Module_Project11_3.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Module_Project11_3.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyMsJpalwCiW+KmLZ87w4a6o",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Ankita-Phd/Ankita/blob/main/Module_Project11_3.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "XLDcr6QKMlEg"
      },
      "outputs": [],
      "source": [
        "# Imports\n",
        "import pandas as pd\n",
        "import numpy  as np\n",
        "import scipy.integrate\n",
        "\n",
        "from matplotlib import pyplot as plt\n",
        "plt.style.use('ggplot')\n",
        "from matplotlib.pylab import rcParams\n",
        "rcParams['figure.figsize'] = 10, 8\n",
        "\n",
        "import copy\n",
        "from   sklearn.metrics       import mean_squared_error, mean_absolute_error\n",
        "from   sklearn.linear_model  import LinearRegression, BayesianRidge\n",
        "from   sklearn.tree          import DecisionTreeRegressor\n",
        "from   sklearn.preprocessing import PolynomialFeatures"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "path = \"https://raw.githubusercontent.com/PranavTadimeti/Regression_Project/main/covid_19_data.csv\"\n",
        "df_orig = pd.read_csv(path)\n",
        "print(df_orig.head())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xBXqMVk8Po4k",
        "outputId": "cefcd202-3038-4899-8366-1528862ca804"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "   SNo ObservationDate Province/State  Country/Region      Last Update  \\\n",
            "0    1      01/22/2020          Anhui  Mainland China  1/22/2020 17:00   \n",
            "1    2      01/22/2020        Beijing  Mainland China  1/22/2020 17:00   \n",
            "2    3      01/22/2020      Chongqing  Mainland China  1/22/2020 17:00   \n",
            "3    4      01/22/2020         Fujian  Mainland China  1/22/2020 17:00   \n",
            "4    5      01/22/2020          Gansu  Mainland China  1/22/2020 17:00   \n",
            "\n",
            "   Confirmed  Deaths  Recovered  \n",
            "0        1.0     0.0        0.0  \n",
            "1       14.0     0.0        0.0  \n",
            "2        6.0     0.0        0.0  \n",
            "3        1.0     0.0        0.0  \n",
            "4        0.0     0.0        0.0  \n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df = copy.deepcopy(df_orig)\n",
        "df.drop('Last Update',inplace=True,axis=1)\n",
        "date_lst  = df.ObservationDate.unique()\n",
        "date_dict = {}\n",
        "\n",
        "for i in range(len(date_lst)):\n",
        "    df_temp = df.loc[df['ObservationDate'] == date_lst[i]]\n",
        "    date_dict[date_lst[i]] = df_temp"
      ],
      "metadata": {
        "id": "Pb7xZNiuQGrf"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "date_tot_tup_dict = {};\n",
        "for date, df in date_dict.items():\n",
        "    tup_temp = (df['Confirmed'].sum(), df['Deaths'].sum(), df['Recovered'].sum())\n",
        "    date_tot_tup_dict[date] = tup_temp"
      ],
      "metadata": {
        "id": "yAtbOw_fQSI5"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "df_date_tots = pd.DataFrame(date_tot_tup_dict)\n",
        "df_date_tots = df_date_tots.transpose()\n",
        "df_date_tots.columns = ['Confirmed', 'Deaths', 'Recovered']\n",
        "print(df_date_tots)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9LV80w1wP7Sg",
        "outputId": "c1566471-cfd5-43ae-8225-f8f016dafbd7"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "            Confirmed   Deaths  Recovered\n",
            "01/22/2020      555.0     17.0       28.0\n",
            "01/23/2020      653.0     18.0       30.0\n",
            "01/24/2020      941.0     26.0       36.0\n",
            "01/25/2020     1438.0     42.0       39.0\n",
            "01/26/2020     2118.0     56.0       52.0\n",
            "...               ...      ...        ...\n",
            "03/27/2020   593291.0  27198.0   130915.0\n",
            "03/28/2020   660706.0  30652.0   139415.0\n",
            "03/29/2020   720117.0  33925.0   149082.0\n",
            "03/30/2020   782365.0  37582.0   164566.0\n",
            "03/31/2020   857487.0  42107.0   178034.0\n",
            "\n",
            "[70 rows x 3 columns]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.figure(figsize = [15,5])\n",
        "plt.plot(df_date_tots['Confirmed'], label = \"Confirmed Cases\")\n",
        "plt.plot(df_date_tots['Deaths'], label    = \"Deaths\")\n",
        "plt.plot(df_date_tots['Recovered'], label = \"Recovered Cases\")\n",
        "\n",
        "plt.legend()\n",
        "plt.xlabel(\"Day\")\n",
        "plt.ylabel(\"Count\")\n",
        "plt.xticks(rotation = 90);\n",
        "start_date = str(date_lst[0]);\n",
        "fin_date   = str(date_lst[len(date_lst) -1])\n",
        "plt.title(\"Total Cases, Deaths, and Recoveries : \" + start_date + \" - \" + fin_date);\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 401
        },
        "id": "arvXSgJ2PtRg",
        "outputId": "592e483a-705c-4638-943d-54b858fa1b49"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1080x360 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAA5EAAAGACAYAAAA0x+aoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdd3wVVfrH8c+5SQgdkShNsIIFFAUUFrNINyCKoA4KKqCCro11VfaHvaCia0NdEBVBFISxICBNBETXgoi6IgorCopSJAEpCQkk9/z+mEm8hJQbSDIp3/frNS9yZ54588xMbrjPPTNnjLUWERERERERkWiEgk5AREREREREyg8VkSIiIiIiIhI1FZEiIiIiIiISNRWRIiIiIiIiEjUVkSIiIiIiIhI1FZEiIiIiIiISNRWRIhI4Y8wxxhhrjEkMOpfKxhjTyT/2RwWdy8GqCPtQXIwxk4wx7wedh4iIVGwqIkUkh/9BvKBpfSHrv2SM+aAE84s1xtxkjPncGLPLGLPTGPOVMeZOY0zdktpuSTDGDI44rmFjzA5jzNfGmCeNMUeX0DbXGmPuK4m2yzq/uMo+3lnGmF+NMZONMY2Dzq2YDQcuCToJAGNMnDHmMWPMJmPMHmPMf4wxbXLF9DXGzDPGbPbPzeUFtHe/MWaG/3fgYf+9v8sYk2yMWWCMaZcr/nZjzKfGmO3GmD/87Sfl0/YQY8xXRVnPGNPOGPOJMSbd38dHjDExEcv/aox5y/9d22OM+cEYc58xJj5XO7WMMS8aY1KMMan+8Tg+uqNcOGNMc//4pPnH6nljTI2I5dWMMe8aY37x92WLMeYdY8wp+bS33hjTJ9r1/L/PH/l/rwv8sscYs9QYM9wY09QYM94/Znv8Yzgxr/erMWaEMeZnY0yG/zvRI9fyYjmfIrI/FZEiEqlhxHSRP691xLwzA8oLY0wcMAd4CHCBLsBpwJ1Ae2BQULkdgiy849oIaIu3b+2Bb416ZUvCR3jHuykwADgDeCPQjIqJ//7AWrvDWrs96Hx8/wKuBq7F+9vxE/C+MaZBRExN4HPguija6wu8DcQDfwGewHu/dAI2+21HFl9dgJeBzsBZwCfAu8aYswtoO6r1jDFNgIXAGqAN8Dd/Px+KaPNs4Ee837VTgH8C1wNP59r2q0BX4GIgETDAQmNMtcIPScGMMTWBRUAm0AFwgCRgQkSY9ffFAU4EzgNigUXGmKq52msNJAALirBePDCL/Y9NXrkegXfMZvjt1QD+DrQELgVaAPNzFep/B+4H7gZO9/OZbYw5LaLp4jqfIhLJWqtJkyZNB0x4H8wscFTEvF7ACiAD+B0YC9Twl93nx0dOg/1lw4Gvgd14H/amAQ0j2j3Gj08sIJ9bgTDwl3yW1/X/PRbvw+BGIA1YCVyRKzYR+BjY5U//Bc6NWF4fmARs9Zd/DHSMWB4HPAn86h+LTcC0Ih7fwUBmHvNj/e2tBWIi5nf35+8BfgMmAvUilrcG5vnnZTewHEiKWP5BHufnmIjz3B340D9m3wE9c+V1B14RkOEflwVAtSLsb13gNeAXfx/W+OfURMRMAt4HhgE/AzvxPnzWz9XWTf6xT/PzuJJcv6t5bH8S8H4e7VigdrTH2Y/pj/c+SAdS/OOe/fsXB4z2193rH8sBEetOAd7LI795wGtFON/Zx+omYD3ee6NaPvt5Kd77L92PfRL/fRvN++Eg/37U9rc3LGJeDN77/7581rHA5fksO8E/nnXzWR4DbAduKiSvb4Ancs2r6R/nltGuBzzs/w6GIubdAKRGHts82vkHkBLxurm/3z1yvVcy8P9+HuJ5GObvW52Ieef52zy2gPVa+TGtcs0fBbxZ1PX8ZZ0o4H0KXAN8UUDbrf31T/VfG/+98XCuuOXApKL8Hhzs+dSkqTJP6okUkaj43+zOwis0WuH1/PUGnvdDHgemAp/yZ8/l9IgmbgNOxfvGvyleIVkUVwCLrbWf5rXQ/tn7UhNYDPT0t/cCMNEY09nfj1h/P5bhfShpjVcAp/nLqwFLgFp+G2cAc/F6Bk72t3ET3rfvlwPNgAuAz4q4P3my1mbifcg/3t82xpguwEy8Y3YacCFeAfi2Mcb4q9bGO96d/X1aAMwyxjT3l/fDKyCe4M/zsyFi04/jfZBqhXdsphv/EmFjTD/g//C+DGiGV+DMK+KuxQPf+rmfAjyI14MwOFfcmf4+nAeci3cOH89eaIzpAzyFd4xOx+uV/lcRc8EY0wiv5yfLn6I6zsaYIXjF8Dt4x7kzMB+viAHvGA7lzx6U14DXjDFd/eWvAF397Wfn0hDvmE6ONg/fWXi9LH3wztvePPZzMDAO77yfgldwd8N/3xb2fsjn2GXfwzw4vxi83px4/9gAYK3NwuvtOZhe9r7ABzb/XtZqQBW8D/15MsaE8N4nuWN6Ahustd8WYb2z8b4MCEfMmw9Ux3/f5uOwPNrZh9dbCOT8LfucgztOuZ0NfGqt3REx7z28Lx3y6pHFGFMLr6D7Dfgh1+LIHtuirBeNfNv2Heb/m/27eQzeVRzzc8XNp4BjV8znU6TyCrqK1aRJU9mcyPWtMd4lV5/niumD92HkaP/1S3gf9Apr+wy/7cb+62MovCcyDXjmIPdlJvCi/3Ndf1ud8okdjPeNdGyu+YuBp/2fx/ivzcHkE7GdA3oi/WUn+Tk6/usPgNG5Ypr6MacXsI3/AndGvF5Lrl6giPPcL2JefX/euf7rW4D/AXHF/Ds2BlgY8XoSXk9qfMS8fwKbIl7/B5iSq53Hia4nMhOvlzaNP3tjH4+IKfQ44/WkPpfPNqrj9SBdn2v+DLwvQMC7jeQ34PaI5bcR0QsSZR6TgD+Amnns5/sRr9cD1+WK6ei3Vbew90M++9kYWA30LSBmgN9ulVzz/wWsymedgnoiPwH+VsD2XvL3tWYBMXf5x+yoXPOnAo8WZT3//ZC7B6yGvw+X5NPOyXi96zdGzLsD2JhH7BvAnGJ4j70HTM1j/tbI30F/3qP++8Pi9aCfkGt5c7wvKuoUZb2IuE7k8z7F+9IuHTg5n3Vr4vWmvxkxr4PfXvNcsTcAqSV9PjVpquyTeiJFJFot8HohIy3Fu6QozwEYshlv9MwFxpgNxphdeIUAwNFF2L4pPASMMdWNMaONMauMMduMMbvxLsM9GnK+5X8JWOAPYPF/xpgTI5o4E2gA/GGM2Z09AX/F64UD79LCU4G1/iAVFxljqhRhXwrdDf9fG5HT33Pl852/rJm/30cYY8YaY1b7g0fsxjtn0R7jr7N/sNZuweudq+/PcvEu0/zZeAPUXOH3OkS/Q8aE/GP9tfEG99iNdx9c7vxWW2szIl5vjMgDvN+1T3Kt8x+iswyv9/IsvJ7QT/E+UGYr8DgbY44EmuB9MM/LCXi9YXm9T1oAWK+n4zW8nvVsV+AVxtm9IIWeb9/31trd+e2sf4/Z0cCTudrK7kU+IYr3wwGstb9Za0+y1s4oKK64+D21Z+H1/ua1fDReb+0F+R0PY8z1eAXbxdbaXyPmV8Hr9c5zX/Jb7yD2oRne7800a+1zB9tORHurIs7pqkNtz/cvvC/4OuNduj4j1/u8H96XITuKuF40zgPWW2u/z73AeIMAzcL7EujqIrabu61iOZ8i4t17IyJSYowxTfEuB30VeABIBo7Cu5+rKIXXGgopVn3/wush/Ye/TirepXx1sgOstUONMWOAHniXET5ojLnRWjser6foe7xLq3JL89f/2hhzrL9uZ7wetQeNMe2ttTuLsE/5aeH/+5P/bwjv2/5X84jd7P87Ca+3agSwDu8+qGlEf4wPuBTS3y7W2t+MMSfh7WsXvEEsHjXGtLPWbshjvbzcCozE69X8Cu/eu1vwPjwWlIclyi8QorDHWrvW//lbfxCWZ/EuP4XCj3P1YspjMjDCGHO6//o04LKI5dGcbyjg0s2IdsC7DHlJHst/hULfDwdrk/9vA7ze22z1I5ZF60K8qyD2W8+/tHcM3rHraq39Jq+VjTG34V06fYG1NvfjT7ri9aItK+J6m/D2LVL9iGWR7bTEu4x3Jt6ALbnbSTDGxFjvct/Itv6X1/74euF9sQPe5bD52YT3xUdkPnHA4bnztNYm4/19/sEY8wne/b4D+fOWhb7sPyBPtOtFI89LWY0xdfAGVIsDuuUqYCN/xyKPVZ6/Y8V1PkXEo55IEYnWKrzL4CKdg/chP/ub8L38eW9YtjPx7lf6u7X2Y2vtGvbvWYrWa0AXY8xf8lpo/nzER0e8Xh3XWvtfvEKsee54a+231tonrbU98T4YDfMXfQEcB+y01q7NNW2MWH+3tXaGtfZmvJFVT8Y7HofEv0ftH3iXnn4VkVOLPPJZG9Hz0hEYa62dZa1diffB57hczed1fqJirc2w1s631o7A64WtjvfhPlodgfnW2pettV/5xVyzwlbKw3d4l7FFyvPerijcBwwxxrT1Xxd4nK21v+MVXj3yaW8t3uWseb1Pcu63s9auwhuY5wq8exRXWGu/i4iP5nwXyu9R3gCcmE9b6RGx+b0fDlb2AFznZs/w70XrRvQ9x9n6kaun0HgjdL6M9ziTTv57/QDGmAeAe4FeeRQO2W2/Y621RVzvY6C7v0/ZkvC+aMp+32KMOROvJ9rFuxx3v+347cThfTmTvc5hQDsKOE7W2p8jzuPP+cX57f/FGFM7Yl53vM9/HxewHnhf3lT1czoK7z7XmYWss9960TDeI096ceA5TuDPLz+6W2v/yLXqerwrFc7NNT+JXMeuuM6niEQoretmNWnSVL4mDrwn8jS8y4mewrtnLwmvh+HViHVux7vXpgXeMPDx/nphvMsGj8UrPFYTcR8W0d0TGYf3bf5OvHvI2uJdqpeEd5nbcD/uTb/9s/B6Ll8CduDfq4l3yeGjeAMvHI33qIBV2fuB9+HnW7wR/nr4ubXD60W7MGI/B/r7eSze5VGZeB/Woz2+g/11GvhTM7wPxJ/g9dIlRsR2xuttyB5M5nj+HKa/mh/zBd4Hp1P9mFn+fk+KaGcO3r2cTf3zE8p9niNiM/lzdN2r8XrrWvnH7Cq8y127FmF/Hwe2+PvSHG+Uxx14l7Blx0ziwJFFLwdsxOu+fm7Zg/wMweudi+aeyPfzmD8DWFCE43yNH3M33hcHLYAbgQR/+WN4PTGX+Pt5B97vf9dc270Zr9DfRK4RRaPMI7/92W8+XqG6F+9ROC3xHp1wITA+mvdDPsey0Hsi/bin8f4e9PaP0yS8EVQjR2Y+3N/H0/1zeIf/c1N/eV0//+Mj1onFu2cwBe8y8wYRU81c29/j729kTB1/eQjvdzL3uSlwPT+mCd7fogn+vl3g5zM6IqajH/NyrnYa5NreO3hfQJzj7/tcvC+/oh79uIBzUBPvi4R38d6/nfGuVJgWEdMJr4e0Fd7fhrPx/n7sjDgPNwIf5Wq70PX8uKb+fl3jn+Me/uvD/eW9gV9ytd0Q7wujFXi//5HHr0pE3N/xCr3L8f5fGo335UWriJhiOZ+aNGnafwo8AU2aNJXNicIf8bEVb9THyEcFHO5/ANrB/o/4uMH/ILMHr9BJoohFpB8Xi1c8fIF3Kd9OvG+J7wAO82Oa4I1Mmor3Af1+/4PBB/7yhniXTWU/nmMj8GKuDxT1/H3LfkzDb3jFxhn+8mv947CTPx+n0SePY9epgH0ZzJ+Du4T9tv6LVzgcnUf8X/EuAd7l79v3/oejWH/5qXgF6B68b+iv9+MnRbTRFvjSj7Hs/4iPgorIfn7b2/E+sH0LXB0Rm33+Bhewv3XwemN24n04+zfefYnrI2ImUUgR6c8b7p+TPf4+DsprH3Ktc0Db/vzswTmyfxcLPM5+zED/XGX4+zKHP3//CnzER0QbCf7yvfgFaBHPd377k9cxvBDv/s80//h/DdwT7fshj20Uer4jjsVjeEV+Ol5vT9sC3geR0yR/+RXAN/lsP6/pvoi4/GImRRzjbRw4iFaB60XEtcd7X6T7+/gI+z+WZ1J+beVqp5Z/zLf552g++QxOczAT3hcH7/ltpwDj2f/vdju8wZxS/N+Bn/Gu/DglImYR8I9c7Ra6XiHHIfvvywRyDZpWwO/FAX9X8Qbf+sXP4WtyPZ6muM6nJk2a9p+MtRYRESk+xpir8D6AnGgPvASrwvEfSTEH7xLMnwqLF4mWMWYGXhF5bwm0/RReb9ig4m67IjHG1MMrqppba9cVc9vZzw51rLVLirNtESlZGlhHRKT49Qb+WRkKSF9vvEckqICU4vYp3iXqJeF7vKsapGD18P6eFWsBGdH2sxw4orGIlHHqiRQREREREZGoaXRWERERERERiZqKSBEREREREYmaikgRERERERGJmgbWyZtuFBURERERkcrO5DVTRWQ+Nm7cGHQKB0hISCA5OVmxilVsJYktK3koVrGKVaxiy35sWclDsSUbW5oaNWqU7zJdzioiIiIiIiJRUxEpIiIiIiIiUVMRKSIiIiIiIlHTPZFRstaSnp5OOBzGmDzvLy1xW7ZsISMjQ7HFHGutJRQKYa3GUxIRERERKYyKyCilp6cTFxdHbGxwhyw2NpaYmBjFlkBsZmYm27Zti6pNEREREZHKTJezRikcDgdaQErJio2NJTMzM+g0RERERETKPBWRUQrqElYREREREZGyREVkOfL777/zt7/9jQ4dOpCUlMQVV1zBjz/+eFBtLVu2jM6dO9O9e3c2bdrE0KFDiznbvDVr1izP+cW5byIiIiIiUnJ0fWY5Ya1l8ODBXHzxxYwbNw6AVatWkZyczPHHH1/k9t5++21uvPFGLrroImJjY3nxxRcPiMnMzCyVS3iLe99ERERERMqD8jqwo4rIcuLjjz8mLi6OK6+8MmdeixYtAO+Xb9SoUSxZsgRjDDfffDMXXXQRn3zyCU8++SR169ZlzZo1nHbaaTz77LO8/vrrvPvuuyxdupQlS5Zwxx13cPnll7N48WKmT5/OvHnzSE1NJRwO4zgOCxYsIC0tjXXr1nH99deTnp7OW2+9RZUqVXj11VepW7cu69ev58477yQlJYVq1arxr3/9i5NOOolffvmFG264gbS0NHr06FHkfUtNTWXIkCHs2LGDzMxMRowYwXnnnUdaWhrXXnstmzZtIhwOM3z4cPr06cM333zD/fffT2pqKocffjjPPvss9erVY8KECbz66qvExsbSrFmznGJVRERERCQINhwm/MSdpHU5D9okBp1OkaiIPAjhaS9iN6wr1jZNk2MJXZr/JaXZRWBe5s6dy6pVq1i4cCHbtm2jV69eJCZ6v4jffvstixcvpkGDBvTp04fly5czYMAAPv/8c7p160bv3r3ZuHHjfu2tXLmS999/n7p16zJ9+nTWrFnDggULyMjI4Oyzz+aOO+7gvffe49577+XNN99k6NChjBgxgtGjR3Pcccfx5ZdfMnLkSGbMmME999zDlVdeySWXXMKkSZOKvG/x8fFMmDCBWrVqsW3bNs4//3x69erFkiVLaNCgAa+++ioAO3fuZN++fdx1111MnDiRevXqMXPmTB5++GGeeOIJ/v3vf/Ppp58SHx/Pjh07CjsdIiIiIiIl66tP4X+rML0uDjqTIlMRWQF8/vnnXHjhhcTExHDEEUfQvn17vvrqK6pXr87pp59Oo0aNAK93b8OGDZx11lkFttexY0fq1q2b87pDhw7UrFmTmjVrUrt2bbp37w7AySefzHfffUdqaiorVqzg2muvzVln7969ACxfvjznUtmLLrqIhx56qEj7Zq1l9OjRLFu2DGMMmzdvZuvWrZx00kk88MADPPTQQ3Tr1o127dqxevVq1qxZw6WXXgp4I+rWr18/J9cbb7yRpKQkkpKSipSDiIiIiEhxsuEswjOnQsMmVE3sRur27UGnVCQqIg9CQT2GJaV58+bMnTu3yOtVqVIl5+eYmJioHmNRvXr1fNsIhULEx8cD3oi1WVlZhMNhateuzcKFC/Nsr7CRbQvat7fffpuUlBTmzZtHXFwc7dq1Iz09neOPP5758+ezePFiHnvsMRITE0lKSqJ58+bMnj07Z/3sR3dMnjyZzz77jIULF/LMM8+waNEiPbJFRERERAJhl/8HNm3ADBuBifL552WJRmctJxITE8nIyOC1117Lmffdd9+xbNky2rVrx6xZs8jKyiIlJYVly5bRunXrUsutVq1aNGnSJKd4s9ayatUqAM4880xmzpwJeAVhXgrat127dpGQkEBcXBwff/wxv/76KwCbN2+mWrVqXHTRRVx33XWsXLmS448/nm3btvHFF18AsG/fPlavXk04HGbjxo2cffbZ3HnnnezatYvU1NQSOx4iIiIiIvmxWVnY2dOg8dGYNh2CTuegqCumnDDGMGnSJO68807Gjh1LfHw8Rx11FPfffz9nnXUWK1asoHv37hhjuPPOOznyyCNZvXp1qeX33HPPMXLkSMaMGUNmZiZ9+vShVatWPPDAA9xwww2MHTs234F1Ctq3fv36MWjQILp27cppp53GCSecAMDq1asZNWoUxhji4uJ45JFHqFKlCuPHj+eee+5h586dZGVlMWzYMI4++mhuuukmdu3ahbWWq666ijp16pTasRERERERyWaXLYUtvxH620hMqHz26amILEcaNGjA+PHj81x29913c/fdd+83r0OHDnTo8Oe3G5H3Iz799NM5Pzdt2pTFixcD0L9/f/r375+zLPfrL774IueS2MhlTZs2ZcqUKQfk1bRp0/0uL/3nP/9Z5H2LXB+8S1QbNWpEp06dDoht2bLlfj2e2ZezvvPOO3m2LSIiIiJSWmxmJvbdadD0ODijfdDpHLTyWfqKiIiIiIiUM/bTxbB1M6ELBhQ6bkhZpiJSRERERESkhNnMfdg5LhzTDE47M+h0DomKSBERERERkRJm//M+pPxOqE/57oUEFZEiIiIiIiIlyu7bi537Bhx/ErQovacolBQVkSIiIiIiIiXIfvQebE8m1Gdgue+FBBWRIiIiIiIiJcbuzcDOfROat4STTgs6nWKhIrIcadiwId27d6dz585069aN559/nnA4fFBt7dixg0mTJuW8/uSTT7jyyiuLKVMREREREQGwS+fDjm0V4l7IbKX2nEjHcW4BrgEssBIYAjQEpgH1gBXAFa7r7nUcJx6YDLQBUoD+ruuu99sZCVwNZAE3u667wJ+fBIwBYoCXXNcd7c8/Nq9tlMY+F7eqVauycOFCAJKTk7nhhhvYvXs3t912W5Hb2rlzJ5MnT2bw4MHFnKWIiIiIiADYjHTsvDfh5FaY5i2DTqfYlEpPpOM4jYGbgbau67bEK/QuBR4FnnJd9wRgO15xiP/vdn/+U34cjuOc4q/XAkgCxjqOE+M4Tgzwb6AncApwmR9LAdso1xISEnjssceYOHEi1lqysrJ48MEH6dWrF926dWPy5MkApKam4jgO5557Ll27dmXBggUAPPzww/z88890796d+++/H4C0tDSGDh1Kx44dufHGG7HW5sR26tSJbt26cd999wWyvyIiIiIi5Y1dMgd27SB0wYCgUylWpdYT6W+rmuM4+4DqwCagC5B9RF8B7gPGAX38nwHeBJ5zHMf486e5rpsBrHMcZy1wlh+31nXdnwAcx5kG9HEc5/sCtnHQXvpiC+u2px9KEwc4tm5Vrmlbv0jrHH300YTDYZKTk1mwYAG1atVi7ty5ZGRk0LdvXxITE2nUqBETJkygVq1abNu2jfPPP58ePXpwxx13sGbNGhYuXEhsbCwffvgh3377LYsXL6ZBgwb06dOH5cuXc8IJJzBv3jw+/PBDjDGkpqYW636LiIiIiFRENj0Nu+BtaNkac8LJQadTrEqlJ9J13d+Ax4Ff8IrHHXiXlv7hum6mH/Yr0Nj/uTGwwV8304+vFzk/1zr5za9XwDYqlKVLl/Lmm2/SvXt3evfuzbZt21i3bh3WWkaPHk23bt3o378/mzdvZuvWrXm2cfrpp9OoUSNCoRAtWrRgw4YN1K5dm/j4eG699Vbmzp1LtWrVSnnPRERERETKH7voXdi9i9AFA4NOpdiVSk+k4zh18XoRjwX+AN7Auxy1zHAcZxgwDMB1XRISEvZbvmXLFmJjvcN1Xfvg6tDsHADWr19PTEwMDRo0wBjDI488QufOnfeLnzZtGtu2bWPhwoXExcXRtm1bMjMziYmJ2a+9mJgY4uPjc17HxcVhraVq1aosWLCAjz76iNmzZzNp0iTefvvtg8q3rMeGw+EDzntBbSpWsSUdW1byUKxiFatYxZb92LKSh2I9ofQ9sHAmVc5MpO6Zfym2dsuK0rqctRuwznXdrQCO47wNnA0c5jhOrN9TeBTwmx//G9AE+NVxnFigDt4AO9nzs0Wuk9f8lAK2sR/XdV8AXvBf2uTk5P2WZ2Rk5BReQcrM9DpVU1JSuP322xk8eDBZWVl07NiRiRMn0r59e+Li4vj555854ogj+OOPP6hXrx7GGJYuXcqGDRvIysqiRo0a7N69m8zMTGJjY8nKysJam9N+OBwmKyuLHTt2sGfPHjp16kTr1q3p0KFDTkxhYmNjy1WstZbc5z0/CQkJilVsiceWlTwUq1jFKlaxZT+2rOShWE/V99/Bpu4iM+miQtcp6nkuLY0aNcp3WWkVkb8A7R3HqQ7sAboCXwBLgIvxRk8dBMz042f5rz/1ly92Xdc6jjMLmOo4zpNAI6AZ8DlggGb+SKy/4Q2+M8BfJ79tlDvp6el07949pyfx4osvZtiwYQAMGDCADRs2kJSUhLWWevXqMWHCBPr168egQYPo2rUrp512GieccAIAhx9+OGeeeSZdunSha9euB/RgZtu9ezdXXXUVGRkZWGtzBuEREREREZED2dRdpM2eDq3/gml6fNDplIhSKSJd113mOM6bwJdAJvAVXq/fHGCa4zij/HkT/FUmAK/6A+dswysKcV13leM4LvCd384NrutmATiOcyOwAG/k15dd113lt/XPfLZR7mzatCnfXrVQKMTIkSMZOXIksH8P3OzZs/Nc59///vd+sR06dMhZ9tBDD+X8PGfOnJyfi9KzJyIiIiJS2dj33sHuSatwI7JGKrXRWV3XvRe4N9fsn/hzdNXI2HTgknzaeQh4KO/eSL8AACAASURBVI/5c4G5eczPcxsiIiIiIiLFyW5PwS6aTfzZXchsfHTQ6ZSYUhmdVUREREREpKILT38RwmFqDrwu6FRKlIpIERERERGRQ2RXroAVn2DOc4htUCGfKphDRaSIiIiIiMghsHszCL8+HhochenRN+h0SpyKSBERERERkUNg574BWzcTGngdJi4u6HRKnIpIERERERGRg2Q3/Yqd/zamfSfMSacFnU6pKLXRWeXQNWzYkJNOOomsrCyaNGnCM888Q506dYJOq0ieeOIJatSowXXXHXiz8RtvvMG4ceMwxhATE0O/fv3yjBMRERERKQustYSnjIP4eMwlQ4JOp9SoJ7IcqVq1KgsXLmTx4sUcdthhTJo0KeiUAP/NEw4fUhuLFi3ipZdeYurUqSxatIjZs2dTq1atYspQRERERKT42WUfwJqVmL5XYmrXDTqdUqMispxq06YNmzdvBmD9+vUMHDiQpKQk+vbty9q1awHYunUrV199Nd26daNbt24sX74cgPHjx9OlSxe6dOnCiy++CMDDDz+8X1H6xBNP8PzzzwMwbtw4evXqRbdu3XjssccA2LBhA3/961+5+eab6dKlCxs3btwv7vHHH89pa8yYMSQmJnLhhRfy448/5rk/zzzzDHfffTcNGjQAID4+noEDBwIwZcqUnHaHDh1KWloaALNnz6ZLly5069aNfv36AZCVlcWDDz6YEz958mQAtmzZQr9+/ejevTtdunRh2bJlh3gGRERERKQys6m7se7LcGxzTMdzg06nVOly1oPw7Zdp7Pwjq1jbrH1YDC1bV48qNisri//85z9cdtllAIwYMYLRo0dz3HHH8eWXXzJy5EhmzJjB3XffTfv27ZkwYQJZWVmkpqbyzTff4Lou7777LtZaevfuTWJiIhdccAH33nsvgwcPBrwCbcqUKSxdupR169YxZ84crLUMGTKEzz77jMaNG7Nu3Tqefvpp2rRpc0Dc4MGD+fTTT4mPj2fWrFksXLiQzMxMkpKSOO20A68VX716dZ7zAXr27JlTUD766KNMnTqVwYMH8/TTTzNlyhQaNmzIjh07AHj99depVasWc+fOJSMjg759+5KYmMjcuXM555xzGD58OFlZWezZs6eop0hEREREJId951XYvYvQ3+/DhCpX35yKyHIkPT2d7t27s3nzZpo1a0bHjh1JTU1lxYoVXHvttTlxe/fuBeDjjz9mzJgxAMTExFC7dm0+//xzkpKSqF7dK1h79uzJZ599xpAhQ0hOTmbz5s2kpKRQp04dGjduzIQJE1i6dCk9evQAIC0tjXXr1tG4cWOOOuoo2rRpA8DSpUsPiPvpp5/YuXMnSUlJVKtWDYDu3bsXeb/XrFnDY489xs6dO0lNTaVz584AtG3blltuuYXzzz+fnj175uTx/fffM2fOHAB27drFunXrOP3007n11lvJzMzk3HPPpWXLlkXOQ0REREQEwK77H3bpfEyX3pimxwedTqlTEXkQou0xLG7Z90Tu2bOHAQMGMGnSJBzHoXbt2ixcuPCQ2+/duzdz5szh999/54ILLgC8+x1vvPFGrrjiCgBiY2PJzMxkw4YNOYVoXnHZsePGjYtq2yeeeCLffPMNiYmJByy75ZZbmDBhAi1atGD69Ok5l6I++uijfPnllyxatIiePXsyb948AEaNGkWnTp32yxfgrbfeYtGiRdxyyy0MGzaMSy65pIhHSEREREQqO5uVRfi1sVCnLqbPwKDTCUTl6netIKpVq8aDDz7I+PHjqVatGk2aNGH27NmAV8ytWrUKgMTExJx7ArOysti5cyft2rVjwYIF7Nmzh7S0NObPn0/79u0BuOCCC5g5cyZz5syhd+/eAHTq1Inp06eTmpoKwKZNm0hOTj4gp7zitm7dSvv27XO2t3v37nyL3ZtvvplRo0bx+++/A15v6tSpUwHYvXs39evXZ9++fcyYMSNnnfXr19O6dWtuv/126tWrx8aNGznnnHOYPHky+/btA+DHH38kLS2NX3/9lSOOOIKBAwcyYMAAVq5ceQhnQEREREQqK/vBXPjlJ0L9r8FUC6ZzKWjqiSynWrZsycknn8w777zDc889x8iRIxkzZgyZmZn06dOHVq1a8cADDzBixAimTZtGKBTikUceoW3btlxyySWcd955AFx22WWceuqpZGZmcuKJJ5KamkqDBg2oX78+AOeccw4//PBDTs9kjRo1eOaZZ4iJidkvn9xx1atXZ9y4cZx66qmcf/75dO/enYSEBE4//fQ896dbt25s2bKFSy+9FGstxhj69+8PwO23307v3r2pV68eZ5xxRs7AOqNGjWLdunVYa0lMTKRFixaccsopbNiwgaSkJKy11KtXjwkTJvDJJ5/w/PPPExsbS40aNXIu8xURERERiZb9IwX7zmvQ4gxoc3bQ6QRGRWQ5sm7dupxLMwFeeeWVnJ+nTJlyQPwRRxzBxIkTD5h/7bXX7ncPZaRFixYdMO+aa67hmmuuAfa/PHTx4sX5xkXGDh8+nOHDhxe0awD0798/p3CMNGjQIAYNGnRAuy+99NIBscYYRo4cyciRI/eLdRwHx3EKzUFEREREJD92+gTIzCQ04DqMMUGnExhdzioiIiIiIlKIjK8+w37xH8x5l2CObBh0OoFSESkiIiIiIlIAuzeDXS88AfUbY869KOh0AqciUkREREREpAB23ltkbf6N0MDrMHFxQacTOBWRUbLWBp2CiIiIiIiUMrtpA3b+m1Tt2ANzcqug0ykTVERGKRQK7TeojVQsmZmZxMZqnCkRERER+ZMNhwlPfg6qVKXmkJuDTqfM0KfmKFWtWpX09HQyMjICG4kpPj6ejIwMxRZzrLWWUChE/fr1SUlJiapdEREREan47NJ5sPZ7zJDhxBx2OOTxvPTKSEVklIwxVKtWLdAcEhISSI7yF1exRY+tzMM0i4iIiMj+bMpW7FuT4ZQzMH/pEnQ6ZYouZxUREREREYlgrSU8ZRzYMKErrldnQy4qIkVERERERCLYzz+ElV9g+l6OSagfdDpljopIERERERERn921EzvtRTi2OaZL76DTKZNURIqIiIiIiPjs9BdhTyqhQTdhQjFBp1MmqYgUEREREREB7MoV2GVLMT0vxjQ+Ouh0yiwVkSIiIiIiUunZ9DTCr42Fhk0wvZyg0ynTVESKiIiIiEilZ2e8BtuTCV15IyYuLuh0yjQVkSIiIiIiUqnZH1djl8zBdOqFOeHkoNMp81REioiIiIhIpWX37SP8yrNQtx6m3xVBp1MuqIgUEREREZFKy859AzZtIHT5DZiq1YNOp1xQESkiIiIiIpVS5s8/Yue9iWl3DubUNkGnU26oiBQRERERkUrHhrPYOXY0VKuO6T806HTKFRWRIiIiIiJS6djF77Lvf6swlw7F1KoddDrliopIERERERGpVOy2ZOyM16jS+i+YszoGnU65oyJSREREREQqFTvjVQhnUXvYrRhjgk6n3FERKSIiIiIilYZd9wP2syWYbn2Iqd8o6HTKJRWRIiIiIiJSKVhrCbsvQa06mF6XBJ1OuaUiUkREREREKgX7xcew9nvMhZdjqumZkAdLRaSIiIiIiFR4dt9e7FuT4KhjMIndgk6nXFMRKSIiIiIiFZ59fxak/E7IuRoTigk6nXJNRaSIiIiIiFRodsd27Jw3oNVZmJNbBZ1OuaciUkREREREKjQ7cwpk7iN0yVVBp1IhqIgUEREREZEKy/7yE/Y/CzGdz8PokR7FQkWkiIiIiIhUSN4jPSZAjZqY3v2DTqfCUBEpIiIiIiIV03+XwZqVmAsGYGrUDDqbCkNFpIiIiIiIVDg2cx/hNyZCwyaYjklBp1OhqIgUEREREZEKxy6eA79vInTJVZgYPdKjOKmIFBERERGRCsXu2ol9dzq0bI05tU3Q6VQ4saW1IcdxDgNeAloCFrgKWANMB44B1gOO67rbHccxwBigF5AGDHZd90u/nUHAXX6zo1zXfcWf3waYBFQD5gLDXde1juMcntc2SnZvRUREREQkKHb2VMjYo0d6lJDS7IkcA8x3XfckoBXwPfB/wCLXdZsBi/zXAD2BZv40DBgH4BeE9wLtgLOAex3HqeuvMw4YGrFe9oXP+W1DREREREQqmMwN67BL52M6JmEaNQ06nQqpVIpIx3HqAB2BCQCu6+51XfcPoA/wih/2CnCh/3MfYLLrutZ13c+AwxzHaQicCyx0XXeb35u4EEjyl9V2Xfcz13UtMDlXW3ltQ0REREREKphdE5+F+GqYCwYEnUqFVVqXsx4LbAUmOo7TClgBDAfqu667yY/ZDNT3f24MbIhY/1d/XkHzf81jPgVsYz+O4wzD6/XEdV0SEhKKuIslLzY2Nuq8FKtYxZb/2LKSh2IVq1jFKrbsx5aVPIKOzVjxKX989Rk1B99EjWOPCySHkowtK0qriIwFWgM3ua67zHGcMeS6rNS/f9GWZBIFbcN13ReAF/yXNjk5uSRTOSgJCQlEm5diFavY8h9bVvJQrGIVq1jFlv3YspJHkLE2M5PwhKeJaXgUae06sSeKtsvLvgWhUaNG+S4rrXsifwV+dV13mf/6Tbyicot/KSr+v7/7y38DmkSsf5Q/r6D5R+UxnwK2ISIiIiIiFYRdMgc2baDm4JswsXFBp1OhlUoR6bruZmCD4zgn+rO6At8Bs4BB/rxBwEz/51nAlY7jGMdx2gM7/EtSFwA9HMep6w+o0wNY4C/b6ThOe39k1ytztZXXNkREREREpAKwf6RgZ02FU9sSf2Zi0OlUeKX2iA/gJmCK4zhVgJ+AIXhFrOs4ztXAz4Djx87Fe7zHWrxHfAwBcF13m+M4DwLL/bgHXNfd5v98PX8+4mOePwGMzmcbIiIiIiJSAdg3JkFmJqFLh2KMCTqdCq/UikjXdb8G2uaxqGsesRa4IZ92XgZezmP+F3jPoMw9PyWvbYiIiIiISPln16zEfr4U0/tSzJENg06nUijN50SKiIiIiIgUG5uZSXjqeKh3JKbnRUGnU2moiBQRERERkXLJLn4XNv7iXcZaJT7odCoNFZEiIiIiIlLueIPpvA6ntoVWZwWdTqWiIlJERERERMod+8ZEyNJgOkFQESkiIiIiIuWKXf0N9vMPMT0v0mA6AVARKSIiIiIi5UbOYDoJ9TFJGkwnCCoiRURERESk3LCLZ8OmDRpMJ0AqIkVEREREpFyw21Ows6bBaWdiNJhOYFREioiIiIhIuWDfeDlnMB0JjopIEREREREp8+zqb7DLP8L0vBhzRIOg06nUVESKiIiIiEiZljOYzhENMEn9gk6n0lMRKSIiIiIiZVrau643mE5/DaZTFqiIFBERERGRMstuSyZ1+gRodRam1ZlBpyOoiBQRERERkTLMvjkRG84i1P+aoFMRn4pIEREREREpk+x3X2OXf0SNfldoMJ0yREWkiIiIiIiUOXbfXsJTnocjG1Kj7+VBpyMRVESKiIiIiEiZY+e9Bb9vJDTwOg2mU8aoiBQRERERkTLFbtmInfcG5qyOmFPOCDodyUVFpIiIiIiIlBnWWsJTxkFcPMa5Ouh0JA8qIkVEREREpMywn38I3/8X0/cKTJ26QacjeVARKSIiIiIiZYJN2411J8AxzTDnnBt0OpIPFZEiIiIiIlIm2Bmvwa6dhC6/HhOKCTodyYeKSBERERERCZxd9z/s0nmYLudhjj4+6HSkACoiRUREREQkUDYri/BrY6HO4ZgLBwadjhRCRaSIiIiIiATKLpkDv/xE6NKhmKrVg05HCqEiUkREREREAmO3p2DfmQKntoXWfwk6HYmCikgREREREQlMePqLEM4idNkwjDFBpyNRUBEpIiIiIiKBsCu/gBWfYHr3xxzRIOh0JEpRF5GO41ySz/yLiy8dERERERGpDGxGOuGp46FhE0yPC4NOR4qgKD2RE/KZ/0JxJCIiIiIiIpVH6huTIHkLocv/homNCzodKYLYwgIcxznO/zHkOM6xQOSFyscB6SWRmIiIiIiIVEx24y+kzpyK6dAV07xl0OlIERVaRAJrAYtXPP6Ya9lm4L5izklERERERCooay3hKeMwVatjLh4cdDpyEAotIl3XDQE4jrPUdd1zSj4lERERERGpqOz8t+F/q6h1w0hSa9UJOh05CFHfE6kCUkREREREDoVdsxI741XMmX+latfeQacjBymay1kB8O+HfAg4HagZucx13abFnJeIiIiIiFQg9o8UwuMfg/qNMFfeqGdClmNRF5HAVLx7Im8F0komHRERERERqWhsZibh8f+CvRmEbnsIU7Va0CnJIShKEdkCONt13XBJJSMiIiIiIhWPffsVWPsd5ppbMY10EWN5V5TnRH4InFFSiYiIiIiISMVjV3yMXTgT0/k8Qu00zEpFUJSeyPXAfMdxZuA92iOH67r3FGdSIiIiIiJS/tnNvxGe9Awc2xzjXBV0OlJMilJE1gDeBeKAJiWTjoiIiIiIVAQ2I53w86MhNo7Qdf/ExMYFnZIUk6iLSNd1h5RkIiIiIiIiUjFYa7GvjYWNvxD6+32Yw48IOiUpRkV5xMdx+S1zXfen4klHRERERETKO7t0PvazDzB9BmBO0bAqFU1RLmddC1gg8oEu1v83ptgyEhERERGRcsuu+wE7/UVo2QbTywk6HSkBRbmcdb+RXB3HaQDcC3xU3EmJiIiIiEj5E965w7sPss7hhK6+BRMqysMgpLw46LPquu5m4O/AI8WXjoiIiIiIlEc2HGbH0/fDzu3eQDo1awedkpSQQ/1q4ESgenEkIiIiIiIi5Zed47L3q88w/YdijmkWdDpSgooysM5H/HkPJHjFYwvggeJOSkREREREyo/wsqXY2a9TtVMSe89JCjodKWFFGVjnpVyvU4H/uq77QzHmIyIiIiIi5Yj98hPsy09Bs1Oofd0IUnbtDjolKWFFGVjnlUPdmOM4McAXwG+u6/Z2HOdYYBpQD1gBXOG67l7HceKByUAbIAXo77ruer+NkcDVQBZws+u6C/z5ScAYvJFiX3Jdd7Q/P89tHOq+iIiIiIhUdnblF4RfeByObU7oprsx8VVBRWSFF/U9kY7jxDmOc7/jOD85jpPu/3u/4zhVirC94cD3Ea8fBZ5yXfcEYDtecYj/73Z//lN+HI7jnAJcincZbRIw1nGcGL84/TfQEzgFuMyPLWgbIiIiIiJykOz3/yU89hFofDShm+/BVNVQKZVFUQbWeQzoBlwHtPL/7YJf4BXGcZyjgPPwL4t1HMf467/ph7wCXOj/3Md/jb+8qx/fB5jmum6G67rr8J5deZY/rXVd9ye/l3Ea0KeQbYiIiIiIyEGw/1tF+LlRUL8RoVvux1SvGXRKUoqKck/kJUAr13VT/NdrHMf5EvgvcEsU6z8NjABq+a/rAX+4rpvpv/4VaOz/3BjYAOC6bqbjODv8+MbAZxFtRq6zIdf8doVsQ0REREREisiu+x/hZx+AwxMI/eMBPcqjEipKEWmKOD+H4zi9gd9d113hOE6nImyz1DiOMwwYBuC6LgkJCQFndKDY2Nio81KsYhVb/mPLSh6KVaxiFavYsh9bWnns+2kN28fcT0ydutQdNZaYekeUeg6VObasKEoR+QYw23Gc+4FfgKOBu/z5hTkbuMBxnF5AVaA23iA4hzmOE+v3FB4F/ObH/wY0AX51HCcWqIM3wE72/GyR6+Q1P6WAbezHdd0XgBf8lzY5OTmK3SpdCQkJRJuXYhWr2PIfW1byUKxiFatYxZb92NLIw/72C+HHR0J8PPaWB9huDeRqoywci4ocW5oaNWqU77Ki3BM5AngfbwCbFcCzwGLg9sJWdF13pOu6R7muewzewDiLXdcdCCwBLvbDBgEz/Z9n+a/xly92Xdf68y91HCfeH3W1GfA5sBxo5jjOsf5AP5cCs/x18tuGiIiIiIhEwW7+jfCTd0FMHKFbR2HqHRl0ShKgQnsiHcc5G7jAdd1/Avf4U/ayR4HW7H+fYlH8E5jmOM4o4Ctggj9/AvCq4zhrgW14RSGu665yHMcFvgMygRtc183yc7kRWID3iI+XXdddVcg2RERERESkEFlbNhJ+8m6wltBtD2KOzL+HSiqHaC5nvQMYm8+yJcCdwPnRbtB13Q+AD/yff8IbWTV3TDreQD55rf8Q8FAe8+cCc/OYn+c2RERERESkYHbbVrY9cRdkpBO67SFMwyaFryQVXjSXs54OzM9n2ftAm+JLR0REREREygK7czvhJ+7G7t7pPcajybFBpyRlRDRFZG2gSj7L4vjzkR0iIiIiIlIB2LTdhJ+6D/5I4bC7nsAc0yzolKQMiaaIXA30yGdZD3+5iIiIiIhUADYjg/CzD8KmDYSuv4MqJ58WdEpSxkRzT+RTwHjHcWKAd1zXDTuOEwIuxBup9R8lmaCIiIiIiJQOm5lJePyj8ONqzNDbMS3OCDolKYMK7Yl0XXcq8BjwCpDuOM5GIN1//S/XdV8v2RRFRERERKSk2XAYO/FpWPkF5vK/ETozMeiUpIyK6jmRrus+CTTGG4X1Nv/fxv58EREREREpx6y12NdfwH7+IabflYQ6JgWdkpRh0VzOCoDrujvxnsMoIiIiIiIViJ01FfvBXEyPvpiki4JOR8q4qHoiRURERESkYgq/PxP77nRMYnfMxYMxxgSdkpRxKiJFRERERCqp8CeLsdMnQOu/YC6/XgWkREVFpIiIiIhIJWS/XoZ95Rk4uRWha27DxMQEnZKUEyoiRUREREQqmb3ffkl4/GNw9AmErh+JiYsLOiUpR6IeWEdERERERMo/+/Na/njiLjiiAaGb78FUrR50SlLOqCdSRERERKSSsD+vJfz0fZiatQnd8gCmZu2gU5JySD2RIiIiIiKVgP3yU8ITnoSatal73xj+qFIt6JSknFIRKSIiIiJSgVlrse/NwL71ChzTjNANdxLbqAkkJwedmpRTKiJFRERERCoom5mJnfo89qP3MG0TMUOGY6rEB52WlHMqIkVEREREKiCbupvw86Nh9TeY8xzMBQMwIQ2JIodORaSIiIiISAVjf99I+NkHYesWzJC/E+rQJeiUpAJRESkiIiIiUoHY/60iPO5hsBD6xwOY5i2DTkkqGBWRIiIiIiIVRPjTJdhXnoWE+oRuvhtzZKOgU5IKSEWkiIiIiEg5Z8Nhdk99AfvGJDjxVEJ/G4mpUTPotKSCUhEpIiIiIlKO2b0Z2EnPkLr8I0xid8zA6zCxcUGnJRWYikgRERERkXLK7thO+N8PwfofqHnlDaQl9sAYE3RaUsGpiBQRERERKYfsr+sIPzsKdu8kdN3/UaPH+exJTg46LakEVESKiIiIiJQz9pvlhF94HKpVIzRiNObo44NOSSoRFZEiIiIiIuWEtRa7aDbWfRmaHEvoxrswdesFnZZUMioiRURERETKAZuZiZ3+IvaDeXBGe0JX/wMTXzXotKQSUhEpIiIiIlLG2bTdhMc/Bt99jUm6CNP3CkwoFHRaUkmpiBQRERERKcPs1s2En30Qft+IGXQTocTuQacklZyKSBERERGRMsr+8B3hsQ+BhdAtD2BOPDXolERURIqIiIiIlEV7lswjPPYRqFef0E13Y+o3CjolEUBFpIiIiIhImWIzM7FvvcLO92fCiacS+tv/YWrUCjotkRwqIkVEREREygi7PYXwC4/B2u+p1utiMs6/DBMbF3RaIvtRESkiIiIiUgbY7/9L+MXHYW8GZuht1O7Vj+Tk5KDTEjmAikgRERERkQDZcBg7703szKnQoLF3+WrDJkGnJZIvFZEiIiIiIgGxqbsJv/wUfLMcc1ZHzBU3YKpWCzotkQKpiBQRERERCYD9+UfCz4+G7SmYy4ZhOp+HMSbotEQKpSJSRERERKQUWWuxH72Hff0FqF2H0IhHMMedGHRaIlFTESkiIiIiUkpsRjp24hjsp4vhlDMIXXMrplbtoNMSKRIVkSIiIiIipcBu/pVtDz2J/flHzPmXYnr3x4Rigk5LpMhURIqIiIiIlCBrLfaTRdip47HxVQndfA+mZZug0xI5aCoiRURERERKiE1Lxb42Frv8IzjxVOrdPortVoPnSPmmIlJEREREpATYH1cTfvFx2J6M6XsFJqkfMfWOgOTkoFMTOSQqIkVEREREipENZ2Hnv42dOQXqJhAaMRpz/ElBpyVSbFREioiIiIgUE/tHCuEJT8HqbzBn/hVz+fWY6jWCTkukWKmIFBEREREpBva/ywlPehr27sUMvhnToSvG6P5HqXhURIqIiIiIHAK7N4Pw6y9gF78LTY8jNPQ2TIOjgk5LpMSoiBQREREROQjWWlj9Ddvenoxd/wOmWx9MvysxcXFBpyZSolREioiIiIgUgQ1nwZefEp7/Nvy81hs85+Z7MKe2DTo1kVJRKkWk4zhNgMlAfcACL7iuO8ZxnMOB6cAxwHrAcV13u+M4BhgD9ALSgMGu637ptzUIuMtvepTruq/8P3vnHSfJUd79b/fEzWluL+kUEJIBSSCQRDBJAgMCjIlbyK8BY2Twa0xweDHJrxMGY2xjYzDYssEimFA2Sa+tRJIAG4EQYEAISUgg6fJt3tnJ0/X+UTUzvbOzdzu317c9e8/38+lPd1f/pubpqk5PV/VTLv0i4GqgD7gWeL3W2qz1HxHvsiAIgiAIgrDFMNUK5r+/jLnxs3D4AEzuwnvpb5H7xRcxs7i02eYJwknDP0n/UwN+T2v9MOCxwG8ppR4GvAn4ktb6HOBLbh3gmcA5bnoV8AEA5xD+EfAY4NHAHymlxtxvPgC8MvS7y136Wv8hCIIgCIIgCMfELOcJ/lMTvPFKzMfeDwND+L/5Jvy3/T3+k56Bl85stomCcFI5KS2RWusDwAG3vKSUugPYDTwXuNTJPgzcBLzRpX9Ea22AW5RSo0qpnU77Ba31LIBS6gvA5Uqpm4BhrfUtLv0jThe1bAAAIABJREFUwPOA647yH4IgCIIgCIKwJmZ2GvPFz2O+eiOUi3D+o/AvfyGce75EXRVOaU76N5FKqTOBRwLfBLY7BxPgILa7K1gH84HQz/a6tKOl7+2QzlH+o92uV2FbPdFak8vlut21yEkmk+u2S7SiFW3va+Nih2hFK1rRnmra6r13svTJqwhuvh4CQ/YJT6X/+S8hdeaDY2lvnOwQbbTauHBSnUil1CDwaeC3tdaLSqnmNvf9oony/4/2H1rrq4Cr3KqZnp6O0pTjIpfLsV67RCta0fa+Ni52iFa0ohXtqaA1lTLm1q9jbr4OfnoXpDN4T7oc72nPpZrbzgLAGr+Nw77FxQ7RRqs9mezatWvNbSfrm0iUUimsA/mvWuvPuORDrpsqbn7Ype8D9oR+fppLO1r6aR3Sj/YfgiAIgiAIwimOObiX4FP/TPCGl2Oufg8UC3hXvJJtH/w8/i+/Ci/XsRObIJzSnKzorB7wQeAOrfW7Q5uuAX4VeKebfz6U/hql1CexQXQWtNYHlFI3AO8IBdN5OvBmrfWsUmpRKfVYbDfZlwHvPcZ/CIIgCIIgCKcgplaD791CcNN1cOcPIJHEe9Tj8J78TDj3PDzPwx8chlL8WocEIQ6crO6sjwdeCvxAKfU9l/YWrGOnlVJXAvcBjf6t12KH9/gJdoiPXwNwzuLbgFud7k8bQXaAV9Ma4uM6N3GU/xAEQRAEQRBOIepHDhJ87pOY//oCLMzBxCTe81+K94RfwBseO3YGgiAAJy8669eBtUJYPbWD3gC/tUZeHwI+1CH928D5HdJnOv2HIAiCIAiCsPUx5TLme7dg/vvLTP/4f8AYuOBi/EufCec9Es9PbLaJwilMEBiq1WCzzeiakx6dVRAEQRAEQRCixAQB3H075htfxnz7v+3wHBOTDLzwVyle9Hi8icnNNlE4hTHGMDdTZ999FfY/UOX8CxPsPnOzreoOcSIFQRAEQRCELYE5uA9zy1cwt9wEM4ch24d38ePxHvcUOOdhDE5OUophFEzh1GBpoc7e+yrsv79KYTnAT8D2XSm27chiv+DrHcSJFARBEARBEHqWYGmR4KZrMd/4Ctx7J3g+POwR9lvHCx+Ll8lstonCKUxhOWD//RX23V9hcT4AD7ZtT3LueVl2nJYilfLI5fqZnhYnUhAEQRAEQRAiw8wcxnz/Vsz3b+XIj38AtSrsOh3vRS/He8yT8UYnNttE4RSmUg748Q8XuPNHS8weqQMwNpHg/Ef1sWtPikz2pI2yGBniRAqCIAiCIAixxgR1uPdO5zh+G/bdZzdM7qT/WS+k9IjHwJ4H4XlrxXEUhGgxxjBzpM5995Q5uLdKEMDgsM9DLsiy6/QUA4NbK4CTOJGCIAiCIAhC7DCFPOb278L3b8X88DbIL0EiAQ9+GN7UK/Aefgnejt0M5XKU5TtHYZOolAP2/qzCffdUyC8FpFIeZ5yd5oJHbidgccu+2BAnUhAEQRAEQdh0jDGw/37MD29j9sffJ/jR93DNOXgXXAwXXIJ33oV4/YObbapwitOIrnrfPWX2P1AlqNvuqhc+up+de1Ikkx4TuQzT01vTgQRxIgVBEARBEIRNwhQLcMf3MD/8DuaH34E526Joznww3uUvxHv4JXDWOTKWoxALKpWAn91d5r57yiwuBCSTsOfMNGecnWFk7NQ6RsWJFARBEARBEE4KxhjY9zPMD75ju6jecwfU69DXDw+9EO85V+CdfxET5/wc09JFVYgBpWLA9OEa0wdrHNi7QK1mGBlL8PCL+9h9eppkauu2Nh4NcSIFQRAEQRCEyDDGwN23s/CpbxB8+79hfsZuOO0svKc/D+/8i+BBD8FLymOpsPmUS9ZpnDlcY/pwjeWlAIBkCs46Z4gdpxlGx+VYlRIQBEEQBEEQTjimVsXc+nXMF6+B+++h3D8AD30E3vkX4Z3/KBmGQ4gFpWKd/Q9Umk5jftE5jUkY35bkjAelmZhMMjKaYNvkNmkhd4gTKQiCIAiCIJwwzNIi5qvXY75yLSzMws49eC/9LbY9+0XMLC1ttnnCKYwxhvxSwNx0jbmZOnPTNZYW5wFIJGE8l2TPmWlyk0mGxxL4/qnZVXU9iBMpCIIgCIIgbBiz/37MF6/B3HITVCtw3iPxX/46OO+ReJ6Hl8mAOJHCSaRaMczN1pibrjM3U2N+pk61agBIpTxGJxKc+7BR+gYqjIyL09gN4kQKgiAIgiAIx4UxBvPD2wi+eA3c/l1IpfEedxneU5+Dt+v0zTZPOIUItzLe+YPDHNibZ8l1TQUYGvHZuSfF2ESCsVySwSEfz/PI5cali+pxIE6kIAiCIAiCsAJTrUBhGQp5WM5DcRnj5iznbXphmZn7fkKw92cwMo73vJfgPelyvKHhzTZfOAWo1wzzc3Vmp2vMTdeYna5TrdhWxnSmzOi4z67T04xNJBgdT5JKSyvjiUScSEEQBEEQhFMUs5yHvT/F3H8v3H8v5oF7OXTkAFQqR/9hOgP9g/g7dhNc+Tt4Fz8BL5k6OUYLpySF5Rr7H6gwN20dx4X5OsY1NA4M+ezYnWI8Z1sZzzxrkpmZmc01eIsjTqQgCIIgCMIWxxiDmZ2GB+7FOGeR+++FmcMt0eg47HkQ/Rf/PEU/Cf0D0D+I1z/YXLbzgabDOJ7LSVdA4YQT1A2L83Ub/GbGBsEpLNsAOL4Po+MJzj43w1guyVguQSbjr/i950mrY9SIEykIgiAIgrAFMEEA87MwfRAzfQimD8GRQ5jpQxw5vB/jolDieTC5C+9BPwdPvhxvz4Pg9LPwhscAGMrlKItjKJxEioWg6SzOTddYmKsTuFbGbJ/H6ESS8x4xRqavzPBYgkRCnMTNRpxIQRAEQRCEHsCUCjA/BwtzmIVZlstFgvvvxRw5CNOHYeYQ1GqtH3gejOUgt53so59IaXKXdRhPOxMv27d5OyKcspjAUCgE5BcDlhbqfH/5AIf2FygV7beMvg8j4wnOPCdjA+BMJOnrt62MudyYtHrHCHEiBUEQBEEQNhFTLMD8DMzNUKxXCfbeb8dXdM4iC9ZxpFxa8bs8wMAQ5LbjnXYmXPgY2LYDL7cdtm2H8W3NbqfDuRwVeQAXThL1mo2Uml+qk1+sW6dxsc7yUtBsYQQYGk4xMZlkbCLJ2ESC4ZEEvrQy9gTiRAqCIAiCIESAqdVgcY7KzEHMz+7FzM/Y7qbzM5i5xvIslIvN3yw2FjJ9MDIGo+N4ZzwYRsZhdAxGxvBGxmFkjImzz2W2WOr434JwMqjXTbNVcXGhztJCncJynvxiuEUc+gd8hoZ9JnekGBz2GRxOMDjks2v3pLQu9ijiRAqCIAiCIHSBMYZgOY/Zf79zCK1juNI5nIHFeTCGufCPE0kbwGZ03LYeXnCRW5/AG51g7KyzmQtYV3dTf2AQxIkUTgLGGArLdRbnQw7jfJ38UoCxPVHxfBga8tm2vZ/dp9cYHLLO4sCQL98wbkHEiRQEQRAEQXCYctl2JZ2ftV1J52Za642Ww4VZjpQ7OG8DQ9YhHJvA23MWjE7A2Dgjp5/Fop+CsQkYGMLz/dW/dSRzOTxpmRFOIsYYKmVDqRhQKhnKxYByya6XS4ZiIWB5aYFq1TR/0zfgMzzis313iuFR2w11YMjH9z1yErH3lECcSEEQBEEQthwmCKBUpH64inngPlhewuSXYDkPy0vNyeSXoJCH/BKH84uY5aXVmaXSrdbDM86GkUsY2L2H5VQGb2zCOouj43jpTEdbMuIYCpuIMYZiwbC8ZL9JzOcDCvk6tWqR5XyFcsk0WxPDJFOQzfpk+3zOfsgwqXSF4dEEQyMJUilpWTzVESdSEARBEITYYYIAigXr4BXy1vkr5DGFPMseBNNHoLAMxWUbmMYtUyzYeakIxrCm65bpg4FBGByyrYN7cmRzk5Sy/bYlcXQcRqxzSP/AqnHnBnI5iuIYCjHBBIZSyXBgX4H9e8ss5wPrMC7VKeRXBrNJJKB/0Gd4JE3/YIpsn0cm65PJemT7fLJZu55Ito55aV0U2hEnUhAEQRCEE4YJAhtFtLBMdXkBc/CAdehKRYybUyrYYDKlIpRKmHKRuWqF+sKcdRYbzmCn5hFcVFLfhz478D19/XZ52w68FWn9DE3uIG8829V0YMg6jf2DeKnUqnwlgqkQR4wxtntpIaBYDCgWwssBpUJAqdhoTbShmXzfOooDQz7bd6YYGLLLA4MJsn0enifdToWNIU6kIAiCIAjW+auUmw4f5VLI+Ss01/N+qxXQFJdbLYDhlkDn/M0e7Q8TCdsamLWTGR6BkXG8XadD/6CdBgasw9dct/PcntOZXsqvah3sRF8ux7I8KAubiDGGoA61uqFegzmvzOyRGtWqoVoxHee1iqFSMQT1PMvLNUywMk/fh2y/T1+/z/g2O5ZiX7/Pzt1j1IMl+vp8PF+6nArRIU6kIAiCIMQMYwzUa3bg+Fq1Oa+V8pjDh6BSgWoFKhVMpQzVciitDNUKS8kEwcJCc5tx6e06KmUOVyqYcnHNlr8wy9Bs5Wu2BI5vw9t9pk3rb7QEDjA8uYOlarXpKIadRpKpFU7geBetIl62Dy+/fHyFKwjrxBhDrdrByVuxHFCtGnyvSqFQpl4z1GrWWazX3XIdWHFqLXb8v0QCUmmPVMojmfbI9nkMDffh+RWyfdZJzPZ59PX7pDNex5couVw/09OFSMpDEMKIEykIgiBsWYwx1lkql9xku09SKUGpRGmg3zpaQWAdqCAAE9h5W9pyX5ZgcQmCutteh3rdzd26S1tIJgmWlzF15wDW684prNrlWs2u12tMG0O9XAo5jO43HZhZ7457HsVMFpNMQToNqYwNDpNOQzpjnb10Gi+VgXSavpFRisZzjl4Wsn142axz+vqhuZwlt3sPM3Nzx7YByOZy5KUVUNhk6nXrDNZqBs+UOXKoStW19FUqhmrZUKkEzTS7bqhW51e1ALaTTEEqZR0+MCSSje8JIZHwSCY9u5z0SCbs8tjYMMVynlTKazqNqZSH32EYDOlyKsQVcSIFQRCEdWOMsU5Os8ujc8pc98dSf59zyuotR2yN5Xw2S7C0FEp32+r1NkctYCGVJCgUMM1tzoGrr16erteoF5ZbzuJRWtcWutj3fKfERBISPvgJOyXsvJJOYTzfbU9CMmm3JVPWoUskIJnESyRJDQ4S1Op2WzLZee7yGJqYIF+qrHQInSPYTEtlIJlk27Zt6374HMrlKK+3FTCRWH+hCcIJwhjb8lcuG8olO/REpWQolwN87zD5paLtBlpzDmOV5nKwwhFcHX03kbQtgOm0TzrtkR2185GRAar14kpnL91aTqY8fNdltBtnL5cbYnq6fAJKRRA2D3EiBUEQTgFMtdr6Xq1UoHLgPsyhAzbQSTEU7KRYaH0D5xzF6WqFeiHf+k6uXl/zf7pxypqdEf2GE+Y3nbBmWsIHz6eaztjeYH7DMQvpU2nI+E0nLjU8QmBotpyRca1omQxeptHSloV0lrFcjrmFRfA8m5/n2Xwa674HXgJ8j4ltk8zMzTedxqON9dfNA+VIF1r5vk/odVpdRGl9/9fhm8CEf4iF+aJ1FsvBmsNQeB5ksjUSCUPSOXZ9/XaeStmWwEZ6MukxPjFMqZQnnWk5hIkOLYAAudyEtAIKwhqIEykIghBjTL2GWc63HDznBJrQciN9IahTn59b4Sw2hzto6x7ZsTOi77vv1frt3H3zltyxi8BLNLs5hr9t80LdH8lkGcttY25hoeXkNR3C9mWf3OR2pmdn1xUcJSqnrJuB3f3+AbxCcV1aQehljDHU64agblwjv3EN/qvXZw8vMj9ftt/+hVsCa4SW7RTUFqlUjtE/FNsAn8nWSaYM2T6PkbEUmaxHJtMaiiKdacy9rlrdc7lBpqdLGy0iQTjlESdSEARhkzHGwNwM7L8Ps/9+2He/nR/ax+HiOgIkeB5k+6kMDkI6a52/kTG87bts4JNmEBTrFHp9fYxs38lCpdpyGLPuG7kODt1ot05Z3/q7Ra7HgRQEoTuMMe6z25YTVykVmD5SWREYplJuCxZTsd8G1qrzXfxb6xrl+diWvyTNlr9U2gaCSSRhaKifWr20smuoayVs7yIq3wIKQrwRJ1IQBOEkEizMYe74H8z+B6zTuO8+2P+AbS1sMDIGu07He+xl9G/fQcEA2X68hiOYbZtnsl2P+ZXuogVOEIRoMUGrW+csZWaO1EJRPlutevVQK1+9ZvD9CoVC2cZrqrVa/Ood4zKt/KrX8yEdcuYyWY/BYZ90OsnwyADlctH1LrfdPVvLNmCM7+a5beMsLc2RTHYODBNGHENB2DqIEykIghAxplrBfOcbmK/dyJE7f9DaMDAEu0/He8yTrdO4+3Q7HxxuSgZzOUry0CUIscaYlWP7hVv3alXDfckZFhcKHb//aziFLVYHfmniuVhLSY9E0iOb9fE8j75+SCZ9GwE06bmYTFbTaBGcyI1QLC6FvgNkzZ4A3XwLODScolxZ+/tgQRC2JuJECoIgRITZfz/mazdivvEVWF6C3HYGrvh1ijtPh91nwPCodOcUhE2m0QpYqRgISkwfqTpnL/RNn+sSWq22nD4TFCkWq1QrAbXq0f/D80qtQC+u5W9g0G+OB5hyw0Sk0h5j4yMUikuuW6jtGtpwDv02x6+7iKAyfqAgCCcOcSIFQRBOIKZcxtz2X5iv3QA/uQMSSbwLH4P3pKfDQx7B4OSktCwKQgSYwFCpujH+3Dh/lXLAgfvnmJ8vtsYAXPX9XziXzq2AjW/9rBNov/frH0jQNxCQTiebjmE63XAS/eZyOu2xfUeOmZn1jfIpgV8EQegFxIkUBEE4AZgHfor52g2YW2623zdO7sJ70cvxHvcUvOHRzTZPEHqG5hAQbU5feNn3D7G4UAw5iza9M0U8j2Y3znTb93823Tp94xO2y2fDUWwMD9FpCIhuWgGlx4EgCFsNcSIFQThlMPU6prDshr5w4yK68RHteImFZtqiB8Hysh3w3hg38H0Axs6NmxMEzOQXCe69E5IpvEf9vG11PPd8eXAUTnlqtYBiIbBOYDloOnxNxy+UVq8vUyrWqFYNrOUPYkeKyWZrJFOGdNpnZNRvjvmXztjldNoO/ZDO+OzcmWN+YWadQ8kMMD0tw7gIgiAcC3EiBUHoOUy1CsuLkF+ksv9nmAP77FiKhTy4uVlegsKy/RbRpa1ruAyAVJpSXz/G82w/tuYg9G1jHXp2MHpvYAjvxb+O99hLVwTFEYStRL1unb2ZI2WOHKquaAGslIMVzmHDYazX1h4qIpHEOXvW8RsbymCMt6LFMJX2Q8u2O2ki2V0k4lTalxc6giBsKsYYlioB08tVjixXObxcZbpQ44hbf9b5FS47Lb3ZZnaFOJGCIGwqplqxjl7eOXvLSxQ8Q3BwP+QXYWkBk1+yy26dUqulYK49w0QSBgZt5NP+ARidwNt1BgwM0p/bRsF4zXERvb6+1hiJfW68xEwfXjLZ1UPquIStF3qQWtVQLgeUS4blxTwz02XK5ZAT6JbLbr01bMTiqrySKawz6LqKDo0kSad9RscGqNaKrlXQI51utRq2dxGV4R8EQehF6oFhvlRjrlhnrlhjrlRjrlgjH8zxwEyeI8tVpgtVSrWVXSzSCY9cf4rJgSSDmcQmWX/8iBMpCMIJxZSKsDgHC/OwOIdZsMuL1RL1mSNNR9HOF6FSWZVHM7RFOgODw3YaGsab3BlaH8EbHGZk92ks1OrQP2Sdx3RmzVYHGS5D2MoEQaNV0E75+TxHjpQpl6yjWC4HVErGLpcC6vXwr1tjCCYSNLuCpjMeA0M+6UyKjHMEJ3IjlMv5pmOYStvB4TuRy42LYygIQs9RDwyL5ToLpRrzJTtfKNcpk2ff7CJzJeswzhdrLJbrHXvgj/WlmOhLsGckzaN2DbBtIMW2/hS5gSSTAymGM4nm80ovvkQTJ1IQhKNigrp1+JYWYGkR8guYpQVYXGCxVqF+6IBzGudgcR7KHaIKej7lkVHoG7AthLnteGecbZfd5DVaDweGGN9zOrOVOl4mc0z70rkcXo9deAXhaNTrNrBMpeLGHXTjD1Yrhr0/nWV+rtDWlbRTlFEIO4bpjG0hzGR9xiZ8MtmUW/dIZ322bx+jUFwknbHBZI6GRA8VBKEXKdUC5l1L4XypznyxxkKpTtlb4OB83jqKpToL5TpL5XrHPJK+x2g2wVhfku2DKR6S62OsL8FoNsl4X5LRPjfPJti5fbLnHMNuECdSEE5BTBDY1sCFWZh3rYXzM7Awx3y5SH36sHUa84u2m6kJOuZTGhyCoVEYGcM76+dgeBRGRmF4DG9krLnM4BDbJrev+2KaEMdQ6HGMsY5dteKGnGiLMGrndrsxJQrLlabjGHR+dnEUSSZt19GUCx4zMOiTziRXpNlhJcYpFBdIZ9ZuKWyQy2WZns4fVSMIghAn6oEhX7FO30KpxmKpblsNyzVKZp6D8/lmK+J8qbaqO2mDkWySobTPSDbB6aMZRjLWKRzJJhjOJhjNNJaTnLVrct3D9Wx1xIkUhC2EMcYGl5mbgfkZzNwMLMyyWCpSP7jPthYuzNmWw3qHJ9W+fmpjOegfhJ2n4Q2OwJDtOsrQCN5QY30UBobYtmPHln7LJmxNajVDqRhQKhpKy8ssLFRdAF4bFdQYCAxgTDMwLwYOPDDH4kKxOQh9vWYHoK/VCC0b6lWoVtcOKAOQTNIcWmJgwGNwOEHajTWYTHukG4PQh5dTHjt3bWNubn0PMOO5DMG0v/ECEwRBiJjAGAqVgKWKbQVcKtdby25eMtMcWSw0u5Yulev2Wt2GB4z0pRhJ+4z0JTg318do1jqGY66VsOEkjmST7JjcJsP1HAfiRApCD2CqlVbL4NIiZmmB5XqVYP8DMDeDmZ+B+Vk7VVd/Y1gaHrWO4Mg43s49MDpml0fsHDf3Mpme7JcvnHoYY5qjr5jArBiJ5Ui9xMEDFeskFoOWw+iWV3b77Kb1zQZ0SiTtwPPJpI0UaoPKePQN+C4dRkZsQJlUKJBMOMpouGWwm3Ou03iFgiAIccAYQ6VuOJIvc99cieVKQL5Sd5NdXq7UWaoELLv0Qu0+5osVliudHUKwTuFA2md8IMNgEnYPp3lopuEEJhjJtBzCkWyCoXSC7V04hsLxIU6kIGwippCH6cMwcxgzc4ilUpHg8EFMOBLp0iKUV49blgdIpWF0HMYm8M461y6PTtiIpGNueWScbTt3ysVUOOkYY6jXW6109drKFry56UXmZ8u2Ba9qqFZNc7lWs9FDay7NmEXq9cAO02k46jiC4eihngeZrEe2z2dwOMG27UkyfT7ZPp9sn8fExBiLC/N4vn3D7Hm0pra0ycluxhuckHNOEITY03D8irWAUjVYMS/WAopVO3npAtPzeQpufblap1gNKDSnOoVqsKYjCC1ncDCdYCCdYDDtc9p4P2mTYSiTYCiTYDCdYDi0PJRJMJDySfjdDe0jRI84kYIQASYIbICZYoHqwgzmnjsxM0esszh9CGas40jbuIWFdNpFHx2BQReNdGikFZ10cKQZmXTizAcxUypL1wphwzRa9RqOXsvps87c4uwSc3Pl5nq7UxjWYpYpl2ut9aPSOv49H1Ip27qXTEEyZYO+DA75JJIeAwN9lMul0LCdXmOYTpfmNYfyzG0bpVJdItvnk8l4eEf5HjCXy+In13crlPEGBUHYLKp1Q9k5duVaQKlm3Dw8GZI/KzGzuES5Zprpq/V2uVy/m0J17RbAdtIJj/6UT3/Kpy+VoD/ls30wRZ9L608l6Ev57BgfhkqRwXTCTdZx7E/7+J4M7bNVECdSEABTqxEsLWJmj1jnr1yCchnKRUy5DJUSlEp2Xi6x5HsEs9OYYhFKBesMlopuXlgxjuFs+I8yfZCbtNFJzzkPcpN4E9tt2sQkuTPOWvcH2/7QMF5ZLrxbGWMMtVpgx+irQ1B3LXtubtdby4f3z7O4UGqm1Wsh/YplwCxTqdSazqA56kPE8oo138d240w25nbq64eBgQy1eqOrJ60un6vWYXJygsWlOZKp1WMGttPNg0YuN8D09OrWe0EQhOPBGEMtgFpgqAeGWmCoGUOt3ppP1/McmS1QrRuqdUMlMG45oNpctunJdJ75/DKVmqFcDyjXDJV6QNk5ipV6w8kzVOt3U6zWqa/T0WuQSXhkkz6ZpE826bm5z0R/srk8NjQA1TJ9Kbvel/Lpc/Pwejbls2fHNhbmZo/9x4hjeKpwSjiRSqnLgfcACeCftdbv3GSThHVijLHjCC4vQcGOLVi+L4OZm4OgjqkHENRxT9Fu2aUFAcupJMHMdNPRM6XiSmev4fxVKxxZr1GeRzHTh8n22QHq+/rt0BXjObzmoPX90GcHsR/euZultHMe+weP2pIhrRzxwHbDbHSjtOvGdaE0pvUdXmM54ZWZn60RBM6xc/Pmuj0cCeqGBzKzLC0VqVVd612ju2e11ZrX+G87uPtCF5a3HCc/Yb+fSzTmycbcI52G/oEM9ZpLD3/ft2IdkgmP3OQ4S0vzzW1Hi/TZzcPD0EiKclUCvwhCnDDGEBjrLDXTCL9oMivWi1X7nVs9MNSMjZhZDwz1xrKxTldgYF95gbn5AoGx62vPYWA6YH5hsZlPYKDu7KoHbtnY5XR2ifxyofm/9v+s49ewoe6W/eQBiuVKc1vNTY28q0FrH2rmLmr1oGsH7lgkfI9Mwk7ppG/nCZ9M0mMwnSCTdOsJj5HBfky1TNY5dlmnX7Xulnfv2MbS/OyqFr9OdHO9TiXkWi2sZMs7kUqpBPD3wNOAvcCtSqlrtNY/2lzL4osJWk4Ybct139humfWQ41avtdZDaeWBAczcLKZex/WBs9uby620JQ+C6SP2G8HGYPQFN2/rE3f0mIcryQMkU86x67NTXz+MjuNld69w9gYmcizX6naw+kwVTMFQAAAgAElEQVQWOk3pLKTTbNu2/g+2s7kc+R56I2e7NrooldDmPNmHh8b6cr5GYTloOlkmYIXT1VgPDFRKBebnq620YPVvGsFRDu2bJ79UImhub2mD0HoQQCZ9iGKxpQ2MCelWBl3x/QLVas1tb+XR/jtLN87bUhfakmvJAz/pNR08P+GRzHqkk+An/GbawGAfpXIRP2G7b3o+K5ab3TsTMDYxysLiQrO7p3H12azbZh3D6Ogg8/PzKx4G6wTUQ+uNui6VSswtl5rHRHibPTxM8zfDhSQLi8srWjZX2eF+N7zks7iYd+tmRZ4r/8MwNAdLS61vHdfCAENzsLi42Oyi1ThuG/8bXh88WCOfP3ZwHYPVLi3lV+xvuPwa6QYYGCiTzy93yMc0y6Dx+4GBEsvLy80dX7H/oQWDob+/SKGwvHI77T+wi/39BQqFQsd8VvyPgf7+ZQqFVvdi02ZLK93Q15+nGM53lab1276+JYrFYvO3a+2f1S5QKJZWZBK2MUwmO0+x2BqrMlyu7XZlMrMUS+WWsu34Dv9POjNDuVReUUarj2WnTR+hXCmvsq/d4QJIpQ5RrlQIOhyD4eXAQDK1j0qlumLfw8cWtL4L9pP322saq/MxxhCEysTz7qFWD5rpQWDzDcL65n/cSa+Q8KxT5nseSd8uJ9yyTbPrCbcti48HZJMeSd+3232PpOeRTOB+a9OGBvqplkvNfJNrTAkfxkdGKC4vkU74pBIeKd8jnfDscsIn7dvlpO91FfSl21a9vlSCZXkhLZwEtrwTCTwa+InW+l4ApdQngecCPeVEfvyjd2BIr3HT7pR6D6vujmvq7179tNIRD9Zsr/Owh1P7IbXtKPmFf+5DyoNRD8bCkS2a0S3shIfn+60H0uZ1su2C6S6gnu9j1tPZvwLeIe8Y2qqbwPdm15cv4HkzTW3ny7qHhwHjWW37rrHW+mzzScVrU3jhNQOeN+uq1mumd5rbrd246cd+qG/RfRTMMIF7rGvNG45H0T4QYZpzg3ujHdK1awKnqYd+Uw+lu1Edmr9tOktuOQitB+63jd/XTGPZUHP51zHUQnmxOojuCeC+KDLtkgciynd/RNo4sO5+EEA0L6Q8Wt3o258/vbZ0z5vreJ3q9Nzqe/MYY0LbVl6tmnkCvr9IYMyqa1Pzeh5K9/xljAlWakO68N8l/CJBELT9p9dcDpudSFQIgvoKK9s1jfVEsk5Qqzc3em1lEP6/ZBUCN6yS7/7f91q/8TyPhDM8nUyQJGjuh+eF8/ZodATwPMikM9SqlZDGww/l2dD6HvRl+6iUS/Z/XcCohGev+n5I73sw0D9AoVhYVZat8vaa64MDA5SKBeukeV7T2Wo4dwm3nPQ9RkdGyC8t4jf/q33eWp4YH2dxYa7pFCY8++1z0yF0Os/rLuBKdNoJpqfX91wgCFuBU8GJ3M3Kp5q9wGPaRUqpVwGvAtBak8vlTo516yRIDeGTWsMJOZG03S03ol3jTZh3lG3g3piGlhu/6bS89qtwpzVg1rlDXn39WvuD1doVtoUT3fdeJpy2QnussuyUtvJNfWO9Pf9Gebfcn7b8wlqvYadpexJraEP74fz6xnLr6cJg3IOJaeyWj30Z0Hi779Pa7of+2+XnJ5x/7LX+B89f8cDZfKDyrfPveba/euOiFu4a3HjA8hsvFdry8UIJjV8lfJ+g1Sx5VHzfB/eg3OhC1HgYwj0Yee7hL5FI2AfaxkMfXujhvG0dqzdB0LStk8Zz5Z1IJKi7B+DG/nd6+AVIOjva9zucd+M3YW3rwdhb8dDcsCuZTFB3D8vtdqzM17P5NsYrDe1X+0Nre77HIplMENRDZRyytbHceLhOJZPUusq3vnL/m/vVeKC1y6lkglq9vrL82xYax10qlaTmely0yneluLEa1q7Is5nQSkklQ/l6K/XtXeeTyZX5Hg3RinZj2rH1a8f61q1d73PbVtbGxQ7RRquNC6eCE7kutNZXAVe5VRO3D4JfcsVpMXnTJlrRivZkabvPe7wL7dga2tVvZrrLdySachvrQjvqtOtoFJgYzjE9vb7W94mhldrQaxkAwq7ocC7H9Pz68s3lciwszK1f20Vwi7nZ9QXqisPxLlrRinZj2rjYIdpotSeTXbt2rbntVPhKdh+wJ7R+mksTBEEQBEEQBEEQuuRUaIm8FThHKXUW1nm8Avhfm2uSIAiCIAiCIAhCb7LlWyK11jXgNcANwB02Sd++uVYJgiAIgiAIgiD0JqdCSyRa62uBazfbDkEQBEEQBEEQhF5ny7dECoIgCIIgCIIgCCcOcSIFQRAEQRAEQRCEdSNOpCAIgiAIgiAIgrBuxIkUBEEQBEEQBEEQ1o04kYIgCIIgCIIgCMK6ESdSEARBEARBEARBWDeeMWazbYgjUiiCIAiCIAiCIJzqeJ0SpSWyM14cJ6XUbaIVrWhPHW1c7BCtaEUrWtHGXxsXO0QbfT2f5Kkj4kQKgiAIgiAIgiAI60acSEEQBEEQBEEQBGHdiBPZW1wlWtGK9pTSxsUO0YpWtKIVbfy1cbFDtNFqY4EE1hEEQRAEQRAEQRDWjbRECoIgCIIgCIIgCOtGnEhBEARBEARBEARh3YgTKQiCIAiCIAiCIKwbcSIFQRAEQRAEQRCEdZPcbAOEziilngE8D9jtkvYBn9daX7+G/iHAc9v012it79igdt12RKiNyt4R4PI27Q1a6/kN2hCVtht7t/K+9dqx3mv2xuHcjEM5bNl6i9jmqK5THvDoNu23tNarogN2qY3K3qj2rdfuA71mbxzuW3EoB6m37vON5DyOM9ISGUOUUn8LvB64GXiXm24GXqeUek8H/RuBTwIe8C03ecAnlFJv2oB23XZEqI3K3pcB3wEuBfrddBlwm9t2vDZEpe3G3q28b712rPeavXE4N+NQDlu23iK2Oarr1NOBu4E/Bp7lpj8B7nbbjlcblb1R7Vuv3Qd6zd443LfiUA5Sb93nG8l5HHekJTKePEtrfW57olLqU8Bd2Bt6mCuB87TW1Tb9u4HbgXcep7YbO6LSRmXvW4GL2t/6KKXGgG8CHzlOG6LSdmPvVt63XjvWe83eOJybcSiHrVxvUdoc1XXqPcAvaK1/1qY9C7gWeOhxaqOyN6p967X7QK/ZG4f7VhzKQeqt+3yjOo9jjbRExpOSUuqSDumXAKUO6QGwq0P6TrfteLXd2BGVNip7PaDTIKmB23a8NkSl7cberbxvvXas95q9cTg341AOW7neIB7X1W60SWBvB+0+ILUBbVT2RrVvvXYf6DV743DfikM5SL11n29U53GskZbIePJy4ANKqSFaB+UeYMFta+e3gS8ppe4GHnBppwMPBl6zAW03dkSljcretwPfUUrd2Jbv04C3bcCGqLTd2LuV963XjvVeszeqfYtDvlHZEAd7u9FGaXNU16kPAbcqpT4Z0u4BrgA+uAFtVPZGtW+9dh/oNXvjcN+KQzlIvXWfb1TncazxjOnkDAtxQCm1g9BHt1rrg0fR+qz+oPdWrXV9I9rjsOOEa6Oy13UdeAarP2ye24gNEWq7sXcr71tPHeu9Zm9U+xaHfKOyIQ72Hoc2DtfVbrQPA36pTXuN1vpHG9RGZW9U+9Zr94FeszcO9604lIPUW/f5RnIexxpjjEwxnaamplId0nJxtiMqbS+Wm0y9N8Xh+I2Dvb12bsZh3+Jgb5T7Fodyk0kmmWSSKT6TdGeNIUqpy4CPAlml1HeAV4U+1r0ReFSb/uHAVdg3GtcBb2y8zVBKfUtr/ejj1K7bjgi1Udm7B/jLUL5/2fh4Win1Oa31805C+Xaj7cberbxvvXas95q9cTg341AOW7beIrY5quvUMPBm4DTgWq31J0Lb3q+1fvVxaqOyN6p967X7QK/ZG4f7VhzKQeqt+3wjOY/jjgTWiSfvAp6htc5hD+AvKKUe67Z1+uj2/diwwhdgo+V9XSl1ttvW/kFvN9pu7IhKG5W9HwJuAl6L/Uj6ZqXUhNt2xgZsiErbjb1bed967VjvNXvjcG7GoRy2cr1FaXNU16l/cf/1aeCXlVKfVkpl3LbHbkAblb1R7Vuv3Qd6zd443LfiUA5Sb93nG9V5HGukJTKepLXWtwNorf9dKXUH8Bllx6zp9BHrkG4N+vxXSqnbgOuVUi/toO9G240dUWmjsneb1vof3PJrlVIvAb6qlPqlDdoQlbYbe7fyvvXasd5r9sbh3IxDOWzleovS5qiuU2drrV/olj+nlHor8GWnbacbbVT2RrVvvXYf6DV743DfikM5SL11n29U53GskZbIeFJVNogBAO4G/lTsG5FzOv1AKTUS0n8FeCG2C9KqtxpdaLuxIyptVPamlFLZkPZj2PHObsC+GTpeG6LSdmPvVt63XjvWe83eOJybcSiHLV1vEdoc1XUqo2yAi4b27cA/AV8FJjagjcreqPat1+4DvWZvHO5bcSgHqbfu843qPI414kTGkzcB28MJWuu9wKWsHNy0wV+wciBTtNbfx97oP7MBbTd2RKWNyt5/Bh7Tpv0iMAX8cAM2RKXtxt6tvG+9dqz3mr1xODfjUA5bud6itDmq69T/A57Spr0a+D2gsgFtVPZGtW+9dh/oNXvjcN+KQzlIvXWfb1TncayRIT5ijlJqHEBrPdsrdkSl7Ya4lJvQe8Th+O0GOTe7z1fq7fiIQ7kJgiAI8UCcyBiilDodG8zgqcA89mPdYeDLwJt0KzJeQz+CjQr1PGAS26f6MPB54J1a6/nj1K7bjgi1UdmbBK4Eng/scsn7XL4f1C5SVsTl2422G3u38r712rHea/bG4dyMQzls2XqL2OZIrlNO/wxnb3hctc/r1jdLXWsjvK5Gsm89eB/oNXvjcN+KQzlIvXWZr9Of8PM47kh31njyKeCzwA6t9Tla6wdj+0l/DvhkB70G5oBLtdbjWusJ4DKXpjeg7caOqLRR2ftR4ELsdz3PctOfAI8APrYBG6LSdmPvVt63XjvWe83eOJybcSiHrVxvUdocyXVKKfW32G+GbsY6tO9yy69TSr3neLVR2RvVvkVobxyu7XGwNw73rTiUg9Rbl/lGeB7Hm80eqFKm1dPU1NTd3Wybmpq68yj6OzegXbcdEWqjsveuo2jv2oANUWm7sXcr71uvHeu9Zm8czs04lMOWrbeIbY7qOtVROzU15a3Xhm60Udkb1b6dABvicG2Pg71xuG/FoRyk3k5Qvhs9j+M+yRAf8eQ2pdT7gQ8DD7i0PcCvAt/toL9PKfX7wIe11ocAlFLbgZeHfn882m7siEoblb2zSqkp4NNa68Dl62M/bJ7bgA1RabuxdyvvW68d671mbxzOzTiUw1autyhtjuo6VVJKXaK1vrUt/RKgtAFtVPZGtW+9dh/oNXvjcN+KQzlIvXWfb1TncawRJzKevAzbX/pPaPWt3ouN/vTBDvoXYyPo3ayUmnRph4BrALUBbTd2RKWNyt4rsJG33q+Uapy0o8BX3LbjtSEqbTf2buV967VjvdfsjcO5GYdy2Mr1FqXNUV2nXg58QCk15P4brCO74LYdrzYqe6Pat167D/SavXG4b8WhHKTeus/35URzHscaCawjnPIopSYAtNYzm23LeujG3q28b4IgxJcorlPKjlfZDFqhtT54IrRR2duNNi72RkWv2dsNUdkbh3KQeus6z8jO41iy2f1pZeo8TU1NPWNqaurKqampM9rSX7GG/iFTU1NPnZqaGmhLv3yD2nXbEaE2KnuHp6amzu6Q/vCTWL7daLuxdyvvW68d671mbxzOzTiUw5att4htjuQ65dJTHdJyG9FGeF2NZN968D7Qa/bG4b4Vh3KQeusyX5d+ws/jOE8SnTWGKKXeAbwVuAD4slLqtaHNr+mgfx02NPBrgduVUs8NbX7HBrTrtiNCbVT2KuDHwKeVUrcrpS4Jbb56AzZEpe3G3q28b712rPeavXE4N+NQDlu23iK2Oarr1GVKqb3AAaXUjUqpM0Obb9yANip7o9q3XrsP9Jq9cbhvxaEcpN66zzeS8zjuiBMZT54DPEVr/dvARcAzlVJ/47Z5HfSvBC7SWj8PuBT4v0qp16+h70bbjR1RaaOy9y0u3wuBXwM+qpR6/gmwISptN/Zu5X3rtWO91+yNw7kZh3LYyvUWpc1RXafeBTxDa50DrgK+oJR67AnQRmVvVPvWa/eBXrM3DvetOJSD1Fv3+UZ1HscacSLjSVJrXQPQdjDT5wDDSql/A9Id9L7WOu/0P8Me7M9USr2b1QdkN9pu7IhKG5W9Ca31Aaf9Fnbsnz9wb57aPxSOqny70XZj71bet1471nvN3jicm3Eoh61cb1HaHNV1Kq21vt1p/x07oPeHlVLP26A2Knuj2rdeuw/0mr1xuG/FoRyk3rrPN6rzONaIExlP7lFKPbmxorWua62vBO4EHtpBf0gpdWFInwd+EchhuyAdr7YbO6LSRmXvklLq7JD2APYC8VzgvA3YEJW2G3u38r712rHea/bG4dyMQzls5XqL0uaorlNVZQNWNLS3A0/FDtZ9zga0Udkb1b712n2g1+yNw30rDuUg9dZ9vlGdx7FGnMh48svAN9sTtdZ/gA0Z3M5LgQNt2prW+mXAkzag7caOqLRR2fubtL1J0lovAZcDr9iADVFpu7F3K+9brx3rvWZvHM7NOJTDVq63KG2O6jr1JmB7m3Yv8GTgnRvQRmVvVPvWa/eBXrM3DvetOJSD1Fv3+UZ1HscaGeIjhiilvo0dZ+Z64Hptm9GPpr8N+DpwHXCT1rp9YNPj1a7bjgi1Udl7g9Ndp7X+8Vq647AhKm039m7lfeu1Y73X7I3DuRmHctiy9RaxzVFdp65ytn7RPWydKG1U9ka1b712H+g1e+Nw34pDOUi9dZ9vJOdx3BEnMqYoG9npcjftpnUg36y1Lrdpk8ATnPYyYAa4AXuA3nW82uOw44Rro7LXdTto6M7Fvmm/HnsBWD7eMotQ2429W3nfeupY7zV7o9q3OOQblQ1xsPc4tHG4rnajfQzwTGz3sAo22uH1Wuv/6WBrN9qo7I1q33rtPtBr9sbhvhWHcpB66z7fSM7juCNOZA+glEoBT8QecJcCR7TWzz6KfhetA/TBwC1a61evQ3s28M2jaNdtR4TabvZtXfkqpXwgfAEoAjdqrd91Amw44dpu7N3gvh3reDjhx9lJtHezjnU5NzeWr9TbCdYex/6d8Otql9oJ4OlO+3DgO9iHNX0M7QXAd4+ijcreqPbtZN0H5L61SfctqbcTsm+bUW+RnMexZLMHqpSp+2lqamp3F1p/amrq8cfQTK5X240dU1NTE8dp87q0J9rekC43NTX1Kyfahgi13di7bQvvWyQ2dHPsdHmcybkZk3yl3k5K3XVzLnejvWhqauqtEWijsjeqfYvKhjhc2+W+JfUWt33rJt9IzuM4TNISGUOUUsPAm4HTsM3mHw9te3/7mw/XNP5HQAD8IXZg1BdgBzN9vXahhJ12vO3vPOA24JGAp7WeDWkv11pf75ZHgb8GLgF+CPyO1vpQSPtO4K+01tNKqYsB7exJAS/TWt8c0n4H+AzwCa31Pccoi4uBvwT2uTL5kLPhbuCVWuvvhbSDwO8DL3RlVwHuAT6gtf5wW75J4Erg+cAul7wP+BzwIa119Wh2hfK5Smv9quPRKqUSwK87W6/XWv9XaNsfaK3/LLTejx3Y2wDvBa7A1vEdwNu0C0N9lP++S2t9bof0h2utv++WU8AbgUdj6/jPtNaFkPY1wCddHT8YWxcXAHcBV2qtfxjSfgZbx59bh20PAv4A2I/9AP1vgMe5fXuDDn2D5d7avZxWHdfd/38gfIw57Qj2mHke9oP3ADiMHTz4ndoOV3BMlFLXaa2fGVpf9/kp52ZT2825KfVGdPXm9JHU3TH+8w+11n/alvYMl+eX2s7zV2itPxRa94Ap7PXv34GnYCMZ3gH8o9Y6OMZ/f1lr/ZQO6Tmt9XRo/SXY698PgH/WWpvQtudju+/OKqUmgb/C1vGPgN/TNohGQ/tu4NPha/pRbBvHXtv3Ax/E1sfPu317h9Z6rk1/GbYu9tC6/v1Te71HdY+T+1ZTG9V9S+qN3qs3p38G9r612yXtczbdcDSb2vJYdZ2MM8nNNkDoyL9gb+afBl6hlHoh8L+0/fbksR30VwP/CQwAXwH+FXg29mD+B+zNtsE0cF/b73djm9sN8KBQ+juw/bTB3jAPYMcIewHwjy7/Bs/WWr/JLf8l8GKt9a1KqXOBjwMXh7RjwCjwFaXUQeATwKe01vs77Nv7sQ9zo8B/Yx+0nqaUeirwAezJ3+Bfgc8CzwCUK49PYsff+Tmt9VtC2o8C89jwy42b/2nArwIfA17cEHZ4SGzgAc8KJ3SjxZZhP/At4O+UUjdrrX/XbXsB8Gch7dXAA0Aftq7vwJbzL2HL4aUhG5ZojTXUiADW30jXWg+35fsot/xOYAL7YNs4dl4W0v6m1vp9bvk9wN9orT+rlLrU7cvjQ9rHYB94/04p9UVsHf+n1rrSoWyudttHgFuwx/+fYruDfAj7wNjgg9jj98+BFwGLwNewgwA/XGv93pBWA18GLtVaH3RlswN7U9Au/0aZPYrOeMCFbWndnJ9XI+cmdHduSr1Zoqo3iK7ujsavY89rAJRS78B+b/Qd4C1Kqb8Nnb+vwZ77Df4emMSOS/lcIANcg62ThwCvD+X7/bb/9YBzG+la64eHtt2Iu/4ppf4A203349gw/g8DfiekfbvW+mFu+b3Ya9VbgF/AHltPC2lfCjxJKbUN+BTWuf/uGuXyMazTehHwErf8Fy6/qwkda0qpPwd2AF9y859iHfp/V0q9Q2v9b6F8I7nHIfetsA1R3Lek3iw9VW9Kqb/Fft/4EVbW2+uVUs/SWr+e9bHiOhl3xImMJ2drrV/olj+nlHor8GWl1C+tod/eOJiVUq/WWv+FS3+vUurKNu0bsDenN2itf+B+81Ot9VnHsOlirXXjoexvlFK/2rY9qZRqDErdp7W+FUBrfZdSKtOmndNa/x/g/yilnogNIf8dpdQd2JvtVSFtSmt9nbPzL7QdxBWt9ZeUUn/Vlu+ZWuur3fK7lVK3aq3fppT6Nezb4vDDzkUd3pTtBW5RSrUHljiCvZCEQzIbtz65Ae2jGw80Sqn3Ae93b9R+ue33AOdqrZV7I38A+AWttVFKfR1o/3D7X7APh29otGwcpY7D//NU4BKtdVUp9dUO+YavF5Na688CaK1vUkoNtWkPa61f5Fp/ngu8ErhKKfUf2Dq+MaQd0lp/wNn5aq31X7v0D7q3kWEu0lr/mlv+ulLqFq31Hzp7v4d9uGtwZuhcwNl6EHinOybC3ArczOpyB1uWYbo5P+XctHRzbkq9WaKqN4io7pRSi2vsq4d9IA3zHOCRWuuaUuqPgY8rpR6ktf4dVtfnE7XWF7jWi4PATq11RSn1CawTGuZn2Ie9P8N+X+RhH/yes4ZdDV7g/mdZKfXxDvkmQssP1lo3HuavVkr9dpt2r9b6Yufwvxj4mLItQZ/A1kf4HrNLa/0sd23fq7W+1KV/TSn1vZXZ8ota6wsAlFKfxLaMvkEp9e9uH8NOZFT3OLlvWaK6b0m9WXqt3p7Vod5QSn0K23oZftHVzXUy1ogTGU8ySilfuy46Wuu3K6X2AV8FBjvow+N9fuQo29Ba/7U7qP9GKfUA9m30Wn2aJ5VSv4s9sIeVUp5ude9pH2P0/cC1ynbBul4p9R5s94KnYE+2jmitv4a9Wb4W+yD2YiD8wFNSSj0d+9bIKKWep7X+nLKDX9fbsltWSj1Ba/1192A46/4jcBfDMLNKqSlsl6MAml0XpoC5Nu29wFO11ve32+/K8Hi16VA51IBXKaX+ENsK06mecRfyaxv14NZNm+Z1SqmLgE8opT4HvI+163hE2W5aPpDRrqtMp3yxb7uvxr4l+6x7cPosto7b97dh3yL2zepHlf3YfAo7nlL4oh64h60R7BvMi7XW31a2C0uiLd+qUupsrfU9yrZCVdz/lDvYe59S6veBD4dubtuxLVrtdXEH8Bta67vbC6hDvXVzfsq5aenm3JR6s0RVbxBd3c1jHwwPtaV3qo+Gg4zWel4p9Rzsg9+/Ebo2Ohq6qrJObOO8rymlVnRl1Vr/krumXYXtDnyNUqqqtW5vLQboU0o9Elv2Ce0iI7r/aS+Hm5RSf4ptlbhJKfV8bVtGLgMW2rSN699dwNuAtymlHo59YL8WGxikga+UGgOGgEGl1Jla65+562V7OQRKqXFtu0jvwl0ftdZzJ/EeJ/ctS1T3Lak3S6/VW0kpdYl2L/tCXAK0Dw3SzXUy1rTftIR48P9Y2aSOtm+Cfw93ELfxeWW/W2kMAA2AOylWhWvXWu/VWk8BNwFfwHZx6MQ/4W5swIeBnMt3B20PMdq+tX8H8BvYN0FPwfZ338fqwVM72VTXWl8feuvT4H9j9/sV2O5Ulyml5rEPWO3dA/439k35PPYbntc6e7dhu0OFuQLbPeGQUuoupdTdwCHs2+gr2rR/i+0u1on2CFrdaL+tlLo8nKBtX/h/Ac7soG3UcbM8lVJnA6vGJNJa34btZgW2pSa7hk03Y7uo/CL2Ted2l+8ObDe9cJ5vxR4znwB+F/twdB1wDvArbfmu+i5Baz2jtf4HvfrbpN/HHvMfwXZrebNS6ifYbnb/t037BmyXvZ9guyW+wdm7DfiPNu2Lsd1lblZKzSqlZp3949hueWH+mLWvh69tW+/m/JRzs6V9t1JqjmOfm1JvRFpvEN119SPAGWvs+8fb1u9xTmvY3iuBO4GHtmkPhuqjec105bbqnqhtq8UzgUuVUp9ntTPW4ADwbmzX4lml1E6X7wTOcQ3xGmyXuTuxD6efVrbL3isJde9zrGoZ11p/X2v9Zq31g9s2/Tn2W9tbsfXxz8p2yfs+9n4S5h3Ad5VSX8AOs/I2ZxNypUEAAAoGSURBVO82VrfktN/j7sK24m70Hif3Lcvx3Lca3emPdt+SeqMn6+3lwPuUUj9SSt3opjuAv3PbwnRznYw1EljnFEcp1Yft5vXDY4q3KO6BAa31zGbbcjyola0ZnbbvxHYbu/YkmrUhlFI5bBe99taARpCNCR0KiLEVkXOzN5F6Wx+unNBaFzts26213reOPAaAAa314aNoHgE8Tmv9D13YlsC2lBTW2D6CbUnteM9QSg3qYwT46PB/nmtZTWK/592nQwGcQtpx7He2P9HrDzIVu3uc3LfWlZ/U20ngRNabc4ibgXW0+65/qyJOZExRSj0EeyB+M3wzUqHofserF20z7dHYD79vVUo9DDv2zx3afSu0Du2PO10onQ3PZWWErmu01neItivt57XWP95IvmuhlPo1rfW/iDae2tB5fIsODb4c42tJT2kjtmMEe30Mn583dHJ4RButdi2UUk/TWn9BtPHUKvtt3za9OuJuM7qpaGOp3QH2G35lWyufiH1O/BFtrKG9U2t9e7s2zkh31hiilHodNpz9a4EfKqXCkQDfsRG9aJvaP8J2M/iAshHv3oeNOvhmZQNurEf7pg7aN2IjF3rYKGjfcsufUEq9SbRdaT+5kXyPwZ+INp7atvP49h64lvSUNmI7XoYNSHMptktvP3AZcJvbJtqTpD0GHxRtPLVKKYXt3vxppdTtSqlLQpuvFm1stb8BfAPbTfc3sd1dn439nvPKdWo/066NOxJYJ568EhsVKq+UOhP7gfGZWuv30DkKYTd60VpehO0ylMF+b3Ca1npR2ciE3wTefpzaK4HzdNtYTsqOG3Y7Nqy1aE+CVq0O9d/Aw44/iGjjpyUe14etrI0y77c67YpWMWUDx3yTlUGKRBuhVil1DZ3xsN8cI9r4abHRji/SWh9QtgfUR5VSb9b2O9/280208dG+BjgPG131PmwE54Pu3PwKK18WdKONNeJExhO/0WVI2yhtl2Jv3GfQ+YGgG71oLTVt+78XlFL3aBvVC611UbVF++tSG2Cj5rVHAtzpton25Gm3Y4OGtEe087Af0Ys2nto4XB+2sjbKvD06R2YMRHvStU/EjjvZ/m2mhx3gXbTx1Ca0+xZWa/0tZaP//odSag+r61608dFWtf2GuvGceND9bk6tjuTajTbWiBMZTw4ppS7UWn8PwL0B/kXsQKgXbFAvWktFKdXvTuSLGonKfm/S7ox0o/1t4EvKRvNqhGo+HRvSvX0MItFGq/0PYLBxPIRRSt0k2thq43B92MraKPN+O3Z8yhtZeX4+DRdJVLQnTXsLUNBa39yWjlLqTtHGVruk3LASANq2gl0KfA7beiXaeGqNUiqlbS+pZzcSlVJZVn862I021khgnRiilNqNbf3qNIbM47XW/3W8etE219PajTXWpsthB7L+wfFoXbqPfbsYDnxwq+4c+Uu0EWqF3iMm14ctqz0JeY9hW53bA7+0t0KLNmKt0HsoG014WWv9k7b0FKC01v8q2lhqTwf2azf2bSh9N/BQrfUXj0cbd8SJjCFKqduwY0BdB9yktW4fqPS49aIVrWhFK9pTVxuxHTcA1wPX6Q6RlUUrWtGKVrTr18YdcSJjiLLjRD0BG8b7MmAGuAF7wK0aVLobvWhFK1rRivbU1UZsxw6nuxw4Fxvs5Xrgizo0VItoRSta0Yr22Nq4I05kD6CU2kXrgHswduy0V58IvWhFK1rRivbU1UaVt7Ldzh8DPBN4KlAEbtRav0u0ohWtaEXbnTaWGGNk6qFpamrKn5qaenwUetGKVrSiFe2pq43YjtzU1NSviFa0ohWtaDeujcMk0Vl7j3/QWr8qIr1oN6hVSiWAXwdOA67XK4NOvEVr/WeiFa1oRbtZ2ojt6MdGSTbAe4ErgBcAdwBv0ysDUYhWtKIVrWiPoo07PRVK9lRBKTW+xjQBPGsjetFGqwX+EXgy9ruhv1NKvTu07QWiFa1oRbvJ2ijzvho7JuhZwH8CFwN/iX3W+IBoRSta0Yq2K22skZbIeHIEO5h6ePBg49YnN6gXbbTaR2utHw6glHof8H6l1GeAX277vWhFK1rRboY2yrzP1VorpZQHHAB+QWttlFJfB/5HtKIVrWhF25U21khLZDy5F7hUa31WaHqQ1vosYNVYXV3qRRutNt1Y0FrXtO3q+j/Al4FB0YpWtKLdZG3UeaP/f3t3D2JHFcZh/FkLBTFqbESDHxAFQcHK2CgWFgoqppA/qDGFiKCVlbHToIXNiqCFBoIkikleC2UlYiMoCEq0iiQKaQQ/UIRAEEU0yVrMXbjCeplwyZ2zzPODZdnZ5y5ve/bMnKlaBT6afF/7ed1T/GxtbW1tZ7etchHZpleBzf/zu/VObDqX3vb8tl8nuXf6QlXtBt4Crre1tbUduD3fc1wyaR5fu5hkK/C7ra2tre05tU3zFR+NSnIT8CCwZXLpJ2Clqr6dt7e1tbW1HW87xBzAd2v/cbe1tbW17de2zJ3IBiV5FjhI97zJkcnXEnAgyXPz9La2tra2422HmgPYZWtra2vbv22dB+u06Qng5qr6Z/piulPxjgEvz9Hb2tra2o63bWUOW1tbW9vZbdPciWzTWeDqda5fNfndPL2tra2t7XjbVuawtbW1tZ3dNs2dyDY9A3yS5ATww+TatcANdC8onae3tbW1tR1v28octra2traz26Z5sE6jklwAbOO/D91+VVVn5u1tbW1tbcfbtjKHra2tre3stmUuIiVJkiRJvflMpCRJkiSpNxeRkiRJkqTeXERKkiRJknrzdFZJkhYkyffAlcBp4AxwHNgP7KmqDXW8uyRpvNyJlCRpsR6oqk3AdXQvlt4F7B12JEmS+nMnUpKkAVTVKWAlyS/Al0mW6RaWLwFbgVPA3qp6ASDJYeDjqnpt7W8kOQo8X1XvL3p+SdJ4uRMpSdKAquoI8CNwJ/AHsBO4HLgPeCrJ9km6D9ix9rkkt9K9Z+zwQgeWJI2eO5GSJA3vZ+CKqvp06trRJAeAu4APgBXgzSQ3VtUJ4DHgUFX9vfBpJUmj5iJSkqThbQFOJrmd7jnJW4ALgYuA9wCq6q8kh4AdSXYDDwMPDTSvJGnEvJ1VkqQBJbmNbhH5OfAu3Y7jNVV1GfAGsDSV7wMeBe4G/qyqLxY8riRJLiIlSRpCkkuT3A8cBN6pqm+ATcDJya7jNuCR6c9MFo1ngWXg7UXPLEkSeDurJEmL9mGS03SLwePAK3Q7jgBPA8tJXgc+A4rukJ1p+4EXge1IkjSApdXV1aFnkCRJPSXZCTxZVXcMPYskaZy8nVWSpA0iycV0u5V7hp5FkjReLiIlSdoAktwD/Ab8SncAjyRJg/B2VkmSJElSb+5ESpIkSZJ6cxEpSZIkSerNRaQkSZIkqTcXkZIkSZKk3lxESpIkSZJ6cxEpSZIkSertX7OpOrRIDXWGAAAAAElFTkSuQmCC\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df_date_tots['Closed Cases'] = df_date_tots['Deaths'] + df_date_tots['Recovered']"
      ],
      "metadata": {
        "id": "waMkhmhpTx7d"
      },
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "df_date_tots['Active Cases'] = df_date_tots['Confirmed'] - df_date_tots['Closed Cases']"
      ],
      "metadata": {
        "id": "rKufQ9yoT6Dk"
      },
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "days_since_lst = []\n",
        "for i in range(len(date_lst)):\n",
        "    days_since_lst.append(i)\n",
        "df_date_tots[\"Days Since:\"] = days_since_lst\n",
        "df_date_tots = df_date_tots[[\"Days Since:\", \"Confirmed\", \"Deaths\", \"Recovered\", \"Active Cases\", \"Closed Cases\"]]"
      ],
      "metadata": {
        "id": "WIx0yDzYSu5X"
      },
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import train_test_split, RandomizedSearchCV\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "\n",
        "X = np.array(df_date_tots[\"Days Since:\"]).reshape(-1,1)\n",
        "y = np.array(df_date_tots[\"Confirmed\"])\n"
      ],
      "metadata": {
        "id": "w4RbdGlpUMMV"
      },
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "X_train, X_test, y_train, y_test = train_test_split(X,\n",
        "                                                    y,\n",
        "                                                    test_size = 0.15,\n",
        "                                                    shuffle = False)"
      ],
      "metadata": {
        "id": "RHhOXDifUQKi"
      },
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def cal_cost(y_pred, X, y):\n",
        "    '''\n",
        "    Calculates the cost for given X and Y.\n",
        "    y_pred = Vector of y_preds \n",
        "    X = Row of X's np.zeros((2, j))\n",
        "    y = Actual y's np.zeros((2, 1))\n",
        "    \n",
        "    where:\n",
        "        j is the no of features\n",
        "    '''\n",
        "    \n",
        "    m = len(y)\n",
        "    \n",
        "    predictions = X.dot(y_pred)\n",
        "    cost = (1 / 2 * m) * np.sum(np.square(predictions - y))\n",
        "\n",
        "    return cost"
      ],
      "metadata": {
        "id": "MVa6MZEhUWJs"
      },
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def gradient_descent(X, y, y_pred, learning_rate=0.01, iterations=1000):\n",
        "    '''\n",
        "    X = Matrix of X with added bias units\n",
        "    y = Vector of Y\n",
        "    y_pred = Vector of y_preds np.random.randn(j, 1)\n",
        "    learning_rate \n",
        "    iterations = no of iterations\n",
        "    \n",
        "    Returns the final y_pred vector and array of cost history over no of iterations\n",
        "    '''\n",
        "\n",
        "    m = len(y)\n",
        "    cost_history = np.zeros(1000, 1)\n",
        "    y_pred_history = np.zeros((1000, 1))\n",
        "    \n",
        "    for it in range(1000):    \n",
        "        prediction = np.dot(X, y_pred)\n",
        "        y_pred = y_pred - (1 / m) * learning_rate * (X.T.dot((prediction - y)))\n",
        "        y_pred_history[it,:] = y_pred.T\n",
        "        cost_history[it]  = cal_cost(y_pred, X, y)\n",
        "        \n",
        "    return y_pred, cost_history, y_pred_history    "
      ],
      "metadata": {
        "id": "WmI651cGUck4"
      },
      "execution_count": 39,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "print(y_pred)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qSxyFyAQbS00",
        "outputId": "0206a8f8-9e48-42b4-c080-b6f809e61b3b"
      },
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[ 3.54606154e+00 -5.94295184e-01]\n",
            " [-1.38399473e+00  1.49069508e+00]\n",
            " [ 2.50473971e-01 -1.90501143e+00]\n",
            " [-7.06877507e-01  9.82567567e-01]\n",
            " [-2.45687735e-01 -9.59811430e-01]\n",
            " [ 4.72959866e-01  6.88933223e-03]\n",
            " [ 4.88514783e-01 -1.53753839e+00]\n",
            " [ 6.23831247e-01 -3.78115803e-01]\n",
            " [ 1.63017207e+00 -1.61142541e-01]\n",
            " [-6.42898641e-01  5.34726478e-01]\n",
            " [-1.01509714e+00  4.10612092e-01]\n",
            " [-3.31414864e-01 -3.92391304e-01]\n",
            " [ 8.13259454e-03 -8.74073671e-01]\n",
            " [-1.79013085e+00 -6.55091021e-01]\n",
            " [-2.11356928e-02  4.16022933e-01]\n",
            " [ 3.02190753e-01 -6.96271628e-01]\n",
            " [-1.22443124e-03  1.73814978e+00]\n",
            " [ 1.00891867e+00  2.34748445e+00]\n",
            " [ 4.84324673e-01  4.85136474e-01]\n",
            " [-9.10397515e-01  9.41121139e-01]\n",
            " [-9.80849513e-01 -3.44348699e-01]\n",
            " [-2.31195794e-01  1.37278406e+00]\n",
            " [-6.42631155e-02  1.33405213e-02]\n",
            " [ 1.38953810e+00  8.35744755e-01]\n",
            " [ 6.31180959e-01 -4.90329467e-01]\n",
            " [-1.00348253e+00 -6.78673992e-01]\n",
            " [ 8.94717972e-02 -3.74298142e-02]\n",
            " [-1.10567652e-01  4.40796095e-01]\n",
            " [ 1.99182272e+00 -8.09310784e-01]\n",
            " [ 2.63117119e-01 -7.27764886e-01]\n",
            " [ 1.18963697e+00  1.66715160e-02]\n",
            " [ 1.13788987e+00 -2.00887992e+00]\n",
            " [-1.16101197e+00  1.90126890e+00]\n",
            " [-1.25343289e+00 -1.47658420e+00]\n",
            " [-4.23062635e-01  8.83168242e-01]\n",
            " [ 1.72421339e+00  5.23668347e-01]\n",
            " [ 7.13876327e-01 -6.82317668e-01]\n",
            " [-2.28828618e+00 -1.21434627e+00]\n",
            " [-3.42885402e-01 -1.37709333e+00]\n",
            " [ 4.73707812e-01  1.06158619e+00]\n",
            " [ 2.21440505e+00  7.46109235e-01]\n",
            " [ 1.22855582e+00  9.88939803e-01]\n",
            " [-6.37645771e-01  1.29018220e-01]\n",
            " [ 3.50955462e-01  1.04242713e+00]\n",
            " [-1.27854264e-01 -1.02144818e+00]\n",
            " [ 2.10032486e-01  2.34106295e-01]\n",
            " [ 7.73603372e-01  3.18316772e-01]\n",
            " [ 7.48584726e-01  1.73789862e+00]\n",
            " [ 1.18503942e+00  1.28154183e-01]\n",
            " [ 1.27223206e+00 -7.36291670e-01]\n",
            " [ 1.41797498e+00 -3.51782739e-01]\n",
            " [ 4.25631641e-01 -1.60154886e+00]\n",
            " [ 4.65237285e-01 -2.91202514e-01]\n",
            " [ 1.21851762e+00  1.48205413e+00]\n",
            " [-9.56406850e-01  7.29513152e-02]\n",
            " [ 7.64769384e-01 -8.82299670e-01]\n",
            " [ 1.76277060e+00  1.63591076e+00]\n",
            " [-2.95828016e-01  3.51066794e-01]\n",
            " [-9.02382557e-02  1.98609730e-01]\n",
            " [ 3.62755439e-01  3.70145854e-01]\n",
            " [-8.22338698e-02  7.19439959e-01]\n",
            " [ 1.11844413e+00  8.26801321e-01]\n",
            " [-2.52181460e-01  4.18829161e-01]\n",
            " [ 2.50867565e+00  1.99362446e+00]\n",
            " [ 9.69421717e-01  2.20582228e+00]\n",
            " [ 1.40219157e+00  1.43711076e+00]\n",
            " [ 6.23357820e-01 -1.19992716e+00]\n",
            " [-5.93548360e-01  1.83312859e-01]\n",
            " [-1.86194026e+00  1.79324776e+00]\n",
            " [-1.09279821e+00 -5.15545866e-02]]\n"
          ]
        }
      ]
    }
  ]
}