer': 'Adamax', 'model__unit1': 80, 'model__unit2': 80}\n",
            "0.837526 (0.023298) with: {'batch_size': 30, 'epochs': 30, 'model__optimizer': 'Adamax', 'model__unit1': 80, 'model__unit2': 60}\n",
            "0.831237 (0.030847) with: {'batch_size': 60, 'epochs': 20, 'model__optimizer': 'SGD', 'model__unit1': 100, 'model__unit2': 80}\n",
            "0.818891 (0.015977) with: {'batch_size': 60, 'epochs': 20, 'model__optimizer': 'SGD', 'model__unit1': 100, 'model__unit2': 60}\n",
            "0.843583 (0.036168) with: {'batch_size': 60, 'epochs': 20, 'model__optimizer': 'SGD', 'model__unit1': 80, 'model__unit2': 80}\n",
            "0.831237 (0.015477) with: {'batch_size': 60, 'epochs': 20, 'model__optimizer': 'SGD', 'model__unit1': 80, 'model__unit2': 60}\n",
            "0.843583 (0.036168) with: {'batch_size': 60, 'epochs': 20, 'model__optimizer': 'Adam', 'model__unit1': 100, 'model__unit2': 80}\n",
            "0.837410 (0.032459) with: {'batch_size': 60, 'epochs': 20, 'model__optimizer': 'Adam', 'model__unit1': 100, 'model__unit2': 60}\n",
            "0.831237 (0.030847) with: {'batch_size': 60, 'epochs': 20, 'model__optimizer': 'Adam', 'model__unit1': 80, 'model__unit2': 80}\n",
            "0.837410 (0.032459) with: {'batch_size': 60, 'epochs': 20, 'model__optimizer': 'Adam', 'model__unit1': 80, 'model__unit2': 60}\n",
            "0.831237 (0.030847) with: {'batch_size': 60, 'epochs': 20, 'model__optimizer': 'Adamax', 'model__unit1': 100, 'model__unit2': 80}\n",
            "0.825064 (0.031652) with: {'batch_size': 60, 'epochs': 20, 'model__optimizer': 'Adamax', 'model__unit1': 100, 'model__unit2': 60}\n",
            "0.825064 (0.031652) with: {'batch_size': 60, 'epochs': 20, 'model__optimizer': 'Adamax', 'model__unit1': 80, 'model__unit2': 80}\n",
            "0.843583 (0.036168) with: {'batch_size': 60, 'epochs': 20, 'model__optimizer': 'Adamax', 'model__unit1': 80, 'model__unit2': 60}\n",
            "0.837526 (0.023298) with: {'batch_size': 60, 'epochs': 30, 'model__optimizer': 'SGD', 'model__unit1': 100, 'model__unit2': 80}\n",
            "0.831237 (0.030847) with: {'batch_size': 60, 'epochs': 30, 'model__optimizer': 'SGD', 'model__unit1': 100, 'model__unit2': 60}\n",
            "0.837526 (0.023298) with: {'batch_size': 60, 'epochs': 30, 'model__optimizer': 'SGD', 'model__unit1': 80, 'model__unit2': 80}\n",
            "0.824948 (0.023857) with: {'batch_size': 60, 'epochs': 30, 'model__optimizer': 'SGD', 'model__unit1': 80, 'model__unit2': 60}\n",
            "0.831353 (0.025899) with: {'batch_size': 60, 'epochs': 30, 'model__optimizer': 'Adam', 'model__unit1': 100, 'model__unit2': 80}\n",
            "0.849988 (0.015462) with: {'batch_size': 60, 'epochs': 30, 'model__optimizer': 'Adam', 'model__unit1': 100, 'model__unit2': 60}\n",
            "0.849872 (0.027278) with: {'batch_size': 60, 'epochs': 30, 'model__optimizer': 'Adam', 'model__unit1': 80, 'model__unit2': 80}\n",
            "0.837526 (0.023298) with: {'batch_size': 60, 'epochs': 30, 'model__optimizer': 'Adam', 'model__unit1': 80, 'model__unit2': 60}\n",
            "0.831237 (0.030847) with: {'batch_size': 60, 'epochs': 30, 'model__optimizer': 'Adamax', 'model__unit1': 100, 'model__unit2': 80}\n",
            "0.837410 (0.032459) with: {'batch_size': 60, 'epochs': 30, 'model__optimizer': 'Adamax', 'model__unit1': 100, 'model__unit2': 60}\n",
            "0.843699 (0.023817) with: {'batch_size': 60, 'epochs': 30, 'model__optimizer': 'Adamax', 'model__unit1': 80, 'model__unit2': 80}\n",
            "0.837410 (0.032459) with: {'batch_size': 60, 'epochs': 30, 'model__optimizer': 'Adamax', 'model__unit1': 80, 'model__unit2': 60}\n"
          ]
        }
      ],
      "source": [
        "# Fitting models of entire traing data\n",
        "\n",
        "inner_cv = KFold(n_splits=3, shuffle=True, random_state=1)\n",
        "grid = GridSearchCV(estimator=model, param_grid=param_grid, n_jobs=-1, cv=inner_cv)\n",
        "grid_result = grid.fit(X_train, y_train)\n",
        "\n",
        "# summarize results\n",
        "print(\"Best: %f using %s\" % (grid_result.best_score_, grid_result.best_params_))\n",
        "means = grid_result.cv_results_['mean_test_score']\n",
        "stds = grid_result.cv_results_['std_test_score']\n",
        "params = grid_result.cv_results_['params']\n",
        "for mean, stdev, param in zip(means, stds, params):\n",
        "    print(\"%f (%f) with: %r\" % (mean, stdev, param))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 57,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 345
        },
        "id": "Xu7BDtbjNLs2",
        "outputId": "5768ed31-0b6b-4adb-ca66-8a445d4fba62"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "    Outer Loop:\n",
            "        ACC 83.12% +/- 10.57\n",
            "\n",
            "    Inner Loop:\n",
            "        ACC 85.45% +/- 1.88\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "       batch_size  epochs model__optimizer  model__unit1  model__unit2  \\\n",
              "model                                                                    \n",
              "0              60      30              SGD            80            80   \n",
              "1              30      30             Adam           100            80   \n",
              "2              60      30              SGD            80            80   \n",
              "3              30      30             Adam           100            60   \n",
              "4              30      30             Adam           100            80   \n",
              "\n",
              "       best_outetr_score  best_inner_score  \n",
              "model                                       \n",
              "0                0.87500          0.835364  \n",
              "1                0.62500          0.890181  \n",
              "2                0.84375          0.851421  \n",
              "3                0.90625          0.851606  \n",
              "4                0.90625          0.843854  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-9a242d81-eb56-4677-b2e9-c00fcd8b585d\">\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>batch_size</th>\n",
              "      <th>epochs</th>\n",
              "      <th>model__optimizer</th>\n",
              "      <th>model__unit1</th>\n",
              "      <th>model__unit2</th>\n",
              "      <th>best_outetr_score</th>\n",
              "      <th>best_inner_score</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>model</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>60</td>\n",
              "      <td>30</td>\n",
              "      <td>SGD</td>\n",
              "      <td>80</td>\n",
              "      <td>80</td>\n",
              "      <td>0.87500</td>\n",
              "      <td>0.835364</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>30</td>\n",
              "      <td>30</td>\n",
              "      <td>Adam</td>\n",
              "      <td>100</td>\n",
              "      <td>80</td>\n",
              "      <td>0.62500</td>\n",
              "      <td>0.890181</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>60</td>\n",
              "      <td>30</td>\n",
              "      <td>SGD</td>\n",
              "      <td>80</td>\n",
              "      <td>80</td>\n",
              "      <td>0.84375</td>\n",
              "      <td>0.851421</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>30</td>\n",
              "      <td>30</td>\n",
              "      <td>Adam</td>\n",
              "      <td>100</td>\n",
              "      <td>60</td>\n",
              "      <td>0.90625</td>\n",
              "      <td>0.851606</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>30</td>\n",
              "      <td>30</td>\n",
              "      <td>Adam</td>\n",
              "      <td>100</td>\n",
              "      <td>80</td>\n",
              "      <td>0.90625</td>\n",
              "      <td>0.843854</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-9a242d81-eb56-4677-b2e9-c00fcd8b585d')\"\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-9a242d81-eb56-4677-b2e9-c00fcd8b585d 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-9a242d81-eb56-4677-b2e9-c00fcd8b585d');\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": 57
        }
      ],
      "source": [
        "# Implementing nested cross-validation\n",
        "# Fitting model on inner loop and outer loop of training set\n",
        "\n",
        "outer_scores = []\n",
        "best_params = []\n",
        "best_inner_scores = []\n",
        "outer_cv = KFold(n_splits=5, shuffle=True, random_state=1)\n",
        "\n",
        "\n",
        "for train_idx, valid_idx in outer_cv.split(X_train, y_train):\n",
        "\n",
        "    grid_result.fit(X_train.iloc[train_idx,:], y_train[train_idx]) # run inner loop hyperparam tuning\n",
        "    best_inner_scores.append(grid_result.best_score_)   \n",
        "    best_params.append(grid_result.best_params_)\n",
        "    # perf on test fold (valid_idx)\n",
        "    outer_scores.append(grid_result.best_estimator_.score(X_train.iloc[valid_idx,:], y_train[valid_idx]))\n",
        "\n",
        "print('\\n    Outer Loop:')\n",
        "print('        ACC %.2f%% +/- %.2f' % \n",
        "            (np.mean(outer_scores) * 100, np.std(outer_scores) * 100))\n",
        "print('\\n    Inner Loop:')\n",
        "print('        ACC %.2f%% +/- %.2f' % \n",
        "            (np.mean(best_inner_scores) * 100, np.std(best_inner_scores) * 100))\n",
        "\n",
        "best_params_ = pd.DataFrame(best_params)\n",
        "best_params_['best_outetr_score']  = outer_scores\n",
        "best_params_['best_inner_score']  = best_inner_scores\n",
        "best_params_[\"model\"] = best_params_.index\n",
        "best_params_ = best_params_.set_index('model')\n",
        "best_params_"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 58,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 738
        },
        "id": "SB5X6l7eNLs2",
        "outputId": "73202ede-de6d-4dac-e4ba-884b4c75faa5"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Best model:\n",
            "  model 3\n",
            "Best model's parameteres:\n",
            " {'batch_size': 30, 'epochs': 30, 'model__optimizer': 'Adam', 'model__unit1': 100, 'model__unit2': 60}\n",
            "Training Accuracy: 90.00%\n",
            "Test Accuracy: 87.50%\n",
            "legitimat_prediction_rate: 0.12%\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "        Spam       0.67      0.57      0.62         7\n",
            "    Non-spam       0.91      0.94      0.93        33\n",
            "\n",
            "    accuracy                           0.88        40\n",
            "   macro avg       0.79      0.76      0.77        40\n",
            "weighted avg       0.87      0.88      0.87        40\n",
            "\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Based on the above table we can decide on which model to choose. The model with\n",
        "# a higher score in the outer part of training data, could be chosen for the test set.\n",
        "models = best_params_.copy()\n",
        "\n",
        "models_ = models.iloc[np.argmax(models.best_outetr_score),:5].to_dict()\n",
        "grid_result.best_estimator_.set_params(**models_)\n",
        "\n",
        "best_model = grid_result.best_estimator_.fit(X_train, y_train)\n",
        "print(f\"Best model:\\n  model\", np.argmax(models.best_outetr_score))\n",
        "print(f\"Best model's parameteres:\\n\", models_)\n",
        "model_accuracy(X_train, y_train, X_test, y_test, best_model)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UYX8snCANLs2"
      },
      "source": [
        "### Models comparison"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 59,
      "metadata": {
        "id": "s7ZdFCLpNLs2"
      },
      "outputs": [],
      "source": [
        "# Function for plotting area under cureve (auc) of each model\n",
        "\n",
        "def models_auc(X_train, y_train, X_test, y_test,grid_result, best_params_):\n",
        "    for i in  range(best_params_.shape[0]):\n",
        "    \n",
        "        best_params__ = best_params_.iloc[i,:5].to_dict()\n",
        "        grid_result.best_estimator_.set_params(**best_params__)\n",
        "        grid_result.best_estimator_\n",
        "        best_model = grid_result.best_estimator_.fit(X_train, y_train)\n",
        "\n",
        "\n",
        "        y_pred_proba = best_model.predict_proba(X_test)[::,1]\n",
        "        fpr, tpr, _ = metrics.roc_curve(y_test,  y_pred_proba)\n",
        "        auc = metrics.roc_auc_score(y_test, y_pred_proba)\n",
        "        plt.plot(fpr,tpr,label=\"model\"+str(i)+ \" auc=\"+str(round(auc, 3)))\n",
        "        plt.title(\"ROC Curve\")\n",
        "        plt.legend(loc=2)\n",
        "        \n",
        "    return         \n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 60,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 452
        },
        "id": "7K4OWKtNNLs3",
        "outputId": "35bea929-a94e-4e4a-a0ea-26e61350f80c"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Polt shows each models performance according to its auc rate. The cureve closer to top and left side of the figure has higher auc.\n",
        "models = best_params_.copy()\n",
        "models_auc(X_train, y_train, X_test, y_test,grid_result, models)"
      ]
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.9.7"
    },
    "vscode": {
      "interpreter": {
        "hash": "addb8bc39cdd9526d9c14036c250468c073d597d5675ade8e31643e7a6a9befd"
      }
    },
    "colab": {
      "provenance": [],
      "include_colab_link": true
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}