Black_Friday_Model_Training.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Black Friday Model Training.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm",
      "authorship_tag": "ABX9TyOOd51TiOhCdFgWHDR3AHhS",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Puru35/Black-Friday-Sales-Prediction/blob/master/Black_Friday_Model_Training.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6dynwK7vtUAC",
        "colab_type": "text"
      },
      "source": [
        "This is the Third and final notebook in this project. In this notebook we will train various models to get the best accuracy while predicting the price of an item on black friday sale. \n",
        "\n",
        "The conclusion of the full project will be at the end of this notebook."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_ln1xnyEdmuo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.preprocessing import OneHotEncoder, MinMaxScaler, StandardScaler\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "from sklearn import datasets, linear_model, metrics\n",
        "from sklearn.linear_model import LinearRegression\n",
        "from sklearn import preprocessing, svm \n",
        "from sklearn.tree import DecisionTreeRegressor\n",
        "from sklearn.ensemble import RandomForestRegressor, GradientBoostingRegressor\n",
        "import tensorflow as tf\n",
        "from tensorflow import keras"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HhXJvLixuINS",
        "colab_type": "text"
      },
      "source": [
        "## Importing the cleaned Dataset"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BBXSj-dvdxjd",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "data = pd.read_csv(\"Black Friday cleaned Data set.csv\")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8Zmgq5OPeRos",
        "colab_type": "code",
        "outputId": "55c82b77-0107-434e-914b-ea3052d62b07",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 422
        }
      },
      "source": [
        "data"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Occupation</th>\n",
              "      <th>City_Category</th>\n",
              "      <th>Stay_In_Current_City_Years</th>\n",
              "      <th>Marital_Status</th>\n",
              "      <th>Product_Category_1</th>\n",
              "      <th>Product_Category_2</th>\n",
              "      <th>Product_Category_3</th>\n",
              "      <th>Purchase</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0</td>\n",
              "      <td>55</td>\n",
              "      <td>10</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>8370</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0</td>\n",
              "      <td>55</td>\n",
              "      <td>10</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>6</td>\n",
              "      <td>14</td>\n",
              "      <td>15200</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0</td>\n",
              "      <td>55</td>\n",
              "      <td>10</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>12</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1422</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0</td>\n",
              "      <td>55</td>\n",
              "      <td>10</td>\n",
              "      <td>2</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>12</td>\n",
              "      <td>14</td>\n",
              "      <td>0</td>\n",
              "      <td>1057</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1</td>\n",
              "      <td>55</td>\n",
              "      <td>16</td>\n",
              "      <td>0</td>\n",
              "      <td>4</td>\n",
              "      <td>0</td>\n",
              "      <td>8</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>7969</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>550063</th>\n",
              "      <td>1</td>\n",
              "      <td>53</td>\n",
              "      <td>13</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>20</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>368</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>550064</th>\n",
              "      <td>0</td>\n",
              "      <td>30</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>3</td>\n",
              "      <td>0</td>\n",
              "      <td>20</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>371</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>550065</th>\n",
              "      <td>0</td>\n",
              "      <td>30</td>\n",
              "      <td>15</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>20</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>137</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>550066</th>\n",
              "      <td>0</td>\n",
              "      <td>55</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>2</td>\n",
              "      <td>0</td>\n",
              "      <td>20</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>365</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>550067</th>\n",
              "      <td>0</td>\n",
              "      <td>48</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>4</td>\n",
              "      <td>1</td>\n",
              "      <td>20</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>490</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>550068 rows × 10 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "        Gender  Age  ...  Product_Category_3  Purchase\n",
              "0            0   55  ...                   0      8370\n",
              "1            0   55  ...                  14     15200\n",
              "2            0   55  ...                   0      1422\n",
              "3            0   55  ...                   0      1057\n",
              "4            1   55  ...                   0      7969\n",
              "...        ...  ...  ...                 ...       ...\n",
              "550063       1   53  ...                   0       368\n",
              "550064       0   30  ...                   0       371\n",
              "550065       0   30  ...                   0       137\n",
              "550066       0   55  ...                   0       365\n",
              "550067       0   48  ...                   0       490\n",
              "\n",
              "[550068 rows x 10 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kTqW6BO4uOwE",
        "colab_type": "text"
      },
      "source": [
        "Here we are going to import the dataset that we cleaned in the first notebook."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Nb9NTrsEeTSY",
        "colab_type": "code",
        "outputId": "d857121b-b462-4502-f641-c2704c462ec2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 202
        }
      },
      "source": [
        "data.isnull().any()"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Gender                        False\n",
              "Age                           False\n",
              "Occupation                    False\n",
              "City_Category                 False\n",
              "Stay_In_Current_City_Years    False\n",
              "Marital_Status                False\n",
              "Product_Category_1            False\n",
              "Product_Category_2            False\n",
              "Product_Category_3            False\n",
              "Purchase                      False\n",
              "dtype: bool"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5rJl-VxGucCI",
        "colab_type": "text"
      },
      "source": [
        "Just to verify if there are any null values in our dataset after cleaning, we use isnull().any()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oUp4PH4WvNum",
        "colab_type": "text"
      },
      "source": [
        "## Pearson Correlation between the target and rest of the features"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_879eFmVho_D",
        "colab_type": "code",
        "outputId": "1d02ced4-1f47-4cd5-b270-94fd54e207bd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 737
        }
      },
      "source": [
        "corrdf = data.corr('pearson')\n",
        "\n",
        "fig, ax = plt.subplots(figsize=(10,10)) \n",
        "sns.heatmap(corrdf, annot=True, linewidths=.5, ax=ax)"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f8dfae3deb8>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAArAAAAK/CAYAAACREv5jAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOzdd1yV5f/H8dfFAUQcLBE0B87MCYYjG2IpKuVIzdRylGW5cuHMlZY7cpSalZpaqaVlFuZI0Epz4jbFgYoKIhtFUc71++Mc8YBoKKvz+36ejwcP5Zx7vO+L61xefM51bpXWGiGEEEIIIayFTWEHEEIIIYQQ4mHIBFYIIYQQQlgVmcAKIYQQQgirIhNYIYQQQghhVWQCK4QQQgghrIpMYIUQQgghhFWRCawQQgghhHgkSqnFSqkrSqkj93leKaXmKqVOKaUOKaXq58V5ZQIrhBBCCCEe1VKg1QOebw1UM3/1ARbkxUllAiuEEEIIIR6J1no7EPeATdoBy7TJ34CzUqpMbs8rE1ghhBBCCJFfHgMuWHwfaX4sV2xzewAhsqFvXT1T2BlyzK5UZQZ5dSnsGDk2J2Kl6c8KrxdykpwbdH6FVbUxmNp5iBVl/sTcL/p5dS7kJDk3P2I1Pb06FnaMh/J1xBre83q1sGPk2NyIVQCM9OpayElybnrEd0ys+Fphx8ixiee+AVAFec5bV8/ogjqXvXuVdzC99X/HIq31ooI6//3IBFYIIYQQQmTLPFnNzYT1IlDe4vty5sdyRSawQgghhBDWxJhe2Akexs/AAKXUSqARkKi1vpzbg8oEVgghhBBCPBKl1HeAH1BKKRUJTADsALTWC4FgIAA4BVwH3siL88oEVgghhBDCmmhjYSfIoLV+4AJrrbUG+uf1eeUuBEIIIYQQwqpIBVYIIYQQwpoY/zsV2MIiFVghhBBCCGFVZAIrhBBCCCGsiiwhEEIIIYSwIvo/9CGuwiIVWCGEEEIIYVWkAiuEEEIIYU3kQ1xSgRVCCCGEENZFKrBCCCGEENZE1sBKBVYIIYQQQlgXqcAKIYQQQlgTY3phJyh0UoEVQgghhBBWRSqwQgghhBDWRNbASgVWCCGEEEJYF6nACiGEEEJYE7kPrFRghRBCCCGEdZEKrBBCCCGEFdGyBlYmsOK/b+yUILb/tRtXF2d+WrGwQM9do2k9OozviY3Bhr9XbWXLgp8zPW+wt+X1oP6Ur12JawkpfD1gDnGRMQA079eOxp2bYUw3svaDpfyz/RAA4/+cx82UVIxGI8bb6Xzc9n0AWg3uxFNdniclLgmAX2es5FjogTy7lopN69J0YneUwYajK0PZO399pufLNnycphO6U+qJ8mwY8CmngvcAUKpmBZ7/6A3sSxRFpxvZ/ek6wtfvyrNc+dHGAMpGEbh+ColR8SzqPQOAbrP6UrXRE6QmXwfg28AFXDx2Ltf5Xx7fE2WwYdeqrfyeTf7XgvpTrnYlrpvzx0fG4OhcnF4LhlChbhV2/7CNtROWZOzTf+V4Sro7c+tmGgALu08hJTYpVzmzemXCG9Rq5sOt1JssC5zPhaNn79mmfO1K9JjVHzsHe46GhPH9B6aMPgGNeXHwK3hWfYwZ7cZw/vAZU1s8U4f2I1/DYGdL+q3brJ2ynJM7j+ZJ3tcmvEm9ZvVJS03ji8B5nMsmr1ftyrw1awD2DvYcDNnPNx8szvR8q7fa0HVsL/r79CIlPtmUuXEtuo1/A1tbW5Ljk5j66viHztZxQi9qNvMhLfUm3wQuIPI+bfnarH7YOdhzLCSMNR8sBcDRqRi9Ph2Mazl34iJjWNJ/NqlJ1x543LajulGzWX0ANs5bQ9gvO01tNKsvVRvVzOjf3wTOf6jrqN60Hm3H90AZbNizKoTQbPryq0H9eMzcl78dMIf4yKuUq1eFjlPfMm2kFFtm/8DRjXsBcCjpSKdpffB4vBxo+H7E55zfH/5QuXKiatO6tJrQHRuDDftXhvLngszjm8HelpeD+lK2jhfX41P4YcA8EiKvAuBRozwvTe1NkeJF0UbNF23HcfvmrTzPKPKWTGCthFLKA/gEaAzEA2nADK31j7k8rh8QqLV+Kdch80n7gBZ069iWMZNnFeh5lY3ilUlvMv/1j0iIimXYz1M4vHkf0acuZmzzVOdmpCam8KHfYHzaPEWbUd34esAcPKo+Rv02TZjqH4hTaRf6fzOWD5sNRhs1AJ92ncw18z+glkK/Cibki1/y5Vr8PuzJj69NI+VyHF3WT+LM5n3EhV/K2Cb5Uiybh31O/XcCMu17OzWNTUMWkhARTTEPZ7r++iHnth0mLel6nuTKrzZu+kZrok9dwqF40UznXDflGw5uyJsJuLJRdJz0JgvN+Yf8PIUjWfI3NuefYpF/2YA53L55iw0fr6bM4+XxrF7+nmOvGPwpF8wTw7xWy8+H0pU8mej3Hl4+1ejy0VvMbP/+Pdt1/fBtvhn9ORFh4fRfOpqaft4cCz3A5RMXWPTuLLpN6ZNp+5T4ZBb0nk7ilXjKVC/PwGXvM6bxu7nOW9evPp6VyjDCbwBVfKrR86M+TGo/+p7ten7YhyWjF3A6LJxhS9+nrp8Ph0LDAHAt40bt57y5av7lB8CxpCM9Jr/NrJ4fEnfpKiXcSj50tpp+3rhX8mSy3yC8fKrR+aPeBLUfe892nT98i5WjFxERFs67S0fxhJ83x0MP0Lxve07uOMKWBeto3rcdLfq14+dp3973uDWb+VCuViVmBIzA1t6OgSsncDz0ADdSUgFYN2UFBx6hfysbRftJb/Dl61NIjIplwM8fcWzzPq5Y9OUGnZuRmniNmX5DqNfmKVqP6sa3A+YSfeIC89q8jzHdSAl3ZwZvmMbxLfsxphtpO6EnJ7YdZEW/2RjsDNgVLfLQ2XKSPWByL5a/NpWkqDje/nkyJ7bsJyb8bvb6r/pxI/Eac5sOo3abxjQf1ZUfBszDxmBDh9n9WDtkAdHHz1PUuTjpt27necY8J2tgZQ2sNVBKKeAnYLvWurLW+kmgC1CuELIU+C89vt51cCpZoqBPS0XvqsSciyL2whXSb6Wzf/0O6vj7Ztqmtr8vu9dsB+Bg8C6qN6kFQB1/X/av30F62m3iImOIORdFRe+qBX4Nd3h4VyExIpqk8zEYb6Vzcv3fVPZ/MtM2yZFXufrPhYwJ4B0JZ6NIiIgG4Fp0AtevJuLomjc/j/xqYydPV2o9X5+dK7fmSc77qeBdlasW+cPW76D2v+SvZs6flnqTs3tPcKsQKj11/X3ZtdaUKSIsHMcSxSjp7pxpm5LuzjiUKEpEmKlatmvtdur5NwAg6vRFrpy5fM9xI49GkHglHoDLJy9g52CPrX3uh4z6/g34a+02AE6b8zplyevk7oxDCUdOm/P+tXYb9f0bZjzfbdwbrJq6DM3d/t247bPs+20XcZdMlbjkR6hy1/FvwG6Ltiyag7bcvXY7dc1tWaeFL7t/MF3b7h+2UadFgwce17NaOU7vPo4x3Uha6k0u/XOOJ5rWe+jcWZX3rkrsuSjizH354Pqd1MzSl2v5P8k+c18+HLyLqk1qA3DrRhrGdNOEyraIHdrcxA4lilKpYQ32rAoBIP1WOjfy4BffrB7zrkJcRDTxF2JIv5XOkfV/83iLzOPb4y2e5IA5+7Hg3VR+2vQ6rPJcHaL/OU/08fMApCak3DMGiv8mmcBah+eBNK11xvvnWutzWut5SimDUmqmUmqPUuqQUuodMFVWlVKhSqkflFL/KKW+MU+EUUq1Mj+2H+hw55hKqWJKqcVKqd1KqTClVDvz472UUj8rpbYCvxfolRciJw9XEi7FZnyfcDkOJw/XTNs4e7gSb97GmG7kRnIqxVxK3LNvouW+WtN3+RgC10/hqa4vZDresz1bMnLDdLrOeIeiJYvl2bUU93Qh+VJcxvcpl+Mo7uHy0MfxqFcZg50tCeeu5Emu/GrjDuN7sm7qN2h97z9ELwa+ysgN03l5XA8MuZxcOT/o55zNNVrm/zddZr5LYPA0Wgzs8K/bPixTm17N+D4+KhZnzyzt7ulKwuW71xZ/ORbnLNf2ID6tG3HhyBlup+W+muXi4UqsRd64qFhcPN0yb+PpRrxF3rjLsbiY8/q0aEB8dBwXjmdeLuJZuSyOTsUYtfIDPlg/g6c7NH3obE4eLpn7cFQsTlna0snTlYTLd19/pn5uev2VcHciKSYBgKSYBEq4Oz3wuJeOn+OJpt7YOdhTzKUE1Z6qhXOZUhnbvRjYhZEbZvDyuB4P9ctD1vMlXo7NyHhHSQ9XEjP15es4mvtyee8qDN00kyEbZ/Dj2C8xphtxKV+aa7FJvDLrXd77dSodp72dLxXYkp6uJFn87JMux1HSM0t2TxeSzGPg3ezFcatUBq3h9WUjeefXD3n6nf/sm5EiC1lCYB1qAfvv81xvIFFr3UApVQT4Sym1yfycj3nfS8BfwNNKqb3AF5gmxaeAVRbHeh/YqrV+UynlDOxWSm0xP1cfqKu1jkPkypxOE0iMjqe4W0n6rXifK6cvcnr3P/y1YjMb564BDQHDOtN+7Ot8N+Lzwo6bwbG0My1n92XT0IWQzcTwv6LW8/VJiU0k8shZqjaumem5X6Z/R1JMAgZ7W7pMfZvm77Zl49y1hZT0/lYMmkdidDxFijnwxoKhxHd4lr1r/yjsWDlWplo52o96jXndPyrsKNg72NOmfwdmdp98z3MGgwGvOlWY3m0i9g72jFs7lVNhJ4k+e291ucD8y2vrnz8OUaFuFYasnUxKbBIR+8PR5reT15v7t629La9O7UPzd9sVRGIALhw4TZD/cEpXKUvnj/tyIvQgNgYDZWtXYt3EpVw4cJo2E3rQrG9bNgV9X2C5/o2NrQ0VGlTnizbjuJWaRo/vxnDpyFnO/pU3a7fzjXyISyqw1kgp9ZlS6qBSag/gD/RQSh0AdgFuQDXzpru11pHa9HHFA4AXUAM4q7UO16by1AqLQ/sDo8zHCgUcgArm5zY/aPKqlOqjlNqrlNq7aNGiPLvWwpQYHYdz2btVHucyriRGZ26ChOg4XMzb2BhscChRlGvxyffs62Sxb2K06S3WlNgkDm3cQ4V6pre9k68moo0arTU7V26lYr28W3KQEhVPibJ3q0LFy7iSYs6RE/bFi9JuSSA7Zq4mKux0nuXKjzau5Fud2s2fZPyf8+g57z2qNalF90/6A2RUutLTbrPr+225buOEB/ycs7tGy/wPcqeP3Lx2g30//5XRR3Ljue4tGR08g9HBM0i8koBL2btVOxdPNxKisrR7VBzOZe5em0sZNxKi//33V2dPV/p8HsjXQz/j6vnoR877QvdWTAqexaTgWSRcicfNIq+rpxvxUbGZto+PisXFIq9rGTfio+MoXdET93IeTN7wMbP+XICrpxuTfpmJk7szcVGxHNl+gLTUm6TEJ3Ni9zEqPOH1r9me7e7PiODpjAieTtKVhMx92NONxCxtmRgVh3OZu68/Uz83/YyTYxIzlhyUdHcm+appGUNidPx9j7vpsx+ZETCS+d0/AgVXzpjWst/p37fTbrPr+1Aq1Kvyr9eSkTHL+ZzKuGVkvCMpOg6nTH3ZketZ+vKV05e4ef0mHtXLkxgVS2JUHBcOmMaMw8G7KFu7Uo4z5VRSVBwlLX72Jcu4khSVJXtUPCXNY+Dd7CkkXY7j3K5/uB6fwq0baYSHHKBMba88zyjynkxgrcNRTBVQALTW/YEXAHdAAQO11t7mr0pa6zsV2JsWx0jn3yvuCuhocawKWuvj5ueuPWhHrfUirbWv1tq3T58+D9rUapw/eBp3L09cy7ljsDNQv00Tjmzel2mbI5v30bDjcwDUC2hE+I6jGY/Xb9MEg70truXccffy5NyBU9gXLUKRYg4A2BctQo1n63L55AWATOvm6rZskPF4Xog+eAbnSp6ULO+OjZ2B6m0ac2bz/Yr6mdnYGXjpi8EcX/tHxp0J8kp+tPEvM1Yy4an+THpmIF8PnEv4jqMsH/IZkLmN6/j75rqNL2TJ79OmCUf/Jf+pHQ+u7NgYbDKWGNjYGqj1fH2i8qAvbF++kakBI5gaMIJDm3bTqIMpk5dPNVKTr2dMfu5IikngRnIqXj6m34cbdXiOQ5v2PvAcRUs60m/JKNZN/5Yz+07kKu/vy39jfEAg4wMC2b9pd8bb+1XMeROz5E2MSeBG8nWqmPM+3aEp+zftIfLEeQb6vkngM30JfKYvcVGxjH9pOIkxCezftJtqvjWwMdhg72BPFe9qXDoV+a/Z/li+iRkBI5kRMJJDm/bQ0KItb+SgLRt2eI7Dm0yvpSNb9tKwk+naGnZqyuHNpjY+vHlvtsdVNgpH5+IAlK1RgbI1KvLPH6a7b2QaQ/wfbgyJPHgaNy9PXMx9uV6bpziepS8f27yPJ819uU5AI06b+7JLOXdsDKbphPNjpShdpSzxkTGkxCSSeCmWUpXLAFD16dpcCf/39n1Ylw6ewa2SJ87lTdlrt2nMiSzZT2zZj7c5e82Ahpw1Zz+17RAeNcpj52CPjcEGr0ZPZPrw13+WMb3gvv6jZAmBddgKTFFK9dVaLzA/5mj+cyPQVym1VWt9SylVHXjQq+8fwEspVUVrfRroavHcRmCgUmqg1lorpXy01mF5fTEPa/iEaewJO0RCQhIvtH+dfr2707FNy3w/rzHdyJrxS+i7bIzpFk+rQ4gKj6T1kFe4cPgMR7bs4+/VIbwe1J+xobNNt0gaOBeAqPBIwn7ZyZjNH5N+O50fxi9BGzUlSjnRe9EwwDRR2bfuL/7ZdhCAtqNf47GaFUFrYiNjWD3myzy7Fp1uJHTc17RfPgJlsOHYqm3EnbxI46EdiT58lrOb9+NRtzIvfjEYBydHKjX3ofHQjqxoPopqLzWmbMPHcXAuTs1Opn8ANg37nKvHzuc6V3608YN0nzOA4q4lUUpx8VgEq97PXRvfyf+OOf8uc/5W5vxHt+xj1+oQXgvqzxhz/uXm/ADj/pxHkeJFsbWzpY6/Lwu7TyH+4lXeWTYag60BG4MNJ/86ws7v8nbp+ZGQMGo1q88H2+aSlprG8uF3b7c0OngGUwNGALBy3Jf0MN/66WjoAY6aP9Ffr2UDOk98k+KuJem3eBSRxyP4tMcUmvZohXtFT1oP6kTrQZ0AmNf9w1zfAuxgyH7qNqvPzG2fcTP1Jl8O/yzjuUnBsxgfEAjA1+O+4G3zbbQOhYZxKPTBv6RdPn2Rw9sO8OFvQWijZtuqLVx8yF8WjoWEUauZD+O3zSEtNY1vhi/IeG5E8HRmBIwEYPW4r3htVj/sHew4Fnog4xZ5mxes443PBtO4czPiL15lSf9PHnhcg50tg7//AIAbKaksHzIv4wNUPeYMpLhrScjo31/Qom/7HF2HMd3IuvFL6b1sNDYGG/asDiU6PJIWQzoRefgsx7fsY8/qUF4N6sfw0E9ITUjh24HzAPBq8DjN+rYj/fZttFHz47jFGZXZdROX0nX2AAx2tsRdiOb7wLxfFmVMNxI8findl41EGWwIW72NmPCLNBvakUuHznJiy37CVoXy8id9eW/bx6QmXOOHAabsN5Kus/PLDby9fjJoTXjIQcK35t3tC0X+Udl9yEH89yilymC6jVYjIAZTRXQh8D3wIdAGUwU1BmiPaf1rxu2xlFKfAnu11kuVUq2A2cB14A+gitb6JaVUUfPjTTBV58+aH+8F+GqtB+Qwrr51NX9u/5Mf7EpVZpBXl8KOkWNzIlaa/qzweiEnyblB51dYVRuDqZ2HWFHmT8z9op9X50JOknPzI1bT06tjYcd4KF9HrOE9r1cLO0aOzY0wfcxhpFfXf9nyv2N6xHdMrPhaYcfIsYnnvgHTv78F5ubxkAKbvBV5olmBXltOSQXWSmitL2O6dVZ2xpi/LIWav+7sP8Di779hWgub9RypwDvZPL4UWPpwiYUQQggh8odMYIUQQgghrIn8RwbyIS4hhBBCCGFdpAIrhBBCCGFN5D6wUoEVQgghhBDWRSqwQgghhBDWRNbASgVWCCGEEEJYF6nACiGEEEJYEa3/u/9DVkGRCqwQQgghhLAqUoEVQgghhLAmchcCqcAKIYQQQgjrIhNYIYQQQghhVWQJgRBCCCGENZHbaEkFVgghhBBCWBepwAohhBBCWBP5EJdUYIUQQgghhHWRCqwQQgghhDUxyn9kIBVYIYQQQghhVaQCK4QQQghhTWQNrFRghRBCCCGEdZEKrBBCCCGENZH7wEoFVgghhBBCWBeltS7sDOL/H+lUQggh/peogjzZjZ3fFdi/sw5PdS3Qa8spqcAKIYQQQgirImtgRb4Y5NWlsCPk2JyIldy6eqawY+SYXanKAAy1ojYOilhJP6/OhR3jocyPWM17Xq8WdowcmxuxCoDRXt0KOUnOTY34lm4VXy7sGA/l23M/EujVtbBj5NisiO8AGGJF48UnESutrh8XOFkDKxVYIYQQQghhXaQCK4QQQghhTaQCKxVYIYQQQghhXWQCK4QQQgghrIosIRBCCCGEsCJapxd2hEInFVghhBBCCGFVpAIrhBBCCGFN5ENcUoEVQgghhBDWRSqwQgghhBDWREsFViqwQgghhBDCqkgFVgghhBDCmsgaWKnACiGEEEII6yIVWCGEEEIIayJrYKUCK4QQQgghrItUYIUQQgghrImsgZUKrBBCCCGEsC5SgRVCCCGEsCayBlYqsEIIIYQQwrpIBVYIIYQQwprIGlipwAohhBBCiEenlGqllDqhlDqllBqVzfMVlFIhSqkwpdQhpVRAbs8pE1ghhBBCCPFIlFIG4DOgNVAT6KqUqplls7HAaq21D9AFmJ/b88oSAlFgajStR4fxPbEx2PD3qq1sWfBzpucN9ra8HtSf8rUrcS0hha8HzCEuMgaA5v3a0bhzM4zpRtZ+sJR/th8CYPyf87iZkorRaMR4O52P274PQKvBnXiqy/OkxCUB8OuMlRwLPVBg1zp2ShDb/9qNq4szP61YWGDnzapG03q0t2jzrdm0eTeLNl82YA7xkTE4Ohen14IhlK9bhT0/bGPthCV397Ez0OGDN6nSuCZaG9kwcxWHftudb9fwyoQ3qNXMh1upN1kWOJ8LR8/es0352pXoMas/dg72HA0J4/sPTHl9Ahrz4uBX8Kz6GDPajeH84TN5kumJpvXoML4XNgYbdq7aypYF6zI9b5vRlytzLSGZpRZ9uUW/9hl9ec0HS/ln+0Gcy7jRPag/JUo5obVmx3e/s23JhkzHbPbWS7w8tjujfd7iWnxyrvJXb1qXl8b3wMZgw55VIWxbsD7T8wZ7WzoH9eWx2pW4npDCtwPmkhB5NeN5p7JuDNk8k99nr+GPL34FoOOMPtR43oeU2CTmtByZq3zZ6TGxN97NniQt9SYLA+cRceTen2Wl2pV55+P3sHew50DIPpZN/CrjOf9eAfh3b43RaCRs6z6+m7qMKvWq0XtqXwCUgjWzV7F34648yft403q0M7fxrlUhhGTz2usa1I9y5jZePmAO8ZFXqfZMHV4c2QWDnS3pt27zy5RvObXzaKZ93/giELcKpZnVckSeZL2jRtN6vDy+J8pgw65VW/k9m8yvBfXPyPx1lvGiQt0q7M4yXgQEvopvh+dwdCrGqFq98jTvo/bjcvWq8PLU3gAopdgyew3HNu4FoMkbrWjQpRlKKfas3Mpfi3/L08x55r+1hKAhcEprfQZAKbUSaAccs9hGAyXNf3cCLuX2pFKB/R+klGqvlNJKqRoFdk4bxSuT3uTzXtOY2mIY9ds+jUfVxzJt81TnZqQmpvCh32BCv/qVNqO6AeBR9THqt2nCVP9AFvacyiuTe6NsVMZ+n3adzMyAURmT1ztCvwpmZsAoZgaMKtDJK0D7gBYsDPqwQM+ZlbJRdJj0Jot6TWP6fdq8kbnNp/gNZttXv/KSuc1v37zFho9X8/OUFfcct/mAl0mJTWTa80OY0TyQ07uO59s11PLzoXQlTyb6vcc3YxbR5aO3st2u64dv883oz5no9x6lK3lS088bgMsnLrDo3Vmc2p13Ge/05YW9pjKlxVCebPs0nlnatXHn57meeI3JfoMI/SqYtuZ29czoy8NY0HMKnSe/ibJRGG+n8+OHy5nSYhhBL4/l2e7+mY7pXMaNGs/VzZgE5zZ/20lvsKTXDD5pMZx6bZtQOkv+Bp39SE28xiy/ofz51QZaj+qa6fkXx77OydCDmR7b98N2lvScnut82fFuVh/PSmUZ2rQfX45ewJsfvpPtdm9+9C5fjprP0Kb98KxUlnp+9QGo+VRtfFs0ZFTrIYxoMYhfF5l+4bhw4hxj2wQyJmAo03tOpveUvtgYcv/PorJRvDzpDb7sNZ2ZLQLxadvkPq+9a0zzG8L2r4J50dxHrsUns7j3LD5uNZKVwxbQ9ZN+mfar3bIBN6/fyHXG7DJ3tBgvfLIZLxpnGS/a5GC8OPr7Pma3e/+ex/Mi76P24+gTF/iszVjmBYxhSY/pvPxRb2wMNnhUL0eDLs2Y324cc1uPosbz9XGr6JHn2a2NUqqPUmqvxVefLJs8Blyw+D7S/JilicDrSqlIIBgYmNtcMoH939QV+NP8Z4Go6F2VmHNRxF64QvqtdPav30Edf99M29T292X3mu0AHAzeRfUmtQCo4+/L/vU7SE+7TVxkDDHnoqjoXbWgoj8SX+86OJUsUagZKnhX5eq5KOLMbR62fge1s2nzPeY2PxS8i2rmNk9LvcnZvSe4ffPWPcdt+Eozfp9vmgBorXNdDXyQuv6+7FpryhcRFo5jiWKUdHfOtE1Jd2ccShQlIiwcgF1rt1PPvwEAUacvcuXM5TzNZOrL0Vn6coNM29Tx92X3mm0AHAj+m+pNapsfb8D+9Tu4ndGXo6noXZWkmAQizZXlm9duEH36Ik6erhnH6zCuB+umfoNG5zp/ee+qxJ6LJt6c/+D6nTzh/2SmbZ7w92X/mj8AOBK8iyrm/AA1/X2JvxBDdHhkpn0idv/D9cSUXOfLzpMtGvLHmhAAToWdxLFkMZxLu2Taxrm0C0WLF+VU2EkA/lgTgq9/QwCav96Kn+ev5XbabQCSYhMBSLuRhjHdVMmyK2IHOvftC6bXXqzFa+/A+k8jRE4AACAASURBVJ3UyvLaq+X/JHszvfZMbXzpaARJV+IBiDoZiZ2DPQZ705ul9o5FaPpWAL/P+zFPcmbNfNVijL7feGE5RmcdL25lM16cCztFUkxCnufNTT++ZfFzty1il/Fjd6/6GBcOnMp4/uyu49Rqlfm1/Z+hjQX2pbVepLX2tfha9AiJuwJLtdblgABguVIqV3NQmcD+j1FKFQeeAXpjWoeCUspGKTVfKfWPUmqzUipYKdXJ/NyTSqltSql9SqmNSqkyj3JeJw9XEi7FZnyfcDkOJw/XTNs4e7gSb97GmG7kRnIqxVxK3LNvouW+WtN3+RgC10/hqa4vZDresz1bMnLDdLrOeIeiJYs9SmyrlpM2t9zGss3vx6GkIwCthnVm6C9T6fHZYIqXcsqH9CamPnH3rev4qFicPbP0G09XEi7fvc74y7E4Z7nOvM6UuV1jcfLIPJm6t12vm/uyS6brScgmq2s5dx6rWYlzB04BUKeFLwnRcVw6fi5P8pf0cCHRIn9SNv2ipIfLPfkdXUqYJlDvtuH3OWvyJEtOuXi6EWeROS4qFpcsmV08XImLstjmciwunm4AeFYqy+MNazLpp+mMW/Uhleve/QW4inc1Zmyew/SNs/nq/YUZE5vccLJoP8hZH0k1t7Gluq0bEnnkLOnmiXerYZ3Z9uWvpN24meuMWWXt14l5MF7kp9z0Y4Dy3lUYvGkGgzZO56exX2FMNxJ94gKVGtTA0bk4dg72PN7MG6cybgV3UdbrIlDe4vty5scs9QZWA2itdwIOQKncnFQmsP972gG/aa1PArFKqSeBDoAXpsXX3YGnAJRSdsA8oJPW+klgMfBRdge1fIth0aJH+eXs0czpNIFZL41mYa9pPNvDnyoNTasi/lqxmcnPvceMgFEkXUmg/djXCyzT/2cGgwGXsm5E7DtJ0EujObf/JG3HSNvmFXvHIvReMJS1k77mRkoqdg72tOjfnuCg1YUdDYAXBnfkz6+CSbue9xOo/GSwNVDcuTjj24/k2ylf8978wIznTh8IZ0SLQYxtO4J2/TqaKrH/AR7VyhEwqhtrxnwJQNmaFXGr4MER81pNkTsXDpxmtv8IPms7Fr++7bAtYkfM6UtsW7ieN5eP5o2vR3Lp2Dn0f2ut6V1GY8F9/bs9QDWlVCWllD2m4tjPWbY5D7wAoJR6AtMENldrouRDXP97ugJzzH9faf7eFvhea20EopRSIebnHwdqA5uVUgAGINv3Y81vKdyZuepBU7Zmej4xOg7nsnd/k3Uu40pidFymbRKi43Ap60ZiVBw2BhscShTlWnzyPfs6WeybGG16qy0lNolDG/dQoV5VTu/+h+SriRnb71y5lT5f5e2HHaxBTtr8zjZZ2/x+rsUnc/P6DQ6bP7R1MHgXjV5tlqe5n+vekqfN1fRzB0/jUrYUcAIwVeISorL0m6g4nC2qJC5l3EjIcp15KeGednXL6Id33GnXhIx2dTT35Xjz9dzd905WG1sDvRcOY+9Pf3Joo6l9S1X0wK1caUZumGHa3tON4b9M4+P2Y0iOSeRRJEXH42SRv2Q2/SIpOh7nsm4kWeS/Hp9Mee+q1AloROvR3XAo6Yg2am7fvMXOZZseKcuDtOjRmmZdWgBw5tApXC0yu3q6EZ8lc3x0HK6eFtuUcSPeXJGNu3yVPb/9DcDpg+Foo6aEa0mSzR/yBLh0KpIb129QrnoFzh4+navsieb2u+NBfeTOa6+ouY0BnDxd6fX5UFYOnU/s+SsAVKxfjXJ1KzPmz7nYGGwo7uZE35XjWNBlcq6y3pG1XzvlwXiRn3LTjy3FnL5E2vUbeFQvx8XDZ9m7OpS9q0MB8B/+KkkW7+6I7GmtbyulBgAbMc0TFmutjyqlJgF7tdY/A8OAL5RSQzB9oKuX1rlbsyMV2P8hSilX4HngS6VUBDAc6Ayo++0CHNVae5u/6mit/R/l3OcPnsbdyxPXcu4Y7AzUb9OEI5v3ZdrmyOZ9NOz4HAD1AhoRvuNoxuP12zTBYG+Lazl33L08OXfgFPZFi1CkmAMA9kWLUOPZulw+aVpHbrlOsm7LBhmP/y+5kKXNfbJp86Ob99HA3OZ1AxpxasfR7A6VybHf91OlsekOKdWerk1UeNZ3inJn+/KNTA0YwdSAERzatJtGHUz5vHyqkZp8/Z71dEkxCdxITsXLpxoAjTo8x6FN+Velyq4vH96c+XxHNu+lYcemAHgHNM7oy4c376V+mybYZunLAN2mv0v0qYuEfPVrxnEun7jA+759+OCZgXzwzEASomKZ+dKoR568AkQePE0pL09czPnrtXmK41n6xfHN+6jf8VkAagc04rQ5/6LOk5jxzCBmPDOIvxb/Ruhn6/Jl8gqwedkGxgQMZUzAUPZu2sWzHU2/KFX1qU5q8nUSrmSeECZciSc1JZWqPtUBeLZjM/ZtNv0isHfTbmo+VQcwLSewtbMlOS4J9/KlMz60Veoxd8pWeYyrkVdynf2CuY3v9BHvNk9xNJvXnm82rz2Hko70XjKCX6d/R8S+kxnb71yxhcmN+jHlmff47JWJXD17Oc8mr3cyZx0vsmbOOkbnZLzIL7npxy7l3DN+7s6PlcK9SlnizXfZKOZm+qC8U1k3arVqwIGfdxTUJT2cAlwDm6M4WgdrratrratorT8yPzbePHlFa31Ma/201rqeeT6R64FDKrD/WzoBy7XWGR/hVUptA+KAjkqprwF3wA/4FlPZy10p9ZTWeqd5SUF1rfVDj1rGdCNrxi+h77Ixpls6rQ4hKjyS1kNe4cLhMxzZso+/V4fwelB/xobONt2iZeBcAKLCIwn7ZSdjNn9M+u10fhi/xFRBKeVE70XDALAx2LBv3V/8s830yei2o1/jsZoVQWtiI2NYbX4brqAMnzCNPWGHSEhI4oX2r9Ovd3c6tmlZoBmM6UbWjl9CH3Ob714dQnR4JK3MbX50yz52rQ6hW1B/xpjbfJm5zQHG/jkPh+JFMdjZUtvfl8+7TyH61EV+mfYt3YL60358D1Liklk5fEG+XcORkDBqNavPB9vmkpaaxvLhd28dODp4BlMDTJX1leO+pMesfqbbaIUe4GhoGAD1Wjag88Q3Ke5akn6LRxF5PIJPe0zJVSZjupEfxi+mX0ZfDiUqPJKAIa9w3tyXd64OoXvQAMaFzuF6QgpLB5re9Mjcl418P34x2qip7Ps4DTs+x8Xj5xgRbPok/y8zvsuXu2cY0438PH4pby4bhTLYsHd1KFfCL9J8SCcuHj7D8S372bs6lM5B/QgMDeJ6wjW+GzjvX4/bZe4AKjV+gmIuJRi1cx5bPlmTUcnKrQNb9+Hd7Ek+2b6Am6k3+Tzwbp4pwUGMCRgKwOKxn/Ou+TZaB0P3cyBkPwChq3/nnZkDmL5pDrdv3WLBMFM/f9z3Cdr268DtW+lobWTJ2M9JzoOKojHdyI/jl/L2stEogw17VocSHR5JyyGduHD4LMe27GP36lC6BvVjVOgnXE9IYYW5jZ/u0ZJSFT1oMagDLQZ1AOCL7lNJiU160CnzJPOa8Ut4x9yvd5nH6KzjxWsW48Vyi/Fi3J/zKFK8KLZ2ttTx92WhebxoM6ob9ds9jV1Reybs/Iy/V4WwcfYPeZL3UfuxV4PHadq3Lem3b6ONmnXjlmRUZl9bMBhHl+IYb6fz87gl3Ei6nuusIn+oXFZwhRUxLw2YrrX+zeKx94AnMFVb/TDdCkOZt9uslPIG5mK6b5stMFtr/cW/nEoP8uqSD1eQP+ZErOTW1by5P2hBsCtVGYChVtTGQREr6efVubBjPJT5Eat5z+vVwo6RY3MjVgEw2qtbISfJuakR39Kt4suFHeOhfHvuRwK9CuwGLrk2K+I7AIZY0XjxScRKq+vH3P+dzHyR+uO0Apu8FX15VIFeW05JBfZ/iNb6nsWKWuu5YLo7gdY6RSnlBuwGDpufPwA8V6BBhRBCCCEeQCaw4o5flFLOgD0wWWsdVdiBhBBCCJGNHK5N/f9MJrACAK21X2FnEEIIIYTICZnACiGEEEJYk//q/WkLkNxGSwghhBBCWBWpwAohhBBCWBOpwEoFVgghhBBCWBeZwAohhBBCCKsiSwiEEEIIIayJ/CdUUoEVQgghhBDWRSqwQgghhBDWRD7EJRVYIYQQQghhXaQCK4QQQghhTaQCKxVYIYQQQghhXaQCK4QQQghhTbRUYKUCK4QQQgghrIpUYIUQQgghrImsgZUKrBBCCCGEsC5SgRVCCCGEsCbyP3FJBVYIIYQQQlgXqcAKIYQQQlgTWQMrFVghhBBCCGFdlJZ1FCLvSacSQgjxv0QV5MlSvwossH9ni/aeVaDXllOyhEDkizkVXi/sCDk26PwKhnp1KewYORYUsRKAW1fPFHKSnLMrVZk3vToVdoyHsjjiB0Z4dS3sGDk2I+I7AAKtKPOsiO/o59W5sGM8lPkRqxlkRePFHPN4McSKMn8SsdLq+rEoeDKBFUIIIYSwJvI/cckaWCGEEEIIYV1kAiuEEEIIIayKLCEQQgghhLAi2iiflZYKrBBCCCGEsCpSgRVCCCGEsCbyHxlIBVYIIYQQQlgXqcAKIYQQQlgTuY2WVGCFEEIIIYR1kQqsEEIIIYQ1kbsQSAVWCCGEEEJYF6nACiGEEEJYE7kLgVRghRBCCCGEdZEKrBBCCCGENZEKrFRghRBCCCGEdZEKrBBCCCGENdFyFwKpwAohhBBCCKsiFVghhBBCCGsia2ClAiuEEEIIIayLTGCFEEIIIYRVkSUEQgghhBDWRP4rWZnAiv+Gik3r0nRid5TBhqMrQ9k7f32m58s2fJymE7pT6onybBjwKaeC9wBQqmYFnv/oDexLFEWnG9n96TrC1+/Kl4w1mtaj/fie2Bhs+HvVVrYu+DnT8wZ7W7oF9ad87UpcS0hh2YA5xEfG4OhcnF4LhlC+bhX2/LCNtROW3N3HzkCHD96kSuOaaG1kw8xVHPptd77kf5CxU4LY/tduXF2c+WnFwgI/f1bdJrxJnWY+pKWm8VXgp5w/evaebSrWrkzvWf2xc7DncEgY336wGICXh3bBu0UDtDaSdDWJxYGfknAlPmM/r7pVeH/tFBYO/IR9G/7O09zVm9aj3fgeKIMNu1eFEJpNH+kS1I/HalfiekIK3wyYQ3zkVcrXq0LHqW+ZNlKKzbN/4OjGvXmazdLj5pw2Bht2rQohJJucXYP6Uc6cc7k5Z7Vn6vDiyC4Y7GxJv3WbX6Z8y6mdRzPt+8YXgbhVKM2sliNynfOVCW9Qq5kPt1JvsixwPhey6Qfla1eih7kfHA0J4/sPTK8vR6di9P50CG7l3ImNjOHL/p+QmnSN5n3a0KD9s6brNNjgWbUcI+r35nriNYqWdOS1ae9S9vHyoDXLRyzg7P7wHOet0bQeHSzGiC3ZtOvrFmPE1wPmEBcZA0Dzfu1o3LkZxnQjaz9Yyj/bDwFQtKQjXaa9Q5nHy6E1fDdiIRH7w+n56SBKVy5j3qYYqUnXmBkw6uEbOUv+l8f3RBls2LVqK79nk/+1oP4Z/eLrLGNchbpV2G0xxtk52NNr/mDcKnqg040c/X0/v0z/LlcZLeVHP24V2BnfDs9R1KkY79d6I8+yirwnSwjykVKqnFJqnVIqXCl1Wik1RyllX4h52iulalp8P0kp1byw8mTksFH4fdiTn3rOYPkLI6jetjGu1cpm2ib5Uiybh33OiXU7Mj1+OzWNTUMWsqL5KH7qMYOmE7pjX9IxXzJ2mPQmi3pNY3qLYdRv+zQeVR/LtE2jzs1ITUxhit9gtn31Ky+N6mbKePMWGz5ezc9TVtxz3OYDXiYlNpFpzw9hRvNATu86nufZc6J9QAsWBn1YKOfOqo6fDx6VyjDabyBfj1lIj4/6ZLtd9w/fZunohYz2G4hHpTLU8fMBYMOidUxoPYyJAcM5tHUfbQa9krGPsrHhlVGvc/SPg3meW9koXp70Bl/1ms7HLQLxbtuE0ln6SMPOzUhNvMYMvyH88VUwAeY+EnXiAnPbvM/sgNF81WMaHT96CxtD/gzPd3J+2Ws6M1sE4tO2yX368jWm+Q1h+1fBvGjOeS0+mcW9Z/Fxq5GsHLaArp/0y7Rf7ZYNuHn9Rp7krOXnQ+lKnkz0e49vxiyiy0dvZbtd1w/f5pvRnzPR7z1KV/Kkpp83AC37tufEjsNMbDaIEzsO07JfewC2LFrP1IARTA0YwboZ3xG+6xjXE68BpgnzsW0HmPTCED5qPZyoUxdznFfZKF6Z9Caf95rG1PuMEU+Zx4gP/QYT+tWvtDG3q0fVx6jfpglT/QNZ2HMqr0zujbJRAHSY0JPj2w4w5YVhzGg9gmhzpq8HzGFmwChmBozi0IZduf7FV9koOlqMcT7Z5G+cZYxrk4MxLuSLX5j2wjBmvTiKSk8+Tg3zzye38qsfH/t9P3Pajc2TjPlKGwvu6z9KJrD5RCmlgLXAT1rrakB1oDjwUSHGag9kTGC11uO11lsKMQ8AHt5VSIyIJul8DMZb6Zxc/zeV/Z/MtE1y5FWu/nMBneVtk4SzUSRERANwLTqB61cTcXQtkecZK3hX5eq5KOIuXCH9Vjph63dQ29830za1/X3Zs2Y7AIeCd1GtSS0A0lJvcnbvCW7fvHXPcRu+0ozf568DQGvNtfjkPM+eE77edXAqmfft9ih8/BuwY20oAGfCwnEs4YiTu3OmbZzcnSlawpEzYabq2I61ofj4NwDgRkpqxnb2jkUy3S+xea/W7Nuwi6TYxDzPXT5LHzm4fie1svSRmv5PstfcRw4H76Jqk9oA3LqRhjHd9A+FbRG7fL3FYwXvqsRa5DyQTc5aFjlNfdmU89LRCJLM1eyok5HYOdhjsDe9kWfvWISmbwXw+7wf8yRnXX9fdq01ZYgIC8exRDFKZukHJd2dcShRlAhzP9i1djv1zP2gbosG/P3DNgD+/mEb9Vo0uOccvm2fZu/PfwHgUKIoVRs+wY5VWwFIv5VOatL1HOet6F2VmHNRxJrbdf/6HdTJZozYbW7Xg8G7qG4eI+r4+7J//Q7S024TFxlDzLkoKnpXxaFEUao0fIK/V4U8MJP3i0+x/+cd9zz+MO6McbH/MsZZ5s86xt3KMsbdupHGqZ3HMrJHHj2Ls6drrnJa5s2Pfnw+7BTJMQl5klHkL5nA5p/ngRta6yUAWut0YAjwplKqmFJqllLqiFLqkFJqIIBSqoFSaodS6qBSardSqoRSqpdS6tM7B1VK/aKU8jP/PUUp9YlS6qhS6nellLv58beVUnvMx1mjlHJUSjUB2gIzlVIHlFJVlFJLlVKdzPu8oJQKU0odVkotVkoVMT8eoZT6QCm13/xcjbxuqOKeLiRfisv4PuVyHMU9XB76OB71KmOwsyXh3JW8jAeAk4crCZdiM75PuByHk4frfbcxphu5kZxKMZf7TwodzJXiVsM6M/SXqfT4bDDFSznleXZr4+LhRpxFW8dFxeHi6ZZ5G0834i9bbHM5DhePu9t0COzKrB0LadzuWX4KWgWAs4cr9Vs2JGTFxnzJ7eThQqJF7sTLsZTM0o+dPFwztjH1kes4mvtIee8qDN00k6EbZ7B27JcZE9r8yJm5L8filE1Oy76capHzjrqtGxJ55CzpabcBUz/e9uWvpN24mSc5nT1cib90NeP7+KjYeyY/zp6uJFj0g/jLsTibX5cl3J1IMk9EkmISKOGe+bVl52BPzabehJmXkZQqX5qU2CS6z+rH6F+n89q0d7AvWiTHeXMyRpiu6d4xIuu+ieZ93cyZus3qy/Bfp9JlWp97MlVpWIPkqwnERETlOGt2nO+T4X7XmJMxzpJDSUdqvVCf8L+O5Crn3Sz504+thlEX3Nd/lExg808tYJ/lA1rrJOA88BbgBXhrresC35iXFqwCBmmt6wHNgVQerBiwV2tdC9gGTDA/vlZr3cB8nONAb631DuBnYLjW2ltrffrOQZRSDsBS4FWtdR1Ma6P7Wpznqta6PrAACMwuiFKqj1Jqr1Jq76JFi/4ldt5zLO1My9l92Ry4yGr+hxKDwYBLWTci9p0k6KXRnNt/krZjXi/sWP8vrJ31HYFN3uXvdX/wfM9WAHQd/wbfT1uB/o/2jwsHThPkP5x5bd+nWd922BaxK+xI9+VRrRwBo7qxZsyXAJStWRG3Ch4cycd1u7mW5edet/mTnNl7ImP5gI3BQPnalfhjxSamvjiStNSb+PdtXxhJM9gYDJSrXYm/Vmxm5oujSUu9SfO+7TJtU7/t07muvuY3G4MNPea+x/alvxF7Ie8LDI8qaz8W1kUmsIXDD/hca30bQGsdBzwOXNZa7zE/lnTn+QcwYpr0AqwAnjH/vbZS6g+l1GHgNUyT6Qd5HDirtT5p/v5r4DmL59ea/9yHaeJ9D631Iq21r9bat0+f7Ncs3k9KVDwlyt79Tb94GVdSouMfsEdm9sWL0m5JIDtmriYq7PS/7/AIEqPjcC57t8LnXMaVxOi4+25jY7DBoUTRBy4JuBafzM3rNzhsXrt2MHgX5Wp75X14K/B891ZMDJ7JxOCZJFyJx9WirV09XYmPis20fXxULC5lLLYp40p8dOZtAP7+6Q+ebNUYAK+6lXl33hBm/Dkf39aN6T757YxlB3khMToeJ4vcTmXcSMrSjxOj4zK2MfURR65n6SNXTl8i7fpNPKuXz7NsWXNm7stuJGaT07IvF7XI6eTpSq/Ph7Jy6Hxiz5smIxXrV6Nc3cqM+XMu/b+fSKlKZei7ctxDZ3uue0tGB89gdPAMEq8k4FK2VMZzLp5uJERlfs0lRMXhbNEPXMq4kWB+XSbHJGYsOSjp7kzy1aRM+z7Z5mn2/PynxbFiSYiKJeLAKQD2B/9NhdqVcpw9J2NEQnQcLtmMEVn3dTLva8oUxzlzpgNZxggbgw31WjZg/y87c5zzfhLuk+F+15iTMe6OzlPfJubsZbYv3pDrnHez5H0/tibaaCywr/8qmcDmn2NApoWcSqmSQIWHPM5tMv+cHB6w7Z0Sw1JggLma+sG/7JMTd94TTCcf7lwRffAMzpU8KVneHRs7A9XbNObM5v052tfGzsBLXwzm+No/Mu5MkB8uHDyNu5cnruXcMdgZ8GnThCObMxXYObp5Hw06mub9dQMacWrH0ewOlcmx3/dTpbFpWXK1p2sTFZ7zD438f7J1+W9MDBjOxIDhhG3aTZMOfgBU9qnG9eTrJGZZk5YYk0Bq8nUq+1QDoEkHP8I2mX7+pb08M7bzadGAqNOmNh35bH9GPNOPEc/0Y++Gv1k+7ouMffJC5MHTlPLyxMXcR+q1eYpjWfrIsc378DX3kToWfcSlnHvGh7acHytF6SplMz6dntcumHPe6cvebZ7iaDZ92TebvuxQ0pHeS0bw6/TviNh3MmP7nSu2MLlRP6Y88x6fvTKRq2cvs6DL5IfOtn35xowPWB3atJtGHUwZvHyqkZp8PWNJwB1JMQncSE7Fy9wPGnV4jkObTFXgQ1v20rhTUwAad2rKoc13f9YOJYpSrVFNDm3ea3GsROIvxWZ8sr/G03W4HB6Z4+zns4wR9bMZI45s3kdDc7vWC2hEuLldj2zeR/02TTDY2+Jazh13L0/OHThFckwiCRaZqmcZI6o/U4foM5dIzDKxfxTZjXFZ+0XW/DkZ41oP64xDCUd+mrQs1xmz5s3rfiysi9xGK//8DkxTSvXQWi9TShmAjzFNLsOBd5RSIVrr20opV+AEUEYp1UBrvUcpVQLTEoIIoJ9SygZ4DGhocQ4boBOwEugG3CknlAAuK6XsMFVg74x4yebnsjoBeCmlqmqtTwHdMS1JKBA63UjouK9pv3wEymDDsVXbiDt5kcZDOxJ9+CxnN+/Ho25lXvxiMA5OjlRq7kPjoR1Z0XwU1V5qTNmGj+PgXJyanUwD1aZhn3P12Pk8zWhMN7J2/BL6LBuDjcGG3atDiA6PpNWQV7hw+AxHt+xj1+oQugX1Z0zobK4npLBs4NyM/cf+OQ+H4kUx2NlS29+Xz7tPIfrURX6Z9i3dgvrTfnwPUuKSWTl8QZ7mzqnhE6axJ+wQCQlJvND+dfr17k7HNi0LJcuhkP3UbVafads+JS31JouHz894bmLwTCYGDAdgxbgveXNWf+wd7DkcGsbh0DAAOo18Hc/KZdFGTezFGJa9XzBLWozpRtaNX8pby0ZjY7Bhz+pQosMj8R/SicjDZzm2ZR97VofSJagfI0I/4XpCCt8OnAdApQaP49e3Hcbbt9FGzY/jFt9Tmc3LnD+OX8rby0ajLHK2HNKJC+acu1eH0jWoH6PMOVeYcz7doyWlKnrQYlAHWgzqAMAX3aeSEpv0oFM+kiMhYdRqVp8Pts0lLTWN5Rb9YHTwDKYGmG7TtXLcl/SY1c90G63QAxw194NNC36i92dDaNL5eeIumm6jdYd3y4Yc/+MgaamZ1+uunriYN2a/h62dLVcvXGFZ4HxyyphuZM34JfQ1jxF/rw4hKjyS1uYx4siWffy9OoTXg/oz1jxGfG0eI6LCIwn7ZSdjNn9M+u10fhi/JOMDq2smLqH77AEZmb4NvHubu/ptmuTZ8oE7+d8x599lzp91jHvNYoxbbjHGjftzHkWKF8XWzpY6/r4s7D6FGymp+A/sQPSpiwz7dSoAf3y9kV3mD6XlNm9+9OMXR3XDp10T7IraM3bnp+xeFcKm2WtynTfP/YfXphYU9V9dD/b/gVKqPDAfqIFpshmMaQ1pOjADaAXcAr7QWn+qlGoAzAOKYpq8NgeuYVoe8CSm9awuwEStdahSKgVYBPgDVzCtYY1RSvUFRgAxwC6ghNa6l1LqaeALTBXVTsA44Bet9Q9KqReAWZh+qdkD9NVa31RKRQC+WuurSilfYJbW2u9fLl3PqWA9azkHnV/BUK8uhR0jx4IiVgJw6+qZQk6Sc3alKvOmV6fCjvFQFkf8wAivroUdZSLNOgAAIABJREFUI8dmRJjurxloRZlnRXxHP6/OhR3jocyPWM0gKxov5pjHiyFWlPmTiJVW148BVZDnvPbR/7F35/ExXe8Dxz8nQVFkIRJqSWzfliCJpaotSdsEaS21Fa1dF/uuaK3VoJZWlaBViiraaovyRdsErdpjCWqPopKS1RJCcn5/zE1M9pHV/L7P+/Wal8y959773DNnzhzPnHunR4EN3h5/b0WBnpulJAObj7TWl4DWmaweYTzMy+8HmmRQ9vUsjjEig2WBmC64Srv8D8xuowX0Mlv3K+CZwTauZn8fwDR/VwghhBCF5RG+P2tBkTmwQgghhBDCqkgG1opprUsVdgxCCCGEKGAyB1YysEIIIYQQwrpIBlYIIYQQwpo8wvdnLSiSgRVCCCGEEFZFBrBCCCGEEMKqyBQCIYQQQghrIhdxSQZWCCGEEEJYF8nACiGEEEJYE/khA8nACiGEEEII6yIZWCGEEEIIayJzYCUDK4QQQgghrIsMYIUQQgghrIhOSiqwhyWUUi2VUqeUUmeVUmMzKdNZKXVCKXVcKbU6t3UgUwiEEEIIIUSOKKVsgQWAL3AZ2K+U2qC1PmFWpiYwDnhWax2tlCqf2+PKAFYIIYQQwpo8WnNgGwNntdbnAZRSa4C2wAmzMm8CC7TW0QBa639ze1CZQiCEEEIIIXLqCeCS2fPLxjJztYBaSqk/lFJ7lFItc3tQycAKIYQQQliTAszAKqXeAt4yW7REa73kIXdTBKgJeAOVgJ1Kqbpa65icxiUDWCGEEEIIkSFjsJrVgPUKUNnseSVjmbnLwF6t9T3gglLqNKYB7f6cxiVTCIQQQgghrIlOKrhH9vYDNZVSbkqpYkAXYEOaMj9iyr6ilCqHaUrB+dxUgQxghRBCCCFEjmit7wODgK3ASWCd1vq4UmqqUqqNUWwrEKmUOgEEAaO11pG5Oa5MIRBCCCGEsCaP1l0I0FpvBjanWTbR7G8NjDAeeUKZ9ilEnpJGJYQQ4n+JKsiD3RzRpsA+Z0vN3VCg52YpycCKfDHUtUthh2CxeWFrGODaubDDsNjCsHUA9HHtWMiRWO7LsO+4dz1X050KXNFy1XjbtVNhh2GxxWHfAtDbtUMhR2K5ZWHfM9yK+gqAj8PW0N+K+otAK+0vulV9tbDDsNjqiz8Udgj/k2QAK4QQQghhRfQjNoWgMMhFXEIIIYQQwqpIBlYIIYQQwppIBlYysEIIIYQQwrpIBlYIIYQQwpokWfQDA/+vSQZWCCGEEEJYFcnACiGEEEJYE5kDKxlYIYQQQghhXSQDK4QQQghhTSQDKxlYIYQQQghhXSQDK4QQQghhRbSWDKxkYIUQQgghhFWRDKwQQgghhDWRObCSgRVCCCGEENZFMrBCCCGEENZEMrCSgRVCCCGEENZFBrBCCCGEEMKqyBQCIYQQQggromUKgWRghRBCCCGEdZEMrBBCCCGENZEMrGRghRBCCCGEdZEMrCgwTzavT/uJPbGxtWHP2t/4JXBDqvW2xYrwxtyBVHZ341bMTb4aNI+oy9cAeGlAW5p09iEpMYn1U5bz186jKdspG8WojQHEhkezpO9HAHSb3Z8aTz9F/I3bAKweFciVExfz9Hw6TepNHR9P7sXfZcWohVw6fiFdmcrubvSYPZCixYtxPCiEb6csA8DTvwkvD+uES40n+KjteP4+dj5PYzPXbVIf6vp4khCfwNJRn/F3BnFWda9GXyPOY0EhrJ7yJQCvjuiCh28jtE4i7nocX476jJh/o1O2c61XnffWB7Bo8Mcc3LIn384hI+8HzGXnH/twdLDnx1WLCvTYWXltUm/cfbxIiL/L8lELMmwXVdyr0cuo79CgQ6w12kWHcd2p91ID7ifc59rfEXw1egHxcbfzPeZuk/pQz8fLaCPzuZhJG+k3exBFixfjaNChlDaSrEW/1nR5vxeDPXtxM/pGnsb3ZPP6vDqxJ8rWhr1rf+PXDPqO1+cOpJK7G7eNviP68jVK2peiV+BwqtSrzr7vdrB+0rJ0++77+SjKVnHmoxaj8zRmgM5GH5GQRR9RJU0fsc5oC15mfcTMNH3EE09WoVvAWxQvVQKdpJnRdhz3797Lk5jzq7/4T5M6dJ3YC9siRbgZHcfM1yblSbw9JvfFw6cBCfF3WTRqPmGh6ftSN/dqvD1nCMWKF+Nw0EFWTF6ass6vlz9+3VuRlJREyG8H+Wb6Ctyfq0/Xsd2xLVqExHv3+TrgK07sPpYn8eaZpMIOoPBJBlYUCGWj6DS1D4t7zWC670i82jyLc40nUpV5prMP8bE3meY9jOClP9N6bDcAnGs8gVfrpkz3G8WintPp9EFflI1K2a5571ZEnP0n3TF/CviaWf5jmeU/Ns8Hr3W8PSnv5sJk7yF8PX4JXT7sl2G5rtPe5Otxi5nsPYTybi7U9vYA4OqpSyx5ZzZn953M07jSquvtibNbBcZ5D+ar8Yvo8eFbGZbrPu1Nlo9bxDjvwTi7VaCutycAW5b8xKRWI5nsP5qjvx2k9dBOKdsoGxs6jX2D47uO5Os5ZKadvy+L5k4rlGNnxt3bk/JuFZjgPZhV4xfz+odvZliu27Q3WTluERO8B1PerQJ1jHZx4vcjTPEbwQetRvHvhX9oNeDVfI+5nrcXzm4VGOs9iOXjA+meSRvpMe0tlo0LZKz3oFRtBMCxQlncm3lw3fgPZ15SNooOU/uwpNcMZvqOxDODvqOJ0XcEeA9jh1nfcf/uPbbMWceGgFUZ7rtui0bcvX03z2OGB33EJO8hrB6/hK7Z9BGTjD4iuS38k0kfYWNrQ6+PB7P6vc/5wG8kH3eZTOK9+3kSc371FyXKlKT7B/34tN9MJvgNZ+GAOXkSr4ePFy5uFRnRfABfjAukz7S3MyzX58N3+GLsQkY0H4CLW0Xqe3sBUPsZdxr6NmZsq+GM8R3Kz0t+AuBGdByz+nzI2BbDCBzxKQM+Hpon8Yq8JQNYCyilXJRSa5RS55RSB5VSm5VSzZRS3xnrPZRS/rnYfw+lVKhS6phSKkQpNSqb8u2UUrVzerzCUNWjBtcuhhN56V8S7yVyaONu6vo1TFXG3a8h+77fCcCRzXup1bQOAHX9GnJo424SE+4Tdfka1y6GU9WjBgB2Lo7UecGLP9f8VqDnU8+vIXvXm2INCzlDydKPU8bJPlWZMk72FC9dgrCQMwDsXb+T+n6NAAg/d4V/z1/N9zg9/Rqxe30wAOdDzlCydEns0sRp52RPidIlOW/EuXt9MJ5GnHduxqeUK1byMdAP5l291KsVB7fsJS4yNp/PImMNPepiV6Z0oRw7M/X9GrFn/Q4ALoScoUQm7aJE6RJcMOp7z/odePg1BuDkrqMkJZpSK+dDzmDvUjbfYza1kR0pxyxZ+nEL2sgOvIyYAbpM6M266SuAvJ+XV8WjBtfN+o6Qjbtxz6bvqGn0HQnxd7lw4BT3MshOFiv5GN79Xmb7/PV5HjNAfb+G7DH6iAvZ9BEP2kLqPiIigz7iqefrc+Wvv7ly0vSf8lsxN/PsivT86i+atHmeg//dS9Q/1wG4ERmXJ/E28G3Mru+DADgbcpqSZR7HvrxDqjL25R0oUaoEZ0NOA7Dr+yAaGm33pTdasmHheu4nmP4DkNyXXTx+ISVzfPn03xQrXowixR6tL6x1ki6wx6NKBrDZUEop4AcgWGtdXWvdABgHaK11R6OYB5CjAaxSqhUwDPDTWtcFmgDZjQjaAfk6gFVK5em71c7ZkZh/IlOex1yNws7ZMVUZe2dHoo0ySYlJ3LkRz+MOpdNtG2u2bfuJPflp+tdonf5N9vKo13h3y0xendAD2zzufEyxXk95Hh0eib1LmvNxcSTm6oO4o69GYp/mnPObg3NZoszqLio8Coc0gyIHl7JEm8UZdTUKB+cHZdqP6srs3Yto0vZ5fpy7FjCdv1eLxgSt2prPZ2Bd7J0dU9V3THgkDmnahYOLY6r6zqxdPNvJh+PBIfkXrMEUc+q2nFEbiUrVRh7E7OnbiJiIKC6dzNtvOczjy+z9n8y8jzDvO7LiP/I1gr/4mYQ7CXkfNDnrI2Is6COcq1UArRm8YjzjNs3A9+02eRZzfvUXLtUq8LhdKcasmcLEjTNp2r553sTrkjbeSBzS1J+DsyNR4anbbvI5ubhV5D+NazP1x5lMWDuNavVqpDtGY/9nCAs9nzLIFY8OGcBmzwe4p7VOmWSntT4CXDKypsWAqcBrSqnDSqnXlFJnlFJOAEopG6XU2eTnGRgHjNJa/2Ps+67W+nNj2zeVUvuVUkeUUt8rpUoqpZoCbYBZxvGqG4//GtnhXUqpJ43tqyul9hiZ3WlKqZvGcqWUmmWW9X3NWO5tbL8BOKGUmqqUGpYcqFLqQ6XUI/NdSp0XvLgZGcvl0PRztDbN/IaAF0cwu+17lLR/nJfeybtO/n/N+tnfMKrpO+z5aRcv9GwJQNeJvfl2xqoM/+Mgcq/VwPYkJiax98ddhR1KlooVL8bLA9vzw9w1hR3KQ6lYuyplq5Tn2Nb9hR3KQ7OxtaV6oyf5cuh8ZneciEeLxvynqXthh5Uio/7CxtaWqnWr8UnvAOb2mEbrwR1xdqtQyJGCbRFbStmXYmK7d1kd8BVDFqb+8vOJmpXpOrYHX4x7dObYp0jSBfd4RD1aOfFHkztwMLOVWusEpdREoKHWehCAMYB8HfgEeAk4orXObHJYVvtfbzaYnQb01VrPNwaYm7TWyVMYfgXe0VqfUUo9DSwEXgDmAfO01t8opd4x2297TFnj+kA5YL9Saqexzgtw11pfUEq5AuuBT5RSNkAXoDEZUEq9BbwFsHjx4nTrYyOisK/44H/p9hUciY2ISlUmJiIKh4pliQ2PwsbWhuKlS3Ar+ka6be2Mbd1faoD7Sw14yseToo8VpXipEnT/eCArhy8g7loMAIkJ99n77Q5eePOVTKrYcs26t+DZri8CcPHIORwqlgNOAaZMQEx4mvMJj8K+woO4HSqUJSbNOeeHF7q3pJkR54Uj53A0qztHF0eizbIRYGTczOJ0rOBIdETqMgB7ftzFsGXj+enjdbjWq8Y784cDUMqhNPW8vUhKTCRkm/UNCHLLu3sLnuv6EgBhR87iWLEs54x19i5liU7TLqLDo1LVd9p28UxHb+q92IC53abkW8wvdG9JcyPmC0fO4lix3IN4XMpm2EYcU7URU8zlq7rgVMmZqVvmpGw7edMsprYbm/IezK2YTN7/5pL7iLR9R2ZcvWpRuV41Jvw+HxtbG0qVtWPgmoks6DI1V7E2z2UfYW9BHxETHsnZfSdTzi80KIQq7m6c2h2ao5gLor+IDo/kZswNEuLvkhB/l9P7TlD5KVciLjz8NCrfHq3w6eILwPmjZ9PEW5boNPUXHRGFo0vqtpt8TlFXr7P/v6aLT88dOYNO0pR2LMONqDgcXcoyYslYAkfM49+/wx86TpH/ZACbP74EfsI0gO0DpL/01TLuxsDVHigFpPu+VilVCmgKfGua7QDAY8a/z2CabgCwGpht/P0c8I3WOhGIUErtABoBccA+rfUFAK11mFIqUinlCTgDIVrr9D2VqewSYEny06EBqeek/n3kHE6uLjhWciI2Igqv1k1ZMWR+qjKh2w/SuEMzwg6dob7/05zZfTxleY9PBxP0xc/YlXfAydWFi4fPEnboDJs+MmV+ajSpzQtvvsLK4QsA09yy5A/Qun4NuXr6UoYV/DB2rtzKzpWml8Ddx5PmPVtyYMMfuHrWJP7G7XQf2HHXYrhzIx5Xz5qEhZzh6fbNCF7+31zHkZ3fVv6X31aajlPPx4sXe7Zi74Y/qOZZk9s3bhObJs7YazHE37hNNc+anA85Q9P23vyyfDMA5V1d+DfM1Hl7+jYi/NwVAN59fmDK9n1mD+TIrwf/JwevAMErtxKc0i688OnZkv0b/sAti3YRfyMeN8+aXAg5Q5P2zQlavgWAOs098Hu7LXNem8S9fPpqGzJrI79TzYg5+zbSnF+Xb+Hyqb8Z2rBPSrlZvwcypfWYPL0LwaU0fYdn66asyqTvuGj0HWeNviMzu1dtZ/eq7QA4VHLizaVjcj14Bdixcis7zPoIb6OPyKot3EnVFpoRlE0fcWLHEXzfbkPR4sVIvHefWk8/xa9Lf85xzAXRX4Rs28/rU/thY2tDkaJFcPOoybalm3IU7/YVW9i+wvR+8XihAX49/flzw+/U8KxF/I3bqe6SAhDzbzTxN+Op4VmLsyGneb6DD9uWm+rrwLZ91H6mLif+DMXFrSJFihbhRlQcJcuUZPSy91gzcyWnD/yVozjzndyFQAawFjgOdMy2lBmt9SWlVIRS6gVMGcvXs9l/AyCjq5CWA+201keUUr0A7wzK2AAxWmuPh4kxC7fSPP8C6AW4YBqY50hSYhLfT1xG/xXjTbfRWhdE+JnLtBreiUvHzhP6y0H2rAvijbkDeT/4E9OtcAZ/CkD4mcuEbPqT8dvnkHg/ke8mLst2Ynn3eYMo5VgGpRRXToSx9r0vchp6hkKDQqjj48WUHZ+SEJ/AytELU9aN2/wR0/3HALBmwhf0mD3AdIuc4MMpcxrrt2hE58l9KOVYhgFfjuXyyTA+6xGQpzECHA06RD0fL2bs+IyE+Lt8aRbn5M2zmOxvunXQqglf0Gf2QIoVL8ax4BCOGXF2fPcNXKpVRCdpIq9cY8V7SzI8TmEYPWkG+0OOEhMTx4vt3mBA3+50aN2iUGMKDTpEXR9Ppu2YT0J8Al+NXpCy7v3Ns5hm1Pc3Ez6np1HfocGHCTXqu8uUvhQpVoRhqyYAcD7kNKvf+zxfY05uIzN3LCAh/i5LzWKesnk2k/xNX6uunPA5fWcPSmkjR4MP5WtcyZL7jreNvmOv0Xe0NPqO478cZO+6IF6fO5DxRt+x0ug7ACb8Pp/HSpWgSNEi1PVryKLuAUScvZLvcYcGheDu48VUo49YYfbeG7/5IwKMPuKbCV/QM5M+4jWjjxho9BHzewRwO+4Wv37xM2M3TAetCQ0KITQob+ZK51d/cfXcFUJ3hDD1v3NIStLsWvsrV/IgqXD4t4N4+DTg452B3I2/y+JRD/5jE7B5LuP9RwDw5fuLece4jdaR4EMcDjK13eB1v/L2rEHM3DaP+/fuETjS1G78evrj7FqBV4d05tUhnQGY0X1KoV2wKjKmZA5b1oyLuPYAS40sI0qpeoAdEKi1dldKdQDaaK17mm3XAZgPrNRav5vF/v2BD4CXtdbhxpzaHlrrL5RS1zFdrBUNbAauaK17KaXmA4e01suMfewGPtZaf2vEW88Y9P4MrNBarzW+4p+rtS6llGoPvI3pwjNH4ADwNPAkpvm4r5jFVww4BhQFahpZ2+zooa5dLCj2aJgXtoYBrp0LOwyLLQxbB0Af14f6f1Wh+jLsO+5dz7973eaHouWq8bZrp+wLPiIWh30LQG/XDoUcieWWhX3PcCvqKwA+DltDfyvqLwKttL/oVjX/byGXV1Zf/AFAZVcuL0V38i6wwZvDt8EFem6Wkou4sqFNI/xXgZeM22gdB6YD5pNigoDayRdxGcs2YPraP8vpA1rrzcBnwC/Gvg8BZYzVE4C9wB+A+fcYa4DRxi23qmPK8PZVSh3BlNFta5QbBoxQSh0FavDg7gY/AEeBI5gyv2O01hlO8tFaJxjnt87CwasQQgghRL6SKQQWMO4QkNF/ud2N9VGY5pCaq4/p4q1sJ9AYmdR0A12tdSAQmMHyP0h/G62WGez6CtBEa62VUl2A/xjba2C08TDfbzAQbL7MuHirCWA9qSghhBDi/zOZAysD2PyglBoL9Cfrua8FoQHwmTGtIAbTBWUWM34sYRPwg9b6TD7EJ4QQQgjx0GQAmw+01jOAGebLlFLvkT6L+a3W+sN8jGMXpkxwTrc/AVTLu4iEEEIIIXJPBrAFxBio5ttgVQghhBD/Gx7ln3gtKHIRlxBCCCGEsCqSgRVCCCGEsCZyEZdkYIUQQgghhHWRDKwQQgghhBXRkoGVDKwQQgghhLAukoEVQgghhLAmkoGVDKwQQgghhLAukoEVQgghhLAiMgdWMrBCCCGEEMLKSAZWCCGEEMKaSAZWMrBCCCGEEMK6SAZWCCGEEMKKyBxYycAKIYQQQggrIxlYIYQQQggrIhlYycAKIYQQQggrIwNYIYQQQghhVWQKgRBCCCGEFZEpBKC01oUdg/j/RxqVEEKI/yWqIA8W4dO8wD5nnYN2FOi5WUoysEIIIYQQ1kQ/kmPKAiUDWJEvhrt2KewQLPZx2BqGuL5W2GFY7NOwtQCMce1ayJFY7qOwb3jbtVNhh/FQFod9y73r5ws7DIsVLVcNgP6unQs5EssFhq1jkBW99wA+C1trlf2FtcU81Io+Q+aFrSnsEP4nyQBWCCGEEMKKyBxYuQuBEEIIIYSwMpKBFUIIIYSwIjpJ5sBKBlYIIYQQQlgVycAKIYQQQlgRmQMrGVghhBBCCGFlJAMrhBBCCGFFtNwHVjKwQgghhBDCukgGVgghhBDCisgcWMnACiGEEEIIKyMZWCGEEEIIKyL3gZUMrBBCCCGEsDIygBVCCCGEEFZFBrBCCCGEEFZE64J7WEIp1VIpdUopdVYpNTaLch2UUlop1TC3dSADWCGEEEIIkSNKKVtgAdAKqA10VUrVzqBcaWAosDcvjisDWCGEEEIIK6KTVIE9LNAYOKu1Pq+1TgDWAG0zKPcBMBO4kxd1IANYIYQQQgiRIaXUW0qpA2aPt9IUeQK4ZPb8srHMfB9eQGWt9c95FZfcRksIIYQQwooU5G20tNZLgCU53V4pZQPMBXrlVUwgGVghhBBCCJFzV4DKZs8rGcuSlQbcgWClVBjQBNiQ2wu5JAMrhBBCCGFFLL07QAHZD9RUSrlhGrh2Abolr9RaxwLlkp8rpYKBUVrrA7k5qGRghRBCCCFEjmit7wODgK3ASWCd1vq4UmqqUqpNfh1XMrCi0DzZvD6vTuyJsrVh79rf+DVwQ6r1tsWK8PrcgVRyd+N2zE2+GjSP6MvXKGlfil6Bw6lSrzr7vtvB+knLUrYZuGYiZZzsuXc3AYBF3QO4GRmX4xifal6f9hN7YWNrw59rf+OXwJ9SrS9SrAhvzB1IZfdq3Iq5wfJB84i6fA0A3wHtaNLZh6TEJL6fspy/dh7BvkJZus8dSOlydmit2f3Nr+xYtiXVPn36vcKr73dnnGc/bkXfyHHsadVqXp+2E3ugbG3YtzaI4Azqu8vcATxh1PfXg+YRffk6letXp8P0fqZCSrH9k+84vjVX/3F+KK9N6o27jxcJ8XdZPmoBl45fSFemins1es0eSNHixQgNOsTaKaY20WFcd+q91ID7Cfe59ncEX41eQHzc7QKLPa33A+ay8499ODrY8+OqRYUWB0DnSb2p4+NJQvxdVoxamEm9utHDqNfjQSGsM+rVy78JLw/rhEuNJ5jZdjx/HzsPQNX61Xl9+tsAKAWbPvmWI1v35zjGjpN6pcS4clQglzOIsbK7G91nD0iJ8bspywEoafc4fT4bhmMlJ6IuX2PpwE+Ij7v14NzqVWfk+g9YNngeh7fs5YnaVekyrR/FS5UgKTGJrQt+4NCmPx8q3rzuLwC6ffQOdV7w4kZkHDNajErZl/+IztT1bYjWmpvXY1k1KpC4f6MfKt78jt+S/u5hPdm8Pu0n9sTG1oY9a3/jlwz6MVOMbtwyPjeSY3xpQNuUGNdPWc5fO48CMPH3+dy9GU9SUhJJ9xOZ0+a9VPv06fcy7d7vznjPN/O0T86NR+2nZLXWm4HNaZZNzKSsd14cUzKwolAoG0WHqX1Y0msGM31H4tnmWZxrpLpokSadfYiPvUmA9zB2LP2Z1mNN30jcv3uPLXPWsSFgVYb7XjXsM2b7j2W2/9hcDV6VjaLT1D4s6jWdAN8RNGjzLC7pYnyB27G3+MB7KMFLN9PGiNGlxhN4tW7KdL+RBPYMoPMHfVA2iqT7ifwwbSUBviOZ++r7PN/dL9U+7SuU5clm9VI63LyibBSvTu3N0l4zmeM7Co82TSmf5lwad/YhPvYWH3kPZ9fSzfgb5xJ+6hKftn6PT/zHsbTHDDp82A8b24LpOty9PSnvVoEJ3oNZNX4xr3/4Zobluk17k5XjFjHBezDl3SpQx9sDgBO/H2GK3wg+aDWKfy/8Q6sBrxZI3Jlp5+/LornTCjUGgDrenpR3c2GS9xBWj19C1w/7ZViu67Q3+XrcYiZ5D6G8m0tKvf5z6hJL3pnN2X0nU5X/59QlZrQeS4D/GOb3COD1D9/KcVup7e2Bk5sLU7yH8s34z+nyYd8My702rR+rxy1hivdQnNxcqG3E6Nu/Had2hzLVZxindofiN+DBXX2UjaLt2G78tetoyrJ78QmsGLGAD/1GsbDndDpM7EmJMiUtjjc/+guAvd/tILDn9HTH+23JRma2GsNH/u8S+tshWg7tYHGsBRV/dv1dTmNc3GsG031H4pXB58YzxufGNO9hBJt9bjinxDiKRT2n0+mDvil1DPBZ1w+Y5T823eDVvkJZ/pMPfbLIPYt6FqXUe0qp40qpo0qpw0qpp5VSw5RSlr+7LTtOmFKqXPYlM9x2lFLqLyO+/UqpHnkZmwXHt1dKDbCgXC2l1Gal1Bml1CGl1DqllLNSqqFS6lOjjLdSqmkOYnhTKbXW7HkZpdQ5pVS1h91XfqviUYPrF8OJvPQvifcSCdm4G3e/1PO53f0asu/7nQAc2byXmk3rAJAQf5cLB05x7+69fI2xqkcNrl2MSInx0Mbd1PVrlKpMXb+G7Pt+BwCHN++hVlN3Y3kjDm3czf2E+0Rdvsa1ixFU9ahB3LWYlCzS3Vt3iDh3BTsXx5T9tZ/Qg5+mf40mbyc4VTbqO8o4lyMb/6ROmvqu7deAA0Z9H9u8lxrGudy7k0BSYhIARR4rWqBzr+r7NWLPelP9Xgg5Q4nSj1MyLZfjAAAgAElEQVTGyT5VmTJO9pQoXYILIWcA2LN+Bx5+jQE4uetoSuznQ85g71K24ILPQEOPutiVKV2oMQDU92vInvWm1/pCyBlKZlKvxVPV607qG+0//NwVIs5fTbdf87ZS9LGi6Fw0lnp+jdhnxBiWxWtfvHQJwowY963fST0jxnq+Ddn7nant7P1uB/V8H7x3m/dqxZEte7kRGZuy7N8LV7kWFg5A7L/R3IiMo5RjGYvjzY/+AuDcvpPcjr2Z7nh3bsan/P1YyeK5nhRZGP1dzmIMTxNj1p8btYzPjbp+DTm0cTeJKTGGp9RxVl6d0IMN07/O4x4597RWBfZ4VGU7gFVKPQO8AnhpresBL2G639cwIE8HsDmllHoH8AUaa609gBcBi2tdKVUkq+cWsgeyHMAqpYoDPwOBWuuaWmsvYCHgpLU+oLUeYhT1Bh56AAt8AVRWSr1kPJ8KfKm1Pp+DfSXHbJvTbbNi7+xIzD+RKc9jr0Zh55y6Y7MzK5OUmMSdG/E87pD9h3+XWe8wavMMfAe3z9MYY65GYufskE2Mt3ncoTR2zg5E/3M91bb2ac7PsZITT9R24+LhswDU9W1ITEQU/5y8mKu4M2Ln7EBsqvqOpEwG5xKb5lxKGvVd2aM6I7bNYsTWj1j//hcpg5T8Zu/sSJT5axAeiUOaD0AHF0eirz4oE51BXQM828mH48Eh+ResFbF3dkzVPqPDI7FPU6/2Lo7EXE3d/jOq17RcPWowYdsc3t86h9Xvf57jtmLv7EB0mtc+4xijzGKMwt5o16Wd7Ii7FgNA3LUYSjvZAab3Qv0Wjdi1anumx65avzpFihbh+sWIh4g3f/uLjLw86jWm7F5Ag7bPsXnuOotjLYz40/Z3OWGXLsb0nxumtp3+cyPttqk+c7Sm/8rxjNoYwDNdX0wp4+7bgNiIKP45+XeOYxb5x5IMbAXgutb6LoDW+jrQEagIBCmlggCUUoHGDW6PK6WmGMteUEr9mLwjpZSvUuqH7A6olHJVSp1USn1u7G+bUqpEFpuMB/prreOMGOO01l8Z+0rJ6hpZzmDj78lKqZVKqT+AlRk8d1JKfW9kc/crpZ412+5LpVSwUuq8Uip50DkDqG5kgGdlEmc34E+t9cbkBVrrYK11qJF13aSUcgXeAYYb+3peKXVBKVXUOH4Z8+fmtCnd8Q7wiXF7iheBWUqpN5RS+4z9LU4elGb0mpnV2Uyl1CGgk1JqiFLqhJGBX5PJa5Zyo+MlS3J8u7hcWzV0PrNajmF+p8lUb/QkDds/X2ixZKVYycfoGziC9VO/4s7NeIoWL4bvwHa5/hDKL5cOn2Ou32jmt3kPn/5tKfJYuub3SGs1sD2JiUns/XFXYYfy/17Y4bN84DeSmW3G0bL/q49OWzEylB0m9uKnGaszzQ6XcbKnx9xBrBodmKsMckH4efZaJjUdyMGffuf5ni0LO5xMpe3vHjXzOk5i9ivjWNRrBs/38KN64yeNPvnVR7ZP1kkF93hUWZJp3AZMVEqdBn4B1mqtP1VKjQB8jAEtwHta6yhjcPSrUqoeEAQsVEo5aa2vAb2BLy2MrSbQVWv9plJqHdABSDfpUSlVBiidwyxjbeA5rXW8UmpymuergY+11r8rpapgurruKWO7JwEfTPc2O6WUCgTGAu5GBjgz7sDBrALSWocppRYBN7XWs41zDAZeBn7EdHuK9VrrDL8/11ofVUptBX7F9FNu1YHXgGe11veUUguB14EVZPCaaa2TJ4VFGhlilFL/AG5a67tKKft0ByXdjY718IDfsjpNYiKisK/44OtcuwqOxEZEpSoTa5SJDY/CxtaG4qVLZDuBPjbCdBHD3Vt3OLjhD6rUr8GB9TkbtKSN0b5C2ZT9p40xJiXGktyKvkFsRDQOFcul2jbGOD+bIrb0XTSSAz/+ztGt+wAoV9WZspXK8+6Wj0zlXcoyetMM5rQbz41rseRWbEQ0dqnquyxxGZyLXar6LsntNPX977l/SLh9F5dalbl8LMeJ/Sx5d2/Bc11NXyKEHTmLY8WynDPW2buUJTo8dTuJDo/CocKDc3Mwq2uAZzp6U+/FBsztNoX/Zc27t+BZI7t08cg5o32eAsDBxdSGzcWER2FfIXX7j0nzHs1K+Lkr3L19h4q1Kqdc5JWdZt39aJoqRrPjZxrjgwycfQVHYox2feNaLGWc7Im7FkMZJ3tuXDfNh69Srxq955vyDqUcylDH25OkxESObjtA8VIl6L9sLBtnr0mZlmCp/OovLHHgx128vWwcWz7+9qFiLoj4M+rvcio2XYzpPzdiIqJwyOBzI+225p85yed5MzKOo1v3U6V+DW7H3qJsJSfGpPTJjozeNJ057d7Lkz5Z5F62GVit9U2gAfAWcA1Yq5TqlUHRzkbGLgSoA9Q2MoIrgTeMgc8zgKWXIF7QWh82/j4IuFq43cPYoLWOz+T5S8BnSqnDwAagjFKqlLHuZ631XWPw/i/gnA+xmfsC0+Af499lWZQFWABc0VoHY8rCNgD2G+fyIpA8Jzbda2a2j7Vmfx8FvlZKvQHcz8V5pLh05BxOri44VnLCtqgtnq2bcnx76rF96PaDNO7QDID6/k9zdvfxLPdpY2uTMsXApogtdV7wIvz0pSy3ycrfaWL0at2UY9tTX30fuv0AjTs0B8DDvwlnjBiPbT+AV+umFClWBMdKTji5uqR8ddZt5jtEnL1C0NIHv6h39dQl3mv4FlOeG8yU5wYTEx7JrFfG5llHefnIOcq5uuBgnEv91s9wIk19n9h+kIZGfdc1q2+HSk4pF+LYP1GO8tUr5usFDcErtzLNfzTT/EdzeNt+mrQ31a+bZ03ib9xO+Vo4Wdy1GOJvxOPmWROAJu2bc2Sb6cr3Os098Hu7LQv6zeTenYR8i9ka7Fi5lQD/MQT4j+HItn00aW96rbOq1zup6rUZR7ZlffeJsmZtxfGJcjhXr0jkQ7SVnSu3McP/XWb4v8vRbftpbMTomk2MrkaMjds346jx2h/75QBPdzS1nac7Nueo8d6d/PxgJj1neoRs2cPaCUs5uu0AtkVteXPxSPau38nhLXstjjlZfvUXmXFydUn5u65vI/49dyWL0oUXf0b9XV7GGJrN50ZyjKHbD+LVuim2aWIsVuIxHnu8OADFSjzGk8/X4+rpS1w9dYn3G77N1OcGM/W5wcSERzHrlXGPzOA1SasCezyqLJrrqbVOBIIx/YrCMaCn+XplunntKKCR1jpaKbUcKG6sXgZsBO4A3xr3C7PEXbO/E4EMpxBoreOUUjeVUtUyycLe58FAvXiadbeyeG4DNNFa3zEvoJTKKDZL58weB5pbWDaF1voPY1qFN2CrtQ7NZpMk4wGmucBfaa3HmRfI5jWD1HXxMtAMaA28p5Sq+xCvY8YBJibx/cRlvL1iPDa2NuxdF0T4mcu0HN6JS8fOc/yXg+xdF8TrcwcyPvgTbsfcZOXgT1O2n/D7fB4rVYIiRYtQ168hi7oHEH3lOm+vGIdtEVtsbG04/Ucof37za65i/G7ilwwwYtyzLpjwM5fxH96Jv4+dJ/SXg/y5LojucwcxIXget2NusnzwPADCz1wmZNOfjN8+h8T7SXw78Ut0kqZaw//QuEMzrpy8yJjNMwHY9NE3nAg+nFUouZaUmMRPE5fTb8U4bGxt2L8umIgzl/Eb3pHLxy5w4peD7F8XTJe5AxgT/DG3Y26yevB8ANwa/Qfv/m1Jun8fnaT5YcKX6TKz+SU06BB1fTyZtmM+CfEJfDV6Qcq69zfPYpr/aAC+mfA5PWcPpFjxYoQGHybUmOvaZUpfihQrwrBVEwA4H3Ka1e99XiCxZ2T0pBnsDzlKTEwcL7Z7gwF9u9OhdYsCjyM0KAR3Hy+m7viUhPgEVoxemLJu/OaPCPAfA8A3E76gZ/ItqoIPp8whrt+iEa9N7kMpxzIM/HIsl0+GMb9HANUbPUmL/u1IvJ+ITkpizYSlOb7t0PGgEOr4eDJpxzzuxSewanRgyrqxm2cyw/9dANZNWMobswdQtHhRTgQfTnkvbQ/8iT4LhvFMZx+irlzny4EfZ3k8r5efoUbjp3jcoTRNjIHvylELuXLCsjnp+dFfAPT8dAg1mtSmlENppv65kM0ff8uedUG0frcb5atVRCclEX3lOmtz2a6tob9L/tzonxKj6XOjlfG5EfrLQfasC+KNuQN53/jc+Mr43EgdYyLfTVyGTtKULmdH3yUjAVMS5OBPf/DXjiO5qktRMFR2c3yUUv8BkrTWZ4zn0zBdsNQcaKO1vqCUqo/pK2lPwAlTxu5drfVyY5uNgBfwktb6ZPqjpBwrDGgIlAI2aa3djeWjgFJa68mZbDcA0+DqNWNAWwpor7VeoZT6BZijtd6ilPoY8NRaextTBsy/pk/7fDUQorWeZTz30FofzqBcKKaL3G4Ah7TWVbM4vxLAMWCo1vpnY1kzIArTr1SM0lq/opQaCZTRWk8y23YkMBL4QGsdmH7vqY7jmlx/SqnawE+YphD8q5RyxDT1wZ5MXrPk10FrfV2ZfsO4ijG1oShwEVN2PSbdgR/Qw127ZBXiI+XjsDUMcX2tsMOw2KdhpuT4GNeuhRyJ5T4K+4a3XTsVdhgPZXHYt9y7nj/TJPJD0XKmL1b6u3Yu5EgsFxi2jkFW9N4D+CxsrVX2F9YW81Ar+gyZF7YGHuLC8bxw+qmWBTZBu9bJ/z6SaVhLLuIqBXyVfBEPpq+ZJ2Oa7/hfpVSQ1voIpq+h/wJWA3+k2cfXwKWsBq+5FIhpvu1+Y0C5iwcZyCnAPKXUAUzZUksNARoaFy6dwHRxVKa01pHAH0qp0Mwu4jKmJ7wCDFam22idwHTngrTfsW0EXk2+iMtY9jXgAHzzEOeA1voE8D6wzXj9tgMVLHjNktkCq4zMewjwaTaDVyGEEELkI7mNlgVffWutD5LxLZ3mG4/kcr2y2M1zQLbfb2itXY0/r2O64Cl5+exsttPAR8Yj7bpdQK0Mlk/O5vl1TBc/ZbedeZzd0pbPYPu/gIwuF43ANE0DrfVpoF6a9c8B31kyeNRah5G6/taSek5r8vJemWzvavb3PePYQgghhBCPhHz/KVml1EFM8ylH5vex/r9SSs0HWgH+hR2LEEIIIQrXo/ZTsoUh3wewWusGaZcppfYCj6VZ3F1rfSyrfSmlFgDPplk8T2ud3VX5BUopVRfT3RfM3dVaP52T/WmtB2dwDKuoCyGEEEKIvJbvA9iM5GIgNzCvY8kPxkA8q/vB5sUxrKIuhBBCCJG3HvHf2CgQllzEJYQQQgghxCOjUDKwQgghhBAiZ2QOrGRghRBCCCGElZEMrBBCCCGEFXmUf+K1oEgGVgghhBBCWBXJwAohhBBCWJFH+ReyCopkYIUQQgghhFWRDKwQQgghhBWR+8BKBlYIIYQQQlgZycAKIYQQQlgRuQuBZGCFEEIIIYSVkQGsEEIIIYSwKjKFQAghhBDCishttCQDK4QQQgghrIxkYIUQQgghrIjcRksysEIIIYQQwspIBlYIIYQQworIbbRAaclDi7wnjUoIIcT/kgIdUR6o1K7APmcbXv7xkRwtSwZW5IsBrp0LOwSLLQxbxzjXboUdhsWmh60GYJRr10KOxHKzw76ht2uHwg7joSwL+57+VtSOA8PWAXDv+vlCjsRyRctVY4Rrl8IO46HMDVvDENfXCjsMi30athaAQVYU82dha62yjguS3IVA5sAKIYQQQggrIxlYIYQQQggrInNgJQMrhBBCCCGsjGRghRBCCCGsiFwpLRlYIYQQQghhZSQDK4QQQghhRWQOrGRghRBCCCGElZEMrBBCCCGEFZH7wEoGVgghhBBCWBnJwAohhBBCWJGkwg7gESAZWCGEEEIIYVVkACuEEEIIIayKTCEQQgghhLAiGrmISzKwQgghhBDCqkgGVgghhBDCiiTJb8lKBlYIIYQQQlgXycAKIYQQQliRJJkDKxlYIYQQQghhXSQDK4QQQghhReQuBJKBFUIIIYQQVkYysEIIIYQQVkR+SlYGsKKQdZrUmzo+ntyLv8uKUQu5dPxCujKV3d3oMXsgRYsX43hQCN9OWQaAp38TXh7WCZcaT/BR2/H8few8AE8+V5d2776ObdEiJN67z/qAlZz+83iuY63VvB6vTOyBja0N+9cGsSNwY6r1tsWK0Hluf55wd+N2zE1WD/qUmMvXU9bbVSzL8O2z+PWT79n1+c8AdPjoLZ58wZObkXHMa/FurmNM6z/N69PWiHnv2iCCAjeki7nr3AFUMmJeOWge0ZevU/O5urz8bpeUOtwUsJqzaeqw9+ejKFulPLNbjMnzuM11m9SHej5eJMQnsHTUfC5m0Eaqulej3+xBFC1ejKNBh1g95ctU61v0a02X93sx2LMXN6Nv5HmMnY12nJBFO66Sph2vM9qxl1k7nmnWjqvWr87r098GQCnY9Mm3HNm6P89jz8r7AXPZ+cc+HB3s+XHVogI9dlaebF6fdhN7YmNrw561v/FbBu2629yBVHZ341bMTVYMmkf05WvUeq4uL7/blSJFi3D/3n02Bnydrl3nxlPN69N+Yi9sbG34c+1v/BL4U6r1RYoV4Y25A6nsXo1bMTdYPmgeUZevAeA7oB1NOvuQlJjE91OW89fOIwB0++gd6rzgxY3IOGa0GJWyLw//JrQa1hHnGk8wp+17XDLaTU50nNQrpf2uHBXI5Uz64e6zB6S03++mLAegpN3j9PlsGI6VnIi6fI2lAz8hPu5WynZV6lVn5PoPWDZ4Hoe37AXAoWJZus14G4eK5dBaE9h7Rko9ZKbDpF7UNmL8OosYXzdiPBEUwvdmMfYyi3GZWYxZ7bd4qRKM3z6Ho9v2890k0/vVtqgtHaf0oWaT2mit2TRrjeUVLfKUTCEQhaaOtyfl3VyY7D2Er8cvocuH/TIs13Xam3w9bjGTvYdQ3s2F2t4eAFw9dYkl78zm7L6TqcrfjL5BYN+ZfNhyFF+NXECvjwfnOlZlo2gztTfLen3Ex76jqd+mKeVrPJGqTKPO3sTH3mK29wh+X7qFVmO7plr/8vtvcDr4SKplB7/bybKeM3MdX2Yxvzq1N1/0msks31F4tmmKc5qYn+7sQ3zsLWZ4D2fn0s28PLYbALeib/Bl39nMafkua0YG0vXjAam2c2/RiLu37+RL3ObqeXvh7FaBsd6DWD4+kO4fvpVhuR7T3mLZuEDGeg/C2a0Cdb09U9Y5ViiLezMPrmfzAZlTye14kvcQVo9fQtds2vEkox3XMdrxP5m0439OXWJG67EE+I9hfo8AXv/wLWxsC7bLbufvy6K50wr0mNlRNor2U/uwpNcMZvqOxKvNs5m065sEeA9jx9KfecWsXS/tO4tZLcfwzciFvP7xwDyNq9PUPizqNZ0A3xE0aPMsLmniatL5BW7H3uID76EEL91MGyMulxpP4NW6KdP9RhLYM4DOH/RB2ZjmOO79bgeBPaenO97VU5dY+s4czqVpNw+rtrcHTm4uTPEeyjfjP6fLh30zLPfatH6sHreEKd5DcTLrh337t+PU7lCm+gzj1O5Q/Aa0TVUnbcd2469dR1Ptq8fcgfy6ZCPTXhrBrLbjuXE91qIYP/Aeytrxn9M5kxg7T+vHmnFL+MCI8Skjxpf6t+P07lCm+Qzj9O5QfI0Ys9uv/8jO6d6XfoPaczMyjmkvDCfgpZGc3Zu7+s8pjSqwx6NKBrBZUEpppdQqs+dFlFLXlFKbHnI/FZVS3xl/eyil/C3Yxjur4yilnJVSm5RSR5RSJ5RSm43lrkqpbhbs36Jy+ameX0P2rt8JQFjIGUqWfpwyTvapypRxsqd46RKEhZwBYO/6ndT3awRA+Lkr/Hv+arr9Xj4eRuy/0QBcPX2JosWLUaRY7r5sqOxRg8iLEURf+pfEe4kc2fgnT/k1SFXmKb+GHPp+FwChm/dSval7yrrafg2JvnSNiDOXU20Ttu8vbsfezFVsmaniUYPIi+FEGTEf3vgndfwapipTx68BB743vQZHN++lphHzP8fDiDPqMPz0ZYoWL4atUYfFSj5G837+/Dr/h3yJ25ynXyN2r98BwHmjjdilaSN2TvaUKF2S80Yb2b1+B15+jVPWd5nQm3XTVwD5c+fv+n4N2WO04wvZtOMLRox70rTjiAza8b07CSQlmr4oLPpYUbQu+DuXN/Soi12Z0gV+3KxU8ajBdbN2HbJxN+5p2rW7X0P2p2rXdQC4kkW7zq2qHjW4djGCSCOuQxt3U9d4jZPV9WvIvu9N7fnw5j3UMt5vdf0acWjjbu4n3Cfq8jWuXYygqkcNAM7tO5lhHxGRSf/3sOr5NWKfWT9cwoJ+eN/6ndQzzq2eb0P2fmc6p73f7aCe74Nzbt6rFUe27OVG5IMBqkuNJ7CxteWv348BkHD7LvfuJGQZY91cxljXtyH7jBj3fbeDukaMWe23srsbpcvZpxt8N+nkzfaFPwKgteZWPnyjIywjA9is3QLclVIljOe+wJWH2YFSqojW+h+tdUdjkQeQ7QDWAlOB7Vrr+lrr2sBYY7krYMnA1NJy+cbe2ZHofx58xR4dHom9i2PqMi6OxFyNfFDmaiT2zqnLZMWz1dNcCj3P/YT7uYq1jLMDsf88iCPuahR2aeIo4+xAjFEmKTGJOzduU9KhtGnA905rfp33fa5ieFh2ZvEAxFyNxM7ZIU0Zx1Qxxxsxm6vXqjGXQy+QaNRhy5Gd2fHFzyTcuZvPZ2BqI1Fp2oiDS9lUZRxcyhJl1kaizNqIp28jYiKiuHTyYr7G+LDtOMbCduzqUYMJ2+bw/tY5rH7/85QB7f8y8zYLEJPBezFtu75zI57H07Xrp1O169yyTxdX9u+3Ozdu87hDaeycHVK1IUvbR97E7UC0edyZtt8os/iisDfOrbSTHXHXYgCIuxZDaSc7wNT/1G/RiF2rtqfaV/lqFYiPu0W/RSN59+cZtBv3ekq2OTPp+rLwSOzSxGiXQYx2FsSY0X6VUrR7vzs/fbgy1TFKlCkJwMsjOzN60wx6LxhO6XJ2WcaeX5IK8PGokgFs9jYDLxt/dwW+SV6hlGqslPpTKRWilNqtlPqPsbyXUmqDUuo34Fcj2xmqlCqGaeD5mlLqsFLqtcz2YYEKQEo6T2ud/N/EGcDzxv6HG8fepZQ6ZDyaZlKul1LqM7Nz22RkgW2VUsuN+I8ppYY/fBUWjgo1K9Fu7OusHv95ocbx4rAO/L50Mwm383/Al9eca1bCf2w3vh//BQAVa1elbBVnQrceKOTIsleseDFeHtieH+Za7xy1sMNn+cBvJDPbjKNl/1cp8ljRwg7p/wXnmpV4ZWw3vjXatchDxjcFHSb24qcZq9N9c2Bja0v1Rk/xw4crmdVmPOWqONOko3ehxJiZ57r7cSLoMDHhUamW29ja4lCxHBcOnmbWK2MJO3SaduPfyM9IRRbkIq7srQEmGl/n1wO+BJ431v0FPK+1vq+UegkIADoY67yAelrrKKWUK4DWOkEpNRFoqLUeBKCUKpPFPrKyAFirlBoE/AIs01r/gykTO0pr/Yqx/5KAr9b6jlKqJqYBeMMMyvXK5DgewBNaa3ejnH1GhZRSbwFvASxevDjToJt1b8GzXV8E4OKRczhULAecAkyZtLQdRkx4FPYVHmTcHCqUJSYidZmM2Ls48tbiUXw1YgHX/47Itnx24iKisav4II4yFRyJTRNHXEQ09hXLEhcehY2tDcVLl+R29A0qe9Sgrv/TtBrXjeJlSqKTNPfv3uPPFdtyHVdWYo14ktlXKEtsRHSaMlHYVyxLrBFzCSNmMGU0ei0ewZoRC4n8+18AqnrVpFK9aoz//VNsbG0oVdaO/msmENjlgzyL+4XuLWne9SUALhw5i2PFcinrHFzKEh0emap8dHgkjmZtxNFoI+WruuBUyZmpW+akbDt50yymthubko3Jqea5bMf2FrbjZOHnrnD39h0q1qqccpHX/6rkNpvMPoP3Ytp2Xbx0iZSveu1cHOm9eCSrRywgMg/6hmQx6eLK/P0WY9ZH3Iq+QWxEtNGGHmz7MO3jYTXr7kfTVO3XLO5M2++DjKd9BUdijHO7cS2WMk72xF2LoYyTPTeuxwFQpV41es8fAkAphzLU8fYkKTGRmPAoLp8MI/KSqU85sm0/bp41+XNdUKpjPt/dj2eMGP8+ci513bqYXltzsRnEGJtNjOn6SGO/bl61qNboSZ7r7stjJYtTpGgR7t6+w8aZ33D39h2O/HcfACGb99DkNR8LajzvPcqZ0YIiA9hsaK2PGgPQrpiysebsgK+MgaEGzNMj27XWlvRAWe0jq7i2KqWqAS2BVkCIUso9g6JFgc+UUh5AIlDLkv2bOQ9UU0rNB34GMhx1aa2XAEuSnw4I+CXDne1cuZWdK7cC4O7jSfOeLTmw4Q9cPWsSf+N2uoFF3LUY7tyIx9WzJmEhZ3i6fTOCl/83y4BLlCnJgGVj+Wnmas4fPPUQp5q5y0fOUc7VBYdKTsRFRFG/9TOsGfJZqjIntx/Eq8Pz/H3oDO7+T3Nut+nq5iWdp6aUeXFYBxJu3cn3wSvAJSNmx0pOxEZE4dH6Gb5OE/Px7Qdp2KEZFw+doZ7/05w1Yi5epiR9l43h55nfEHbwdEr5P1f9wp+rTK+tQ6Vy9F06Jk8HrwC/rfwvv600vcb1fLx4sWcr9m74nWpGG4lN00Zir8UQf+M21Txrcj7kDE3bN+fX5Vu4fOpvhjbsk1Ju1u+BTGk9Jk/uQrBj5VZ2mLVjb6Mdu2XTjt08a3Ih5AxN2jcjKJt2XLaSE9FXI0lKTMLxiXI4V69IZD5diGZNLh05h5NZu/Zs3ZSVQ+anKnN8+0EaZdKu31z2Lj/PXJ2qXeeFv9PE5dW6KV8N+TRVmdDtB2jcoTlhh87g4d+EM0Zcx7YfoOenQwj6YhNlyjvg5OrCxcNn8zQ+c5nRK64AACAASURBVDtXbmPnSlMfVMfHk2Y9W3Bww26L++HG7Zuxw2i/x345wNMdm7M98Cee7tico9tN385Mfv7BxbNvzO5P6K+HOLrtAMpGUaLM45RyLM3NqBv8p6k7fx89ly7GXSu3scuIsbYR4yEjxjsWxrjTiDH0lwM07ticXwJ/onHH5hwzYjy2/UCG+10x7EF7atyxOVXqVmPjTNOXr6G/HqJGk9qc+fM4tZ51J/zMFSrUqpzzF0PkmAxgLbMBmA14A+YT8D4AgrTWrxqD3GCzdbewTFb7yJIxQF4NrDYyxM2AyDTFhgMRQH1MU0Yyu3T8PqmnlBQ3jhGtlKoPtADeAToDfdJv/vBCg0Ko4+PFlB2fkhCfwMrRC1PWjdv8EdP9TbdnWjPhC3ok374l+DDHg0MAqN+iEZ0n96GUYxkGfDmWyyfD+KxHAM17tMSpqguthnak1VDT1OP53adxMzIux7EmJSaxYeJy+qwYi7K14cC6YP79P/buO66r6n/g+OuwRFyIA1wJjoYTt5klZEhijtRMLVeW5k4lV6ml5U7TUkpzZs60cpYjUMutKK7MhQmKA2QJivI5vz8+VwQERUDg8/u+nz0+j/jce+6973s4Xs7nfc89nzOhvDa4PaHHznNq22EOrgqgw/S++AZMJy7yFssHfPPY/Xac1R+3hi9QoGghRuz5hm0z1nBwVUCm40wd8y9jFvHBkpEoaysOrArg6pkQvAe359KxC5zcdoj9qwLoNL0vIwJmEBcZy1Ij5pe6elO8vDNeg9riNagtAPO6TMxSHWZGkP9hanjWZvKO2STE32H+x7OT1n2+aRpjfczTCv04eh49p/XHzt6OYwGBBAUczrEYj/sHUs2zNuOMdrwkWTsetWkKE4x2vHz0D3RLpx2/bbTjfkY7/qbrBCrWex7vPm1IvJeINplYMXp+jj8w8vHYSRwIDCIyMpqmbd6lb88utGvpnaMxpGZKNLF2zEJ6LRmFlbUV+1f5c/VMCK8PfotLx85zYtsh9q3yp/P0fowK+Jq4yFiWDDB3JBt39aZYeWeaDWpHs0HmG13fd5mQLe3alGji5zEL6GvEtXdVAGFnQvAZ/Bb/HTvP8W2H2LPKny7T+zM6YCZxkbEsGjATgLAzIQRu2MOorV+ReM/E6jEL0Cbzbe5uswZSqWEVChYtxLg9c9g0YzV7V/lTw7se7T/rQUGnwvReMJzQUxfx6zrhieM+4R9IVc9ajN0xk7vxCSz92C9p3YhNk5nkY57eb9Xo+bw7rS+29racDDjCyYAjAGz1+433Zn/Eix08iQi9wYJ+Mx55PG3S/Prljwz4aTRKKf47fp6/V2x/5DYnjRjH7JhJQnwCPyWLcdimyUxJFuM70/pil0aMPWZ/RMMOntwMvcFCI8ZH7Tc96yb9RJfp/Wk7phuxEdEs+9iPWi0aPnY7kf1UbjzZaimUUrFa64JKqbJAW631LKWUB8atd6XUL8BSrfUapdRnQHettatxOz75MAFXYIPWuppSqh3QSmvdzViX3j6SjpNObK8Ce7XWcUqpQsB+oCvmOwvTtdZNjHIzgBCt9VdKqR7AAq21UkrVSVWuMTAFaAyUAU4ArYDjQILWOtrI8C7VWrs/pup0X9cOGa/oXDYneBUjXXP1ebYnMjF4GQC+rp0eUzLvmBa8nB6uGRkZk3csDF5DHwtqx37BqwC4e8NyhhjYFq/AENeOuR3GE5kevIKBrm/ndhgZNit4JQD9LSjmb4NXWmId5+h8UxudO+VY563F1eV5ci4teYgrA7TWIVrrWWmsmgJMVEoFkvFstj9Q5f5DXJncB0Ad4KBSKgjYA/ygtT4ABAGJxvRag4E5QDel1FHgeR5khlOX+xu4AJwEZgH3U1hlgACl1BFgKTDyCWIUQgghhMh2MoTgEbTWBdNYFoBxm19rvYeUY0o/NZYvAhYl2yYYqGb8HAGknBww7X0kHSed2KYCU9NYfhd4NdXiGsl+Hv6Icu+kc7ja6cUhhBBCiJxlypM50ZwlGVghhBBCCGFRJAObxxnjVgelWvy31jr7vgNRCCGEEBbDlIe/4jWnSAc2j9NaLwQW5nYcQgghhBB5hXRghRBCCCEsiMwfJWNghRBCCCGEhZEOrBBCCCGEBTHl4CsjlFKvK6VOK6XOKqVGpLF+iFLqpFIqSCm1XSlVPlMnnox0YIUQQgghRKYopayB2Zi/1r4K0EkpVSVVsUDMX/BUA/gZ8xz4WSJjYIUQQgghLIhJ5alZCOoDZ7XW5wGUUiuA1pi/GAkArbV/svJ7gXezelDJwAohhBBCiDQppXoppQ4me/VKVaQMcCnZ+xBjWXp6ApuzGpdkYIUQQgghLEhOzkKgtZ4LzM2OfSml3gXqAk2yui/pwAohhBBCiMwKBcole1/WWJaCUuo14BOgidb6TlYPKh1YIYQQQggLktHZAXLIAaCyUsoNc8e1I9A5eQGlVC3ge+B1rfW17DiojIEVQgghhBCZorW+B/QH/gBOAau01ieUUuOUUq2MYlOBgsBqpdQRpdS6rB5XMrBCCCGEECLTtNabgE2plo1J9vNr2X1M6cAKIYQQQlgQU56aRSt3yBACIYQQQghhUSQDK4QQQghhQUxIClYysEIIIYQQwqJIBlYIIYQQwoLk5BcZ5FWSgRVCCCGEEBZFaS39eJHtpFEJIYT4X5Kjg1KXlHk3x/7Odg1dmicH3MoQAvFUdHNtl9shZNji4DV0Lv9mboeRYcsu/gJAX9cOuRxJxs0JXsVg1465HcYTmRG8gv6ub+d2GBn2bfBKAIZYUD1PD17B3RvnczuMJ2JbvAKdyrfJ7TAybPnFXwGwsSuTy5Fk3L2EUAo6uOV2GBkWG3cht0P4nyQdWCGEEEIIC5LHvko2V8gYWCGEEEIIYVEkAyuEEEIIYUHkQRPJwAohhBBCCAsjGVghhBBCCAtiypPzAuQsycAKIYQQQgiLIhlYIYQQQggLIrMQSAZWCCGEEEJYGOnACiGEEEIIiyJDCIQQQgghLIgMIZAMrBBCCCGEsDCSgRVCCCGEsCBaptGSDKwQQgghhLAskoEVQgghhLAgMgZWMrBCCCGEEMLCSAZWCCGEEMKCSAZWMrBCCCGEEMLCSAZWCCGEEMKC6NwOIA+QDKwQQgghhLAokoEVQgghhLAgJpkHVjKwQgghhBDCskgGVgghhBDCgsgsBNKBFbnsnbHvUdOzNgnxCczz/YaLJy48VMa1WgXen9YfO3s7jvof5qfPF6RY//r7Len0aXf61epO7M0YAJ5vWJXOY3pgY2NDzM1oJr49Jlvi7fpZT9w965AQf4fvfL8h+Pj5h8q4VatA768GYmdvxxH/Qyz5bH7SumbdfWjWpTkmk4nAPw+xfOISKtasTM+JfQBQCtZ8vZKDf+zLUpxvje1BVc9a3I2/wxLfOVxKo17LVXOj67R+2NrbccI/kNWfLwTAoUgBen47mGJlSxAecp0f+s0gPvoWr/VqSb02LwNgbW2FS6WyDKvdk7ioW+Qv7MA7kz6k9HPlQGt+HObHhcNnMhX7801q8uaYbihrK/at/JPtfutSrLe2s+Gd6f0oW82NuMhYFvefyc2Q6zg4FqS732CeqVGR/T/vYO3YhQ/tu+c8X4o948wU748zFVty7cd2p6pnLRLi7/Cjrx8h6dRxl2l9k+r4588XAeY6fu/bj3AqW4KIkOvM7/c18dG3krZ7pkZFhq4dz8IBMzmyeR9lqpSn4xfvY18wP6ZEE3/M/oXDG/Zk+RzAXN9txnTDytqKvSv/5M806rvz9H6Uq+bGrchYlhj1/Wzj6rQY3gkbWxvu3b3H+gk/cXbPiWyJKSs+nTCdnX/vx6moI78u/S5XY+n22ftJ1ws/31npXC8q8mGy68Xiz35IWufdvQVeXZqjjevFsomLsbaxptfkfrhWq4i1jRW71gTw25w1TyX+GdPH0fz1V4mLj6dnz8EEHjn+UJmN65fiUsoZGxtr/vprPwMGjsJkMlGjRhXmfDuJAgUduHgxhC5d+xMTE/tU4kxu6rSxNPP2ID7uNr17+3L0yMNt8pffFuHiXBIbG2t27z7A4I/GYDI96BIOGPg+Eyd9QvlytQkPv/nUYxaZY1FDCJRSiUqpI0qp40qp1UophyzsK0ApVTcT2zkqpfpmoNyzSqlNSqkzSqnDSqlVSinnR5R3VUp1ftJ4njalVH+l1FmllFZKFc/OfdfwqI2LWymGefRn4Sg/un3ZK81y3b7oxcKRfgzz6I+LWylqeNRKWudUqhjVXnHnRsj1pGUOhR3oOv4Dvn5/EqOafcS3fadlS7zunrVxcSvNkCZ9+WGkH+990TvNcu99+SE/jJjDkCZ9cXErTU2P2gBUebEadb3qM6L5YIZ5DWLj3N8AuHT6Ip+29GWUzxAmdxtPzwl9sLLO/D/Nqh61KOnmwmceA/lp1Fw6fvl+muU6ffEBP438ns88BlLSzYUqHu4AePdpw+ndx/jMcxCndx/Du28bALbNXc9En2FM9BnGb1OWc2bfSeKizJ2ut8b24OSOI4xrOpgvm39M2NnQTMWurBTtxr3H3O6TmOw1lFqtXsK5UpkUZRp28CQ+KpYJHh+xY/5GWo4w/7O5d+cum79axboJS9Pcd3XvetyJu5OpuFKr4uFOCTcXPvcYxPJR8+j4Zc80y739xfssGzmXzz0GUSJZHXv1acPp3ccZ5/kRp3cfp1nf1knbKCtF6xGd+WdXUNKyu/EJLBkymy+b+TKn20TajelG/sKZvvylOFbbZPVdO436bpCqvt8w6vvWzRjm95zK1NeHsXzoHN6Z0S/L8WSHNj5efDf9i9wOA3fPOri4lWJwkz7MGzmHnl98mGa5977szbwRsxncpA8ubqVSXC/qeNVnRPOP+NhrIBvm/gpAgxYvYWNny3DvQYxqMZSmnb0pXrZktsff/PVXqVzJjeerNKZPn+HM/nZimuU6dv6QOnW9qOn+KiVKONG+/RsAfP/dVEZ9MoFatV/j11834zu0T7bHmFozbw8qVnKlZnVPBvQfydcz024HXd/tz4sNfahX15vixZ1o29YnaV2ZMqVo2vRl/vsvc9ewnGLKwVdeZVEdWCBea+2uta4GJAAprghKqZzIKDsCj+zAKqXsgY2An9a6sta6NjAHKPGIzVyBp96BVUpZP+EmfwOvARezO5bazerx99odAJwLPINDoQIUKeGYokyREo7YF3LgXKA5m/f32h3UblY/aX3n0T1YOXEJOtmkIg1bvcyh3/cRcfkGADHh0dkSbx2v+uxa4w/A2cB/cShcAMeSRVOUcSxZlPwF83M28F8Adq3xp64R72vvvs66OWu5l3APgOjwKAASbidgSjRfJmzz2YLO2gQpNZrVZd/anQAEG/VaOFW9Fi7hiH2h/AQb9bpv7U5qNqtn3t6rHnt/Nv9e9v68g5pe9R46Rt1WL3Fw3d8A2BfKT6X6L7B75Z8AJN5NJD46LlOxP+NeiRsXwwi/dI3Eu4kErt9NtWYpP2dWa1aX/WvM53d00z4qN6oKQEL8HS4cPM3dO3cf2q+dQz483m/B1m/WZiqu1Go0q8f+ZHWcPwN1vH/tTmok1XFd9hl1vO/nHdRIVsdNujfn6OZ9xBjtA+DahStcDw4DIOraTWLCoynoVDjL53G/viMeU98HjPoOSlbfoSeCib5mzk6F/RuCrb0d1na5f1Ovrnt1ihQulNthGNeLAOBx1wuHZNeLAOo2awCA17vNWTdnzUPXC7Qmn4M9VtZW2Nnn497du8THZO7f26O0bOnNjz/9DMC+/Ycp4lgEF5eHO8r3s6o2NjbY2dklXb6erVyBnbv2ArBt+y7efNPnoW2z2xtveLH8J/O/8QMHjlCkSGGcXR7+s5s8Zls7O3Sya+7kKaP59NNJKZaJvMnSOrDJ7QIqKaU8lFK7lFLrgJNKKXul1EKl1DGlVKBSyhNAKZVfKbVCKXVKKfULkP/+jpRSscl+bq+UWmT87KyU+kUpddR4NQImARWNTPDUdGLrDOzRWq+/v0BrHaC1Pm5kWncZWdnDxj4x9vuysd/BSilrpdRUpdQBpVSQUqq3EZOVUmqOUuofpdRWI8vb3ljX1DjnY0qpBUqpfMbyYKXUZKXUYWCE8f/751s5+fvUtNaBWuvgjP1KnkxRZyfCjU4mQERYOEVdiqUs41KMm1fCH5S5Ek5RZycAannV4+bVCC6dStm3dqlQGociBRix4nM+Xz+Fl9o2yZ54XYoRcTlZLGEPYkl+ThFhqeI1zsnFrTTP1a/CuF8nM3rlF1SoUSmpXEX3ykzZOpPJf3zN/E++S+rQZoajsxM3k9XrzbBwHF1Sxuno4kRksnq9eSUcR+NcCpUoQvT1SACir0dSqESRFNva2ttRpYk7gZvNf5yKlytJbHg0Xab1ZeTGybwzqTd2+fNlOvbIZHUcdSWCIqnquEiyMqZEE7dj4ilQ9NEdFp+hbxPww0YSbidkKq6H4yzKzWRxRqZbxxEPylyJwNHZ3IFJr46LOBelpnc9di3dmu6xy9esiI2tDTcuXs3yeRRJVd+RmazvGs0bEHL8AolGZ0uAk8vD1zenVHXrlOp6EX4lHCejHbm4leb5+lUY/+sUxiS7XuzbtJs7cbfxO7CQb/bMY8Pc37gVlf235suUdiHk0uWk96EhVyhT2iXNsps2/MSV0KPExMSyZs0GAE6e/JdWrbwBaN/uDcqVLZ3tMaZWqrQzISFXkt5fDr1C6XRi/vW3xVy4eJDYmFh++WUzAC3e8OLy5TCOHzv11GMVWWeRHVgj09ocOGYsqg0M0lo/C/QDtNa6OtAJWGxkRPsAcVrrF4CxQJ0MHGoWsENrXdM4xglgBHDOyASnN5CuGnAonXXXAC8jK/u2cQyM/e4y9jsD6AlEaa3rAfWAD5RSbkBbzNnaKkAX4EWjTuyBRcDbxrnbGOd8X7jWurbW+ksgSinlbizvATw8WPAJKaV6KaUOKqUOzp07N6u7eyw7ezta9mvL2ukrHlpnbW2Na/WKTO8xgaldx9NqwFs4u5V66jE9jrWNNQUdCzKmzXCWTVjMwDm+SevOHTnDMK9BfNpqGK37tjNnYvOKVJmIGq/V4fzB00nDB6ysrSlXzY1dS7cwscVwEuLv0KxPm9yINE2lq5Sn2DMlOfbHgdwOJX1GHbcb053fJi1LN/tTuIQjXaf3Z+nHfnkmQ+RcuSxvjOjM6lE/PL6wyDBrGysKOhZidJth/DRhMYPmmP/cVHSvjMlkom/99xjUuDctPmhNyXLpjk7LET5vvEPZZ2qTL58dr3q+BMD7vYbQp3c39u3dTKFCBUhIePjOSG5q07oblSrUJ18+O5p4NCJ/fnt8P+7LF+Nn5HZoGaJz8JVX5f79nieTXyl1xPh5FzAfaATs11rff4KiMfANgNb6H6XUReBZ4BWMzqLWOkgpFcTjvQp0NbZJxNzxK/roTR7LFvjW6EAmGrGlpRlQ4352FSgCVMZ8fqu11iYgTCnlb6x/Drigtf7XeL8Yc2f+a+P9ymT7/gHooZQagrkTXZ8s0lrPBe73XPXfE/5Is1zTLq/TpNNrAFw4epZipYtz/1EfJ5di3EyWjQBz9rBoqQdZWadSxbh5NYKS5V0oUdaZ8Zu/Stp23IapfN5mBBFh4cRGxpAQf4eE+Duc3n+SZ15w5eqFKzwpr67N8ezoBcD5oLM4lU4Wi4s5lhTxXo3AySVVvMY5RVy5wYHfzVnLc0fPoE2aQk6FiYl4MMTh8tkQbsfdpuyzz3Dh2LkMx/lKF29e6tQUgItHz1G0dHHgNGDOHEeGpYwzMiwCx2T1WrRUMSKNc4m5HkXhEo5EX4+kcAlHYm6kHIJRp+VLHFj3V7J9hRMZFk7wkbMAHN60F+9MdmAjr0bgmKyOi5RyIipVHUcZZaLCIrCytsK+UH5uGQ/vpcW19rOUq1GB0X99g5W1FQWLFaHfijHM7jjuiWJ7pUszGqWo4wdxOqZbxw8ybo6lnIi8ar7lnl4dP1OjAj2+GQhAwaKFqepRC1NiIkFbDmJfMD99Fo5g/bQVScMSsioqVX07PmF9F3Fxosf3Q1k2ZDbh/2U9I2zpvLo259WOzQA4H3SGYqUfPDbg5FKMiFR1G5HqelGsVDEijHYUcSWc/b+bH9RLfr14qfUrHA0IJPFeItHhUfx76BQValTi2qWs13+fD7vRs+c7ABw8eISy5R5kTcuULUXo5bB0t71z5w7r1m+hZUtvtm3fxenT52jewjwqrnLlCvg0b5rl+NLSq3cXuvfoCMChQ0GULfsgWVG6TCkuPzLmBDZs2MYbb3hx7ep1XMuXZc++TQCUKePCX7vX0+SVNly7eiPdfYjcY2kZ2PtjYN211gO01vfvB9565FaPl/xDhn0W9wXmTG16Gd7BwFWgJlAXsEunnAIGJDtfN631lizElLyO1mDOYL8BHNJah6e9Sfbb/uPvjPHxZYyPL4e37E+6vV+xVmXiY+KIMm6r3hd1PZLbMXFUrFUZgJfaNuHwlgOEnP6PAXXfw7dxH3wb9yEiLJwxb3xM1PVIDm/ZT+W6zxtjxOyo6F6Zy2dDMhXv1iWbGeUzhFE+Qzi4ZR8vt/MEoFKtZ4mPiSPyWsonVCOv3SQ+Np5KtcyfS15u58mhrfsBOLhlP1VerA6Ybw/a2NoQExFNiXIlkx7aKl6mBKUrluFGyLUninPnj38kPWAVtGU/Ddq+AoCrUa/Rqeo1+nokt2PicTXqtUHbVwjachCAoG0Hadje/Htp2L4JQVsfZC7tC+WncoMqBG09mGxfUdy8HE7JCuY/HM+/VJ0rZzJX35eOnqOEqwtOZUtgbWtNrZaNOLE15c2M41sPUb+d+fxq+jTg7O5HP/m+e+lWPmvQl/GNBzDrrc+4fuHKE3deAXb+uIVJPsOZ5DOcoC0HqP+EdVy/7SsEbTHX5bFtB2lg1HGD9k2S6vOzlwcwtrH5Fbh5LytHzydoy0Gsba354Puh7Fu7kyObszZDRXJp1ffxVPV9Yush6hn1XSNZfdsXduCDhcPZOHkZwYf+fWjf/4u2LtnMSJ/BjPQZbFwvPADz9SIu5lY614u4ZNcLj2TXi31pXi9uhF6naiPz8nz581Gp1nNcPpe5f2+p+X23mLr1mlG3XjPWrfuDLu+Y8ycN6tcmOiqasLCU16UCBRySxsVaW1vj07wpp0+bP8iWKGHumCulGDVyEN/P/TFbYkxt7vc/0qhhCxo1bMGG9Vvo9E5bAOrVcyc6OoarYddTlC9QwCFpXKy1tTWvv+7Jv/+e48SJ07i51qPqCy9T9YWXCQ0No3Gjlnm282pSOffKqywtA5sRu4B3gD+VUs8Cz2BORe3EPDb1T6VUNaBGsm2uKqVeMMq9CdxP52zHfBv+a+Php4LGusc9IbAMGKmUaqG13giglHoFiMCcSQ3RWpuUUt2A+w9Vpd7vH0AfpdSfWuu7xrmEYn6oqptSajHmh8I8jOOdBlyVUpW01mcxDy/YkVZwWuvbSqk/AD/MQxVyxVH/w9TwrM3UHbO5E3+HHz6enbRu3KZpjPEx32JfPHoeHxjTaAUFBBIUkO6QXQCunAvl2I4jfPH7dLRJs2PlNkL/vZTleI/8eQh3zzrM2OnHnfg7fO/7TdK6CZumM8pnCAALPv0+aVqcowGHOeJvjjdg1XZ6T+3P5C0zuXf3Ln5DzaNHnqv7Aq36tuXe3US0NrHw0++JeURG8XGO+wdS1bM2n++YRUJ8Aj9+PCdp3chNU5joMwyAFaN/oOv9KZ4CjnAiIBCALX6/0nP2YBp1eJWIUPM0Wve5e9fn1K6jJMSnfJp/1WcL6PH1QPPYzEvXWOI7h8wwJZpYM2YhvZeMwsrain2r/Ak7E8Lrg9/i0rHznNh2iH2r/Hlnej9GBXxNXGQsPw6YlbT96L++IV/B/NjY2lC9WV2+6zKBq5mcEeFRTvgHUtWzFmN3zORufAJLP/ZLWjdi02Qm+QwHYNXo+bw7rS+29racDDjCyQDzDaStfr/x3uyPeLGDJxGhN1jQ79G3LWu3eJFK9V+gQNFCSR8ufvSdQ+jJrD1baUo0sXbMQnoZ9b1/lT9X06jvzsnqe4lR3427elOsvDPNBrWj2aB2AHzfZQKx2fTQZGZ9PHYSBwKDiIyMpmmbd+nbswvtWnrneByBxvXi653fGdeLB+104qYZjPQZDMDCpOtFPo4EHOKIv/kDhP+q7Xw4tT9Ttszk3t17+A2dCcCWJZv5cNoApm6dBUqxY/V2/vsn25+xZdPm7bz++qucPvU3cfHxvP/+kKR1Bw9soW69ZhQo4MAvaxeSL58dVlZWBATsTuqodny7DX36dAfg1183sWjxyrQOk63++N0fb29Pgo4HEB8Xz4cfDktat3vvRho1bEGBAg6sWj2PfHb5sLJS7Ny5lx/m/fTUYxPZT+WVcVQZoZSK1VoXTLXMA/DVWr9hvLfH3DGrC9wDhmit/ZVS+TGP9awJnALKAP201geN2/STgevAQaCg1rq7Me3VXKAC5tv9fbTWe5RSyzB3gDenNw5WKfU85tv3FYG7QBAwCCiMOQOqgd+NGAoqpWwxd1qLYR7LOhP4AmiJORt7HWiDuaM7B3PH9ZKxbrLWeqtSqikwDfMHkwNGvHeUUsFAXa110kdJpVRD4GegvDE8Ir06HwgMA1wwj9/dpLVOe16mB3Q313aPKZJ3LA5eQ+fyb+Z2GBm27OIvAPR17ZDLkWTcnOBVDHbtmNthPJEZwSvo7/p2boeRYd8GmzsIQyyonqcHr+DujYfnRs3LbItXoFP5vDPG+3GWXzRPv2VjV+YxJfOOewmhFHRwy+0wMiw27gKY/xbnmEnl382xztuIi0vzZB7WojKwqTuvxrIAICDZ+9uYH0xKXS4eSPPKrrX+GXNnLvXyq0DrNJY/drorrfU/wOtprLpKyuzvcKP8XcxjbpMbZbxSt4RuAgAAIABJREFUUEr5aq1jlVLFgP0YD7NprbcDtVKX11q7phFHY2DhozqvxrazePCgmRBCCCFErrOoDqxIskEp5Yh5/Ox4rXX6o9TTYEwjVpGHO8xCCCGEyOMs59750yMd2CxQSlUHUo9Mv6O1bvA0j6u19sji9g/dLzc6tanv2QzXWqc9nYAQQgghRC6RDmwWaK2PAe6PLWgB0urUCiGEECLvMUkO1uKm0RJCCCGEEP/jJAMrhBBCCGFBMv9l4/9/SAZWCCGEEEJYFMnACiGEEEJYEBkBKxlYIYQQQghhYSQDK4QQQghhQWQMrGRghRBCCCGEhZEOrBBCCCGEsCgyhEAIIYQQwoKYVG5HkPskAyuEEEIIISyKZGCFEEIIISyIfJWsZGCFEEIIIYSFkQysEEIIIYQFkfyrZGCFEEIIIYSFkQysEEIIIYQFkS8ykAysEEIIIYSwMJKBFUIIIYSwIDILASitpRJEtpNGJYQQ4n9Jjn61wHDXTjn2d3Zy8PI8+bUJkoEVQgghhLAgkiWSDqx4Sga6vp3bIWTYrOCV+Lp2yu0wMmxa8HIABrl2zOVIMm5m8Ar6uHbI7TCeiF/wKotrx2B5//Y6lW+T22E8keUXf+XujfO5HUaG2RavAMCU8u/mciQZN+ziUj517ZzbYWTYF8HLcjuE/0nSgRVCCCGEsCAyC4HMQiCEEEIIISyMZGCFEEIIISyIzEIgGVghhBBCCGFhpAMrhBBCCCEsigwhEEIIIYSwIDKAQDKwQgghhBDCwkgHVgghhBDCgphy8JURSqnXlVKnlVJnlVIj0lifTym10li/TynlmonTTkE6sEIIIYQQIlOUUtbAbKA5UAXopJSqkqpYT+Cm1roSMAOYnNXjSgdWCCGEEMKC6Bz8LwPqA2e11ue11gnACqB1qjKtgcXGzz8DTZVSKit1IB1YIYQQQgiRJqVUL6XUwWSvXqmKlAEuJXsfYixLs4zW+h4QBRTLSlwyC4EQQgghhAXJya+S1VrPBebm4CEzRDKwQgghhBAis0KBcsnelzWWpVlGKWUDFAHCs3JQ6cAKIYQQQlgQEzrHXhlwAKislHJTStkBHYF1qcqsA7oZP7cH/tRaZ2k6WxlCIIQQQgghMkVrfU8p1R/4A7AGFmitTyilxgEHtdbrgPnAj0qps0AE5k5ulkgHVgghhBDCguS1b+LSWm8CNqVaNibZz7eBt7LzmDKEQAghhBBCWBTJwAohhBBCWJAMjk39f00ysEIIIYQQwqJIBlYIIYQQwoLk5DyweZVkYIUQQgghhEWRDKx46tqN7U4Vz1okxN/hJ18/Qk5ceKhMuWpuvDOtL7b2dpz0D2TN54sAcChSgO7ffoRT2RJEhFxnYb+viY++9cj9thrRmSqetQH445s1BG7YA8A70/pQqUEV4mPiAPjJdw6hJy9m+Dyea1KT1mO6YmVtxb6V/vj7pZzmztrOhk7T+1K2mhtxkbH82H8mN0NuULlxdVoM74i1rQ2Jd++xYcIyzu45kWLbHvN8KfZMSaZ5D8twPOl5vklN2o7phpW1FXtX/sm2NOJ8d3o/ylVz41ZkLIv7zyQi5DoAr/VtTcMOnpgSTaz9fBH/7AwCIH9hBzpO6k2p58qiNSwf9h3Bh8/Q7dtBlKxQyihTgPjoW0z1GZGl+DuM7UFV4/e6xHcOl9JoL89Uc6PrtH7Y2ttxwj+QVZ8vBKC2T0NafPQWLpXKMLn1KP47dj5pmzLPP0PnCb2wL5gfbdJMaj2Se3fuZirGF5rUpO2Y7lhZW7Fn5Z9s8/stxXqbpDquwK3IGBYlq2Ovvm2S6njN54v4Z+dRADpP+ZCqr9YmJjyaSd6+SfvyGdKB6l510VoTeyOKpb5+RF+7madjdvdpSPOP2uNcqQxftf6ES8l+D5nV7bP3cfesQ0L8Hfx8ZxF8/OF9ulWryIdfDcTO3o4j/odY/NkPSeu8u7fAq0tztMlE4J+HWDZxMdY21vSa3A/XahWxtrFi15oAfpuzJsuxPqlPJ0xn59/7cSrqyK9Lv8vx49/n1qQGTcd2QVlbEbQigH1+61Ost7azocX0D3Gu7kb8zRjW9f+W6JAbWNla4z2hJy413NAmE9s/X8qlvacA6LjiEwqWdOTu7QQAVneZTFx4dLbEW7lJDXyMa/Khlf7sTCPe9tP7UNq4Jq/sP4vIkBs4li3OoG3TuHH+MgCXAs+y7pMF2Nrb0XHOIJzKO2NKNHF6+2G2TF6RLbGK7CcZWPFUVfFwp4SbC+M9BrFy1Dw6fNkzzXIdvnifFSPnMt5jECXcXHjBwx2A1/q04d/dx/nC8yP+3X0cr76tH7nfKp61KFvVjSk+w5je5hNe/aAl9gXzJx3ntwlLmeIznCk+w5+o86qsFG+O68EP3Scz1cuXWq0a4Vwp5Vc9N+jgSXzULSZ5DGbn/E20GNEZgFs3Y1jQcxpfvT6cFUP96DSjb4rtqnnX407c7QzH8rg43xr3Ht93n8REr6HUbvXSQ3G+2MGT+KhYvvD4iID5G2lpxOlcqQy1WzZiYjNfvus2kbfG90RZKQDaju3GqR1HmNB0KFOaD+PqWfOXrCzuP5OpPiOY6jOCoM37CPp9f5bir+pRi5JuLoz1GMiyUXPp9OX7aZbr9MUH/DTye8Z6DKSkmwtVjfZy+fQl5n44jbP7T6Uob2VtRfcZA1j2yTzGNxvKjI6fkXj3XqZivF/H33WfyASvIdRp9RIuqeq4YYdXiYu6xXiPQQTM30Qro45dkup4KH7dJtBh/HtJdbzv5x34dZv40PH+nLueyc2HMcVnOMf/PMzrg9rl+ZivnL7E/A+/4lyq30NmuXvWwcWtFIOb9GHeyDn0/OLDNMu992Vv5o2YzeAmfXBxK0VND/MH2SovVqOOV31GNP+Ij70GsmHurwA0aPESNna2DPcexKgWQ2na2ZviZUtmS8xPoo2PF99N/yLHj5ucslK8Nr4bq7tNYf5rw3ihVUOKVS6dokz1tz24HXWLeU2GcnD+73iMME/lWbOTJwALvUey6t3JeH7aGZRK2m7DoDks9vmExT6fZFvnVVkpWo7rwZLuU5jl9THVWzWiRKo2XaeDB/FRt5jhMYTd8zfjPaJT0rqIi1eZ7TOK2T6jWPfJgqTlf83byMymvsxpMZJn6jxLZY+a2RJvdtM5+F9eZVEdWKVUolLqiFLquFJqtVLKIQv7ClBK1c3Edo5Kqb4ZKPesUmqTUuqMUuqwUmqVUsr5EeVdlVKdnzSep00p9ZNS6rRR5wuUUrZPsn31ZvXYv3YnAMGBZ8hfqACFSzimKFO4hCP2hfITHHgGgP1rd1KjWT3z9l512f/zDvPyn3dQ3aveI/frUrks5/afwpRoIiH+Dpf/ucgLTbJ+AXrGvRLhF8OIuHSNxLuJHFm/h6rNUjafqs3qcHCNOaagTfuo3KgaAJdPBCdlzML+DcHW3g5rO/PNDzuHfDR534ft3/yS5RgByrtX4vrFMMKNOA+v3031VHFWa1aX/UacRzft49lGVQGo3qwuh9fvJjHhHhEh17l+MYzy7pWwL5SfivVfYO9KfwAS7yYSHx330LHdW7zI4XW7sxR/zWZ12Wv8Xi8EnsHhEe3lgtFe9q7dSU2jvYSdC+Xq+SsP7feFl2sS+s9/hJ4yf2i5FRmLNmXuwmyu46up6rheijLVm9Vl/xpzuz2yaS/PGm2herN6HF6/m3tJdXyV8u6VADi3/xRxUbEPHe92bHzSz/kc7CETX16T0zFfPRfKtTR+D5lVx6s+u9YEAHA28F8cChfAsWTRFGUcSxYlf0EHzgb+C8CuNQHUbdYAAK93m7NuzhruJZg/tESHR5k30pp8DvZYWVthZ5+Pe3fvJt2hyUl13atTpHChHD9ucqXcKxIZfJWoS9cx3U3k1Pq9VPKqk6JMZa/aHF+zC4DTm/bzzEvma0exymW4uNt8VykuPJo70XG41HB7qvGWda9E+MWr3DTa9LH1e3ihWcp4X2hWl0Aj3hOb9lHBaNPpuXs7gQt7TgLm69zlE8EUcXF6OicgssyiOrBAvNbaXWtdDUgAUnwMN75f92lzBB7ZgVVK2QMbAT+tdWWtdW1gDlDiEZu5Ak+9A6uUsn7CTX4CngeqA/mBtFNi6SjiXJTIyw++7jgyLPyhC0IRFycir0Q8KHMlgiLO5j9OhUoUIfp6JADR1yMpVKLII/d7+dRFXmjijq29HQWKFqLyi1VxLFU8qVwL344M3zyFN0d3xcYu483loeNdCU+K8UEZp6QypkQT8TFxOBRN+UepRvP6hBy/QKLxh/T1oR3Y8cNGEm7fyXAsj47TKVWcERRxTlnfjs5O3EwW5+2YeAoULfTQtlHGtsXKlSQ2PJrO0/rw8caJdJzUC7v8+VLss2L954m5Ecn14LAsxW+O7UbS+5th4Timai+OLk5EXkn5u3B0fvQfGecKpUBrBiwZxcgNk/Dq3SpLMT5pW7gdE2fUcdEU55eR2AFa+L7N57tnU6d1YzZNX2URMWcnJxcnwpPFEBEWjlOqGJycnYgIe3CO4VfCcTLajotbaZ6vX4Xxv05hzMovqFDD3AHft2k3d+Ju43dgId/smceGub9xK40O+f+Cgi5FiUl2HY65EkEhl6IPlYm+bC6jE03ciYkjf9GCXD/5H5W8aqOsrShSrgTO1VwpXLpY0nbNp/Wi26YveXFgm2yLt7BzUaKStenoKxEUTtUmkpcxGfHevyYXLVeCvhsn0HPlaMrXe+6h/dsXduD5prU59/eJh9blBaYcfOVVltaBTW4XUEkp5aGU2qWUWgecVErZK6UWKqWOKaUClVKeAEqp/EqpFUqpU0qpXzB3xjDWxSb7ub1SapHxs7NS6hel1FHj1QiYBFQ0MsFT04mtM7BHa500IEdrHaC1Pm5kWncZWdnDxj4x9vuysd/BSilrpdRUpdQBpVSQUqq3EZOVUmqOUuofpdRWI8vb3ljX1DjnY0a2NJ+xPFgpNVkpdRgYYfz//vlWTv4+Na31Jm0A9gNl0yqnlOqllDqolDo4d+7c9HaXdY/JPv2zK4iT/oEMXjuebrMGEnz4DNpk/ie4fvJyvmw6mK9aj8LBsSCvfdj66cWZBufKZfEZ0Zk1o8zj8kpXKU+xZ5w5/sfBHI3jSVlZW1O2mht/L93K1BYjSYi/w2t9UtZd7VYvZTn7+jRZWVtTsd7zLBj0DdPaj8Hduz7PPSYbk5dsnLaSsY36cei3v3i52+u5HY7FsbaxoqBjIUa3GcZPExYzaM7HAFR0r4zJZKJv/fcY1Lg3LT5oTcly6d4oE+kIWrWD2CsRdF0/nlfHvEvo4TOYEs3X3Q2D5rDQeyTL3xpP2XrPUbVt41yOFmKuRTK10UDmtBjF5vFL6TCzP/mSDTWzsraiw6z+7Fn0OzcvXcvFSMWjWORDXEamtTnwu7GoNlBNa31BKTUU0Frr6kqp54EtSqlngT5AnNb6BaVUDSDdTlsys4AdWus3jcxlQWCEcSz3R2xXDTiUzrprgJfW+rZSqjKwHKhr7NdXa/2GcY69gCitdT2jI/q3UmoLUAdztrYKUBI4BSwwsr6LgKZa63+VUkuMc/7aOG64kQlGKfWaUspda30E6AEsfFxFGEMHugCD0lqvtZ4L3O+56pDGzQD47+g5HJN9End0KUZUWESKbaPCInAs9eCTs2MpJ6Kumm+5x1yPonAJR6KvR1K4hCMxN8zjp6Ku3kx3v1tm/8KW2eZb8l1nDuCaMVD/fib3XsI99q0O4NUP3njcaT+IMfXxShVLivFBmQgcS5vjsLK2In8hB+JuxgDmLHP374ewYsgcwv8zXxDL165M2RoVGPXXLKysrShYrAh9VozGr+P4DMf1cJwRqeJ0IupqyvqOvBpB0WRx2hfKz62bMQ9tW8TYNjIsnMiwCC4eOQvAkU37eK3PgwymlbUVNb3rMbXlqEzF3KSLNy91agrAxaPnKFq6OHAagKIuxYhM1V4iwyJwLJXydxGZ6hxTiwwL5+z+U9wyfh/H/QN5ppobp3cff+J4Ix+q4/TbQmRSHTsYdXzTOL+Mx57cwV930XvhSDbPWG0xMWeWV9fmvNrRfB05H3SGYslicHIpRkSqGCKuRuDk8uAci5UqRoTRdiKuhLP/d/PDnOeOnkGbNIWcCvNS61c4GhBI4r1EosOj+PfQKSrUqMS1S1ef9unlObFhNymU7DpcqJQTMWE3HypTuLQTsWERKGsr8hVyIP6mOf/z5/ifksq9s3YMNy+Yh5DEGu0s4dZtTv22m1LuFTix9q8sxxt99SZFkrXpwqWciE7VJu6XiTbadL5k1+T4BHPcl49fIOK/qxRzc+HyMfMDo60nvk/4hTD2LPidvCovj03NKZaWgc2vlDoCHAT+A+Yby/drre8/qtwYWAqgtf4HuAg8C7ySbHkQEJSB470K+BnbJGqto7LhHGyBeUqpY8BqzB3RtDQDuhrnuw8oBlTGfH6rtdYmrXUY4G+Ufw64oLX+13i/GPM537cy2c8/AD2MTvnbwLIMxD0H2Km13pWBskkPSgVtOUD9tuYwXGtV5nZMXFJH8r7o65HcjonHtVZlAOq3fYVjWw4AcHzbQeq3b2Je3r4Jx7aas5XHth5Mc7/KSuHgWBCA0s8/Q+nny/PPLvOvOvlYyhrN6nHl30sZORUALh09R3FXF5zKlsDa1hr3li9yYmvKzygnth6ibjtzTDV8GnDWGBNmX9iBnguHsXHycoIP/ZtUfs/SbYxv0JcJjQcy+63PuHHhSpY6r2D+wFAiWZy1WzbieKo4j289RH0jzpo+DThjxHl86yFqt2yEtZ0NTmVLUMLVhYtHzhJzPYrIy+FJsw08+1I1ws6EJu3v2cbVuXr+8kMfTDJqx49/MMFnGBN8hnF0y34aGr9Xt1qViX9Ee3Ez2kvDtq9wdMujs9gndxyl9HPlsLW3w8raimcbvMCVMyGZijetOr7fLu87vvUg9duZ2627T8OkOj629SC1WzbCJlUdP0oJV5ekn6t71ePaudBHlM4bMWeHrUs2M9JnMCN9BnNwyz5ebucBQKVazxIXc4vIVDMxRF67SXxsHJVqPQvAy+08OLTV/FDhwS37qPJidcA8nMDG1oaYiGhuhF6naiPz8nz581Gp1nNcPpe5dmHprhw9T1E3F4qUK4GVrTUvtGzI2a0p8zxntx2mWruXAXjOpz7/7TaPF7Wxt8PWGFZUvnE1TPdMhJ+5jLK2In9R8/XYysaaik1rceN09tRv6NFzFHN1oajRpqu3fJF/Ul3r/tl6iFpGvFV9GnDeaNMOToWSHkQsWq4kxVxduGkkFl4b+hb2hRzYNO7HbIlTPD2WloGNT535VOYnHW9lcb/JP8rYZ3FfACeAJumsGwxcBWpi/gCR3uPnChigtf4jxUKlfDIZU/I6WgOMBf4EDmmtw9PeJOmYYzGP3+39pAc96R9IVc9ajNkxk4T4BH762C9p3bBNk5niMxyAVaPn8860vtjZ23Iy4AgnA44AsNXvN3rM/oiGHTy5GXqDhf1mPHK/1rY2fLT6c8D88MuPg79JupXVdeYACjoVBqUIPRnMyk/mZfg8TIkmfhmziA+WjERZW3FgVQBXz4TgPbg9l45d4OS2Q+xfFUCn6X0ZETCDuMhYlg74BoCXunpTvLwzXoPa4jWoLQDzukwkNpuexk0d55oxC+mzZJR5Gq1V/oSdCaH54Le4dOw8x7cdYu8qf96d3o9PA74mLjKWxQNmARB2JoTADXsYtfUrEu8l8vOYhUkPOq35bCFdvu6Pja0NNy5dY5nvg6l+ardslG3DB477B1LNszbjdswiIT6BJR/PSVo3atMUJviYpxlbPvoHuhnTrp0IOMKJgEAAanrX4+3P3qOgU2H6LRhByKlgvuk6gbjoW2z/YSMj1k0ErTnuH8hx/8BMxWhKNPHzmAX0TarjAMLOhOAz+C3+M+p4zyp/ukzvz+iAmcRFxrJowEwgdR2bWD1mQVIdd5s1kEoNq1CwaCHG7ZnDphmr2bvKn5bDO1OyQmm0ycTN0BtP1G5zK+Ya3vVo/1kPCjoVpveC4YSeuohf1wmZqm+AwD8P4e5Zh693fsed+Dt87zsrad3ETTMY6TMYgIWffm9Mo5WPIwGHOOJv7tD4r9rOh1P7M2XLTO7dvYffUPO5bVmymQ+nDWDq1lmgFDtWb+e/fzI+O0l2+XjsJA4EBhEZGU3TNu/St2cX2rX0ztEYdKKJbWMW89aSYShrK46t2kH4mVAaD2lHWNAFzm47TNDKHbSY8SEf7PiK25GxrOv/LQAOxQvTYclwtDYRE3aTjYPN12MbO1ve+nE4VjbWWFlbEfzXCY4u939UGBlmSjSxYcwiui0ZYZ5Ga1UA186E0nRwe0KPneefbYc5tCqA9tP7MjhgOvGRt1hpXJNd6z9P0yFvYbp3D23S/PbJAuKjblHYxQmPAW9y7WwofTd+CcDexVs4tDIgW2LOTnl5bGpOUToTT7TmFqVUrNa6YKplHqS89T4EqKq17mkMHdiKOQPbD6iitX5fKVUNOAI01FofVEqdBVpivm+5GojRWndXSq0A9mqtv042hMAGOKy1Lv+IOPMDx4BBWuuNxrJXgAigJxCitf5KKdUDWKC1VkqpOsB0rXUTo3wvwAd4S2t91ziXUGNZN6AV5k7lKaAXsAH4F3hVa33WGMcbqLWeqZQKBupqrZOeglBKfQO0A3pqrTc/4lzeB97DPDQhPr1yqeiBrm9nsGjumxW8El/XTo8vmEdMC14OwCDXjrkcScbNDF5BH9cOuR3GE/ELXoWltWPA4mLuVD77HuzJCcsv/srdG1mf1zan2BavAMCU8u/mciQZN+ziUj51zXOT8qTri+BlYE465Zhuru1yrPO2OHhNjp5bRlnaEIKMmANYGbfoVwLdtdZ3MA8FKKiUOgWMI+UY1RGYO4C7geRzvwwCPI19HcLcAQ7HPB71eHoPcRkdvTeAAcY0Wicxz1xw3Yivm1LqKOan++9nRoOARONhscGYb/OfBA4rpY4D32PuPK8BQox1SzGP5Y3SWt/GPJ51tRGvCXjUjNg/GWW2PKIMxj6cgT3GA2ZjHlNeCCGEEE+RSesce+VVFjWEIHX21VgWAAQke3+/I5e6XDyQZspKa/0z8HMay68CDz2qrrV+7EdDY/xtWo8LXwVqJHs/3Ch/F/OY2+RGGa8UlFK+WutYpVQxzDMDHDP2sR2olUYsrmnE0RhYqLVOfMx5WFQbEUIIIcT/f9I5sUwblFKOgB0w3niYK8OMacQq8nCHWQghhBB5XN7Ni+Yc6cBmgVKqOpD6UcU7WusGT/O4WmuPLG7/ZuplRqc29VenDE/9EJkQQgghRG6TDmwWaK2PAY+aD9ZipNWpFUIIIUTeY5Ic7P/Lh7iEEEIIIcT/Y5KBFUIIIYSwIPJNXJKBFUIIIYQQFkY6sEIIIYQQwqLIEAIhhBBCCAsiXyUrGVghhBBCCGFhJAMrhBBCCGFBZBotycAKIYQQQggLIxlYIYQQQggLItNoSQZWCCGEEEJYGMnACiGEEEJYEJmFQDKwQgghhBDCwkgGVgghhBDCgmgtY2AlAyuEEEIIISyKZGCFEEIIISyIzAMrGVghhBBCCGFhJAMrhBBCCGFBZBYCUDIQWDwF0qiEEEL8L1E5ebCWz7yRY39n1/+3IUfPLaMkAyueiuGunXI7hAybHLycwa4dczuMDJsRvALA4mJ+z7V9bofxRBYE/8xA17dzO4wMmxW8EoD+FhTzt8ErsbErk9thPJF7CaFMKf9uboeRYcMuLgXg7o3zuRxJxtkWr8DuUu1yO4wMa3RlTY4fU76JS8bACiGEEEIICyMdWCGEEEIIYVFkCIEQQgghhAWRabQkAyuEEEIIISyMZGCFEEIIISyIzCAlGVghhBBCCGFhJAMrhBBCCGFB5IsMJAMrhBBCCCEsjGRghRBCCCEsiHyRgWRghRBCCCGEhZEMrBBCCCGEBZF5YCUDK4QQQgghLIxkYIUQQgghLIjMAysZWCGEEEIIYWEkAyuEEEIIYUFkDKxkYIUQQgghhIWRDKwQQgghhAWReWAlAyuEEEIIISyMZGCFEEIIISyISWYhkAysEEIIIYSwLNKBFUIIIYQQFkWGEIhc82yTmrQa0xVlbcWBlf4E+K1Lsd7azoa3p/elTDU34iJjWdZ/JjdDblC2ZkXaTXzfXEgptn39Myf+OAiAfWEH2k/qhfNzZUHD6mHf89/hM9kS7/NNavLmmG4oayv2rfyT7WnE+870fpQ14l3cfyY3Q67j4FiQ7n6DeaZGRfb/vIO1YxcmbePj+zZ1276CQ5ECjKjaPVvifJox29rb0X3ORxQr74xONHFi+2E2TF6e7XF3Hvse1T1rkRCfwHzfb/nvxIWHypSvVoGe0/pha2/HMf9Aln2+AIA3h3TE3aseWpuIvhHNAt9vibx2E4DnGlal05juWNvYEHszmslvj81yrC80qUnbMd2xsrZiz8o/2eb3W4r1NnY2vDu9H+WqVeBWZAyL+s8kIuQ6AF5929CwgyemRBNrPl/EPzuP4liqGF2m96NQ8SJordm9fDs7Fm7OcpwA7cd2p6pnLRLi7/Cjrx8hadRruWpudJnWF1t7O074B/Lz54sAcChSgPe+/QinsiWICLnO/H5fEx99K2m7Z2pUZOja8SwcMJMjm/cBULR0MTpP6k3R0sXRWuPXY1LSuWfVjOnjaP76q8TFx9Oz52ACjxx/qMzG9UtxKeWMjY01f/21nwEDR2EymahRowpzvp1EgYIOXLwYQpeu/YmJic2WuO5za1KDpmO7oKytCFoRwD6/9SnWW9vZ0GL6hzhXdyP+Zgzr+n9LdMgNrGyt8Z7QE5cabmiTie2fL+XS3lMAdFzxCQVLOnL3dgIAq7tMJi48OlvjzohPJ0xn59/7cSrqyK9Lv8vx46fF0dMdt3E3l2tVAAAgAElEQVTvgbUV15ZtJ/TbX1KsL9W7Jc6dm6LvmbgbHsW5IXO4Y7TF8p+8S9HX6gBwacZqwtftzvH4n5QMIJAMrMglykrRZlwPFnSfzHQvX2q2akTJSmVSlKnXwZP4qFtM9RjMX/M30XxEZwCunr7ENy0/YabPSBZ0nUTbL9/HytrclFuN7cbpHUf5qqkvXzcfzrWzodkWb7tx7zG3+yQmew2lVquXcE4Vb8MOnsRHxTLB4yN2zN9ISyPee3fusvmrVaybsPSh/Z7YfoivW3+SLTHmVMz+8zYwqelQprUYgVud53jewz1b467uUQtnt1KM9BjA4lHf0fXLXmmW6/LFBywa+R0jPQbg7FaK6h61ANg89zfGNh/KZz4fE/TnIVoOeguA/IUd6PJ/7N13eBTV28bx75NA6JAEIqFJR0W6gIgNpGkUUUCsYMGfCooIgiICdsWGYgE7CjZQbCgqoARQpHdQOkgLLR1CS573j5mETUgDQmbX9/lcV65kZ89u7plsZs+eOeWZu3nj7hcZ3nEAY/q+etpZJUi44em7eOeOF3i+w0AuuPZiIk84xldwMOEAz7TpT/SHU7nWPcaRdarQrHNrXuj4MGNvf54ez9yFBAlpx1L59tkJPN/hYUZdP4xLe3Y84TlPRf02TYioGclTbfrzxdD3uem53tmWu/HZu/n8sfd4qk1/ImpGUt/9+3bocx1r567i6bYPsXbuKjr27ZLpOHQZcgv/zFmR6bl6jbqf396bwrPtB/Jyl6Ek7Us47f0AuOrKK6hbpybn1r+EPn0e5e23Xsi23E233McFzTvQuMkVRESE0737NQC8+87LDH38eZo2a8933/3MoIf7FEiudBIktH/mdr66/SU+bP8I513bivJ1K2cq0/DGNhxKOMD7lz/Mog9/oc2QmwBofHNbAMZ1eoxJt71I22G3gEjG437sP4ZPoh7nk6jHPam8AlwX1YF3Rj3rye/OVlAQtZ7/H2tufY5llz9EhesuoUS9qpmKHFi5mRVXPsLydgPZ/+M8qg/rCUBYu2aUaliLZe0fZkXUEKr06UJw6RJe7IU5SQFVgRWRVBFZJiKrROQrESl5Gs8VLSLNT+FxoSLSNx/l6onIVBFZLyJLRGSSiFTMpXwNEbnlZPOcaSLyoYgsF5EVIvK1iJQuiOet1qQO+7fGELttD6lHU1k+5S/qd8z85zi/4wUsnjwbgJVT51OndQMAjh46QlpqGgBFihUlvS978TIlqNnyXBZOnAlA6tFUDiUeLIi4nN2kDvu2xrDfzbt0ylwaZMnboGNzFrh5l0+dT93W5wNwJOUwmxet5ejhoyc879alG0jcG18gGQsj89FDR9jw1xrAOb7bV28mNDK8QHM37diCud9EA7Bp6XpKlilJuYjQTGXKRYRSokxJNi11WtfnfhNN044tADiUnJJRLqRkMdJfIK2uvZTFv8wnduc+AJIK4M2/epM67N26O+MYL5kyl4ZujnQNOzZnweRZACybOo967uu4YccWLJkyl2NHjhG7fS97t+6mepM6JO6Nz2gZPXzgELs37qBcARzjRh1bsOAb52+9Zel6SpQpRdksx7VsRCjFy5Rgi3tcF3wzm0bu/jTq0Jz5Xzv7Mf/rWTTqcHw/L7/jKpb/PJ+k/ccrqJF1qhAUHMw/f6wE4MjBwxkth6erc+dOTPjsayfLgiWUCy1HZORZJ5RLb1UtUqQIISEhGeeKenVrMXvOPABm/DaH66+PKpBc6So1qU38lt0kbNtL2tFU/p4yjzodLshUpm6HZqyaPAeAtVMXcPbFzv9e+bpV2Dp3NQAH9ydyOPEgkY1qFmi+09W8SUPKlS3jdYwMpZvWIWVLDIf/3Y0ePca+7/8gvFPm/8PEuatIS3Fef8lL1hFSqTwAJepVI3H+GkhNIy3lMAfWbCW0bdNC34eTlYYW2pe/CqgKLJCiqk1UtQFwBLjP904RKYwuEaFArhVYESkO/ASMVdW6qtoMGANE5PKwGsAZr8CKSPBJPmSAqjZW1UbAv8ADBZGjXMUw4nfuz7idsGs/5SqGZSpTtmI4CW6ZtNQ0DiUdpGSYc9Ks1qQ2A6e9zIBfX+LbYR+QlppGWLWzOLA/kRteuY8Hf3qBbiP/R9ESxQoiLqEVw7PkjaVcxcyVinI+ZZy8KZQK8+4kf6YzFy9bkvPbNWP9nydeuj0dYRXLE+uTOzYmlrDI8pnLRJYnbpdPmV2xhFU8XqbroJt5Ze47tOpyKd+NmghAZK1KlCpXmke+fIoRU16kddfLTztr1mMcn83r+MRjfJBSYWUoVzGMOLcynf7Y0Cx/n/CqEVSpX5OtyzYUQNYw4nyzxuw/4cNHaGQ48btifTLFEuruT5mIchkfthL3xlMmopy7f2E07tSCOZ9Oz/RcZ9WqREriAe5+52Ee/Wkk1z12KxIkFIQqlSPZvm1nxu0d23dRpXJktmWn/vgZu3YsJykpmcmTfwRgzZp1XHttJwC6d7uGalUrZ/vYU1U6Mowkn+OYtCuWMpFhJ5RJ3OmU0dQ0DicdpERYafau+Zc6HZohwUGUqxZBxQY1KFv5+Gv7qlfu4fapz3HRg9cVaOZAViwynCM7jv8vHdkVS0iWc4avs25uR/zMJQAcWLOF0DZNCSoRQpHwMpS7uAHFKuf8WOM/Aq0C62sOUEdE2ojIHBH5AVgjIsVFZJyIrBSRpSLSFkBESojIlyLyt4h8C2RcIxCRZJ+fu4vIx+7PFUXkW7cFcrmItAZGArXdluCXc8h2C/CXqmZ0elLVaFVd5ba0znFbZZe4z4n7vJe6zztARIJF5GURWei2ft7rZgoSkTEi8o+ITHdbebu797Vz93mliHwkIsXc7VtE5EURWQIMcb+n729d39tZqWqiW07cY+YXH8e2LdvIqI6Deevax2nbpwtFihUlKDiYyg1qMu/T6bxx9WMcSTlM2z7Xeh31PykoOIhebzzI7I9/Yf+2PV7HOcE3r3zBoNb3Me/7OVxx+5UABAUHU71hLV6/83lG9XqWzv26U7FmJY+T5iykZDF6jx3IN09/kqlV2W+4zZndRtzB9yM/R7NM6xMUHEztFufx7XMTePnaoVQ4uyKturcp9JhR19xK1bObUaxYCFe0vRiAu+8ZSJ97b2f+vJ8pU6YUR46ceHXEKysmzSJ5Vyy9pjzDFSNuY8eS9RlXnH7sP4ZxnR7jixueoWqLczi/6yUepw08FbpdRunGtdkxxumrnjBrOXG/L6HhD89Tb8wAkhavRdPSPE6ZN2uBDdBBXG5L61XAL+6mZkADVd0sIg8DqqoNReRcYJqI1AP6AAdV9TwRaQTkWGnz8QYwS1Wvd1suSwND3N+VW8e/BsDiHO7bA3RQ1UMiUhf4AmjuPu8gVb3G3cd7gARVbeFWRP8UkWnABTittfWBs4C/gY/cVt+PgXaquk5Exrv7/Lr7e/e7LcGISHsRaaKqy4A7geOjirIhIuOAKGAN8HAOZe4B7gF49913c3s6ABJ2xxHq8ym3XKXyJOyOy1QmcXcs5SqXJyEmlqDgIIqXKcnBuKRMZfZs3Mnhg4epWK8aCTH7SYiJZduyjYDT7aBNny4UhPjdsVnyhpOwOzZTmQS3zPG8JTiQJW9hOpOZe7zwP/Zu3sXsjwpmcNEVPa/kspvbAbB5+UbCfXKHR4YTF7M/U/m4mP2EVfIpUymcuN2ZywDM+24OD40byvevTSIuZj/J8UkcSTnMkZTDrFuwhmrn1WD35l2nnDvrMQ7N5nWcfozjfV7HB+KSSNgdR1jlCpkeG+/+fYKKBNP7nYdZ9N0frPh1wSnnu6xnR1q7x3Xr8o2E+WaNdDJl2p+YWEIrHW+VDa0UTry7P0l7EygbEUri3njKRoSStM/pgnF2o1rc+eaDAJQOK8v5bZqSlppKfEws2//ekvEBZ/m0hdRsWpe/Js08pX3pc9/t9O59KwCLFi2jarXjraZVqlZix86YHB97+PBhfpgyjc6dOzHjtzmsXbuRq652LnjVrVuLqKvanVKmnCTHxFHG5ziWqRROUkzcCWXKVg4nOSYWCQ6iWJmSpMQ5bSm/P/NZRrlbvxlBnPsaTXb/FkcOHOLv7+dSqUktVn/zR4FmD0SHY2IJqXL8fymkUjhHYk48H5S7tBFV+3dj9fXD0SPHMrbvGD2ZHaMnA1D37YdI2Xjq5wRTeAKtBbaEiCwDFuFczv7Q3b5AVdOH014CfAqgqv8AW4F6wGU+21cAmUcbZO8KYKz7mFRVLYgRCEWB90VkJfAVTkU0Ox2BXu7+zgfKA3Vx9u8rVU1T1Rgg/d3gHGCzqq5zb3+Cs8/pJvr8/AFwp1spvxH4PLfAqnonUBmnsnxjDmXeU9Xmqtr8nnuyH3Tja/vyjZSvEUlY1QiCiwbTuPNF/D09c51/zfTFXNDN2YWGURey0e0XFlY1ImPQVmiVCpxVuzJx2/eSvDeBhJ37qVDLaVWrc3ED9qzfnmeW/Ni2fCMRNSIJd/M27dya1Vnyrpq+mJZu3sZRF7LBzeuVM5X5qod7ULxMSb57enyBZf19wi88GTWYJ6MGs3TaAlp3bQNAraZ1OZh0kIQs/YQT9saTknSQWk3rAtC6axuWTlsIwFk1jl9KbtqhBTEbnYF8S6ctpG7z8wgKDiKkeAg1m9Rl14bTe338m+UYN+vcmpXTF2Uqs2r6Ilp2c7orNIlqxXr3GK+cvohmnVtTJKQI4VUjiKgRmdFV4JYX72P3hh3M/PCn08o3e8I0RkY9ysioR1kxbSEtuzp/6xpN65KSdPCE/teJe+M5lJRCDfe4tux6GSvc47pyxiIu7O7sx4XdL2eFu59PXtqPJy5xvpb+PI+Jwz9kxbRFbF2+gRJlS1E63OmSck7rBsScxv/j2Hc+oXmLjjRv0ZEffviVnrd2d7K0bEZiQiIxMZmvBJQqVTKjX2xwcDBRV7Vj7Vrn+EZEOBV5EWHoY/15970Jp5wrO7uWbyKsZiTlqkUQVDSY8zq3YsP0zG0mG2YsoUG3SwE4J6ol/851+pYXKR6S0fWp+iUNSDuWxv71O5HgIEqEOUMQgooEU7tdU/atLZjzW6BLXraBEjUrUazaWUjRIlTocgmxv2b+PyzVoCa1X7qXf24fyVHf/u9BQRRxj2vJ86pTqn514mctK8z4p0RVC+3LXwVaC2xK1pZP58o2B7Ivnm++f6Hip/lcAKuBnDrYDQB2A41xPkAcyqGcAP1U9ddMG0VOdbSB7zGaDDwB/A4sVtUTP6pmoaqpIvIl8Ah5tNjmR1pqGt+P+Jje4x8jKDiIhZOi2b1+Ox0GdGf7ys38PWMxCydFc+OovgyOfo2U+GQ+7/cmADVanEPbPl1IPXYMTVO+Hf5RRsvs909+zM2vP0Bw0SLEbtvNV4Pybg3Ob97JI8Zx7/ihBAUHMX/STGLWb+fKATewbeUmVs9YzPxJM7l11P0MjX6dg/HJTOj3Rsbjh//xJsVKl6BI0SI07Nicd3o+z+4NO+g85BaadbmYoiVCeOKvt5k3cSa/vv6132Y+lJxCx35d2b1hBw//5Iz8nvPJr8yfeGqtatlZMXMJjdo2Y+SstziScpiPBo/JuO/JqS/zZNRgAD4d/gF3vXI/IcVDWBm9lJXRSwHo/uhtRNaqjKYp+3fsZfzj7wGwa+MOVs1aytO/vEpamjJn4m/sWLfttLKmpabx9YiP6Ose43mToolZv52oATfw78pNrJqxmL8mzaTnqAcYHj2ag/HJfNxvNAAx67ez9Me/GDr9VVKPpfHViI/QNKVW83No2e0ydvy9lUemvgjAjy99wZro03tTXT1zKee3bcoTs0ZzNOUInw4em3HfkKkvMjLqUQAmDf+Q217pS9HiRVkTvSzj904f+z13vf0QF/VoS+yOfXx0/2u5/j5NU757bgL9PhuOiPDvqk38+eVvp7UP6ab+/BtXXnkFa//+k4MpKdx998CM+xYtnEbzFh0pVaok334zjmLFQggKCiI6em5GRfWmG6+jT587APjuu6l8/MnE7H7NKdPUNGaM+IQbxj+CBAexctIs9q/fwSUDuxGzYjMbZixhxcRZXP3affxv1qscik/mhwfeAqBkhbL0GP8oqmkkxcTx0wDn71QkpCg3THiUoCLBBAUHseWP1Sz/ouD+707G4CdGsnDpCuLjE2l33W307d2Tbp07eZIFgNQ0Ng39gPpfDEeCg9j95e+krNtGtcE3kbx8A3HTFlF9eC+CShXnnPeci4iHd+zjnztGIkWDafCdM6NCalIK6x4YDan+34XAgPhz7TorEUlW1dJZtrUh86X3gcD5qtrb7TowHacF9n6gvqreLSINgGVAK1VdJCIbgM7AWpxW0SRVvcOtsM1T1dd9uhAUAZaoavVccpYAVgL9VfUnd9tlQCzQG9iuqq+KyJ3AR6oqInIBMEpVL3fL34Nz2f4GVT3q7ssOd9vtwLU4g8L+xrl0/yOwDrhCVTe4/XiXqupoEdkCNFfVjF7uIvIm0A3orarZXgd2+73Wdp9PgJcBVHVQTvvu0kdr3JxHEf/x4pYvGFDjJq9j5NtrW74ECLjMd9Xo7nWMk/LRlq95sEa2Fxz80htbnErYAwGU+a0tEykScvpThBWmY0d28FL127yOkW+PbHWmwju6b5PHSfKvaIVazK3UzesY+dZ612RwGp0KTcvKlxda5W3BzlmFum/5FWhdCPJjDBDkXqKfCNyhqodxugKUFpG/gafJ3Ed1CE4FcC7g2/mlP9DWfa7FOBXg/Tj9UVflNIhLVVOAa4B+7jRaa3BmLtjr5rtdRJYD53K8ZXQFkOoOFhuAc5l/DbBERFYB7+JUnicD2937PsXpy5ugqodw+rN+5eZNA3KbYfozt8y0XMoI8In7fCuBSjjHzhhjjDHGMwHVhSBr66u7LRqI9rmdXpHLWi4FyLbJSlW/Bk64bququ4ETRgGpap7TXbn9b6/M5q7dQCOf24+65Y/i9Ln1NdT9ykREBqlqsoiUBxbgVC5R1d+AEyawU9Ua2eS4BBinqqm57EMacHFO9xtjjDGm8Kkfzw7gS0TCcRoTawBbgB6qGpelTBOcRsayQCrwnKrm2a/nv9gC+//Bj+7grjnAM+5grnxzpxHrBYw+E+GMMcYYY3CucP+mqnWB39zbWR0Eeqnq+TgNf6+LSGg25TIJqBZYfyMiDYGsw1cPq+qFZ/L3qmqb03z89Vm3uZXarMu9PJp1EJkxxhhjvBVA45e6AG3cnz/BuWL+qG8Bn9mTUNWdIrIHZ4xPrstUWgX2NKjqSqBgF4L3SHaVWmOMMcb8/+Y7z7vrPVV9L58Pr6iq6WOLYoCKefyulkAIsDGvJ7YKrDHGGGNMACnMFbLcymqOFVYRmQFkt5bz41meR0Ukx+AiUgnnqvbt7hicXFkF1hhjjDHGnBJVbZ/TfSKyW0Qqqeout4Ka7drjIlIW+Al4XFXn5ef32iAuY4wxxhhzJvyAM3c97vfvsxYQkRDgW2C8OytUvlgF1hhjjDEmgATQUrIjgQ4ish5o795GRJqLyAdumR7AZcAdIrLM/cpzfJF1ITDGGGOMMQXOXfypXTbbFwF3uz9/irMw00mxCqwxxhhjTAApzEFc/sq6EBhjjDHGmIBiLbDGGGOMMQEkUJaSPZOsBdYYY4wxxgQUa4E1xhhjjAkgaYGzlOwZYy2wxhhjjDEmoFgLrDHGGGNMALE+sNYCa4wxxhhjAoy1wBpjjDHGBBDrA2stsMYYY4wxJsBYC6wxxhhjTACxPrDWAmuMMcYYYwKMqPWjMAXPXlTGGGP+P5HC/GX1IpoX2vvsur2LCnXf8staYI0xxhhjTECxPrDmjHiy+q1eR8i3J7d+xmM1bvE6Rr69sOVzAAbVuNnjJPn3ypYvuKX69V7HOCmfb/2W/jVu8jpGvo3e8iUAD9a40eMk+ffGlomULlnT6xgnJfngZoYF0PniWfd8MbdSN4+T5F/rXZM5um+T1zHyrWiFWl5H+H/JKrDGGGOMMQHEBnFZFwJjjDHGGBNgrAXWGGOMMSaA2EIG1gJrjDHGGGMCjLXAGmOMMcYEEOsDay2wxhhjjDEmwFgLrDHGGGNMAFFN8zqC56wF1hhjjDHGBBRrgTXGGGOMCSBp1gfWWmCNMcYYY0xgsRZYY4wxxpgAojYPrLXAGmOMMcaYwGItsMYYY4wxAcT6wFoLrDHGGGOMCTDWAmuMMcYYE0CsD6y1wBpjjDHGmABjFVhjjDHGGBNQrAuBMcYYY0wASbMuBNYCa4wxxhhjAou1wBpjjDHGBBC1abSsBdYYY4wxxgQWa4E1fqHO5Y248omeBAUHseTLaP4YOyXT/cEhRbh+VB8qN6zBwbhkvn7gTeK37wOg4rnVuOaF3hQrXQJNU96/djjHDh8t8Iz1Lm/ENSN6ERQcxMKJM5mVTcYeo/pQpUFNDsYn8/kDbxC/fR9VG9fm+hd6AyAizHh9Mmt+XQRA6zuvpMVNbRERFn75O39+9EuBZj7n8sZ0cTPPnziTmWN/OCHzzaP6UtXNPOGB0cRt30fdSxpy9aM3EVy0CKlHj/Hj85+z4a/VAFw5qAfNu15GiXKlePz8Ows0L0CvJ3vTpO0FHEk5zDuD3mTLqk0nlKnZoBb3vvogIcVDWDZzMeOf/DDjvo53RNGx51WkpaWx9PfFfPHCeBpc0pibh/TM2J/Pnv+ENXNXnnLGcy9vTNcRtxMUHMS8ib8zI5vjetuo+6nWoCYH4pP55IHRxG7fC0D7vl1o1aMtaalpfPPUx/wzewUAI/54k8PJKaSlpZF2LJVXr30803O2vftqrhvWk6FN/8eBuKR8Z+32xB3Ub9uUIymH+WzQWLav3nxCmWoNanLrK30pWjyENTOXMvmpjwEoWa4Ud7z1EOFVI4jdvpdx979OSuKBPJ+3eOkSDJ3+KiumLeTrJ8Y5x6RoMN2fuou6reqjqvz48pf53ofsvPzKE3Ts1IaUg4e4995BLF+2+oQy337/MZEVz6JIkWDmzl3IgIdGkJaWlnF/vwfv5oWRj1O9WjP27487rTxZ1b28EVHu/97iiTOZnc35ovuoPlR2//cmuueL0KoV6D/jFfZt2gnAtqUb+OHxjyhaPISbxvQnvHpF0lLTWPvbEqa9eHrHMDehbZtQ8+m7IDiIPZ//xo63vs10f6V7O1PxlnbosTSO7k9g48AxHHZf49Ufv42w9hc4+V/7iv0/zD1jOfNr2POjmP3nAsLDQvnu03e8jnPabBota4H1CyKSKiLLRGSViHwlIiUL4DmfFJFBBZHvTJMgIeqZO/js9pd4u/0jNLj2IiLqVslUptmNbTiUcIA3Ln+YeR/+TPshNwMQFBxE19f78uPQjxjT4VE+vvFZUo8eOyMZr336Tsbd8RKvdRhM42tbc1adzBlb9GhDSsIBXmkzkD8+/Jmr3Iy7127j7c7DeDNqKON6vcj1z/UmKDiIivWq0uKmtozpMpw3rhrCuVc0o3z1igWa+fqn7+SDO17k5Q6DaHptaypmyXxhj7akJBxgZJsBzP5wKlcPuQWAA3FJfNT7FV698lG+fHgsN7/WN+Mxa35bwuguwwosp68mbZsRWbMyAy/vywePjeWuZ+/Nttxdz93HB0PGMPDyvkTWrEzjNs0AqH9RA5p3aMmQqwbwSIf+/PTe9wAkxSXy8l3PMaTTQ4wd+AZ9X+t/yhklSLjh6bt4946RvNDhYZpde/EJx/WiHm1JSUjm2TYPEf3hT3R2j2vFOlVo1rk1L3QcxDu3v8ANz/RGgiTjcW/d/AwvRw05ofIaWqk851zWKKMSnF/12zQhomYkz7Tpz8Sh79Pjud7Zluvx7N18+dh7PNOmPxE1IzmvTRMA2ve5jnVzV/Fs24dYN3cVHfp2ydfzRj3cgw0L/s60reMDXUnen8izVwzg+fYPs2F+5vtPRsdObahdpwaNG7al3wOP8froZ7Mt1+u2B7ioVRQtmneiQoVwunaNyrivSpVKtGt3Kf/+u+OUc+REgoTOT9/J+Dte4o0Og2l4bWsisrxGLnDPF6+1GcjcD3+mk3u+AIjdupu3o4bydtRQfnj8o4ztf7z/E6PbDWLM1Y9x9gX1qNumcYFnByAoiFrP/481tz7HsssfosJ1l1CiXtVMRQ6s3MyKKx9hebuB7P9xHtWH9QQgrF0zSjWsxbL2D7MiaghV+nQhuHSJM5PzJFwX1YF3RmX/OjGBySqw/iFFVZuoagPgCHBffh8oIsFnLlbhqNKkNrFbdhO3bS+pR1NZNWUe53S4IFOZczpcwLLJswFYM3UBtS4+H4DalzVk9z//svvvfwFIiU9G0wr+k2m1JnXYv3U3cdv2kHo0leVT/uK8jpkzntexOUsmzwFg1dT51G7dAICjh46Qluq0+hQpVpT0D84RdaqwbdmGjPs3z/+b869sUWCZz25Sh/1bY4h1My+b8hfnd2yeqcz5HS9gkXtcV0ydT103887VW0jc47RIxazbTtHiIQSHOBds/l26gaS98QWW09cFHVoyZ/JMADYsXUfJsqUIPSssU5nQs8IoUboEG5auA2DO5Jk079gSgPa3XckPY77h2BHnQ0zi/gQAtq7eTLy7P9vX/UtI8RCKhJzaBajqTeqwd2sM+93jumTKXBpmOa4NOjZngXtcl0+dT73Wzuu1YcfmLJkyl9Qjx4jdvpe9W2Oo3qROnr/z+uG9+OGFz06611vDji1Y8I2TY8vS9ZQoU4qyEaGZypSNCKV4mRJsWboegAXfzKZRR+d12LBDcxZ8PcvZ/vUsGnZokefzVmtQkzIVQvlnzopMv6fVDW2YPuY7wGk9OplW5KyuuaYDX3z2DQALFy6jXLmyVIyMOKFcUlIyAEWKFKFoSEimVqsXXxrOsGEjz0hLVtUs54uVOZwvlrrni9VT51PL/d/LydFDR9j81xoAUo+msnP1FspFhhd4doDSTeuQsiWGw//uRkU7+n4AACAASURBVI8eY9/3fxDeKfO5KXHuKtJSjgCQvGQdIZXKA1CiXjUS56+B1DTSUg5zYM1WQts2PSM5T0bzJg0pV7aM1zEKTBpaaF/+yiqw/mcOUEdE2ojIj+kbReQtEbnD/XmLiLwoIkuAG0TkShFZIiLLReQ3n+eqLyLRIrJJRB70ea7vRGSxiKwWkXvcbcEi8rHbCrxSRAa422uLyC9u+Tkicm5B73DZyHASd+3PuJ24K5aykWFZyoSRuDMWgLTUNA4lHaRkWGnK16yEKtw2/lHu/elZLr73moKO5/z+imEk7MycsVzF8BPKxLtljmd0TpjVmtTmoWkv0f/XF/lu2Iekpaaxe+02arY4l5KhpSlaPIRz2jahnPsmUBDK+eQBiN+1n3IVw7KUCc+UOcUnc7pGV7Vk+6rNpB4p+JbtrMIiyxPrkzk2Zj9hWY5zWMVwYmN8yuzaT1ikc9wia1bmnJb1efq7Fxk+8VlqNTqxctgy6iK2rNqUUck9Wb7HDCA+m9dCaMVw4jK9FlIoFVbmhMcm+D5WlT4ThjJoyvNcdHO7jDINOlxAwu5Ydrof0k4ua5bXQMz+Eyo95SLDid8Vm2V/nNdJmYhyJLofVhL3xlMmolyuzysiXDesJ98/NyHT7yhR1rmodPXDPRj840jufHsAZSqUO+n9SVepckW2b9+VcXvnjl1UrhyZbdnvvv+EzVsXkZyUzLff/uzkuKYDO3fGsGrlqbcC5ya780XZbM4XCT6vkcM+/3th1SLo+9Pz9J44nOotzjnh+YuXLcm57Zqx8c8Tu00UhGKR4RzZsS/j9pFdsYRE5nxuOuvmdsTPXALAgTVbCG3TlKASIRQJL0O5ixtQrHLBndeMSWd9YP2IiBQBrgLy0xFyv6o2E5EIYAlwmapuFhHfs+S5QFugDLBWRMaq6lHgLlWNFZESwEIRmQzUAKq4rcCISHozzXvAfaq6XkQuBMYAV2ST/R7gHoB33333pPf9VAUVCeLsFvV4v/NwjqYcodcXQ9m5ajObz9CJ/VRtW7aR1zs+QkTtytzwah/WRS9n78adzHpnCndNeIwjBw+xc81W1Kd/nj+oWLcqUUNu4f2ez3sdJV+CiwRTOrQ0I657lNqN6/LgmEE8dMnxCxpV6lbj5iG9eOG2pzxMmb3R3Z8gYXccpcuXpe+nj7Nn4w7+XbGJDvdfz9iez3kdz5FHa+UlPTuyZuYy4mNiM20PCg4mrHIFNi9ex7fPTqBt76u5buhtZzJphuu63E6xYiF8NO51Lm/Tmnl/LWLQ4L506dyrUH7/yUraE8/LrR8kJT6Zyg1qcut7A3mj4yMcTk4BnG5TPd54gL8+/oW4bXs8TgsVul1G6ca1WdV1OAAJs5ZTukkdGv7wPEf3J5K0eK3fndf+C6wPrFVg/UUJEVnm/jwH+BBoncdjJrrfWwGzVXUzgKr6vnP8pKqHgcMisgeoCGwHHhSR690y1YC6wFqgloi8CfwETBOR0m6Or0Qy+ukVyy6Mqr6HU9kF0Cefm5VH/OMSY2Ip69PyWLZSOIkxcVnKxFG2cjiJMbEEBQdRvExJDsYlk7grlq3z/+FgnHOpcP3MZVRqUKPAK7CJu+MoVzlzxoTdsSeUCa1cPkvGzJdJ927cyZGDh6hYryo7Vm5m0aRoFk2KBqDj4BsztUSfrgQ3T7rQSuVJ2B2XpUwsoZXLk+BmLuGTuVxkOHe8O5AvB45h/79n7o2yQ6+raHtTBwA2rdhAuE/m8MjyxGU5znG7Ywn3aQ0Kr1SeOLdFNnbXPhb+Mg+AjcvXo2lKmfCyJMUmEh5ZnoHvDWHswNHs+TfmlPOmH7N0odm8FuJ3xxLmc1yLlynBgbikEx5bzuex6X+b5P2JrPh1IWc3rsPBhAOUrxrBIz+/5PyuyHAG//gCr173OEl7E7LNd2nPjhktuP8u35g5a6STKdP+xMQSWun4515nf5wsSXsTKBsRSuLeeMpGhJK0LzEja3bPW7NZPWq1OJdLenagWMniFClahMMHDzHlxS84fPAQy39ZAMDSqfNodWPbPI+1r3vu7ckdd94EwOLFK6hatVLGfZWrVGLnzpz/pocPH+HHH2dwzTUd2LN7LzWqV+Wv+VMBqFIlkj/mTuHyy65jz+59OT7HycjufJGYzfminM/5opjP/17KEed8tnPVZmL/3U35mpHsXOkMkuvywt3s3xzDXwU84NPX4ZhYQqpUyLgdUimcIzEnnpvKXdqIqv27sfr64ajPFY0doyezY/RkAOq+/RApG3ed8FhjTpd1IfAP6X1gm6hqP1U9Ahwj89+neJbHHMjH8x72+TkVKCIibYD2wEWq2hhYChRX1TigMRCN0wf3A/f3x/tka6Kq553C/uVq5/JNlK8ZSWi1CIKLBtOgcyvWTl+cqczaGUto0u0yAOpHtWTzXKeCumHWCiqeW42ixUMICg6ixoXnsXd9wQ/K2L58IxVqRBJW1cnYuPNF/J0l49/TF9Os26UANIi6kI1uxrCqEQQFO3/K0CoViKhdmTh3BoVS5csCUK5yec6/sgXLCnC07jY3c7ibuUnni1idJfPq6Ytp7h7XRlEXssHNXLxsSXqPe4SfXvyCLYvXFVim7Ewf/zNDowYyNGogi6bN59JuTsWmTtN6pCQdzOi7mi5+TxwpySnUaVoPgEu7tWXxdKditGjaAupf1BBwuhMUKVqEpNhESpYtyeBxj/PlixNYt+if08r77/KNRPgc12adW7Mqy3FdNX0xLd3j2jjqQta7x3XV9MU069ya4JAihFeNIKJGJFuXbSCkRDGKlXL+xUNKFOPcSxuxa902dq3dxrDm9/L0Jf14+pJ+xMfE8vI1j+VYeQWYM2EaL0U9yktRj7Ji2kJadnVy1Ghal0NJBzO6BKRL3BvPoaQUajStC0DLrpexctpCJ++MRbTsfrmzvfvlrJzuzJ6xcvqibJ93/ENv8uTF9/PUJf347vlPWfDNbKa8+IXzXL8toU6r+gDUu7gBMSf5f/reuxNo3epqWre6mh+nTOPmW7sC0KJFExITk9gdk3mAW6lSJTP6xQYHB3PllW1Zt24jq1evpWaNFpx/3qWcf96l7NgRwyWtOxdY5RVgx/KNlPc5XzTsfBH/ZHmN/DN9MU3d88X5UReyyX2NlAwvkzGwL6zaWZSvEUmc+wGy/cM3ULxMSaY+nbmLRkFLXraBEjUrUazaWUjRIlTocgmx7swp6Uo1qEntl+7ln9tHcnR/4vE7goIoElba2ZfzqlOqfnXiZy3DFKw01UL78lfWAuu/tuL0YS0GlADaAX9kU24eMEZEaqZ3IcjSCptVOSBOVQ+6/VlbAYhIBeCIqk4WkbXAp6qaKCKbReQGVf1KnGbYRqq6vCB3NC01jakjPqbn+EeR4CCWTprF3vU7aDuwGztXbGbtjCUsnRjN9a/14cFZr5ISf4CvH3gTgEOJB/nrg5/535RnQJX1M5ez/veCP1mmpabxw4iPuWv8ECQ4iEWTotmzfgftB3Rnx8pN/D1jCYsmRdNjVF8GRY/iYPwBvujnZKzR4hwu73MtqceOoWnK98PHZbS03Dr2IUqGlSbtWCo/DB/HocSDBZr52xEf87/xjyHBQSycFM3u9dvpNKA721ZuZs2MxSyYFM3No/oyJPo1DsYn86mb+eJenahQvSId+nelQ3+novB+zxdI3p/I1UNuoWmX1hQtEcKwv95iwcSZTHt9coFkXvb7Ypq0vYDXZo/lcMph3h30ZsZ9z08dxdCogQB8NOxd7nOn0VoevYRlbv+76Em/ce/LD/DitNEcO3qUsQ+/AUDH26OoWKMS1z/Yg+sf7AHAyJ5PZQzyOhlpqWlMHjGOPuOHOtNoTZpJzPrtXDXgBrat3MSqGYuZN2kmt426n2HRr3MwPplP+jk5YtZvZ+mPfzF0+qukHkvl6xHjnFbiCuXo/d7DgHOJePH3f/LPrNP/N1szcynnt23KiFmjOZJyhM8Gj82475GpL/JS1KMATBr+Ibe+0peQ4kVZE72MNdHO/9D0sd9z59sP0apHW+J27GPc/a/l+bw5+WHkZ/Qc9QBdR9xOcmwinw8eS9OrW53Sfv36y0w6dWrLilXRpBxM4b77Hsm4b+68n2jd6mpKlSrJpK/ep1hIMYKChNmz5/HB+5+d0u87WWmpafw44mNuHz/E+Xu654t27vninxlLWDwpmu6j+jIgehQp8QeYmH6+aHku7QbeQFr6+eLxj0hJOEDZyHDa9LuePRt20Pcnp0vJvE+msXhidMHvQGoam4Z+QP0vhiPBQez+8ndS1m2j2uCbSF6+gbhpi6g+vBdBpYpzjvu6PbxjH//cMRIpGkyD75zR/qlJKax7YDSket+FYPATI1m4dAXx8Ym0u+42+vbuSbfOnbyOZU6DWD8K74lIsqqWzmb7S8D1wGYgGfhBVT8WkS1Ac1Xd55a7Cngep8V0j6p2EJEngWRVfcUtswq4BtgFfIfT53UtEAo8CcQB4zje6vuYqv4sIjWBsUAloCjwpao+nccu6ZPVbz2FI+GNJ7d+xmM1bvE6Rr69sOVzAAbVuDmPkv7jlS1fcEv16/Mu6Ec+3/ot/Wvc5HWMfBu9xZkT9MEaN3qcJP/e2DKR0iVreh3jpCQf3MywADpfPOueL+ZW6uZxkvxrvWsyR/edOP+zvypaoRaA5FWuIIWVrlNolbe45A2Fum/5ZS2wfiC7yqu7/RHgkWy218hy+2fg5yzbnsxy23eOlqtyiNIsm9+1Gbgyh/LGGGOMMYXOKrDGGGOMMQHEn+dnLSw2iMsYY4wxxgQUq8AaY4wxxpiAYl0IjDHGGGMCiA3AtxZYY4wxxhgTYKwF1hhjjDEmgPjzAgOFxVpgjTHGGGNMQLEWWGOMMcaYAKI2jZa1wBpjjDHGmMBiLbDGGGOMMQHE+sBaC6wxxhhjjAkw1gJrjDHGGBNAbB5Ya4E1xhhjjDEBxlpgjTHGGGMCiM1CYC2wxhhjjDEmwFgLrDHGGGNMALE+sNYCa4wxxhhjAoy1wBpjjDHGBBBrgbUWWGOMMcYYE2CsAmuMMcYYYwKKWDO0OQPsRWWMMeb/EynMX1YkpEqhvc8eO7KjUPctv6wCawKGiNyjqu95neNkBFrmQMsLlrkwBFpesMyFIdDyQmBmNtmzLgQmkNzjdYBTEGiZAy0vWObCEGh5wTIXhkDLC4GZ2WTDKrDGGGOMMSagWAXWGGOMMcYEFKvAmkASiP2WAi1zoOUFy1wYAi0vWObCEGh5ITAzm2zYIC5jjDHGGBNQrAXWGGOMMcYEFKvAGmOMMcaYgGIVWGOMMcYYE1CsAmuMMYVMRIJEpKzXOYwxJlBZBdb4LREJFpGZXuc4GSJSUUQ+FJGf3dv1RaS317lyIyL1ROR9EZkmIr+nf3mdKzci0k9EwrzOcTJE5HMRKSsipYBVwBoRGex1rtyISH83s7iv6yUi0tHrXCdDRPxy1Ll7frtXRJ4RkYuz3DfMq1y5EZGSIvKIiAwWkeIicoeI/CAiL4lIaa/z5cQ9x/0mIqvc24389Rib/LMKrPFbqpoKpIlIOa+znISPgV+Byu7tdcBDnqXJn6+AJcAwYLDPlz+rCCwUkUkicqWI+OVa3VnUV9VE4DrgZ6Am0NPbSHm6y83cEQjDyTvS20gnEpHwHL7KA1Fe58vBu8DlwH7gDREZ5XNfV28i5eljnP+9msBPQHPgZUCAsd7FytP7wGPAUQBVXQHc5Gkic9qKeB3AmDwkAytFZDpwIH2jqj7oXaRcVVDVSSLyGICqHhORVK9D5eGYqvrzm88JVHWYiAzHqVjdCbwlIpOAD1V1o7fpclRURIriVGDfUtWjIuLv8ximfzCIAiao6mo//bCwF9jK8bwA6t4+y5NEeWupqo0AROQtYIyIfAPcTOb98Cf1VLWH+xrYBbRXVRWRP4DlHmfLTUlVXZDlpXvMqzCmYFgF1vi7b9yvQHHAbfVRABFpBSR4GylPU0SkL/AtcDh9o6rGehcpb+4bZwwQg/NmFAZ8LSLTVfURb9Nl6x1gC84b/WwRqQ4kepoob4tFZBpOi9tjIlIGSPM4U3Y2Ae1U9d+sd4jINg/y5EdI+g+qegy4R0RGAL8Dfns5HjL+96aqO5G8e9ufP4ztE5HaHD8vd8epgJsAZgsZGL8nIiWAs1V1rddZ8iIizYA3gQY4/RwjgO7uJSu/JCKbs9msqlqr0MPkk4j0B3oB+4APgO/cFs0gYL2q1vY0YBZuru6qOslnmwDBbuXFL7m5mwCbVDXe/XBWxd9ezyJyP/CHqp7QCigi/VT1TQ9i5UpEPgU+VdVfsmy/GxirqkW9SZYzEfkAeEhVk7Nsrw18oqqXeJMsdyJSC2cFrtZAHLAZuE1Vt3iZy5weq8AavyYinYFXgBBVrSkiTYCnVfVaj6PlSESKAOfgXAZcq6pHPY70nyMiTwLjVHVrNvedp6p/F36q3InIIlVt7nWOkyEil2W3XVVnF3aWgiAiHVR1utc5TkagZBYRSW+R9dfM7gDKIFVN8jqLOX1WgTV+TUQWA1cA0ara1N22SlUbeJsseyKS3eCLBGClqu4p7Dz54fbL7AOkV1aigXf9teItIsHAalU91+ssJ0NERuK0GE8kc39uv+2qISJTfG4WB1oCi1X1Co8inRYRWaKqzbzOcTIs8+lzr9iMA5JwBnQ1A4ao6jRPg5nTYn1gjb87qqoJWTrf+2MfvHS9gYuA9Om/2gCLgZoi8rSqTvAqWC7GAkWBMe7tnu62uz1LlAtVTRWRtSJydnZ9Hv3Yje73+322KeC3XTVUtbPvbRGpBrzuUZyC4K+Do3JjmU/fXao6WkQ6AeVxznETAKvABjCrwBp/t1pEbgGCRaQu8CAw1+NMuSkCnKequ8GZFxYYD1wIzMY5afqbFqra2Of27yLizyOKwRmwtVpEFpC5NdNvu5aoak2vMxSA7cB5Xoc4DYF4ydEynz7f2TTG+/FsGuYkWAXW+Lt+wOM4o+O/wJlj9RlPE+WuWnrl1bXH3RYrIn55SR5IFZHa6dNPuQMe/H3qr+FeBzgVItIAqI9zOR4AVR3vXaLcicibHK+MpA/oWuJdImNOSaDMpmFOglVgjV9T1YM4FdjHvc6ST9Ei8iPO4gAA3dxtpYB472LlajAwU0Q24bRUVMeZW9Vvqeost3W7hbtpgb/2MU4nIk/gdCmpD0wFrgL+wGmh91eLfH4+Bnyhqn96FaYAbPE6wCnY4nWAU7DF6wBZ9Ob4bBoH3dk0/PocZ/Jmg7iMX3IHj+T44vTXS8XuZamuQPp0MnFARVW9P+dHeU9EiuHMnADOzAmHcyvvNRHpgbMCUDROpftSYLCqfu1lrtyIyEqgMbBUVRu7FfBPVbWDx9FyJCL9VXV0Xtv8hTvo8yPgc1WN8zpPfljmwiHO0tN1yXz1IyBn0zAOW0rW+KtXgFdx5utLwRk5+j7Oylz+utIS7jQym3Baq64H2gJ+N6UTgIhc4X7vClwN1HG/rs5hNgV/8jhO393bVbUXzuh4f+9WkKKqacAxESmL273E40x5uT2bbXcUdoiTcCPOMs4LReRLEekUAH0dLfMZ5s6tOxunC9pT7vcnvcxkTp+1wBq/lt3cmf44n6aI1MNZAvJmjk+VNEhVq3saLBci8pSqPiEi47K5W1X1rkIPlU8islJVG/rcDgKW+27zNyIyBhiKswb7wzgfxpapqt9dyhSRm4FbcK4kzPG5qwyQpqrtPAmWT+7r4Rqc2TRScaZQGu3nU5ZZ5jPEvfrRApinqk1E5FzgeVX19w/qJhfWB9b4u1IiUktVNwGISE2glMeZsvMPzhv9Naq6AUBEBngbKXeq+oT749Oqmmk1Lvc4+7NfRORXnIF94LQITfUwT55Uta/74zsi8gtQ1t9WtPIxF2epzQo4V0LSJQH+mhkAEWmE078xCpgMfIZTEf8dpx+k37HMZ9whVT0kIohIMVX9R0TOyfthxp9ZBdb4uwE4g6B8Bxjd622kbHXFaVmb6VZOvsT/5kLMyWScib19fQ1c4EGWfFHVwSLSDbjY3fSeqn7rZaa8uJdYbwVqqerTInK2iLRU1QVeZ8vKXeFsK86cxgHD7ZsZD3yIM1F9el/u+SJycc6P9I5lLhTbRSQU+A6YLiJxOK9vE8CsC4Hxe+4Ao/RVl/7x5wFG7mwDXXC6ElyBM8L8W39c8cW9jHY+8BLOTATpyuIMiDrfk2D/USIyFmfqnitU9Tx3UMk0VW2Rx0M9IyKtgDdx5n4NAYKBA6pa1tNg2XAvZw9R1ee9zpJflrnwicjlQDngF1U94nUec+qsAmv8noi0Bmrgc8XAn+fOTOdWUG4AbvTHPoMi0gW4DrgW+MHnriTgS1X12wUjRCSJE2epSMCZ9unh9C4n/iR9eU0RWeqzLPLyLItI+BURWYRzZeEroDnQC6inqo95GiwH/tg/Pi+WuXC4S1BXJPP7SCCt5GeysAqs8WsiMgGoDSzj+OT6qqoPepfqv0VELlLVv7zOcTJE5BmcVaE+x+mqcRPO62QJ0EdV23iXLnsiMh9oDSx0K7IROC2wTT2OlqP0ioqIrFDVRu62pf6aWURGcnwQpe8KbX41qMiXZT7zRKQf8ASwm+MLGGj6a9oEJqvAGr8mIn8D9dVeqGeMiBTHmej7fDLPkejPsxCc0HIpIsvcEcZ+1aopIjVUdYuI3Ioz2KwZ8AnQHRimql/l+gQeEpHZQHvgAyAGZ2DXHf50fH2JyOZsNquq1ir0MPlkmc88EdkAXKiq+73OYgqODeIy/m4VEInzxmnOjAk4syh0Ap7GGWjkl3PX+jjoLmaQvnBBd+CQ+7O/fdiZISIf4MxtvBhoh9NqfJ2q+vtx7okzX/gDOAMqq+EMWPRLqurvs2ecwDIXim04XYzMf4i1wBq/JiIzcaZkWQBkDN7y15W4AlH6JeH0y8QiUhSYo6qtvM6WExGpBYzGGSWvwDycCtYO4AJV/cPDeJmIs+760ziD+h5Q1Tl5PMRvBOBKXEWBPsBl7qZo4F1VPepZqDxY5jNHRAa6P56Ps9LgT2R+HxnlRS5TMKwCa/yaO2L0BKo6q7Cz/FeJyAJVbeleLu6Lc6l4gb9eDgxUInIB8BtO3900nFZYv+6Hlz7wLMs2f+4D+wFQFKeLBjgtyKmqerd3qXJnmc8cEXkit/tV9anCymIKnlVgjd8TkepAXVWdISIlgWBVTfI613+Fu8ziZKARzko6pYHhqvqup8Fy4a58NhaoqKoN3EnVr1XVZz2Oli1xlu0djbOE5dscH0iSPueqX8llJa6yOBUVv5tVA3LsG+1XfaKzsszGnBrrA2v8moj8D7gHCMcZZV4FeAenH6EpAKr6gfvjLCBQWl3fx5m79l0AVV0hIp8DfleBFZEvgarALaq60us8+RSoK3GlikhtVd0IGV1NUvN4jNcs8xkmItOBG1Q13r0dhjNVYCdvk5nTYRVY4+/uB1oC8wFUdb2InOVtpP8WESkPPImzqpXitLg94+cjdkuq6gJncasMx7wKk4cZPh8SciQit6vqJ3mVKwxZV+JyXyOXAcmq6q/HGZwPNTOzrNx3p7eR8mSZz7yI9MorgKrG2ftI4LMKrPF3h1X1SHpFRUSK4H+jzAPdl8BsoJt7+1ac+R3be5Yob/tEpDbua0FEuuOnM1Xkp/Lq6s/xPoWeEpEfcVZbWiUilXDm110E1BaR91T1dW8TZk9VfxORujgDdgDW+vPKfWCZC0mqiJydvnCB2y3N3kcCnPWBNX5NRF7CWXO7F9APZ5DRGlV93NNg/yEiskpVG2TZtlJVG3qVKS/uJcv3cBYGiAM2A7f6Y3/S/PKnwVEislrdpYRFZChwrqr2cmdU+NNfB56JSHZTfCUAK1V1T2HnyQ/LfOaJSCecbkezcFqMLwXuUdVfPQ1mTou1wBp/NwRnkv2VOH1hfzqJFi2TP9NE5CZgknu7O85gI3+mqtpeREoBQaqaJCKBNjdlVv7UmuA7HVI7nDd/3OOclv1D/EJvnG4PM93bbXDm3q0pIk+r6gSvguXCMp9BIhIElMNZQCR9asCHVHWfd6lMQbAWWOOXRKQLUFVV33ZvLwAicN7kH1HVr3N7vMk/EUkCSnF8ZHwQx5eHVFUt60mwXOQwvdNiVb3Aq0yny89aYKcA03Cm/PoIqKmq8SJSAliU3jrrb0TkV6CXqu52b1cExgM3A7OzXmnwB5b5zEtfEtnrHKZgWQus8VeP4Kxvny4EuABniqdxHF+ByZwmVS3jdYb8EpFzcSYlL5flMmZZfJbB9UciEqyquY3U/rPQwuStN87iC+2BG30GwLTC+f/zV9XSK1WuPe62WBHxq0n2fVjmM2+GiAzC6duf/uEcVY31LpI5XVaBNf4qRFW3+dz+wz3ZxLqXjU0BEZHLstuuqrMLO0s+nANcA4QCnX22JwH/8yRR/q0XkcnAOFVdk/VOVX3Ag0zZcvsx3pfN9pkcv2yMiLypqv0KM1seot0BaF+5t7u720rh9KX3R5b5zLvR/X6/zzYlcKYNNNmwLgTGL4nIBlWtk8N9G1W1dmFn+q9yLxenK44zbdliVb3Co0h5EpGLVPUvr3OcDHcA1E040w0F4Vya/1JVEz0Ndhqy68rhJXGmK+mKswADOK3ak9WP3+gsszGnxiqwxi+JyGdAtKq+n2X7vUAbVb3Zm2T/fSJSDXhdVbvlWdgjIlIc5zL3+fh0HVDVuzwLdRLEWSL5c5yW5K9x5t3d4G2qk+dvFVgIzJX7LPOZJSK9stuuquMLO4spONaFwPirAcB3InILzhyU4PSBLQZc51mq/x+2A+d5HSIPE4B/gE44fTVvBf72NFEeRCQYuBqnBbYGzgpXn+FM6TMVqOdZuP+IQFy5zzIXihY+PxfHybkEZ+CZCVBWgTV+ye2D19pdQz59xPNPqvq7h7H+k0TkTY5P4RQENOH4hwZ/welZGAAAEO9JREFUVUdVbxCRLqr6ibuM7ByvQ+VhPU7/0ZdVda7P9q9z6occACTvIoUqEFfus8xnWNZ+2iISirOAiwlgVoE1fs2tsFql9cxa5PPzMeALVfWnEfHZSR/pHC8iDYAYwG/fQF29VPUP3w0icrGq/qmqD3oV6jSN9jpAFoG4cp9lLnwHgECfN/r/PavAGmO+Bg6lT/EkIsEiUlJVD3qcKzfviUgYMAz4AWd6tRHeRsrTGziTqft6M5ttnnMH9uVYIVHVa93vHxdWpnya5a4cVkJEOuCs3Dclj8d4zTKfYVlez0FAfY4v3GIClA3iMub/ORGZB7RX1WT3dmlgmqq29jbZf4OIXISz5O1DwGs+d5UFrlfVxp4Ey4U7yCxHqjqrsLKcDHfVpd5AR5zuDb9mHQjqbyzzmZfl9XwM2Kqq273KYwqGtcAaY4qnV14BVDXZHVXsd0RkIJCgqh9m2d4bKKOqr3uTLFchOC3ERQDfRSMScebP9Dv+WkHNh36qOhp36VsAEenvbvNXlvkMcWcruQ+og7Mc+YeqeszbVKagWAusMf/PicifOG9IS9zbFwBvqepF3iY7kYgsBlqp6tEs20Nwljht5E2yvIlIdVXd6nWOkyEidYEX+L/27j7Wsqo+4/j3mQEcKDOUF7XW8KIFRmAAC9MWSIsCmhaxJA1qO7yEaiEtL4UxmlDTWKB/WGx9pRQSWzoCCiookRRQCeKAQoABYZiZCkwBAwJCS4szvDM8/WPvy9y5nHvvmXPn7LX3meeTnNy9176TPLmZnPu766z1W9VHruPblbWyAfwkRwy35ojeXpJ5eCR9k2q9/C3AkVQzr2eWTRWbSmZgI2IxcKWkx6k+DvwN1p9c0zZbTCxeAeoNJW3bEQ+ApC/ZXgxcIOkNMwZj60lbaglwNtXSh8NYfwhDq0haBBwLvEPSNeMezQVaeVxoMjdib9v7Aki6GLijcJ7YhFLARmzmbN8p6V1Ux7QC3N+rSGyJWZLeOuEcdiS9tVSgPlxWf/1c0RSD2dr2jZJUzx6fU8+Ct23D3K3AE8BOVP11x6wBlhdJNL1kHr7X38dsv9rSv3FjQFlCELGZk3Qa8HXb/1ffbw8ssn1h2WRvVJ+ocwbwCTY84OKfqJY9XFIq22QkvRl4s+1VE8b3Bp62/XSZZNOTdCvVcaFXUbWz+wVwnu35U/7DiBaQtI6qZRZUny5tDTxfX9v2vFLZYuZa91FQRDTu5LHiFcD2/wInF8wzqfrox09Tnb71CPAwcC7wd20sXmv/TDVjNdGOtK+P6kRnAttQ/dFwIHA80PNYzjaQdJCkOyWtlfSypHWSflU611SSeXhsz7Y9r37Ntb3FuOsUrx2XAjYiZo9fP1ofebpVwTxTsn297ffY3tH2TvX19eO/R9KnSuXrYXfbN08ctH0L0NpNZ7XdbK+1/Zjtj9o+BtildKgpXAAsojr1bGvgJOBfiiaaXjJHDCAFbER8H/impCMkHUF1xOL3CmeaqQ+XDjDO3CmebdlYisH0+kOgTX8cvIHt1cBs2+tsLwH+qHSm6SRzxMbLJq6I+DTVkoFT6/vvAxdP/u2d0KbdGqslfcD2deMHJR0JPFQo05TqbB8A3i7p/HGP5lE1gm+r5+uWavdI+keqDUdtn6hJ5ogBpICN2EzV55d/hqo10qP18C5URdUsYF2haJtCm3anLgaulfQR4K56bCFwMPDBYqmm9jiwDDia9Zmh2m3+8SKJ+nMC1f/d06ly7gwcUzTR9JI5YgDpQhCxmZL0RaqPtz9ue009NpeqPc4LXW743bam6pLeRNU/c0E9tBK43PaL5VJNT9IWXTi5aIpOD/sAT7Wx00MyR8xMpvwjNl8fpOpAsGZsoL4+herj49aStOM033JlI0H6ZPsl20tsf6J+/fvE4lXSbaXyTSTpW/XlTyUtn/gqGq63yTo97EB7Oz0kc8QMZAY2YjMl6QHbe27sszaQ9CBwD9VJUdd7BN7I2jRrLOlttp+QtGuv5207ElfSMtsLJ3m2wvaCXs9KSuaImckMbMTma1V9MMAGJB0P/KxAno2xJ/AVqrV4D0r6jKTWFtx9ak0RXhevs4Gv2v75xFfpfD10sdNDMkfMQArYiM3XacBpkn4k6fP1aylV0/pTCmebkis32F5E1UHhROAOSUslHVw43kiwvQ54TdJ2pbP0YbWkNyx7aXOnB5I5YkayhCBiMyfpcGCf+naV7RtL5ulHvQb2eKoZ2F9Stf26Bng3cKXtdxSMN5A2LSEYI+m7wG8DN7D+SE5sn1EsVA+S9gCuBW6lR6cH2w+UyjaZZI6YmRSwEdE5kh4ALgOW2H5swrOzbH+2TLLJSfqs7bMmG5O0wPaKMul6k3Rir/E2HtvbxU4PyRwxuBSwEdE5kj5i+1sTxj5su1XdB8aTdLftAyaMLbfd9uNkR4ak22x3aolJMkf0ljWwEdFFf9NjrJVHnEo6RdJ9wPwJ7ageBtrYkup1kvaQdJWkVZIeGnuVzjUDc0oHGEAyR/SQk7giojM6esTp5cD1wD+wYeG9xvYzZSL1bQlwNvBF4DCqU9u6PPHRxY8ckzmihy6/EUXE5mfsiNMXqTaRjL2uAf6wYK5J2X7W9iN1x4THgFeofsFvK2mXsummtXW9qU91C61zgKMKZ4qIyAxsRHSH7XuBeyV9vQtHnI4n6XTgHKquCa/VwwbavAb2JUmzqHrtng78Ati2cKaZUOkAA0jmiB4yAxsRndHBI07HWwzMt72P7X3rV5uLV4AzgW2oegMfSNW2rGdngjaQ9IbuExPGTmgwTl+SOWIw6UIQEZ0h6Q+oZi8fm/BoZ+BJ26ubT9UfSTcB7+/azHGXdLHTQzJHDCZLCCKiS84CPjXxOFNJ86g2Gv1xkVT9eQj4kaRrgZfGBm1/oVyk3iRdM9Vz20c3laUfkk4BTgV+a8JM/Fyqpvutk8wRM5MZ2IjoDEl32v6dSZ7dZ3vfpjP1S9LZvcZtn9t0lulIehp4FLgCuJ0JaxptLy2RazL1cbfb06FOD8kcMTMpYCOiMyQ9aHuPSZ6ttr1705k2lqRtbD9fOsdUJM0G3g8sotpkdi1whe2VRYNNQ9JBwErba+r7ecBetm8vm2xyyRwxmGziioguWSbp5ImDkk5i/dnsrSTpYEmrgJ/V9/tLurBwrJ5sr7P9PdsnAgcBq6mWP5xeONp0LgLWjrtfW4+1WTJHDCBrYCOiSxYDV0s6jvUF60JgK+BPiqXqz5eoetVeA1VLMEmHlo00ufrM+6OoZmF3A84Hri6ZqQ/yuI8Vbb8mqe2/55I5YgD5DxcRnWH7l8Ahkg4DFtTD19r+YcFYfbP9qLTBctJ1pbJMRdKlVD/f64Bzba8oHKlfD0k6g/WzgadSbZ5rs2SOGEDWwEZENEDSVcAXgAuA36PqsbrQ9p8VDdaDpNeA5+rb8b8kBNj2vOZTTU/SW6hmig+nyn0jsNj2U0WDTSGZIwaTAjYiogGSdgK+DLyPqhD8AXCm7f8pGiwiooNSwEZEDFm9q/9S28eVzjLKJC1hwxljAGx/rECcviRzxGCyBjYiYshsr5O0q6StbL9cOs8I+49x13OoNvY9XihLv5I5YgCZgY2IaEC9MWovqi4EY+tLW3kS16iQNAv4se1DSmfpVzJH9CczsBERzfiv+jWL6ujNGL49gLeUDrGRkjmiDylgIyKGrF4Du2fWwA6XpDVUazNVf30SOKtoqGkkc8RgUsBGRAxZ1sA2w3bnZraTOWIwKWAjIprxEPATSVkDu4lJOmCq57bvbipLv5I5YmZSwEZENCNrYIfn8/XXOVRHC99L9fH2fsAy4OBCuaaSzBEzkAI2IqIBts8tnWFU2T4MQNJ3gANs31ffLwDOKRhtUskcMTMpYCMiGiDpJno3fz+8QJxRNX+sqAKwvULSXiUD9SGZIwaQAjYiohmfHHc9BzgGeLVQllG1XNK/AV+r748DlhfM049kjhhADjKIiChE0h22f7d0jlEhaQ5wCnBoPXQzcJHtF8ulmloyRwwmBWxERAMk7TDudhZwIHC+7fmFIo0kSVsB86mWa9xv+5XCkaaVzBEbL0sIIiKacRfrm7+/CjwM/EXRRCNG0nuBS4BHqH7OO0s60fbNJXNNJZkjBpMZ2IiIGAmS7gKOtX1/fb8ncIXtA8smm1wyRwxmVukAERGjTNLxkk7oMX6CpGNLZBphW44VVQC2HwC2LJinH8kcMYDMwEZEDJGk24EjbK+dMP5rwM2Ztdp0JC0B1rHh7vjZtj9WLtXUkjliMClgIyKGSNLdtnsewSlpue39ms40qiS9CTgN+P166BbgQtsvlUs1tWSOGEwK2IiIIZL0n8BC289NGJ8L3Gn7XWWSjRZJs4GVXfp5JnPE4LIGNiJiuC4GrpK069iApN2Ab9TPYhOwvQ64X9IupbP0K5kjBpc2WhERQ2T7c5LWAjdL2rYeXgucZ/uigtFG0fbASkl3AK/PeNs+ulykaSVzxACyhCAioiH1sgFsr+nx7ETblzSfanRIek+vcdtLm87Sr2SOGEwK2IiIFphqs1dMrT7a9K+A3YH7gIttv1o21dSSOWJmsgY2IqIdVDpAh10CLKQqqo4EPl82Tl+SOWIGMgMbEdECmYEdnKT7bO9bX28B3NH2n2UyR8xMZmAjItohM7CDe2XsokMfaSdzxAykC0FERDv8pHSADttf0q/qawFb1/cCbHteuWiTSuaIGcgSgoiIBtSnFx0D7Ma4yQPbf18qU0REV2UGNiKiGd8FngXuAnLkZkTEDGQGNiKiAZJW2F5QOkdExCjIJq6IiGbcKmnf0iEiIkZBZmAjIhogaRVVA/iHqZYQjG182a9osIiIDkoBGxHRAEm79hq3/fOms0REdF02cUVEDJGkHerLNUWDRESMkMzARkQMkaSHAdP7oALbfmfDkSIiOi8FbEREC0jax/bK0jkiIrogXQgiItrhstIBIiK6IgVsREQ79FpiEBERPaSAjYhoh6zniojoUwrYiIiIiOiUFLAREe3wcukAERFdkQI2IqIBkr4j6ShJPd93bR/UdKaIiK5KARsR0YwLgWOBByWdJ2l+6UAREV2VPrAREQ2StB2wCPhb4FHgX4Gv2X6laLCIiA7JDGxEREMk7Qj8OXAS8FPgy8ABwA0FY0VEdE5mYCMiGiDpamA+1YEFX7X9xLhny2wvLBYuIqJjUsBGRDRA0mG2byqdIyJiFKSAjYhoiKQFwN7AnLEx25eWSxQR0U0pYCMiGiDpbOC9VAXsdcCRwI9tf6hkroiILsomroiIZnwIOAJ40vZHgf2B7cpGiojophSwERHNeMH2a8CrkuYBTwE7F84UEdFJW5QOEBGxmVgm6dep+r7eBawFbisbKSKim7IGNiKiYZJ2A+bZXl44SkREJ2UJQUREAyTdOHZt+xHby8ePRURE/7KEICJiiCTNAbYBdpK0PaD60Tzg7cWCRUR0WArYiIjh+ktgMfCbVGtfx6wBLiiSKCKi47KEICJiuG4FDgE+afudwLnACmApcHnJYBERXZVNXBERQyTpbuB9tp+RdCjwDeCvgXcDe+Ugg4iIjZclBBERwzXb9jP19Z8CX7H9beDbku4pmCsiorOyhCAiYrhmSxqbLDgC+OG4Z5lEiIgYQN48IyKG6wpgqaT/Bl4AbgGQtDvwbMlgERFdlTWwERFDJukg4G3AD2w/V4/tCWxr++6i4SIiOigFbERERER0StbARkRERESnpICNiIiIiE5JARsRERERnZICNiIiIiI6JQVsRERERHTK/wNNEqACjaCR7wAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 720x720 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IRax8N0dh49e",
        "colab_type": "code",
        "outputId": "2e75e280-1082-4d16-af2f-9f7fca4d0cf6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 202
        }
      },
      "source": [
        "abs(data.corr(\"pearson\")[\"Purchase\"]).sort_values(ascending = False)"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Purchase                      1.000000\n",
              "Product_Category_1            0.343703\n",
              "Product_Category_3            0.288501\n",
              "City_Category                 0.061914\n",
              "Gender                        0.060346\n",
              "Product_Category_2            0.052288\n",
              "Age                           0.038553\n",
              "Occupation                    0.020833\n",
              "Stay_In_Current_City_Years    0.005422\n",
              "Marital_Status                0.000463\n",
              "Name: Purchase, dtype: float64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BM50qgMfuyuz",
        "colab_type": "text"
      },
      "source": [
        "Using the heatmap, and pearson correlation we try to find the linear correlation between each feature and the target. A few observations such as Product Category 1 attribute being negatively correlated, Occupation and Stay in Current Years being positively correlated can be seen."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OWkzGkA8wKUz",
        "colab_type": "text"
      },
      "source": [
        "## Normalizing the features"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "WKhSw4LSwbKf",
        "colab_type": "text"
      },
      "source": [
        "We will be using MinMaxScaler to bring each data point within the values of 0 and 1."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zly7j4MjwkgZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "scaler = MinMaxScaler()\n",
        "dff = pd.DataFrame(scaler.fit_transform(data), columns = data.columns)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SFwKmoAzj52H",
        "colab_type": "code",
        "outputId": "6cee0524-1341-413b-ad96-528066d2a4c5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 422
        }
      },
      "source": [
        "dff"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Gender</th>\n",
              "      <th>Age</th>\n",
              "      <th>Occupation</th>\n",
              "      <th>City_Category</th>\n",
              "      <th>Stay_In_Current_City_Years</th>\n",
              "      <th>Marital_Status</th>\n",
              "      <th>Product_Category_1</th>\n",
              "      <th>Product_Category_2</th>\n",
              "      <th>Product_Category_3</th>\n",
              "      <th>Purchase</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.0</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.50</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.105263</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.348992</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.0</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.50</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.333333</td>\n",
              "      <td>0.777778</td>\n",
              "      <td>0.634181</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.0</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.50</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.578947</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.058875</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.0</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.50</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.578947</td>\n",
              "      <td>0.777778</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.043634</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1.0</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.80</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.00</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.368421</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.332248</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>550063</th>\n",
              "      <td>1.0</td>\n",
              "      <td>0.955556</td>\n",
              "      <td>0.65</td>\n",
              "      <td>0.5</td>\n",
              "      <td>0.25</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.014865</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>550064</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.444444</td>\n",
              "      <td>0.05</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.75</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.014990</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>550065</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.444444</td>\n",
              "      <td>0.75</td>\n",
              "      <td>0.5</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.005219</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>550066</th>\n",
              "      <td>0.0</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.05</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0.50</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.014740</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>550067</th>\n",
              "      <td>0.0</td>\n",
              "      <td>0.844444</td>\n",
              "      <td>0.00</td>\n",
              "      <td>0.5</td>\n",
              "      <td>1.00</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.019959</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>550068 rows × 10 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "        Gender       Age  ...  Product_Category_3  Purchase\n",
              "0          0.0  1.000000  ...            0.000000  0.348992\n",
              "1          0.0  1.000000  ...            0.777778  0.634181\n",
              "2          0.0  1.000000  ...            0.000000  0.058875\n",
              "3          0.0  1.000000  ...            0.000000  0.043634\n",
              "4          1.0  1.000000  ...            0.000000  0.332248\n",
              "...        ...       ...  ...                 ...       ...\n",
              "550063     1.0  0.955556  ...            0.000000  0.014865\n",
              "550064     0.0  0.444444  ...            0.000000  0.014990\n",
              "550065     0.0  0.444444  ...            0.000000  0.005219\n",
              "550066     0.0  1.000000  ...            0.000000  0.014740\n",
              "550067     0.0  0.844444  ...            0.000000  0.019959\n",
              "\n",
              "[550068 rows x 10 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9nfm_H3AifGW",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "features = dff[[\"Product_Category_1\",\"Product_Category_3\", \"City_Category\", \"Gender\", \"Product_Category_2\"]].values\n",
        "target = dff['Purchase'].values"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6o_1hOW019mx",
        "colab_type": "text"
      },
      "source": [
        "I am choosing these respective features after trial and error, so that they give a better value of test accuracy at the end. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hyrutanv4WcZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "xtrain, xtest, ytrain, ytest = train_test_split(features, target, test_size = 0.2)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GR-oddw-2Ugd",
        "colab_type": "text"
      },
      "source": [
        "We will use Train Test Split to get our respective train and test data."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3rJDkBlk2u8S",
        "colab_type": "text"
      },
      "source": [
        "## Model Training"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4MT58zWw2yp9",
        "colab_type": "text"
      },
      "source": [
        "Using our train and test data, we will now use 4 different methods of model training:\n",
        "\n",
        "1. Linear Regression\n",
        "2. Random Forest\n",
        "3. Decision Tree\n",
        "4. Gradient Boosting\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "H98fDTw-MLNo",
        "colab_type": "text"
      },
      "source": [
        "### 1. Linear Regression"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c6KW5cc7MKxh",
        "colab_type": "code",
        "outputId": "a96ccb36-ad2e-4b27-d431-d97cb70d23c7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "lr = LinearRegression()\n",
        "lr.fit(xtrain,ytrain)\n",
        "ypred = lr.predict(xtest)\n",
        "metrics.mean_absolute_error(ytest,ypred)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.14699073633608867"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TIe_ZlBYMzhv",
        "colab_type": "code",
        "outputId": "c423e714-4dcd-4763-9c1e-f4566d636cca",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        }
      },
      "source": [
        "print(\"Accuracy of Linear Regression on train set is: \",lr.score(xtrain,ytrain)*100)\n",
        "print(\"Accuracy of Linear Regression on test set is: \",lr.score(xtest,ytest)*100)"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Accuracy of Linear Regression on train set is:  15.189956233931468\n",
            "Accuracy of Linear Regression on test set is:  14.915142001848858\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ebw2_vnH36kS",
        "colab_type": "text"
      },
      "source": [
        "As we can see, Linear Regression does not give a good accuracy at all, with highest of 15%, and an MSE or 0.14."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9GRozm1xCsH5",
        "colab_type": "text"
      },
      "source": [
        "### 2. Random Forest"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2pfXbm8i9j3x",
        "colab_type": "code",
        "outputId": "580e6d6c-5c79-4693-82a1-08ad169f1903",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "rfr = RandomForestRegressor(max_leaf_nodes=1200)\n",
        "rfr.fit(xtrain, ytrain)\n",
        "ypred1 = rfr.predict(xtest)\n",
        "metrics.mean_absolute_error(ytest,ypred1)"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.09148673203550609"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XyaUG3RWpbah",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        },
        "outputId": "10214179-a7e9-490c-d187-7fc35a4d1a5c"
      },
      "source": [
        "ytest"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.21833897, 0.32686125, 0.29475135, ..., 0.32807215, 0.65242808,\n",
              "       0.48695144])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JXtOSRRepeUy",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        },
        "outputId": "b41229fa-50b1-4b22-ebdc-8a04aea4a4a6"
      },
      "source": [
        "ypred1"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.24809349, 0.30637157, 0.25355327, ..., 0.56239476, 0.52794613,\n",
              "       0.54006512])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "spPymjKPAYwp",
        "colab_type": "code",
        "outputId": "8d5bff74-9dcb-4103-c961-1104e10dc1a2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        }
      },
      "source": [
        "print(\"Accuracy of Random Forests on train set is: \",rfr.score(xtrain,ytrain)*100)\n",
        "print(\"Accuracy of Random Forests on test set is: \",rfr.score(xtest,ytest)*100)"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Accuracy of Random Forests on train set is:  66.32358014900872\n",
            "Accuracy of Random Forests on test set is:  65.93239861021556\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EPNi8G7K4Upd",
        "colab_type": "text"
      },
      "source": [
        "After a lot of trial and error, I was able to bring up the test accuracy to 65.93%, with an MSE of 0.091. We also observe that the error decrased from Linear Regression to Random Forest, while the accuracies for both test and train increased."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YiD-PVziCxk1",
        "colab_type": "text"
      },
      "source": [
        "### 3. Decision Tree"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-AEu0zQiCwUR",
        "colab_type": "code",
        "outputId": "e9b0d4b2-a82f-4a18-ceb8-9cbcfc82cf8e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "dtr = DecisionTreeRegressor()\n",
        "dtr.fit(xtrain, ytrain)\n",
        "ypred2 = dtr.predict(xtest)\n",
        "metrics.mean_absolute_error(ytest,ypred2)\n"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.09148792777694059"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mBtIstlCpqNf",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        },
        "outputId": "d25e9440-7e19-4ea4-cd1e-a96365e268ae"
      },
      "source": [
        "ytest"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.21833897, 0.32686125, 0.29475135, ..., 0.32807215, 0.65242808,\n",
              "       0.48695144])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CC3ljCh-pr4V",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        },
        "outputId": "26679e32-08b7-4aa0-f6de-d0390dc729cd"
      },
      "source": [
        "ypred2"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.24788501, 0.30654245, 0.2532075 , ..., 0.56213536, 0.5278483 ,\n",
              "       0.53973811])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XCIgZUvEEOnR",
        "colab_type": "code",
        "outputId": "ac43016e-24f9-4467-df4a-009f18c5e741",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        }
      },
      "source": [
        "print(\"Accuracy of Decision tree on train set is: \",dtr.score(xtrain,ytrain)*100)\n",
        "print(\"Accuracy of Decision tree on test set is: \",dtr.score(xtest,ytest)*100)"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Accuracy of Decision tree on train set is:  66.32519477827691\n",
            "Accuracy of Decision tree on test set is:  65.9342091819476\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3JGi5muR47le",
        "colab_type": "text"
      },
      "source": [
        "After training the Decision Tree model, we observe that the MSE value and the accuracies are not that different from those of Random Forest."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ztwb8bsGffI3",
        "colab_type": "text"
      },
      "source": [
        "### 4. Gradient Boosting"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4uhF95wH7ZkZ",
        "colab_type": "code",
        "outputId": "d71751ab-b917-4c86-a702-faa9b7ed2a44",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "gbr = GradientBoostingRegressor()\n",
        "gbr.fit(xtrain,ytrain)\n",
        "ypred3 = gbr.predict(xtest)\n",
        "metrics.mean_absolute_error(ytest,ypred3)"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.09443777754403088"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PBWQdzdmp0YG",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        },
        "outputId": "ceb51034-ce28-4f73-f536-99f7b7cff350"
      },
      "source": [
        "ytest"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.21833897, 0.32686125, 0.29475135, ..., 0.32807215, 0.65242808,\n",
              "       0.48695144])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "h5WC1AQQp2Vd",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        },
        "outputId": "dc838ad9-be55-4aa7-c751-7e809c9683fb"
      },
      "source": [
        "ypred3"
      ],
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0.27779306, 0.30749852, 0.25507114, ..., 0.59518842, 0.54615189,\n",
              "       0.54529712])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MbxUoNv8gFO9",
        "colab_type": "code",
        "outputId": "655331de-4eb6-4f75-cbe8-13f99d60373f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 50
        }
      },
      "source": [
        "print(\"Accuracy of Gradient Boosting on train set is: \",gbr.score(xtrain,ytrain)*100)\n",
        "print(\"Accuracy of Gradient Boosting on test set is: \",gbr.score(xtest,ytest)*100)"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Accuracy of Gradient Boosting on train set is:  64.76804327045475\n",
            "Accuracy of Gradient Boosting on test set is:  64.52218002387201\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AVwbtfjd54bx",
        "colab_type": "text"
      },
      "source": [
        "As seen above the accuracy values, and the MSE are almost similar to the previous two models. \n",
        "\n",
        "Even though Gradient Boosting and Random Forest follow different approches, they still have almost a similar output on this dataset."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rPDss3Nq3jcT",
        "colab_type": "text"
      },
      "source": [
        "## Summary of Model Training"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OplXloMDp6mo",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 402
        },
        "outputId": "8ce6dbb0-af72-4522-dcc5-69593c5183a7"
      },
      "source": [
        "pd.DataFrame({\"Purchase\" : ytest, \"Linear Regression\" : ypred, \"Random Forest\" : ypred1, \"Decision Tree\" : ypred2, \"Gradient Boosting\" : ypred3})\n"
      ],
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Purchase</th>\n",
              "      <th>Linear Regression</th>\n",
              "      <th>Random Forest</th>\n",
              "      <th>Decision Tree</th>\n",
              "      <th>Gradient Boosting</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>0.218339</td>\n",
              "      <td>0.472838</td>\n",
              "      <td>0.248093</td>\n",
              "      <td>0.247885</td>\n",
              "      <td>0.277793</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>0.326861</td>\n",
              "      <td>0.312009</td>\n",
              "      <td>0.306372</td>\n",
              "      <td>0.306542</td>\n",
              "      <td>0.307499</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.294751</td>\n",
              "      <td>0.335336</td>\n",
              "      <td>0.253553</td>\n",
              "      <td>0.253208</td>\n",
              "      <td>0.255071</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>0.170654</td>\n",
              "      <td>0.314397</td>\n",
              "      <td>0.302315</td>\n",
              "      <td>0.302553</td>\n",
              "      <td>0.317690</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>0.296296</td>\n",
              "      <td>0.369886</td>\n",
              "      <td>0.252432</td>\n",
              "      <td>0.252441</td>\n",
              "      <td>0.255762</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>110009</th>\n",
              "      <td>0.408660</td>\n",
              "      <td>0.334614</td>\n",
              "      <td>0.310771</td>\n",
              "      <td>0.310548</td>\n",
              "      <td>0.317690</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>110010</th>\n",
              "      <td>0.024218</td>\n",
              "      <td>0.342636</td>\n",
              "      <td>0.027800</td>\n",
              "      <td>0.026754</td>\n",
              "      <td>0.027213</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>110011</th>\n",
              "      <td>0.328072</td>\n",
              "      <td>0.459079</td>\n",
              "      <td>0.562395</td>\n",
              "      <td>0.562135</td>\n",
              "      <td>0.595188</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>110012</th>\n",
              "      <td>0.652428</td>\n",
              "      <td>0.435700</td>\n",
              "      <td>0.527946</td>\n",
              "      <td>0.527848</td>\n",
              "      <td>0.546152</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>110013</th>\n",
              "      <td>0.486951</td>\n",
              "      <td>0.518942</td>\n",
              "      <td>0.540065</td>\n",
              "      <td>0.539738</td>\n",
              "      <td>0.545297</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>110014 rows × 5 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "        Purchase  Linear Regression  ...  Decision Tree  Gradient Boosting\n",
              "0       0.218339           0.472838  ...       0.247885           0.277793\n",
              "1       0.326861           0.312009  ...       0.306542           0.307499\n",
              "2       0.294751           0.335336  ...       0.253208           0.255071\n",
              "3       0.170654           0.314397  ...       0.302553           0.317690\n",
              "4       0.296296           0.369886  ...       0.252441           0.255762\n",
              "...          ...                ...  ...            ...                ...\n",
              "110009  0.408660           0.334614  ...       0.310548           0.317690\n",
              "110010  0.024218           0.342636  ...       0.026754           0.027213\n",
              "110011  0.328072           0.459079  ...       0.562135           0.595188\n",
              "110012  0.652428           0.435700  ...       0.527848           0.546152\n",
              "110013  0.486951           0.518942  ...       0.539738           0.545297\n",
              "\n",
              "[110014 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EWcLLy9H6lM2",
        "colab_type": "text"
      },
      "source": [
        "After training all the models, I have created a Data Frame which shows us the prediction values for the test data, compares with the actual normalised values of the \"Purchase\" feature."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BlwJ3DUUcLWe",
        "colab_type": "code",
        "outputId": "dba3ea85-c6f4-42ab-9954-ec18a9ca0ecb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 573
        }
      },
      "source": [
        "models = ['Linear Regression', 'Random Forest', 'Decision Tree', \"Gradient Boosting\"]\n",
        "\n",
        "train_accuracy = [15, 67, 66, 65]\n",
        "plt.bar(models,train_accuracy)\n",
        "plt.title(\"Model vs Training Accuracy\")\n",
        "plt.xlabel(\"Models\")\n",
        "plt.ylabel(\"Training Accuracy\")\n",
        "plt.show()\n",
        "\n",
        "test_accuracy = [15, 66, 66, 64.5]\n",
        "plt.bar(models,test_accuracy)\n",
        "plt.title(\"Model vs Testing Accuracy\")\n",
        "plt.xlabel(\"Models\")\n",
        "plt.ylabel(\"Testing Accuracy\")\n",
        "plt.show()"
      ],
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ohjWm2Td67Cq",
        "colab_type": "text"
      },
      "source": [
        "Now when we plot a bar graph of all the accuracies of each model, we see that both training and tes accuracies don't differ that much, from both the same model, and different models, except for Linear Regression. This also goes to show us that for this particular Dataset, Linear Regression is not a suitable method for model training."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "myUvhn3S3oLG",
        "colab_type": "text"
      },
      "source": [
        "## Conclusion"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bhgsvahYF8U6",
        "colab_type": "text"
      },
      "source": [
        "We have finally reached the end of this notebook, as well as the entire project. We finally  have a clear idea of how the various features present in this dataset affect the final purchase attribute, and how the different features compare with each other.\n",
        "\n",
        "During model training we also saw how each model gave almost the same result, the exception being Linear Regression. We know that both Gradient Boosting and Random forest uses the result of a single decision tree to build their prediction for the final output. We also know that they are seperately based on variance and bias. But we still observed that they gave similar results.\n",
        "\n",
        "We also observed that, more than the model training, there were many insights to this dataset, some of which we may not even have covered. The fact that a few features like Occupation and Product Categrories had numerical values without any explanantion to what they mean, made it difficult to extract more information. There may be better results using different approaches, or by changing a few hyper parameters as well. I shall upload them as and when possible."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Oe9c6Xv83qVn",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "Final_Accuracy = pd.DataFrame({\"Linear Regression\" : [15.18, 14.91], \"Random Forest\" : [66.34,65.93], \"Decision Tree\" : [66.32,65.93], \"Gradient Boosting\" : [65.76,64.53]}, index = [\"Train Accuracy\",\"Test Accuracy\"])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1YKm5VTE8RSO",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 106
        },
        "outputId": "243c76fc-1736-4281-fdd0-9cea4381433c"
      },
      "source": [
        "Final_Accuracy"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Linear Regression</th>\n",
              "      <th>Random Forest</th>\n",
              "      <th>Decision Tree</th>\n",
              "      <th>Gradient Boosting</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>Train Accuracy</th>\n",
              "      <td>15.18</td>\n",
              "      <td>66.34</td>\n",
              "      <td>66.32</td>\n",
              "      <td>65.76</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Test Accuracy</th>\n",
              "      <td>14.91</td>\n",
              "      <td>65.93</td>\n",
              "      <td>65.93</td>\n",
              "      <td>64.53</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                Linear Regression  ...  Gradient Boosting\n",
              "Train Accuracy              15.18  ...              65.76\n",
              "Test Accuracy               14.91  ...              64.53\n",
              "\n",
              "[2 rows x 4 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 31
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rKRStkrpTjDy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}