Air_pollution_forecast.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Air_pollution_forecast.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "MXB6ayaPfYxd",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "from matplotlib import pyplot\n",
        "from numpy import concatenate\n",
        "from pandas import DataFrame\n",
        "from pandas import concat\n",
        "from sklearn.preprocessing import MinMaxScaler\n",
        "from sklearn.preprocessing import LabelEncoder\n",
        "from sklearn.metrics import mean_squared_error\n",
        "from sklearn.metrics import mean_absolute_error\n",
        "from tensorflow.python.keras.layers import Dense, LSTM , Dropout\n",
        "from tensorflow.python.keras import Sequential\n",
        "from math import sqrt;"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YX7A5D8apjIU",
        "colab_type": "text"
      },
      "source": [
        "**DATASET PREPROCESSING**\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "Dataset Loading and dropping the rows with NA values.Plotting the graph for each columns. Further preprocessing includes:\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "1. Feature Selection\n",
        "2. Label Encoding\n",
        "3. Data Normalization\n",
        "4. Converting to Time Series data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Lx2bZx7ugNqz",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#dataset preprocessing\n",
        "#datset loading dropping na values\n",
        "\n",
        "dataset= pd.read_csv('/content/pollution.csv');\n",
        "dataset=dataset.dropna();\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OsTy5A67uUq0",
        "colab_type": "text"
      },
      "source": [
        "# **1. Feature Selection**\n",
        "There are 8 features important for the forecast: PM2.5, dew point, temperature, pressure, wind direction, wind speed and the cumulative number of hours of snow and rain. Hence, dropping other features."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7TS4iRJTuSuL",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "dataset=dataset.drop('No',axis=1);\n",
        "dataset=dataset.drop('year',axis=1);\n",
        "dataset=dataset.drop('month',axis=1);\n",
        "dataset=dataset.drop('day',axis=1);\n",
        "dataset=dataset.drop('hour',axis=1);\n",
        "dataset.head();\n",
        "\n",
        "values=dataset.values;"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PKlDY_RWr5ET",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 268
        },
        "outputId": "b539f48e-b40b-47db-f983-2a08fe884432"
      },
      "source": [
        "# specify columns to plot\n",
        "groups = [0, 1, 2, 3, 5, 6, 7]\n",
        "i = 1\n",
        "# plotting each column\n",
        "pyplot.figure()\n",
        "for group in groups:\n",
        "\tpyplot.subplot(len(groups), 1, i)\n",
        "\tpyplot.plot(values[:, group])\n",
        "\tpyplot.title(dataset.columns[group], y=0.5, loc='right')\n",
        "\ti += 1\n",
        "pyplot.show()"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAD7CAYAAACG50QgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0\ndHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOydd3hURdfAf5MeUkggoYYklICEGkCK\nVBFBBEUREQXs9dXP8ooIUkSKIL5iFwtiQ0WaIF1a6DWBECAhBdJ77213vj92s+wm2ZTNbhJgf8+z\nT3LnTtu79547c+bMOUJKiRkzZsyYuT2waOwOmDFjxoyZhsMs9M2YMWPmNsIs9M2YMWPmNsIs9M2Y\nMWPmNsIs9M2YMWPmNsIs9M2YMWPmNqJWQl8IsVYIkSKEuKSV1kIIsU8IEa7+66pOF0KIL4QQEUKI\ni0KIflplnlLnDxdCPGX8r2PGjBkzZqqjtiP9n4H7KqTNAQ5IKX2AA+pjgPGAj/rzIrAaVC8J4H1g\nEDAQeL/8RWHGjBkztzvqQXGAECJHCBEnhFgphLCqJr8UQuQLIfLUnzW1aqe2m7OEEN7ADillT/Xx\nVWCUlDJRCNEW8JdSdhNCfKf+/0/tfOUfKeVL6nSdfPpwc3OT3t7eteqjmdsLhVJiaSEauxtmzBiF\n1NRU7OzscHBwoKysjMjISFxdXWnTpk2V+QMCAujRowd2dnZVnUuTUrpXWVBKWasP4A1c0jrO0vpf\nlB8DO4BhWucOAAOAWcB8rfQFwCw9bb0InAPOeXp6SmPzf38Eyo3nYo1er5mGIy6zQHq9u0N+6x/R\n2F0xc5vj5eUlP/zwQ9m9e3fp4uIin376aVlYWCgPHTok27dvLz/66CPp7u4u27RpI//++2+5c+dO\n6ePjI11dXeWyZcv01vvJJ5/IiRMn6j0PyPDwcH3nzkk9stwoC7nlHTBGXer6vpdSDpBSDnB3r/pl\nVR/+CUpg1sYgo9drpuGIzywEYN+V5EbuiRkz8Pvvv7N3714iIyMJCwtj6dKlACQlJVFUVER8fDyL\nFy/mhRdeYN26dQQEBHD06FGWLFnC9evXq6zzyJEj9OjRo9p2R4wYQZs2bZg8eTJRUVG16mt9hH6y\nWq2D+m+KOj0e6KCVz0Odpi/djJk6I9RandvFc9Tey0mEJ+c2djfM6OG1116jQ4cOtGjRgnnz5vHn\nnyqttbW1NfPmzcPa2ppp06aRlpbGG2+8gZOTEz169MDX15egoMoD0LVr13Lu3DlmzZqlt83Dhw8T\nFRVFaGgo7dq1Y+LEiZSVldXY1/oI/X+Acgucp4BtWulPqq14BgPZUspEYC8wVgjhql7AHatOM2Om\nzpRr8uVt4jDwpd8CuPfTI43dDTN66NDhxnjWy8uLhIQEAFq2bImlpSUA9vb2ALRu3VqT197enry8\nPJ26tm7dyty5c9m9ezdubm562xwxYgQ2Nja4uLjw+eefc/36dUJCQmrsq96VYW2EEH+iWoh1E0LE\nobLCWQFsEEI8B0QDU9XZdwH3AxFAAfAMgJQyQwixBDirzrdYSplRm/bNmKnI7TbSN9O0iY2N1fwf\nExNDu3btDKpnz549vPDCC+zcuZNevXrVqawQolaDoFoJfSnl43pO3VNFXgm8qqeetcDa2rTZGLyx\n/jzbLiQQtWJCY3fFTI2opP5tMtA308T5+uuvmThxIs2aNWPZsmU89thjda7j4MGDTJ8+nb///puB\nAwdWm/fy5cuUlpbSq1cvCgsLmT9/Pu3bt6d79+41tmPekavFtgsJjd0FM7XEPNI305R44oknGDt2\nLJ06daJz587Mnz+/znUsWbKE7Oxs7r//fhwdHXF0dGT8+PGa8+PHj+fDDz8EIDk5mcceewxnZ2c6\ndepEVFQUO3bswNrausZ2ajXSN2OmqaGxzjcP9W8LLsZl0a2NE7ZWlo3dlSq58847mTt3rk7aqFGj\niIuL0xxbWVlVUr8cO3ZM8/+hQ4eqbWP37t2a/0ePHs3Vq1cN6qt5pG/mpkSoh/pmkX/rE5NewINf\nHeeD7Vd00t/fdom/z8fpKWVGH2ahfxMz88fTzFhzusZ8sRkF+F9NqTHfzcTtug/3WHhaY3ehwckq\nLAEgOC5bJ/2Xk9G89VeQ9qbOSiiVkqJShcn7eDNhFvo3AcfC03jsu5OUKZQ66UfD0zgWUbMQGPfZ\nEZ7+6WyN+RqK3cGJbLtQvy0aB0JVL7HbTbsz48eaX/IFJWUUlNRsr32zIfXM6zrO3cXy3aFVnpu7\nJZg7FuwxZbeIiopizJgxJm3DmJiFfhNh1sYgei2qetvCjB9Pc/p6Bml5JZq02IyCWtddUNK0Rjqv\n/B7IG+svVJunqFTBnM0XSc8rrnRu1MeH+OJAOKBfENzKzFhzmoDoDPZdSSYtr5jtQQl4z9nJwm2X\n2HA2Ft+Fe+n7wb7G7qbRELWY131/5FqV6X+di60yvakQFJvFJ/8apps3lNtiIXf+1mC6tXZi5hDv\nxu5KJZRKldDaFBCnObZQOxE7HpHGRa0prbaA22DAzRySmEP3ts716S65RaVYWgia2Rjn1pFSkphd\nRDsXe530u1YcJCO/hBKFklVT++qci0ov0CpvlG40aeKzCnWOj0XcmOF1b+uMQqmaAf56MlqTp6TC\nrNAQLidk0665Pa4ONvWuqzZsCYzD3tqS8b3aUliiwNpSYGV5Y1wqJaTlFROenMeQzi2rrSsiJRc3\nR1tTd7lO7LyYyJ0dXWnldMNB2qSvjwPw9thuDdaP22Kkv+5UDAu2XW7sblTJ5NUn6PTeLs3xuM9u\n7LqcvuY0H+25MW0dsvwgq/0jAcN02uM/PwqoBLeh9Fr0L0OWHzS4vDaZ+SU89dNZ7lpxkMCYTJ1z\nGfmqWc2WwBtqoO8OR3I1SdcVwc0i9DecjeUh9QNeFbM3BXHnsv0olJW/0O7gRL3lotPzSdeaARqT\nCV8cY/LqEyapuyr+uyGIV34PBKD7wj3835/ngRvmuaFJuUz99iSP/3CqxrrGrDpC38WNP9vxnrMT\nv8X/kltUyqt/BPLU2prVrPFZhYz//CgpuUUm6dMtL/Tre+H2Xk4iOcc0Fx/gQmyWznF4Sp6enCrK\nXwJlVQgHbYpKFaTkFnEtVbe+sORcei36VzOzMITsQt2XRqlCyW8noyqtOVRFQPQN4f76+vMcCUsF\nIDIlDyklVxJy2KhnFrN8dygPfHlMJ638KkgpK70Qyhmy/AAfbG/cl/7szRcr/dbabDgXR2puMXnF\nN3TxsRkF5BeXsXSn/q31BSUK0vNNI/QBrqfl16v8G+vP6/ye3nN20nHuThKzVbOXjPwSvjwQrpnx\narP7UpLOsUIpuVbP/jQk5d8ps6CUXov+BSAu88YstYvWYC8oNktjifTLiShCEnN0BjzG5JYX+ici\n0jX/l5TpCiXvOTtZdyq6YhG85+wkOj0fhVLy0m8BPPbdyXr341J8Ns/8dKZOuvjq+EY94tdHvyX7\nGLjsAMk5ujrxCPVL5UCI8bxTvrclmAXbLrPmmK63wNiMgkovzEe0Ro7aL4DswlK+8Y/k/i+O8s6m\ni5XaKH+hVFRblFttbL0Qz7jPjnAgJJnsglL2XFKNjj/dF0ZidhE/HY8y6LsVlJSRUEG9Uh+yCkqY\n+t1Jhq64MVvSrr/c0uRERBrDVx6q1cJtU2bbhYRKv6eUME7tR+jdzRf5ZF+YjkFCRSudqmY/N4PP\npYIqrIZyi2681LUHbpO+Ps5bfwURn1VIaq7qmf3lRJRJ+nXLC/2jWiZui3dUHu3N33qJpOwiPtyl\nO5r693Ky5iVRrkMuUyg1ageA305GceZ61e6DTl1L50hYKlJK3t10kad/Osuhq6kMX3mo3pYV71Yh\nFCtSvnirPRXu0soRK/V6QalCyau/B9JvieFT4LziMmLSC9ionjWs2B2q8zAOX3mIQR8e0BwfCtU1\nG9VeYF66M4SP91a9oDVrYxCr9oVV25crCTkARKbm0Wfxv7y8LpD4rEI+Vy/4VsXmgDjmbw3G/2oK\nH+8NJSWnSDMCLcd34V7uWqGrziooKeOu5QcqvcDzi8v03g/l9F28jzPXM4jPKtTcB6FJOZrzgz48\nwG+nonlCbYp7Pkb/7MBQsgtKefib40Sn6x81VzXyriva3yuroITzWiq8nKIyErMLNa6xvzx443d6\n4Cvd2VxhFcKzqu79eOw6l+KzuRSfXfmkAUSm5mkGSYaQVaB/BlaqZ1Y8d0swMer7KjHbNBqGW3oh\n9+tDEWwOvKHGCIjO4hv/iEr5Bi8/UCktt6iU0Z/4a46zCko0OsLQJfdhZ22pWSco99WTnleMSzMb\nLC0E077Xr3csKFEwd0uwXrcPv5+OrvZmq2iR8O6mi5y+ns7WV4fi0kz/olt7F3uy1KqZnMIyzkSp\nhPC3hyOREib0akvr5rbYWllyKT4bIcDdyRZ3R1uEEHx7+MbsYuiKg7g0s+ZyQo5OG9fT8unk7qiT\nNnDZfs7MG8MzPxtmNropII4WehYTpYQ/Tsfww1HVLENbSFa0z/7zTAyPD/QEYM+lRN5Wx1RYdyoG\ngK8Pqb5f6JL7CIzJ5K7ONzwcBsVm0aeDCwAr91wlIbuI4SsP6fhpentDEHsuJ3Fm3j06i3Wvq3XT\nFckuLKWZjRUhibpqqQVbL1WZv64olZKAmEzu9G4BqFQLXx+KoLeHC+djshj5sT/7/zsC75YOOgum\nAMn1VIteis9mopYqrir9uvbaUEWVYTmf/HuVLw9WfmbL4ylos2SHavNWr/bNK5177+9gPnz4hgMz\npVLy2YFw+nZoztAublXu9L3nk8MAOr9xUakCG0sLjbFFdZSP2CsSlpzL9qCqn/1ydacpuaWFfsWR\no5SSlXtqZx619UKCzps2SGvKee+nh+nb4UZ4368PRei0tfHlIdXWnZJTXK2fn3l/1+2hL38J9F28\nr5LA0aa/lys/q9UcZ6JujEhXqG2cy9cLolZM0H1gO7gQFJels2gan1VYyaoEVKPyt+7tynCfG8Fv\nUvTc/HUhv7jq2dHV5Fze+ztYc6ytB165R9d2e+6WYB4f6EluUSkvrwvU29aCrZfYGBDH3PF3aNIm\nfX2cjS8P4fS1dH6uYtotpWTPZVXb56Iyub9XW7aej2dfSDI7L1a9EPv5/nDmT/TVO8OpL+UGAs52\nVlxcNI5hH6m2+WurS8asUqlZurZ2JCw5j5WP9OaBPu10BPLlhGx6tNMVpFJKypQSa62XRXR6Prsv\nJfH4QE/ORdXNgW5YctWDnKoEPsCIj/W7LAiuYqT/x+kYPny4FwHRmXi42nM1KVdj9jt9kCfL1C+E\ni3FZFJUqGdixhabsn2diaO9iz6BOLbhjwR5eGN6ReRN8KSxRYGdtUcm75Tf+kXi3dNBrTjy2kV1k\n1zpGbmMxYMAAee7cuTqX238lmed/1S3XpZWjwdO1NU8OqFRfU+TZoR0Z7uNW5aj6rTFdKVMq+fJg\nBO1d7KsU2sbi/IJ78auH6shUrH9xMF8cCOdEZHrNmWtB1IoJFJcp6DZfdwPQogd8WVTBbUBVNLOx\nbJB9FFErJuA9ZycAni2aaVQItWX9i4MpVShxsbehhaMNT/54msjUfFZP78f4Xm0BNPUDvHp3Z83M\nqalQ/mIDWD29n8ZSqE8HF7a9OhS48R1srSworrAG+P4DvhpXEKfm3sPg5Qe4u5s77Vzs+f20ara4\n+ZW7dNat6ouhHn+FEAFSygFVnrtVhb72DWhGxZBOLTl5zTjCzoyK4EVj+fJghN7NQU2FA2+P1Kgr\n6jP40ceml4cw5dv6Gzw0Fv97tA8Te7et9+7dO9o4EarHiswQbgmhL4S4D/gcsATWSClXVJffLPTN\nmDFzu2IKod+g1jtCCEvga2A84As8LoTwbcg+mDFjxsztTEObbA4EIqSU16SUJcB6YFID98GMGTNm\nblsaWui3B7TtDePUaUanqfjdmNzPJF/PTAPSx6OyCaCZ25Nys92bmSZpsimEeBF4EcDT09OgOs7N\nV7k6vZqUS0c3B7IKSiguU+LqYMP6MzEM6tiSXh7NKVUoOReVSZdWjrg72ZKRX0IzG0tKFUqi0wvw\naa1a9MrIL8GzRTOup+UzsGMLbCwtsLQQCCFQKiV5JWU42lgRlZ5PdHoBd9/RStOXlY/0Jio9n45u\njiRkFaKUktbOdthZq2yDZ28KwsbKgof92vPv5WT+O7YrWwLjGevbGv+rqeQWlTKuZxuKS5U42Kra\niEjJ4+5urXCwtSS7sJRAtX16x5YOnI/NJD6rkLnju3MgJJl+nq64NLPm5xNRHA1Po28HF+IyCxjY\nsSWlCiWWFgIkjO7eivjMQrq3dSY0KYc9l5J4dlhH9l5Ooke75rRzscPF3obcolJaOtoSGJNJ9zbO\nKKTEzsoCK0sLpJTM/PEMHz/am1ZOdpQqlPx8Iop2LvYolEpGd2vNPxcTyMwvQaGUuDnZ8sRAT9Ly\nisksKCE5p5iiUgXjerQhM7+EfSHJuDazoW1zOzq5OxCalEsnNweEEOy9lISHqz3X0/OJzSikbweV\ncG7lbEfv9s05EJqCi701vT1c2HohnoiUPLq1dsLSQjCwYwvS80tobm9Nc3trXJtZU1ymJCw5l4z8\nEtwcbXlq7RlaOdux9dWhKKVqr4alhSA9vwSl2lzxWEQa43q04XJCNp3cHFFKSUZBCQdCkunt4UJn\ndwfGrDrC59P6cig0hSGdWzLMx52z1zMQAnzbOnM4LJX2LvaUKJQ0t7cmOr2A6YM8yS0q4+/z8Tzk\n155mNpbYWVuSkV+CnbUFcZmFBEZn0trZjrjMAu7s2II1R6/zzrhulCklqbnFZBeW4uZow+7gJLYF\nxbN4Uk+yC0pp6WjD0M5uRGcU4NmiGZYWgtTcYtydbCkqVVBQotC0dy01j1bOdpQplFhZWmBlITT3\nLajs8XMKS7manMvE3u2QSPKKyjgWkYa7oy1uTrZ0cXdkc2Aco+9ohZ21JQlZhVxJzKGNsx2bAuKY\nMdiL8zFZPNCnLV8djGB8r7b4tFI9KwM7tuBoRBpd3B05F51B3w6ufHkwnFZOdkwf5EmHFs1Iyyvm\n78B4Ork70MzGikvx2Uzq2468YlU/nqzgaFGhlKTlFWNrZYGdtSUnItPo4+HCpoA4pvT3IKuwlPDk\nXHxaO9HSwYZV+8L4v9E+XE/L505vV346HoVni2b4ebqw93IyVpYqGZBVWEqLZjaM7OZOZEoeHq7N\nkEh2XExkxiAvsgpLuJqUy8q9V3FtZs0Lwzvh284ZhVLy8rpAJvVtR0pOMQUlZdx9Rytc7GsOfWgI\nDbqQK4QYAiySUo5TH88FkFIur6ZMKlDZV0LtcANuv6gTdcN8jWrGfI2qx3x9aqahr5GXlNK9qhMN\nLfStgDDgHiAeOAs8IaU0iTcsIcQ5fSvYZlSYr1HNmK9R9ZivT800pWvUoOodKWWZEOI1YC8qk821\nphL4ZsyYMWOmMg2u05dS7gJ21ZjRjBkzZswYHZNY7wghOgghDgkhrgghLgsh3lCnLxJCxAshLqg/\n95uifS2+N3H9twLma1Qz5mtUPebro0YIESWEKBRC5AohsoQQJ4QQLwM/qM//LIQoEULkaX2C1Of2\nCiHe1aqrvRBC6klrI4QYJYRQquvIFUJcFUI8U2MfTaHTF0K0BdpKKQOFEE5AAPAQMBXIk1L+r7Z1\nubm5SW9vb6P3saGQqLxvtnSw1UQAutWQqDwKtnJqGmayZsw0FsHBwXh5eeHs7IxCoSA3N5fY2Fic\nnJzw9vYmKioKa2tr2revbMqdmJhIXl4ePj4+AGRkZJCYmIiNjY1OWkJCAj179iQ3N5fr16/Tu3dv\npJRkZ2cTGRmJr68vV65cSWvUhVwhxDbgK2AodRT6hrphaArkFJXSWx0xZ1Q3d1ZO6Y1A4H6LCcf7\nPz/KlURdF8tRKyaQmF2IvbVlte6ebxYKSsq4npbP6WsZTBngwZWEHB332dV5N72ZUSglkal5dG3t\npJO2at9VCkuUbLsQz9l5Y2rlavhWID6rkJj0Ar0xer29vVmzZg1jxozRpJ05c4bBgwdz8eJF/ve/\n/+Hh4cHSpUsrlT169CgPPPAAGRkZWFhY8J///Ic+ffqwcOFCEhMTNWklJSWsWbMGf39/ZsyYQVzc\nDffx7u7urF69mkcffVSvGwaT6/SFEN6AH3AaldB/TQjxJHAOeFtKmVlFmXrb6TcWWwLj6Ofpyqj/\n+euk+19NZeAyld/+VVP7MKhTS6SUtHKyw8bq5ollcyAkme1BCWytxjU0wPO/nGO/OjqXof5Dmgr/\n+T2AXcE3XDYv3lHZe2b5bwvwwYM9eOou74bomsmQUtJxbu2W3s7HZtLfq0XNGW8yyv13TR/kqfGi\nWU7EsvGk5Bbz19lYRnR1p5mNJd3bOldZz8CBA/Hw8ODo0aPVtjdw4ECKi4sJCgrCz8+PI0eOMGvW\nLH799VedtNmzZ1cqq1Qq2bZtG1lZWfTq1auK2m9g0pG+EMIROAwsk1JuEUK0RmWrKoElqFRAz1ZX\nR1Mf6ZeUKek6f3e967m+/H5EE9X/ZBeW8tGeUP6ocOPXFjdHWx4d4MEb9/jobOxpapSUKSlTKrmS\nkMOUb0/y4ohO9fKeGbFsfKXgJE2VMoWSlNxiUnOL+XR/GP5X6x7M4/yCeylVKiksUeBkZ6038E1T\nIr+4jMJSBS721hSVKen5/t561Vf02yus+2WtzkgfYPDgwTzwwAOEh4ezfv167OxuzAonTZrEL7/8\nAsCoUaN4+OGHmTlzJn369CE2NpY5c+bQtm1bZs6ciZubG9evX8fLywt/f39Gjx6Ns7MzFhYWeHp6\nMmfOHKZNm1atwzWTjfSFENbAZuB3KeUWACllstb5H4Adpmrf2GTml+C3ZB/+s0bh7eZAUami3m5Y\ntdkcGM+DfdphbSmahPDfcC6W8ORcTUSq+pCWV8xq/0hW+0cyuV97Xh/tg1fLZo3+PYvLVFGQhBCU\nlCnptmC3TqCY+rpL7rloL0WlSl4f3YX/ju1Wz94al/IAILnFZRoVZH2pGD/h82l9mdi7nWrHdxMk\nMjVP427aWKTlFTN9zWns9xez6AFfzW7g+Ph4WrRQzYZmzZpVpXoHYMSIERw5cgRvb2+GDlX5+B82\nbBg//fQT3t7edOjQAS8vL03+du3a6ah3aoNJhL5QPc0/AiFSylVa6W2llOVhhB4GjBMXrgFYslM1\npa+otqmOlVN6M7Kru06cWH3M2hjErI1B9PFozrbXhhnaTaNwOSGb2bWIw1vOQ33b8dk0P6LT80nL\nK6k2iMSWwHi2BMYDcPidUXi1dEBK2eAvgL/OxvDu5uCaM9ZAdcFCikpVQTi+OBjBsYg0+nu5Mm9C\n4zuVjc0oYPhK/ZGnjMUb6y/wxvoLbH7lLvp7udZcoAHYEhhHqULJ0C5uRhf4FVm0/QqLtl+hODGM\n5Ph4rNp2p1RRfcyBESNG8O233+Lt7c3w4cMBGDp0KM8//zze3t6MGDGi3v0y1Uh/KDATCBZCXFCn\nvYfKlXJfVOqdKOAlE7VvdCwNEEpTB3SgruqzoLhsjS6xoXXhhsYg+GyaHwBeLR3waunAvPu749Pa\nkad/qj4m7siP/TX/39+rDd9M729Q+3VBX8zV2jJ3/B0s3x2Ku5Mtq6f3Y4B3i1pFiAqMySIwJotS\nhWRwpxa8vC6Q8GXjdcINNhSHrqbUnMmIPLL6BGO6t2bNU42zITU9r5i3NgQ1SPxZbZTFBRTFXiLz\nwPc08x3FslP5KMPT6Oilf51yyJAhZGVlsW7dOvbtU82cXF1dcXd3Z926dXpnCHXBJEJfSnkMqEpK\n3nSbsqZ+exI/Lxc2BtRtCuXdshkAQghcmlmTVaAK/Hzg7ZGEJOagUEreWH+huioajPMxmfx2qnbu\njc7OG8Ody/Zjb21JYakCJ9vKt9ALIzoB8MmjfTTBx2tiV3ASUkqupeVrHKqZAkMF/hv3+PD5gXCm\n3enJiyM66fSvqoU+ffx8IkoTY9dv8T72vDkcD9dmBvWprqTkFvHBP1fYGVx1zN6KzB1/B9MHe+Fg\nY8nEL49xOUFloVX+slq+O4TvDtdOBbY/JJnDYan07eBCcxM5EtPm8e9PcfJaep3uwXJeGN6xklrT\nykLw7Yz+3NO9VY0L3KmbF4OFJSCwduuA850P4dh3PAAZ+SV89NFKPvvsM01+Ozs70tJUbnkcHBzo\n378/oaGh9OzZU5Nn+PDhrF692igj/Vs2cpahJGQVMu7TI7z/YA+sLUWtBPOVxePwXai7APTvWyM0\nZm7l0+n/jOrM7PtuBNt+9Y9AvUGzAd4c40NAdCa/PTfIwG9TO+oywi+ffaTnFTP1u5N8/+QAOrs7\n6s0/d0swf56p2wLwpL7t+Fw9e6gvecVlCMBB/XIyZDbz9F3eLJzoS2GpQlNPRQydJbVxtsP/nVFG\nXeAu/84FJQo+3hvKhnNxOjFyq2PF5F7M2aJSe2nPNHOKSnnoq+OsnNKbAd4q3XR8ViFDVxyssp6a\n+GZ6P+5Xx9Y1BfWJnBe1YgLP/HSGoV3cGNWtFSevpTNjkKfmRX84LJXP94fxwYM9eeCrYwa3YUqa\nTLhEdeSsMOBeVL70zwKPSyn1RpA2pdD/93ISfT1dNPbVc7cEU1yqYMv5+FrX0aOdMztfH05ecRlS\nSnqpF8WCF43Fye7GiCYpu4hWTrY69sxSSpJzihm8vHqd/79vjeBb/0hWPda3Ll+vRuZuucimgDhK\nFbW7BwIX3Ftni4wFWy/VehahjbEeivKHf/trw+jl0dwgYVCbvtQ3POfeN0fQrY1TzRn1UFKmRAj4\n7WR0lSalteX7mf1Jyinix2PXOfzO3TXm3x6UQFhyrkEzKGP9xsVlCiyF0FhKRaTkMWZV3fX1s8Z2\n5SG/9nWaeZX/7h9P6c07dVgHC11yH3bWlmTml3AsIo1+Xq60d7Gvc5/10SjWO3rQRM4CEEKUR84y\n/C41kJIyJS/+FgCobr6k7KI6j0gBNr48BABH9Qhw8ytDOBiaoiPwAdo0r7xxRwhRZXpFxn56BAA/\nTxec7a15sE87g9Qfl+KzWXvsOiun9MbK0oI/z8TWXEgLQ0zwJvdrb5DQf3LtGY6EpdLZ3YFPpval\nbx2CV1yKz2bil8c4OXe0JkUljvUAACAASURBVO2Br47xrtYsy1TMu787y3aF1LncuM+O8OljfXjY\nzwOAe1cdZmLvdrwxxqdS3qoWvrvO300LBxsy8ksM67ialo42jO3RppIPen080KcdYJja7IVfz7F8\ncq86BzwqLlNga3VjZtRt/h7uaOPEnjdVqo/nf6l+LUkfr42ufK1rYt9bIzgRmc6jAzpgaSH474ba\nqZLuWLCHh/q20+x3adfcjhNz76lz+4bQJCNnCSFeFEKcE0KcS02t/+LL+M+P8naFHyMpu0jz/xcH\nwmscbeujmY3ue7O/VwveGWca4bJg22XeWH+BwJhK+9mqpKhUQUB0BhfjsvCes5OJXx5jy/l4uszb\nzY6L1W+uqoihVnd+nq4GjejKF90iU/N56Ovjlc5fis8mu7C0UvqfZ2KY+KVqyj2igoXKR3tCa93+\nl4/XTb20amofvp3RjxdGdCJqxQSOzxnNZL/2uDarvf76rb9U92hyThHhKXl8uj9M53x6XjEhiTm8\nvC4A7zk7ORCSjPecnZrRZn0FPmDwJqsDb4/k7LwxNWfUYt+VZAYs3a8xdhj36RGGfVS9uig0KYc7\nFuzh38uqzXJ/n49Tp+eSr55tR6UX1Ln/ni0MW1fxae2k2Yh3X882ANhZW2BfC3Wd9gbHhOwiotLy\nDepDXWlo9c4U4D4p5fPq45nAICnla/rKGEO9o20NU99puDZLJvVgZi1HRNVR3qfnhnXkx2M128V/\n8mgf3J1sKSgpw8rCgjG+ravM5zNvV61VN9Xx+/OD6O/lWi+98zsbg+jSypHlu2sveLWJWjGB62n5\nONlZcffH/uQWl2nStTHG7zuwYwt+e24gszdd5J1x3eq10Prir+f490pyzRm1aNvcjkT1oKTcHNaY\n92111FflciUhh99PR9d6Ybucza8M4ZHVJ3X6oFRKOr23Cyc7K3KLyhjr25ro9AKuJudWKmMo387o\nx8vrAjk7b0y93aNIKflg+xUe9mtPnw4uSCkZ+bE/MRm1fwn9/vwgurVxqne416ak069z5CxDhf4r\n6wJwsLXi4ym9a72dvLYM8HJl/YuDjbbbMiIll7xiBX07uBhN5xwUm8WkKkbIhmDMRadtF+J5Y/0F\nVk7pzdWkXM1LzrNFs2ofjqD3x9Lng8qbiLxbNmOAdwvS84o5ZMAuUoAfnhzAC7+q7rEJvdryydQ+\nRltYLXdH8dQQL345aVgAuC3/uYvJ3+jf+2BMjL2WYmgfdlxM4LU/zhulLzW1ZUo2BcQxq47WQ+Ws\nmtqHyf08DCpbndBvaPXOWcBHCNFRCGEDTAP+MUVDuy8lsSkgDn8T2Oa+dW9Xo26v79LKSaOzvvzB\nuDqrUopKFZXSjCHw+3Zw4cLCe+tdjzaT+rYnasUEpg7owPwJ3XlheEeC3h+L/6xR1ZarSuADRKUX\nsCkgzmCBD+BsZ8Wnj/WhvYs9Xz3hZ1RLGmd7lfpv1B2tOPD2SIPqaCiBP32Q8fxcPa1WeZxfcC8/\nPX1nncp+eSDcJAK/4vX/bqbp94U46rH2qg21XR+oKw0q9KWUZUB55KwQYIOpI2c9U8MGobrw+bS+\nHHv3boZ2cTNanRVxsLXi2aEd61TmjgV78J6zkxMR9QvB+dUTfjovnK2vDjWph0whBPMm+NLc3rpR\nvDQ621lp+vGwnwfH54w2+v6A9x/owez7ujHSx5226kV7n1b6TVwbk8fu7GC0uhY92IOoFRNwdbCh\nXR2tUj7ZF1ZzpjpydPbddHZ35OWRnQGwtBCM69HG6O1U5F616nX0Ha1M3lZtafDtgFLKXVLKrlLK\nzlLKZQ3dfn2Y1Ldu5lyGUq6jH9ypbotqT6w5zZqj13jtj0CD2p3Yux3Xlk/At60zd+lxHXur8McL\ng5jUV2VD0MLBdJuFmttb859RXbCwEDSzsWL/f0ew/f8a181GVUStmEBvj9pbSNUFD1ddoX9/L9ML\n24p0UC/UTumvUpf89eLgBmnX0kIQtWICa+s42zElRtfpCyE+Bh4ASoBI4BkpZZbaxXIIcFWd9ZSU\n8uWa6jNUp2/Mha9z88eQnldSLzvquiCl5Eh4GoM7tcDWyhIpJY99d4ozURkma/OHJwdoRiWNRWhS\nDpfjc+q8g7ImtHdEww1daXGZgotx2dzp3fBugdPyiknPK2HOloucj8lq0LbH92zDc8M6MuVb1SJo\nQ5kL7rmUREZ+CZkFJXy892rNBYxIU3DvfS01D3sbS4Ysv2Gh1N7FnvisQr1lDO13Q+v09wE9pZS9\nUW3Emqt1LlJK2Vf9qVHg14dy+2Fj4OZo22ACH1TqhpFd3TW2yEII5txfPzPQp+/yZkAVTq9aqm3v\n+3RoXq/6jcEdbZx5pL8HPzw5gKtL72PBROM4J9v75gj2vTUCC6HaFFO+OGZrZdkoAh9u3FP/N7pL\nrcsse7hnzZlqYN1zg1g9oz8DvFsQtWICG14awtbXhta73tpwX882PDHIs8Gv+d//uatB29NHJ3dH\n2ja31xHke9+qv1uFumL0zVlSSu0Vt1PAFGO3URu2B9Vsh/7mGB8+2x+uOY5aMYGiUgVvbwyiv6cr\nxyLSeMivclizxqB8t15dNuAM93Hjqyf6YW9tqROo5ZcTUbz/z2XG+rZm9Yz+5BSW4tqEfJ+Xzzie\nG9aRAV6umkXpb2f0Y9W+MHq2a17trumBHVvw+mgfZvx4GoDWzna0drbj2vLGH+1VZPQdrbmyeByW\nFoJu86t31f1IPw+6tXbCwdaK8Z/rD8jR3sWewZ1a4uZkQ18PF344eo03x3RlgLdrpX0lAzs2/EtP\nu80e7Zw1Pn0eG9CBv87VbcPgrteHs3x3CEfD9a9n+Xk2DQ+f2lgIeHtstyoXere9OtRolndVYeod\nuc8Cf2kddxRCnAdygPlSyirvXGNEzqo4pa+KNs52fD6tL4v+uczRd1W7N+2sLfn6iX6qzg+r24Kq\nKWntbMfZeWNo6WBDiUJZyZf/P68N5cGvVDfKcB83joan6fXZM32QJ6UKJTMGe2FpIZqUwK9Inw4u\nHJ19N0WlCnxaO3FfT5W/luqE/oaXVLukn77LG+cGcO5VX8oF8fbXhvHSb+dI0No4WM6Bt0diZ22p\n8Xujj6rUAeNN6OOmvux8fbhGFXv3He51Evq7Xh+ObzvnSr/x4wM9WT65V4PtbTCEqgYgfzw/iLtM\naCRSjkFCXwixH6hqNWaelHKbOs88oAz4XX0uEfCUUqYLIfoDW4UQPaSUORUrkVJ+D3wPKp2+IX3c\n/towlu8O4b/3dqOTm4PGOqSwREFIUg4bz8XySH8PrC0tNAt6TZ3yzSN2Fpa0a26HEIL4rEJNhKZ3\nxnVjTPfWNaqirCwteH54p4boslHoUMVuyb1vjmDcZ0d0Zmt73xxB19Y3LGMWPdijwfpoDHp5NOef\n/xvGgKX7+X5mf9q52Gt2F1d0avfGPT78cPQalxaNIyAmk2Y2ljSw70Sj09r5hkuSyx+MY/qa06Tl\nFROXqdJ5hy65T2ew49tOFZ5w6aSeGseFs+/rxitqC52bDW2Bv+GlIaTmFpukHZNszhJCPI3KV/49\nUsoqd9wIIfyBWVLKaldpm3q4RDNmTElyThElZcoqX3w3M+/9Hcwfp2N0dsnv/+8IxqxS+Zkqn7H8\nfPw6i7Zf4dH+Hnz8aB/NLl3tPAADl+0nJbeYsKXjNarMgOhMgmKzmtSMvSr6LdlH3w4uRrXwaVCH\na2rXybOBkdoCXwjhDmRIKRVCiE6AD1C/eHRmzNziaI9+byU+fLgXHz6sCuC9760R/Hwiik5ujgzt\n0pKntFybtFS7I/BSx6fQt5/jTBV+f/p7uTaZiF3VEbjAuBsga8IUOv2vAFtgn3qjS7lp5ghgsRCi\nFFACL0spTWeDaMaMmZsCn9ZOLFO/AH5/Xtd+fmLvtkjg/p4Nb9t/q2IK650qbdCklJtRBUo3Y8aM\nmVohhOBBI5pfm2l4f/pmzJgxUy8C5o/B3sZ4/pFuNxo8XGJdEUKkAoa5JwQ3oH4OaW59zNeoZszX\nqHrM16dmGvoaeUkp3as60eSFfn0QQpzTt4JtRoX5GtWM+RpVj/n61ExTukYN7nDNjBkzZsw0Hmah\nb8aMGTO3Ebe60P++sTtwE2C+RjVjvkbVY74+FRBC5Gl9lEBvrePpQohFQojSCvmytMpLIUSKEMJK\nK81anSa10vyFEEXq8mlCiC1CiGr9bjR5nb6bm5v09vZu7G7UGokqtqdlIwQFMWPGTN3JKy7D0kLU\nKpi5IQQHB+Pl5YWzs7MmLSEhgeLiYjp2rHq3cEBAALa2tnh4eODioopzkJWVRVxcHMXFxfTvr4r6\ndfXqVVq2bImbmxtlZWVcu3YNKysrMjMz0/Qt5CKlbNKf/v37y5sJr3d3SK93d8isgpLG7ooZM2Zq\nQfkza7L6vbzkvn37dNLef/99OX36dL1lALlkyRI5ZcoUTdojjzwily5dKlViW8XIkSPlDz/8oDn+\n6quvZI8ePSRwTuqRqbe6eqdBkVqzpoiUPABKFUoKSsoaq0tGIbuglNyi6j2W3qp0fm+XJnj7rUxe\ncRlKper+LVMoyS9W3bMRKbmUKpQAJGUX4T1nJ6/+HkhQbBaHTRB/ujFRKCV56u/95vrzPLn2DMt3\nh9DlvV2s2hdGZGoe6XmmcYJWFQ899BBHjhwhKyuLzMxMjh49yqRJk/TmT0tLY/Pmzfj5+VVbr3lz\nlgHsuZTEy+sCWPfcIIb5uKFUSt77O5heHjcCkbyyLoCH/Nrz/RGVe6GoFROIzShASvBsecN51qX4\nbDq4NqN5s6bpAvhIWCpPrj2jk/bZY30JScxh7fHrfDK1r2bH5NwtwewISuDs/DFGDS7eGPx5Joa5\nW4IBWLLjCgHRGewKTuLMvHtIyi7it5PRPNyvPXd1duNkZDoDO7ZgU0AsB0NT+G5mk7DMqxYpJScj\n0xnSuSXFZUp6vr8X37bO7HpjOG/8dYGdFxOZ7Nde48I6asUEBi8/AMDO4ER2Bifq1Ldwom+Td2xW\nEaVS0mfxjfAfj313knPRmUQsG8/WC6p4HEfUL7YvDoTzxQGVN9fHB3oy1rc1Azu24GJcNkt3XmHl\nlN60bW5Pizq4Kd+wYQM7duzQHPv5+XHo0CHNsZ2dHQ888AB//fUXUkoefPBB7Owq+2J6/fXXmTVr\nFg4ODowaNYpVq1axbt06ve02eZ1+U/Gyudo/ktTcYhY+4Kvjp/vMe/cw8MMDdarLt60zCqXk9xcG\nMWDpfqBphHMrR0pJZkEp+cVlDF95qMb8Hq72Gve32qyY3IvWznaM6uZu9IDjxuKD7Zf56XgUzw7t\nyNtju9Lj/b31rvPYu3c3SCzl+lDuvXLhRF8e9muP35J9APz5wmAe/+GUQXVeXXofKTnFHI9IY9pA\nw+JgNBTFZQoCojN54ofTlc599Egv3t0cXOc627vYc3zO6Erp3t7erFmzhjFjbjiFW7RoEREREXqF\nsxCC8PBwEhISmDt3LlJKPvroI9q2bYuPj49GqzBq1ChmzJjB888/X7F8w3nZVDfYAfgVaI1qbfN7\nKeXnQohFwAtA+bzwPSnlLlP0wdh8tCcUgIUP6Ibwq6vAB7iSqAohUC7wG5vw5FzsrC017nvnb73E\n76djal2+KoEPMGfLjQenKb3U1h67jlJKnh/eiZ+OR6nSjl9n7XHjqHGGfXSIqBUT+PZwJKPvaEXX\n1g0XarMmyhRK/j4fz6LtVwBYvOMKi3dc0Zw3VOADOpG/HvJr3yRmewdDkzkQkqJx6FZOdVHKDBH4\nAPFZhRwKTaGflyvNjRS8Z/jw4SQmJiKEYNiwYURGRta7TlOpd8qAt6WUgUIIJyBACLFPfe5TKeX/\nTNSuyTFVNJ7swlJsLC0a3KdIQUkZ936q8mH+z2tD+WhPKMcj0o3eTlBsliYEXOSH9zeaddPFuCyN\nkDNlIJny+2TF7lDGdG/FJ1P7Gk0QGEJKThHuTra8uzmYzYFxJm/vjgV72PjykEaLQQyqWATP/qzS\nEsyb0J1xnx0hNkN/EHJj8MzPZwE4+PZIOlUIfGMIQgi2b9+u+d8YmGQhV0qZKKUMVP+fC4QAN0d4\nqio4GWl8IViRPh/8S/eFe2hodZvvwhvqjAe/Om4SgQ/oxPzs/N4uvjtc/xFLXZi/NZjNAXGakJIA\n074/2SBt7w9Joc8H//L8LyqBsHxXCL+ejGqQtgGi0/MZ+OEBVuwJbRCBX86j357kt1PRjP6fPwpl\nw93XUkoWbrvEIK1ZuO/CvSYX+NqM/uSwRjuglPDbyehKz/Zff/2Fo6OjziclJaVSXT169KBHD+NF\ngTO5Tl8I4Q0cAXoC/wWeRhUj9xyq2UBmFWW0Y+T2j4421N+aYUSk5LIrOIlnhnpjaSF0BKOpGdyp\nBetfVMV4zcwvwdrKosrgycZg9Cf+XEvNN0ndtaWh1D5NJV7q9zP78+JvAQA6UZ6MgUIpSc4pop2L\nvU7614ci+HjvVaO1Yyghi+8z+Ux23alo5m+9ZNI2DGXF5F4NttZRnU7fpEJfCOEIHAaWSSm3CCFa\no/I0J4ElQFsp5bPV1dEYC7nd5u+muExplLrG9WjN3svJdSqzckpvcovKWKJWQxhbMB4OS6Wvh4uO\n5UJjMaqbOz8/M9Dk7TQVoa/Nsod7Mn2Ql9Hqe2djEBsD4jgxZzRujrY88cMpxvZozTf+kWQVNA2T\n2+eHdcTSQuBoa8X/3eNj9Pqb4u+szcopvZm96SI7/m8YPds3r7mAgTT4Qq66UWtUQVN+l1JuAZBS\nJmud/wHYoad4o1FcpqiXwH+knwd3tHEir7iMt+7tCtww8awtszddNLj9mohJL+CpCiaYhtDZ3YFI\n9Szh4ym9eaBPO1JyVDbMIz6u2eKnHP+rqQxYuo83xnSltEzJ6Dta0dbFDlsr440IC0sURqnH0daK\n72b25+tDEXz0SO9aWTZVx6HQVKMJ/Zj0AjYGqFQ3605Fc69va85FZ3IuutJEuk708WjOdzMH8Oh3\nJwDY/PJdHAxN4Y8zMbw+2ofnf63bgGyN1p6HUoWSxwd5UlSqpKObQ736CZhENXph4b0s2HaZN8f4\nsCMokYEdW9Rrsbv82X7m57OcrSLEY0NgqsDoAvgFVUzcN7XS20opE9X/vwUMklJOq64uU430YzMK\niM8qZHCnlpq0o+GpzPzRcIH42IAOfDSld6X01Nxi7ly2nxmDPRnXow0zfzzDb88NrHVbxhzp12ck\n1KOdM9MGenL6WjpfPdGPw2GpDPByxaEK9dPJyHSDH44Jvdry9fR+BveznOIyBXsuJXE8Io0N5wzX\nZc+7vzsvjNC/6HsoNEWzgGcIQe+Prfcir7FHuHW95+rb/sTebdlxMbFe97r/1RSe/snw36GcGYM9\nadHMhof82utdjP3lRBQ/Hb9OVHpBledrw3AfNzILStj6n6FYWRp3ebXB1TtCiGHAUSAYVTxcgPeA\nx4G+qNQ7UcBL5S8BfZhK6JffpNp61breuF887seb689zYs49tGlefQDr5Jwi3BxtdaxWyhRKvjoU\nwWf7w6st+/jADvx5JpZVU/vw7+Vk9lxOImrFBGZvCmLmYG+dTWHaRKbm4dmiGf9cSKC4TElQbBZ/\nnYut03cECFo4FmsrQTObuk8Mt12I5431F+pcbuZgL+ZN6F4vs78lO67Uezdtc3trgt4fW2O++gq9\n+lp7GEPoT+zdlj2XkhjXo02dX7rec3bStrkdidlF9erDtzP6c586Hm5OUSm5RWW0r7BGUZG0vGLe\n3hBUrx3C38/sz7GINOZP8K3TOktRqYKCEgXLd4VoZlp15YcnB3Cvb2vN8fW0fBRKSXsXe4PXQBpN\np28MjCX0lUqJEDfMnsofkin9PejbwcWgxR9jjcANeWB3vzGc8Z8f1duP305Fs2DrJZ4Z6q2xRTeU\n+n7Px78/xclrdbcKWjDRl43nYnllVGcm9a2d8ZdCKYnLLGDqdydJzjF8y/zSh3rS2d2R3h7Nq5zJ\n6MPQlxzAvrdG4GOATf+8v4PrtK+iKtY9N4gB3q71tq0/Fp5GdmEpr/4RaHAd5fdb+XNx7cP7sbAQ\nfOMfwYmIdPJLynhnbDfu6uKmk88QJvVtR0GJgtXT+9V7tF0+ozeEA2+PpLO7I9kFpZq1tiGdWvLn\ni4NrKFk1ZqHPjRvj+vL7EULU60aZPsiT9+7vXidhUB0Z+SX0W7Kv5ox6OPPePbRyvjHTiMssYNhH\n9dM3l2NnbUHokvH1rseQ6+3n6cL5GJW32ZpePGl5xVxNymX6mso7LGti+iBPCksUbDkfT/CisVhb\nWtRb+P12MoqCEgWWFoKlO0NqXe7zaX1r/YLLLSolLDmPR1afMLCX8NUTfni4NqNvBxeD66iK+jxf\nJ+eOZktgvMbiaP6E7jzs157+WpsZ72jjxMKJvqw+HMnRcMOjEG5+ZQj9vYy3lyAxuxB7a0viMguZ\n+OWxOpUNXjSWRf9c0TGrNXTA1aSEvhDiPuBzwBJYI6VcUV1+Ywv9+vDT03dibWnB0C4tTeJWoD59\njFoxgeIyBbZWlvXSMT87tCPPDe9Iexd7AmMy8WnliJNd/TcVbbsQzx+nYzh9PaPedS2c6MviHVfY\n/towurVx4sNdIfx8IqpOdVhaCI3tuCk3i0kpeeHXAHIKSzkTVbvvHrZ0PEKAhRA6/UrOKWLQhwd4\nflhHShRKfj1puCnzqbn3kJRTZHRhX07X+bspKVPy67MDefX3QHKLjet00LWZNZkGWCT99MydPPPT\nWf5vdBeGdGqpmS2YgsXbr9R7l/dNL/SFEJZAGHAvEAecBR6XUl7RV8YYQt9Yi1ymtikvKlXw19lY\n3v/nssF1DOrYgjNRGdT1Z/3l2YFcTcrh2aEdjb6opM2x8DRm/Hialg42pOeXmKyd6nCytSL4g3G8\n+nsgOUWl/PbcIJO3KaXk9fUX2B6UUKdyAfPHkJFfwqaAOL5TO++rLxHLxpv0N67IruBE/vO74eoe\nQ2nhYENGfgnDfdz4v9E+lJQpGebjRnZhKU62VliYeFe4Qin54eg1VuwONbgOUwj9hvayORCIkFJe\nAxBCrAcmAXqFfn3JLqy/ffK3M/obxaSsJuysLXnqLm92X0rk1DXDRsSGjKRfGtmJkV3dGdm16pgL\nxmSYjxubXh6Cn6crD319nOD4bJO3WZFpAzsAGMVCqLYIIVgwoTvbgxLqNErtb2T/TI/082hQgQ/Q\nrU3D+x4qd3oXEJ1Bbw8XrLW+c0O5w7C0ELw0ohOpucVM6e9BcZmSh7R2pjcWDT3SnwLcJ6V8Xn08\nE5XZ5msV8tV7R+4/QQl0dndgwhd106uVc69va54Y6Emb5nZ0b+tccwEjUl/T0drSyd2B6YO8eHKI\nl85D0VBEpeUz6n/+rHtuEIfDUvjhqOn81ltaCCI/vJ+CkjLsrCxNPsrTR35xGc1sLHnxtwD2Xanb\npr36cHT23bg52mJjZdEofo/yistwtLVCoZSUKpTcsUC/w7P68ucLgxnSuWXNGRuB8nu+1vlvAfVO\nrYS+NoaqdwxV6ayY3IuIlDzmT/StObMJuRSfjWfLZvReZPxds/MndOdwWCo/PzOwSYV1nL0pqF72\n9NVRbgHS1Dh1LZ1p3xu+2acqLAQ8dqcniyf1aJSXeW0oKlXgfzW1TpsWa0tT8uhaFZGpedzzyeFa\n5b0VhP4QYJGUcpz6eC6AlHK5vjINLfSb4g1TqlDiM2+30eprit+xnK3n43nzL8NMHvURMH8MLR1t\njVqnMcktKqWXkV7uR2ffjYerfZONX1CRo+GpRKTk8cF242h49745olHUSXWlTKFEqBfqq5NVt4JO\n/yzgI4ToCMQD04AnGrgPOkStmKC56Edn392YXdGLtaUFYUvHcy0tD2tLi1qPErRZ99wg2rva4+7U\ndIUfqPyw92zvjIUQxGUWVoraVVs2vTyE9/4OJiw5r0kLfAAnO2v2vDmct/4KIkQda8EQmupspjqG\n+7gz3MedZ4aqom5VFIAWQuWlsjrOzLuHFs1sCEvOuykEPlCrdZWDb480SduNYbJ5P/AZKpPNtVLK\nZdXlN9ZIf0RXd03os3J2vT4c33bOmtigN8sDE5qUw32fHa0x36FZo7hbrT9syqP76tDerFJbghaO\npXkza3Ug6Jvnd5VSsmDbJdadimF8zzbsvpRUq3LzJ3TnySHeRvXY2Vhoe8kMXjSWg6Epms1un0/r\ny12d3fgnKEHjjLB8383NTEx6AeEpuWTkl5BZUGIUC7omo94xBGMJfe0R/T+vDeX0tYxq/ak0deKz\nCikoVgVAOfLO3USm5fGMlt+Rzx7ry0N+7bkQm0WLZjY6cXlvJqSUfHkwgvE92+DT2omJXx7lUnzV\no+GTc0fTtnn1W/ZvJgJjMpn8zY2NV/f3asOu4CQcba00Abz/eH6QSW3NGxspJc/9co6ZQ7y4u1sr\nTfqZ6xkUlSoY0QAWZzcjZqGPSuhHp+eTklvcqNF8TE12QSmnrqczrkebxu6KSdHe8t7JzYEvHvcz\nqavaxqLfkn1k5JewamofJvfz0KSn5BZxJSGHUVqC0IyZcm5Lof/jsess2XGFts3t2PDSEE38VzO3\nDrEZBbRpbtdkLVSMQalCSVJ2kfn+NVMnbkuhb8aMGTO3Kze10BdCpAKGOhlxQxWpy4x+zNeoZszX\nqHrM16dmGvoaeUkpq1zwaPJCvz4IIc7pe9uZUWG+RjVjvkbVY74+NdOUrtGtqww1Y8aMGTOVMAt9\nM2bMmLmNuNWF/veN3YGbAPM1qhnzNaoe8/WpBiFEFNBbCJEnhEgWQvwshHAUQvgLIYrU6WlCiC1C\niLZa5RYJIUrV58s/WVrnJwkhLgghctTlD6q9HVTfn6au03dzc5Pe3t6N3Y2bilKFEksLgcVNvlPR\njMo7ZV5xGW2cq4/BbKbpEhwcjJeXF87OzpSUlBAeHk7z5s3Jz8+nZcuWuLm5UVZWxrVr17C2tqZj\nR5XcTkhIoLi4WHOsImwcawAAIABJREFUTVFRESEhIXTu3BknJyeUSiU5OTk4ODhgY2NDQEBAmr6F\nXPVW9ab76d+/v7zdUCqV8mBoslQolAaV93p3hxy64oCRe2WmMfBdsFt6vbujsbthph54eXnJffv2\naY5nzZolJ0yYIEeOHCl/+OEHTfrXX38tfX19Ncfvv/++nD59epV1bty4Ufbp00dvm8A5qUem3urq\nnVrx+Pen+MFIUYmMwY6LiTzz01l+O1W9pWpRqYLfT0ejUErWnYrGe85Olqp9ksRlFjZEV03KlwfC\n8Z6zk19PRpGvFW7vSkIOh8NSkVJqroFs4jNWQ5BSkl+i0PyvTU5RKevPxNyS3/tWJjY2ll27duHn\n56eTnp6ezpYtW+jSpUut6unXrx+hoaG89dZbHDp0iLy8vFr3oaG9bDY4gTGZ/B0YT7c2TswY7AWo\nAhm0cLTBWR379eS1dE5eS28yvniSc4oAiE4v0JsnM7+Ez/aH8cvJaHZeTOREZDoAa46ZLhBJQ6Lt\nRmPhtsscj0hj7+XKQUcm92vPlsB4WjvZMca3dUN20aSEJecy9tMjmuPYjEJGfHyIlVN6M3VAB02c\nhdbN7XR80tyK7L+SjEQV2KicX09GYWdtyaP9PW4Kh2sPPfQQVlZWNG/enAkTJvDee+8xfvx4Xn/9\ndd5++21ycnLo06cPa9eu1Sm3YcMGduzYoTn28/Pj0KFDdOrUCX9/f1atWsXUqVPJzc1l2rRpfPXV\nVzg6Olbbl1t6pB+alMPkb07wm5bnvvVnYhj1P3/NQ5NV0DhxWquj/CZWVjOK81uyj1/UgbHLBX5V\nBMVmaQKA3ywExWZVSqtK4ANsCYwHoEShrHTur7Mx/HYyyphdMylFpQouJ6jCR87edFHn3Klrqt/4\np+NROukp6gFCQHQGr/95/qYf+ReUlOE9Zyfec3ay5ug1issUPP/rOV74VXdX/sJtl5m96SId5+66\nKb7z1q1bycrKIjo6mm+++QZ7e5VjwC+++ILs7Gw2/Xuc6/EpxMbG6pSbOnUqWVlZms+hQ4c05wYP\nHsyGDRtITU3l6NGjHDlyhGXLqnVaDNzCQn/9mZhK7oevp+UzZ0uwTtpzvzQ9Fw+W6oGLUkqupeZx\nzyf+nFY/9HsvJ9Hng9q5Gv4nKIFJXx/n2Z/PEhyXTURKrqm6bDQCojOYZEAc0ZOR6QxfeRDvOTvZ\nEhiHQil5d3MwC7YZHmS+oZmz+SITvjjG4u1XuFDhxbcpQBVRLCQxh2yt+LrfHlapJR9ZfZJ/ghIo\na6Iv+D2XkigouaGii07PJzAms1I+7TChS3eG6Bx7z9nJJ/9erSTkz0VXrqcpEpKYQ0hiDoExmVxP\ny9ek772cxNsHMrHq/wgPTX+Op9eeZuCy/VUOfvRx5513MnnyZC5dulRj3ltWvVNRuANsCoitlBbQ\nBG+Y8hCGCqXkXFQmkan5bAqIY1Cnlry7+WKtg72XjwIPh6VyWB1LIGzpeLIKSvj5RBRT+nuw93Iy\nL4/s1GBT5PLwgP6zRtGhRTM6v7eLJ4d4sXhSTwC+PhRpUL3a6x//3RDETTD402HflWS2XkgAYO3x\nyiq6M1E3At5/tDdU8//1tHwdVdjVpFzOx2YxU63KbApcScjh5XUBTO7XnlVT+wIw8mN/oHKch4rP\n45nrGTrHXx6MqPRie/Tbk3Rp5UhESl6TCySz93IScZmFZBaU8PznuoPQpGvpdI/OZGmEKmSkY697\niD/+B7t27qCZz2DCrqYw0bvqcfmxY8cICQlh0qRJtGrVitDQUP755x+eeuqpGvtU40hfCLFWCJEi\nhLikldZCCLFPCBGu/uuqTp8uhLgohAgWQpwQQvTRKhOlTr8ghDD58NrRtvL7rKJAKVMocdLKF52e\nz+Pfn2LWxiAA1hy91qDBqwG+OxypGZ3+fjpGc4Mrpap/BcWKWte1dGdIpbSu83cz8MMDfOMfyehP\nDvPRnlDCU2q/CFRf1qrXHLYExhGToVqz+PWkahH60NUUDoamGKWdt9W/IaiCVDQ1zlzPILuwFCkl\nn/x7tZL6ojr+OB2j99zEL4+xYOslErKazkJ+YalqhH8sPE2juiknM7+EUoWyTr/Rav/KA4MI9T38\n75VkDoelUlxW++fEFBSXKYjNKOB99bO851Jilfk2nLsxEBWW1jj1f5DsE+sBcLC14q+//sLR0VHn\nk5KSgouLC//88w+9evXC0dGR++67j4cffpjZs2fX2Lca7fSFECOAPOBXKWVPddpKIENKuUIIMQdw\nlVK+K4S4CwiRUmYKIcajioc7SF0mChggpayT0yFTxsj98akBetU7/b1cNaOOhow6VbHfHq72JrfE\n+emZO02+GPjVwXAA/vdvmCbttbu78NWhCJO2W86zQzvS0tGGV++unXWEKTkblcGj354E4PyCe/Fb\nss/obXzwYA+eustbJ+1EZBoDvFo0eISt8zGZPKwVDEYfPz19J8/8fLbGfLWlMaPFvbH+PNsuJGjC\nPf733q6s2hdWc8EKmCJGbo2/vpTyCJBRIXkS/9/eecdHVWUP/HtTKaEHQk8AkaagEhCVpiIC6rrF\ntay6lvWn66qLukWwousqYu+KUmRREVQsdJAOQgg1tEASQnohvU0y5f7+eO9NZpIpmcnMpPC+n08+\nmbmv3XfnvXPPPefcc+EL9fMXwG/VfXdLKbXx2R6gL80YV/Z822GmJUB2UkfrowYi9NJ2xS1fU1hR\nQ8ys1byx4ZSdwAcCJvBBMZm8vj4xYNerS8ys1dyqCvqlNqYofwh8gBd+OsZbG2rvd/m+dP702V4e\n/eqAX65XlwNpRVa/RJWxYVq3LwU+wAs/HiVm1uqAvb+2bD6hjFi1S3sj8P2Ftzb9KCmlNl7JARzF\nyv0FWGvzXQIbhBAS+FRK6XTqthDiQeBBgP79+3tZRd8x8Ok1nHhpGm3Dgv16nenvul/31h2x0V24\nclA33tvsmUDNLqnyy1KDWtSJr/jj6L6sUJ2a3rAlMY9Jg7v73e6bVlDJ2qPZPDRpEEY1sigutbBB\nI1Bf8d7mJN7bnESfzm3JVM09GwJkrvx9AzR7f6NFtw18eg1PTRtKbEwXIiPCGRDZ3m/XLKkyWpey\nbK402pErpZSqILcihLgaReiPtykeL6XMFEL0ADYKIU6qowhH55yPms8jNja2WbjkPtySxD+vH8Le\nlAI6twtjSM8OPju3tsShL7h9bH9uvqS3x0L/tk/3sP7xiX7v2Lzlx0euomenNkR1bENJldFr4aWN\nam4Z3Zd5fxjpUPgn55fTo0M4HdR5HN4w8XUltC4xp4zvD2Z6fR5fkBlg+/6SX1MbfY4Vf73CagID\nEAI75/zYAV2JO1PIhVERnMp175N6bV2t8/vF34zgj7F9aRfm2zgWi0U2OLLOEVv+OZmr39jquwo5\nwVvjXq6WGEj9b/W+CSFGAp8DN0sprZJMSpmp/s8DVgJjva20J+x7ZopPzhN/VrFw3TZ/D9e/o/RV\naQWVxMxabXUiucKV72TUSxt46H/7G1W/MHXJwKE9O3i1fGBaYSV/+LhWO/NV7HNIIzTqJfePZdG9\nY3jhpuGM6teZKDX/zId3XsakRi6I/e3+DFYlOHauXfvmNu710OR1LKuE5PxyDEYzT3xzyFrurcB/\n9fcX88s/JgHQPiyY31/Wh7dvG+XmqIbx1oZEJr++hQU7zxAzazWGBppfXFFebeLVNSdYsPMMz/sg\nTNZWqZp4YXduH9PPbntsdBdS597AhicmeXzuF3465hdTX3J+4wIiBkS254dHrvJRbZzjbVf3E3AP\nMFf9/yOAEKI/8D1wt5TSasQSQrQHgqSUZernqcBLjam4O9bOnED/ru1oHx5CcJCwm6DUvUM4+WXV\nHp1vT0ohX+61T4ugaXNT3trm0uFSY7Iw+fUt/GvaEK4f0ZPFu1NpFxrMnJ+PM2FwpEf1cMb+56Zg\nkdCpbX3t9NO7RzeoUzmeXUrMrNWsemw8f/h4N9UmC3tmX0vPTp4n+/rth7sICwniykHd3O5775Ux\nDOze3k5YdG4XyoTBkQ5DSUODg/ji/rG8u+k0b2+qtZV66hT9+9cHmTo8ivCQIKpNFsKCgyhXY8k9\nDeW94b2dHu3vjjvG9reaCUwWyVu3XkJRhf1EwgmDI9lx2vPFmLRR4H/UlB3phZUMjmrcyPWiF9Y3\n6vi6tA2tHXE+M2MYF/SI4Ou42kiXn49k8e9pQwE4/tL1nM4t92h+x6JdqUy8sLtPTH2//2gX9101\noEHKnzPWzpwAKEqbv3Er9IUQXwOTgUghRAbwAoqwXy6E+AvKUoa3qrs/D3QDPlJfVpPqQY4CVqpl\nIcBXUsp1vr0Ve4b16mj9fOKlaVz4rOJeWHL/WKI6trFq657wzMraiQ8T521xsac9pQYjWSUGnvjm\nMHDYbpuzl/a7h6/gDx//6nCbI+qaImyjBSIjwhp8HoBnViZQbVLs0ONe/YUzr86wCl8pJWmFlUR3\nc24XTS+stDrx6sZZO2L2jKH1Zpoeen6q2+NmThnMzCmDmf19AtVGM13ae3afAEOfq30MB0S2t5s0\no1FQXs37m5N4esawgES+JMxR7r1daDD9urblH9cNAaBDmxD6dG7L7BlDCRKCCYMjrak6bnzf+05n\n26l8+nVtx77UQr6OS+OFm0ZYR1UNoaGO0jOvzmDA7DVOt3doE0KZwcR/f3eR3Wi1f9d21rkrGjE2\nz1+7sBBG9evc4Ppq3LdoH7+7tA9v33YJZwsq6N+1nUfzVZLyypm79gQH0oo5kHbQo2u3CwsmPCSI\nInWinSav2oQGs++ZKYz57yaPzucJboW+lPIOJ5uudbDvA8ADDspTAN+MTb0gNLj2h5zYSLOAhhZj\nrpFeWInZIolRnUQfbknirnHRLNp1hkv7d/Ho3OEhQYyO7sqHf7qMR7yMtrhrXLRV6AcHKS9QVMdw\nckvdj3AOZ5TYfd94PJepI3oCyn29seEUqx4bz0V9Ojk8foIHHeKqx8YTHhLMZTZt9PvL+jT4eFBM\nIb7AkcAHeGBJPAfTiunYJoQnpw7hVG4ZnduG0qNjG65+YyvjL2j8aE170cNCgqwdeFCQYMe/r7Hu\nExIcxK5Z19gd5+w38ISXV5+wm9ORUVTFT4+Od3FELTd/sJNLGihwhRCs/vt4+nVtx4msUgoqali6\n5yy7kwu4uE8nfn7M8TU1P1N0t3bWTs6RCXPXrGu4au7mBtVFY+XBTCYP6c7MZYe4MCqCGy7uzcwp\ngx3um15YiUVKq8Iz5a1tHl3Lli8fuJxL+3fhXHk1xZX2ky0dzTHyJa12Rq4tQgievWEY4wbWNzVc\nOaiby9w1DUUTdIdfmMrWxDxeX5/otd3wyesuBOCGkb145Cvv6tNV1XqnDItiZJ9O/G3yIO6+IppN\nx3Otk7/+95exdtPcnfHg//az5u8TKDUYrWGXOSUG2oYFM3ftSebfPRohBBaL5NsDriNr/ji6Lw9N\nGmR9YTShNXZAV3Y+dTWhwUF0aee5xq7x8Z2XERQkeO+X05zOK6dGHbH06BBOnocmPY2Dacqo5ZPt\nKTw5dYg1EVqH8BDKqk1OO4u6LL5vDFcOirSOOjVe+8PFhIc2n4woRzJKiJm1mr9fcwFPTh2i/K77\nM/jtpX3sRjoms4XDGSX1lARXjOit/N6Xq+/i9It68tbGU9xWx2bviO8evpLYlzdZj6tLby/MkAAz\nlyk+mFO55ZzKPWU1GWom21KDkTEvb7KOfj3h4cmD+L8JA/njJ7tJzleeky3/nGyNIIqMCCcyItzu\nmDbqs3DXOP9ELp4XQh/ggQn2GTQfGD+Az3eeoX14iHUKty+4f/E+r1I7/Oe3F/HcD0eZMiyKB22y\nfQ7v1ZHjavz+wMj2pNgImGdvGOZw1q1G0n+nI4QgKEhY7Z+a1n/XuP5MGNyd9mHB1vS9rpjxnn04\naZvQYK59UxHcA2avYd4tI+slCavL3yYP4t/Thjp1HPbt0s5tPdwx/WJl4aHrhkUhgb0pBbQNC+aS\nfp0Z+tw6j17cvFIDS/emMTamK3GphTwy+QIKyms7jrIGhualvDKDGrOFNqqdelTfTgzsHsHKg5n0\n69qW28b0t84gHdG7o6tTBZT3Nifx5NQhPP7NIX46nEVaYSX/vF4Z6QQJUe+Z+N2lfVipOq4v6deZ\nQ+nF1me0YxvHokYIwT+mDnG4bdG9Y+jRsVYgRkaEk/zKDEwWizVwoe65nGH7HjUULbz23itjvBL4\nd43rz7+vH4IQgo/vGs3Ut7fzxf1j3YaMCiFI+u/0eiYtX9GglbOEEAuBG4E8m1m5XYFvgBggFbhV\nnYkrgHeBGUAlcK+U8oB6zD3As+ppX5ZSfoEbvJ2R645Nx3N5YEk8EwZHkllcRUp+rTD95K7L+OtS\nxaySOvcG7lsUx5bEfJ/XQWPN3ycwvHdHtpzM44pB3azCwZZtp/IZE9OFdmEh1ocxde4N1iRqF/Ro\nmAMor8zAQ//bzxt/HMWg7hGcOVcRkDCx+XeP5tphUdYH+e2Np6ioNvHsjcP9fm2N3FIDl7/yS6PO\nERkRzrly1yOGtqHBvHXrKGavTODp6cO41YkWG59aSHS39nTvoAi2X5MLGNarA529GOkczyqtJ4RB\n6Wg/cpC2YNqInqw7luP2vC/+ZgQv/FTrYLeN+a/LwntjuX+x8q5qWnJVjZlhz6/jhot78eGdlzXo\nXhrDofRiIsJDrCPJ2dOV+PwRvTtxz8I49jbAx9QY1j0+gcyiKkwWyeQh3QkPaZoQaFczchsq9D1J\nxTADeAxF6F8OvCulvFztJOKBWJSJWvuB0TYzeB3iL6G/1Cbd8mPXXMD7NnHtqXNvsBOs2oPrK9bO\nnMBLPx9n6QOXczqvjKE9PdPubOvmC/wxYejW2L4sj6819TTllHhbDEaznfPWHyTMmdqoGH9vqfs7\nbv3nZDKLq7jz873Wsk/uuozYmK58uSfNLvLJFyy5fyz/WHGY/LJqu987Ob+cPp3bOlRm/EVRRQ1V\nRjO9O9tPOPT35LjmkvDNldBvkHlHSrldCBFTp/hmlKgeUFIxbAWeUsuXqEt27RFCdFZj+ScDG6WU\nhWqlNgLTgK89uBefMSamK6A4dp+YciF/uKwvk2003s//HEs71YHk6wlLw3p15OsHxwF4LPBBCdXz\nRhsMJH++IoZxA7vRpX0YY9W2bg60CQ1mSFQHEnN9l2Z66z8nExPZ3ipQmkLgOyImsj39u7bjT5f3\ntyZpm3aRYv4yWRybK/Y+fa3XoyGzlOx66hrrDGSNQd1dL+rhD7q0D8NR+MTJ/0zjure3kV7omwlr\nc24aztVDe7AvtYhhvTo0C4HvjsbY9J2lYugD2OYwzlDLnJXXIxBpGIb07GCnjfTpomgEWuhX3VWY\n6s4I1Njx76sdRqus/vt4RvTu5BfNYvF9Y/Hls/Xdw1dSZjBaJyR5KhQ1h6YtF/Xp5JPIEn+w/omJ\nlBqM1oV0zrw6g9IqE1e9tpmZ1w5m3MBu3PRBw0Ig+3Zpa43Yamr+Mn4AC3ae4aZRvXlQ9WEFBQle\n+d3FVNXx26TXiT4bO6Ar1UazR6GaGoO6tyc5vwKLRRIWEhTwhG6e0CY0mB3/vsbte/n2baPUEGvn\nrHt8glVpcxXC3NzwiSPXUSqGRp4v4GkYQoODXJogHrv6AoepDfp1VZyPw3t15IdHrrJGZgypM9nl\nioHdGNC9Pdf4IJulrx08o6PtdaL1T0y0vhTOOjVbjsyZahd/rUUfNWc62mjjQgg6tQvl6IvXO9zX\nVb4f29DGl397EQXlTbcSmxY6eWtsXy7ua9/hvn3bJXbfe9QR7ssfusLr6z5343AW7kpl7IDmM6Lz\nhhV/vYJRfTtbO62RfTvz5wVxZBZXseWfkwkPCaJj21CS88q9mhfQXGiM0M8VQvSSUmbXScWQCdh6\nrvqqZZnUmoO08q2NuH5AefSawby3OYlVj40nJrI9F72wnt+M6g3Y26v3zL6W8mojIWp0geb4eu7G\n4QxvRpEZjpg1fShXDVJizmdPH0qXdmHWTg3goUkDuevyaIxmC8vjM/hkWzL/uO5ChBDNxmbvCTeN\n6s3Ph7McbtPW3gV4/Y+j2J1cYOfAfGD8AO64vL81NBawrsHcVNw0qjej+namfzf3UVCafb19WDBP\n3zCsQefv17UtFkttLp9NT07kh4NZTLqwO5Nb2Dq9x168nqOZJdw2f4+1bEwdM+Sg7hFsfHIi5dUm\nenSo7SRbssAHlBmWDflDidI5avP9dWCW+nkWME/9fANKdk0BjAPi1PKuwBmgi/p3Bujq7rqjR4+W\nzZEak1mazRa3+xWUV8slv6ZKi8X9vs2V7/any+inVslyg9FaVlVjkp9tT5ZGk7kJa9Y4zGaLrHFS\n/8Lyajlx3mZ5MrtUSimlxWKReaUGGf3UKhn91KpAVtMv/O/XVBn91Cq5aGdKvW3xqYVy47Ec+fq6\nk9b7rTYqz7vFYpHVxpb7m9dlbUKWjH5qlUwvrGjqqvgUIF46kakNjd6xpmIAclFSMfwALAf6o6Zi\nkFIWqiGbH6A4aSuB+6SU8ep57geeVk/7XynlInfX9lf0jo6ON+w4nc/IPp3p1K55OGu9RUrJ+mM5\nXDe8p0tz4b2L4rhueBR3Xt58ll/UcU+jQzabEiFEPkqn4g2RgOcZqc4v9DZyj95GrtHbxz2BbqNo\nKaXDnDPNXug3BiFEvLPeTkdBbyP36G3kGr193NOc2qj5xlbp6Ojo6PgcXejr6OjonEe0dqHvdB1e\nHSt6G7lHbyPX6O3jnmbTRq3apq+jo6PTkhBCpAIPSCn9topKo4W+WskywIy6UpY3GTidERkZKWNi\nYhpVx9bO6dxyuncIa/b5eHR0dFyTkJBAdHQ0HTs2biLn/v37zzmL3mnw5CxnfyhCPbJO2TzsJ269\npn6egf3Erb3uzt9cJ2c1FywWS6uZMKSjc74THR0tN27cKPv37y/j4+OllFIuXbpUAvLo0aNSSik/\n//xzefPNN0sppdy7d68cPXq07NChg+zRo4d84oknpJSuJ2f5y6Z/M0rmTdT/v7UpX6Le3x5Ay8Cp\n4yVGs26e09FpbUyaNImtW7cCsG3bNgYOHMj27dut3ydNmgTAzJkzmTlzJqWlpSQnJ3Prrbc6O6UV\nXwh9CWwQQuxXs2OC5xk4dbxES5Hrr1V2dHT8zbnyal78+Rgms+erU7VWJk2axLZtykIwO3bsYPbs\n2dbvtkI/NDSUpKQkzp07R0REBOPGjXN7bl8I/fFSysuA6cAj6oIrVtShhkfqqBDiQSFEvBAiPj/f\nfytWtQZMFqVpdaHf8qgxWeqlPD4fef7HoyzalerX1elaGpMmTWLHjh1kZ2djNpu59dZb2bVrF6mp\nqZSUlHDJJUrW1AULFnDq1CmGDh3KmDFjWLVqldtzN1roSykz1f95wEpgLGoGToAGZuCse875UspY\nKWVs9+6OfRGBRErJ4fRizV/RrDCp5p0ak4WTOZ6tAarTtMx4b4dPV2RrqWgdn663gEXCmoRsBg0a\nRLt27Xj//feZOHEiHTt2pGfPnsyfP5/x48cTpK51PXjwYL7++mvy8vJ46qmnuOWWW6ioqHB5jUYJ\nfSFEeyFEB+0zMBU4CvwE3KPudg/wo/r5J+DPQmEcUGJjBmq2/HAok5s/3MXqhOZXVdsh8ePLDjVh\nTXQ8JSmvvKmr0CzQRqshNoudr03IJmbWanJLDU1VrSahqKKGL/emcTijhEmTJvHBBx9YTTmTJ0+2\n+w6wdOlS8vPzCQoKonNnJeWz1iE4o7GafhSwUwhxGIgDVksp1wFzgeuEEKeBKep3gDVACpAEfAb8\nrZHXDwiJOcrLebag0s2egUd7YYBmvWKRjnck5pSxphkqG75EW14x1EbV/1Jd3jExx3fLWjZXEjJK\nWKv+xmbVmiClZNKkSZSVlTFxomIxr/sdYN26dYwYMYKIiAhmzpzJsmXLaNu2bf2L2NColbOklCnA\nKAflBcC1Dsol8EhjrtkUaNp0aHDzG3+abKJ3Qs6D8XFmcRUR4SF0atuyUxu7YtupfAZGtqdf13Zc\n/44SsdESF6lpKNozbKvpax1BSDN853yNtjRn6twbmPz8co5nlxIaHMRDDz3EQw89ZN3vxhtvrGdi\nXrp0qcfX01XDBmAdfroZNjUFRpsFrssMJgxG947BOT8dY8pb2/xZLb9x1dzNXOek7lnFVexNKQhw\njXzPPQvjuObNrU1djYBhtJp3agW89s6FBTe/d86faNF4/uzszq8W9ZJA/BDeYqvpn84r53ab5d+c\nsXh3aou2J+eVVTssv/qNrXbL39litkj2ny3yZ7V8ii/nX3y4JYlLXtrQbEMirSPpIEeavlJWajBy\n/+J9rcbGX1Vjtt6jLadylffSnwpmqxf6m0/mkpLfOAFnNCkvYGgTaR1Pr0zgs+0pDreZLPYPzqH0\n4kZfb8vJPD7YfLrR5wkURzKK+cfyw1SbnAu19zef5g8f726Wgj9m1mqOZ/kv8ur19YkUVxo514SL\ntrtCU1xsw46tJh+17IeDmWw+mcf7jXwuc0oMxMxazdbEPPc7+5Fhz6/jjjoKyv6zhdbP/jTVtnqh\nf//ieK550ztTRsys1Ty5/FCt1uGnH8Jikby5IZFz5Y412K/2pvHfNSfsyvLLqimpMtpp+hqHGyn4\n71u8jzc2nEJKSXm1qVHn8oas4ios6vDeYpHMW3eS7JIqp/vfv3gf3x3IqFe+43Q+Kw8q5cdUoZrv\nZJTQ1Kw7qjryLP4LC26OI1WwN1FqaMqMpmhp7RIsGncPB9KUTn9ZXLqbPf1PfB0FpLjSaP0c1Mj7\ndEWrF/qN5fsDmdSoQv+ZlUf9co1fUwp4f3MSz6xMqLetruNmWVwaMbNWM+a/mxj3yi/1NH2Amz/c\nRYUXwjo+tZAlv6Zav3+6PYWLXlhPXgCH1GkFlVw5dzMfbEkC4FBGMR9tTebJbw47PcbkRFDevSCO\nJ9TjmsoZbzAtYKUbAAAgAElEQVSaee+X09S4GIUABKkKhaMhv6/wpyDxhqKKGmJmrSYlX4krlzZz\nODXzlqZnaUI/qJGKl3UyYzPsAG2f413J5/hKjWDyNa1W6Mva5G+NxjoByk8vZLVJcb46suPa9v6g\n2OM1qoxmp7ZfZ4LQlrqhgLd88ivP/3jM+v2Hg8q8uUCaBTKLFY1+Z5KynKgmLF1pwOYG2L+bauby\nx1uTeWvjKb7Z5/oFDmmk0N95+hzL4lxfo7lNLjyVax+OWW4wcVDVxLV20Gpskb7R9DUzmsUiScoL\nfDjo48sOcvEL6x1us9g846uOZPH2plN+qUOrFfpf7E5lwOw1Hh8npeT3H+2yDrdBEa7+xJFNU6Ou\n07KuMHdmw5+79qS1M9FIyChhV1Lt2sx/+1LJal1RbXLo99AigZbHB24oLOu83It3pQJwrqLaqdAy\nN0CYaW1s65cpNRj9br4qMyjnd+VvAAhWHXfeOnDvWrCXWd/XHyna4kfLkcccSCvi9fWJdmUPLd3P\n7z7ajcForme21PrCxnban2xLBmDt0RymvLWd4srA+jl+OJRFmZNnzvY5Npqk3bwFX9Jqhf7XdWx2\nriIXkvLKSM4v54eDmWxNzOdAWjEPf1mb5t/WVDLnp2OkF/p2kpbZGhKq/MgGo5mle85SZjDynuq4\nahOq/FR17f5z1550eM6v49JYHm9v577pg53c+fneevvet2ifQ7+H1tkt3p1KUUVgXg7twQ8KUpzw\n647lAJBXWm2n7acXVvLaupNIKR2OakoN9iMkzQx25+d7yStTzFUj52wg9uWNgNLmsS9vZMtJ3zr4\nzA1MiBccBJU1JrtnrbLGtx2SlJKSSmOziOL5/Ue7ndq0LVJafy9NDmqavvCxicpdZxxIbP1NRovF\nbt6CL2m1Qr+urfuCZ9aSkFHicN8pb23n2je38fg3h7hv8T6g9mEDe4fL4t2pzFx20OP6fLn3LDkl\njm3jdU0P89Yl8uwPR7ly7mZWH1FGHAajhcHPrKln7vEFcamFDstzS2sfQl/bgw+lF/PLidx65d/t\nz7Be7/7F8dbytmHBdsL9r0v38/HWZJLyyh2afu626dyOZ5WyL7X2N/x2f21naDBaqDFZOHOugnPl\nNU47UW/ROrEqo5lFu844Ha0ECcHw59dzrU3nO/x5x2YAV7jyHZilZNRLG/j3d0c8Pq87ygy+ey5z\nS6utIx7NgW+xviM+uwzg2XN9tqCCca/84jKooDG8+PNx62ej2eI3x3urFfqlhvpa0k0f7CT1XAUl\nVUae++EoBqPZpw+rM/JKDTyz8igPLNlXb9u+1EIe+1rpRPLLqomZtZrFu88AtaYBDU+H/ok+TMBm\n8bE9+Lcf7uIvX8TXK//hUBZQX6NrGxpsZ+/WNOLgIOFQ6B+26eBnvLfDbltdu3BaYWXtaMvHL5pW\n5XnrEnnx5+NsP33O4X5ah99Qv1FWcRUbj9fvNC98dq1Tc1y1UTm35qvRsFgki3edadDEPkecyi3j\n4jkb7DpTR5jMFg6lF1tHWs54f/NpSqqU9/LeRco7o/3EQUJgMDqOcQcl/t1ZiHbd0Z/GR1uTKGmA\nMrXk17PklBqsipg/MZml3bwFX9KoNAzNGWeheTtO53PmXCX/23OWC3pE8MOhekk+3VJcacRikQ2O\nJDCoL5sjLf39zUnWz5pjy1e216V70hh/QSRThkU5NS/Yxga7wldC/9U1J5R109xQV/amFVba+S+M\nDmz0DaXGZLEz44QECe5esFc9r2+H+3U1+3sWxrH1n5OJiWxvV779lGdphX/zwU7Oldc4TM+wNiGb\nW2OVZLa274FmyrBIrM9vZnEVm0/kMufn42QWV/HMDcMbXIcyg5EXfjzGpdFdANiSmMcto/vW22/5\nvnTOFlbw4RbFnv7UtKEuz7vDQcdoNfsJwdDnlMyks6YP5a+TBgHwypoTBAnByZxStibmk/zKDLtn\nvqC8mtEv1192dv72ZD7bcYZTOWW8c/ulLuuldYrhocEu93N2XBsHx8Wdcfz+1ZgttPXwOg2l1Qp9\nZ+xKKrDaiYsrjRxM8zymPeVcBR9tTeLRawa73fdYVgmbTygCJqOoij8vjGPhPbFWe121jXbljxjt\nvy49wL+uH8Jd46Idbv/Dx7826DwNcZY2hE+dTDID7BzPjobdi1SnLtRG+TjSGt35XN7caB8VERwk\nKFI7ZG1GpCte+PEopQYTb992Sb1t1SYloioiXHm1HPkbJr+xlXuvjLErc5dLPqu4it6d21JYUUNY\nSJDLiKrM4ipiZq2mT+e21nbSzqHx/cFMbhndl6vmbraWfbbjDNcMjeJPn+8h/pkpdGgTypd7z3L3\nuGiH9uVFu1L5/mAmPx/JqrftRHYpXdqF0bNTm3rmpG2nXPtNHClsmnnHdiQ0d+1JHpo4kKwSA/Pr\nPFdphZUMsOlYc5yEHX+2QxlVOwrWkFKyOiGb6Rf1YnfyOX5NVlJ8tPEwseHQ59bRPiyYYy9Nq7dN\nC02uS0p+BZf06+zRdRrKeSf0NYEPNCok6o0Np1i4K5VNT04iIjzEmuHSYpGkFlTQuV0Y09/dbmcX\nB0WjK6yooUfHNgBU2Djr/BVRt+VkXqNn6mrRFP/+9jBXD+lB3y7tWLjrDM/eMIyiSiMp+eVcGNWB\nI5kljOzTyarJVtWYMctaIeiI7/ZnEBYSZDVzgWOH3WYHTlZHndaEeVs8ure6o6DL/rORv4wfwPDe\nHbl6SA8MRjPFlUZ6dlJ+sy9+PQvAk9ddSL+u7bBYJAOfXsOs6UOZt+6kdaT29m2jnNrYbUNvG8KV\nczfzr+uH8Pr6RKI6hlvLH/v6IHNustfOtY7LVuADVn8VKAEJmUX1bdN3fKbMEl22Lx2zRfLWxlMc\nzSxlYPf2PDxpkHV0azCa+Xirorlroy7bVpz+rmJSczQS2ZPSsNGlxr2L4tiqdoqFdTq7r+PSedrB\n/Jar39jKsRevZ3l8Opf178KPh+p3TLasP5ZrN3rfcTqfvSmFfLAliaemVfHaulpfjyeavhYAUVFj\n5h/L6881cTXo9decEhHo2F0hxDTgXSAY+FxKOdfV/rGxsTI+vr7t1x0xs1Z7V0EvuWlUbwrKq9md\n3LCEXzdc3Is+XdrW01D8QV2NL1DMuWk4c1TnVGx0l3rRGgBxz1zL2P/+Uq/80v6dvRqFBZqpw6PY\nmpjv1Bbfu1Mbspw48Fsqr/zuYoeCdkTvjhzLKuXd2y9hprq2w7Z/TWbS61sDXEPv+fTu0Tz0v/0e\nHXPm1RmsiM/g2/0ZTBrSnbvGRRMSJEjKK+fmD3e5PHbykO7WDq0uw3p1ZO3MCR7VRUMIsV9KGetw\nWyCFvhAiGDgFXIeyPu4+4A4p5XFnx7QUoa+jo6Pja7xNqe1K6Ac6emcskCSlTJFS1gDLgJsDXAcd\nHR2d85ZAC/0+gG08WYZaZkdjF0aXUp4XC4ro6OjoeEqzdORKKecD8wGEEPlCiLNenioScBwYraOh\nt5F79DZyjd4+7vGqjcRrXl/PcbgegRf6mUA/m+991TKnSCm7e3sxIUS8M7uWjoLeRu7R28g1evu4\npzm1UaDNO/uAwUKIAUKIMOB24KcA10FHR0fnvCWgmr6U0iSEeBRYjxKyuVBKeczNYTo6Ojo6PiLg\nNn0p5RrA85zH3jE/QNdpyeht5B69jVyjt497mk0b+SVOXwjRD1gCRKGsgzBfSvmuEGIO8H+AFpLz\ntNoJOCUyMlLGxMT4vI46Ojo6LZWEhASio6Pp2LGjw+379+8/58wf6i9N3wT8Q0p5QAjRAdgvhNio\nbntbSvlGQ08UExODN5Oz7lsUx5bEfK8nN+jo6Og0V2JiYvjoo4+YMmWKw+2uIh794siVUmZLKQ+o\nn8uAEziIx/cn7hJY6ejo6LR0kpKSmDRpEp06dSIyMpLbbrvN7TF+j94RQsQAlwLaqhaPCiGOCCEW\nCiG6ODmmUZOzdHR0dM4HnnvuOaZOnUpRUREZGRk89thjbo/xq9AXQkQA3wGPSylLgY+BQcAlQDbw\npqPjpJTzpZSxUsrY7t29DtPX0dFppfx0OKtBC5+0dkJDQzl79ixZWVm0adOG8ePHuz3Gb0JfCBGK\nIvC/lFJ+DyClzJVSmqWUFuAzlFw8Ojo6Og0mJb+cv399kCeWH2rqqjQ58+bNQ0rJ2LFjGTFiBAsX\nLnR7jF8cuUJJhr4AOCGlfMumvJeUUltr7HfAUX9cX0dHp/WiLXiS1QTpwpsbPXv25LPPPgNg586d\nTJkyhYkTJ7o8xl+a/lXA3cA1QohD6t8MYJ4QIkEIcQS4GnjCT9fXQVm169ZPfvV67VMdneaMo4V2\nzjdWrFhBRoayNnGXLl0QQhDkZm1dv2j6UsqdOF4UJlCTsnSAF386TlxqIYfSixk3sFtTV0engRiM\nZmJf3sQbfxzJtIt6NXV1mh0BXvepWbNv3z4ef/xxSkpKiIqK4t1332XgwIEuj2mWWTZ1fEtr0ocO\npBXx+492s+qx8VzUp1NTV8cvZJcYKK82MXftSV3ou6A1PdeekpqaCsCUKVOYN2+eR8cGOuGaTgCR\ntD6VaOPxXAC2nWq9obyBXsK0paE1j27d8Q5d6OvoNFN0m7VjNGVGbx7v0IW+yt6UAtILK5u6Gn6h\nKRZF19HxN+K8NvB4jy70VW6bv4cJ87Y0dTV8ijYMfnL54aatiI5H6MYd1+jWr8ahC32dFkVDXviq\nGjMv/nyMimqT/yvkB6w266atRrNHN+94hy70gX+taJ2acGtUiBrinF68O5VFu1L5dHtKAGrkR5wI\ntWqTmf9bEs/p3LLA1qeZ0Bqf60ByXgn98moTl7+yib0pBdYyKSUr9mc0Ya38R2uOAtG0vJnLDvLy\nquN224xmC9B673/Is+vYeDyXZ1Y6ntAen1pI7MsbKTW07tw0uqLvHeeV0D+WWUJuaTVvbEgElJdj\nwOzzY76YxSIxW1qfEPzxUBaf7zxjV2aRWnRHSxULSv0tdX4vi0VSbbKZXe3k9t7edIpz5TUcTi/2\nVwWblNbamQeK83Jylub1X3kws4lr4l9sX43b5+8hLrWwxSwqYzCaqaox06V9mMPtriI3NFkZ1EJl\n/oKdqQCkFthHk01/dweJNiYdZ7dX6xNooQ3gButz3WI79ablvNL0ax+WpqxF0xCXWgjQYob8d3y2\nh0v/46WJQtP0EZQajCTnl/u4dv7l2/3p1s9lNvefWMeGH+RE6LX2yUu6ot84ziuhbx32q9/Px2en\npUS0HExTTBO3f7rH42NtNf1bPt7NtW9u82XV/I6tVefiORsAHOaO14S6yWxh/bEcq9nDOnnJv9Vs\nclr7/fmL80roa2gvS1Jey9IAPcWRRtTShvzHs0vtCxrQU9vO2DyV2/J+Y0e+l/9bUn+daO05/mzH\nGR76337WHc0BbH73lvVTe4A+I7cxnF9Cv46tM+5MYRNWpmloyfl4lsens/9skdv9LFbzRuuQChaL\n5GROab3yXUlKFJqWVz6vrBqo7RedmX9aEquPZLO9Tp4l6+/bBPVpDQRc6AshpgkhEoUQSUKIWYG+\nvlKHpriqTmP597dHiLcR+pU1jk1VLc2m/f4vp9l+Kp+qGjNPflN/NSiTxXVXHax6rDXzpZStx7zz\nyFcH+PPCON7ZdIqdp88BtVFNB9KK+eeKw+el8tYYAir0hRDBwIfAdGA4cIcQYnigrm8d9bp4G1pT\nnpqWq9M3jJ8OZTks10SkrabbnMP83tx4ij8vjGPVkSy+dxBRVlJlpMzguIN78ptD1udZ04D3pSod\nY2sZ6QC8s+k0dy3YC9j7PL7dn8Gtn/7aRLXyHcezSrnhvR12jnt/IQL5MgghrgDmSCmvV7/PBpBS\nvursmNjYWBkfX9+e6Y6YWasBCA2uffDNFml9YEKDBUaz43u3PaYl4+j+goNEiwhltK279ns4+71s\n93G2X0iQaLaav6v78pS6z3VLf5brtk1osMAi6/s9WtN9avcSGRHOr7Ov9ep8Qoj9UspYR9sCHaff\nB0i3+Z4BXF53JyHEg8CDAP379/fqQt07hNMmNIibRva2K19/LIerh/QgLCSIQ+nF7E5W7KJtQoMw\nGC08MH4AYSGtw9VhMFpYuOsMN1/Sm+xiA3GphTw00fWqOs2FtMJKVh3JZnCPCK4bHgXAruQCukeE\nczKnlJtGKb/rx1uTuWJgNy7t39l6rMkiWRGfzq1j+pFXWs3RzBLrOZojO06fo3fnNgzsHsGK+Awm\nXhjJ9wcUjT8sJIgHxg9g8e5UKmvsl72MjAije4c2TBgcyYr4dG4b058goQQobDiey98mD2qK2/Ep\nB9KKCA0Ooqiyhn5d2jEgsj0Am07kkldWTXGlkSnDenBhVIcmrmnjMBgt/Hgok1vH9LOa5dqH+0c8\nB1rTvwWYJqV8QP1+N3C5lPJRF8fkA2e9vGQkcM7LY88X9DZyj95GrtHbxz2BbqNoKWV3RxsCreln\nAv1svvdVy5zirOINQQgR72yIo6Ogt5F79DZyjd4+7mlObRRoO8Y+YLAQYoAQIgy4HfgpwHXQ0dHR\nOW8JqKYvpTQJIR4F1gPBwEIp5bFA1kFHR0fnfCbgCdeklGuAQKW2nB+g67Rk9DZyj95GrtHbxz0+\nbyMhRCrwgJRyk0fHNef4ZYDIyEgZExPT1NXwKzUmS6uJGNLR0QkMCQkJREdH07FjR7tyKSUHDhw4\n11wcuR4TExODN3H6LYWTOaVMe2cHs6YP5a+TWn6InY6OTmCIiYnho48+IiMjg88++4yxY8eyZMkS\nHn74YQ4cOOA04lFXL5uYNDVnenyq+5wyOjo6Oo7Yu3cvAwcOJDc3l2eeecblvrrQb2IsLSxPjI7v\nWHc0h+NZ9ROp6eh4Su/evXnssccICQmhbdu2Lvdt9uad1k/rSY6l4xl/XbofoMWsZqbTfOnXr5/7\nnVR0Tb+J0fzorSENro6OTtPgSXI9vwh9IUQ/IcQWIcRxIcQxIcRMtXyOECJTCHFI/Zvhj+u3JKyr\nPOndr46OTgDwl3nHBPxDSnlACNEB2C+E2Khue1tK+YafrtvisNis56qjo6Pjb/wi9KWU2UC2+rlM\nCHECJcOmTh0akuNfR6ep+e2Hu7hjbD9uG+Nd1lsd35Oammr9fO+99zb4OL8bFYQQMcClwF616FEh\nxBEhxEIhRBcnxzwohIgXQsTn5+c72qXVoE2Oa+k2fZPZwqJdZ6gxWZq6KgFhRXw6WxPz6i3l11o5\nlF7MU98lNHU1WgSJOWX8ciK3qavhFL8KfSFEBPAd8LiUshT4GBgEXIIyEnjT0XFSyvlSylgpZWz3\n7l4n2fQrRRU1XPnqLxzLKmnUeazmnZYt8/kmPp0Xfz7O/O3JTV0Vv5NbauBf3x7h3kX7+PPCuKau\njk4z4/p3tvOXL5rvhFK/CX0hRCiKwP9SSvk9gJQyV0ppllJagM+Asf66vr/ZkXSOrBIDH21tnJCz\nrufqgzo1Jdpyfs6W9WtNGM3nx2hGp3Xir+gdASwATkgp37Ip72Wz2++Ao/64fiDwlVnG0kpCNq0p\nnBzcRkF5db3l7XRaDs09P5c33D7/V77bn9HU1WgS/KXpXwXcDVxTJzxznhAiQQhxBLgaeMJP1/c7\nvtLQpdW808KFvpNl2Isqahj98ibmrT8Z4Brp+IrW2F/vSSnkHysON3U1mgR/Re/sxLE8DFRKZZ9h\nsUiKq4x0bR9mV64JucYuMl47Oatx52lqajtB+xspqKgBYOOxXGZPHxboaun4AH2U1rrQpwS54aOt\nSVz2n41kl1TZlVtUs64jDX1PSgFHM+s7eA1GM49+dYDEnLLa87QSR65Ga7kPb9l+Kp+kvDK3+wXK\nZHLmXAVbTuY16hwWL+q6JTGPlPzyRl1Xxz/oQt8N648poVd5pdV25S5M2Nw+fw83vr+zXvkb6xNZ\ndSSb69/Z7uA8tWf64WAm9y/e15hqB5wiVaNvrMzXzhMI7l6wl1VHsgAoMxipNpmd7ltqMDbIgfvn\nhXFMeWu72/0CpT1f/cZW7mvgs5RVXMWUt7bVV3C8EPr3LdrHNW9u8/i4QNAafRSeoAt9N1icOGy9\nscUXOBBo0kEahse/OcTmRmpngSQpr4zPd55xvVMDmulAWhGX/mejVRD7mx2nz/HoVwcBuHjOBu6Y\nv8fpviPnbODxZYd8dm2zjeA5kd08Mm1+HZdGUl45y/fZOzhbm3nHF/eTkFHC/rOu06Evi0tzqUg0\nFbrQr0N+WTXT3tlOeqGS5157QOrmxrHasBup2mqdytdx6QCsPpLduBM2EiklJZVGj445nF5ryvKk\nPepeKyFDOc/elEKPru8rDqQVu9y+OiGb69/eTn5Ztcv9QPEFOeJQejHT3tlOoY0CMP3dHZ5V1AUG\no5mqGu8Ezfubk4D6/iVLAyJUDUYzN3+w060gbA74og+76YOd/OHj3YDyHBdX1lfoZn2fwJsbTjX+\nYj5GF/p1+OFgJidzyvhidypQK9yDbd6ElQczKDUowsr2/agxWVi+L92j69Udaj7y1QGP6+xLvopL\nY9RLG0jKa7g91jYKwpMcQsvj0xn10gZO5So28NpRVYNP4TWeDPHXH8uxfk7MLePHQ+5HIvPWJ9Yr\n+zW5gN9+uIuTOWWsO5rj4KjGM/61LQx7fl2jzhEUJFh9JJuCcqVze+5H95HVp3LLOJxRwpyfjjXq\n2oHAlblKSsmK+HSHHafJbGFZXFq9kcI3+9K55KWNDn0587enNL7CPkbPp18Hk/qDBgcLSg1G6zBc\noHzPKzXwxDe1Qi5IKOUd24Ty/ubTVm0J4I+f7Ob5G0dwcd9OvLPpFCsPZta7nu3zc/eCvfW2+xOz\nRVJlNBMRXvsYbD6hmJXOnKvggh4R1nLtHh9cEs+G47lOc8B/sCWJG0b2Ylgvbd1O5QYrq82YzBZC\ngmv1jK2JSgqDpLxyojq2sVlQpvFSv6TKSKe2oU63V9uki3A33H/of/vtvjfExr06IYtZ04fy6poT\ndGoXyt8mX8CBtFot2F9m5XPl7kchjsgrNVg/F1bU8Pr6RNqEBhEsBBU2AvCBL+K5ZXQfpl3Uy+74\n2hGx4FEbxeWBL+L5/J5Yr+rkK4oqavjzwjhGRytZX/51/RDrtrc2JBIUJLjqgkheX5/Iw5MG8a9v\njxCfWsTpvDLm/GYEI/t2BmDpnrPM+fk4BqN9h6ApAUl5FVzQo0OA7sp7dE2/DmZ1LPvToSxGzqnV\neFfEpzNyzoZ6gvsbtXzd0Zx6L9y+1CKrlvTOptN225LyytmTUmCnce44fc7n9+OK/6w6zkUvrHdo\nd7QVu9/sS2PknA28tu4kG44rju0fD2VSXm1id3L9Ok9/dwe7ks6RkFFiFW45pQb+aTMi0I4HxaQ1\n6sUNHMlQzCuNnai2Ij6dUS9uIDGnjLMFFex00K5Dn6vVhoscDM1d0RChr414Pt2ewrx1itZv27k0\nRuafyi0jPtW3JjBHnaDBaLET+ACbTuTy16X2o9HCihpWqWZJs8Vi/azt39SsSsgmIbOExbtTWbw7\n1c6f8t7mJN7ZdJp/rThM3JlCjqv+lR2n8zmQVszzP9aOXIpUU2ShjUkyvbCSX1MKgJYTuRZwoS+E\nmCaESBRCJAkhZgX6+s6oqDZRUmmkWP1Bs0sMdts1R+WHWxynXdBWQaqL0WyxmoJsmfLWNm6fvwdz\nE/rIvjugOOwMNRYqa0yYzBaHwmjjcUX7/9gm5cTMZYe487M9/Okzx6OTOz/fy00f7OTbA7VOwR9U\njehYVgkzlx2ydnLbTysaf4Ia5mprTvEGzQmelFfOpNe3cteCvfW0M1tshXF+WTUxs1Yz/rXNVFSb\n+GZfmsv9QbFn371gr129tQ7N9pgym+egrnnJVCcyyGA0O3UCTn17O7d88qvT+3F0fVuqTWa79qg2\nme2CDBarpk1XxMxaTcys1eSVGvjDx7tZoL4fRzPrO6X9HS1jMJrrJfpbtOsMr645AUB5ndQg0oGP\nQmuvjCIlcilLff8tUmI0W6iqMVNZo+zz3i+1CtxZdY1rUPw4jp6zOz/f06xSdwTUvCOECAY+BK4D\nMoB9QoifpJTHA1kPR4x4YX2jz7HlZP2Mi8eyShk5Z4PTY+ata7qZqppGnZBZwl0L9nLN0B5WLTa/\nvJp1R7OZdlEvkp3EWx/OcJ9s7tNt9W2adV8MLV+P5vzMLK6qd4wnGNWeNCS4VvUa+tw6lj04jmG9\nOtpp/kLUmvQAq5DPKKpy+kxoTnfbc4P9SK2wooYNNp3Asz8c5eu4+h2IxgNL4ll8X20qqqHPrSMy\nIpz4Z6c4v1EXXPTCeo7Mmepw21VzN3OuvMZqohvyrPc+gLGv/OJ2n+T8cj7cksydl/enU9tQBkf5\n1gQy9Ll1DOvVkbUzJ1jLXvxZESmX9OvMa3XeMbODTuhcudLp1f2NzBbJ4GfWOr22prAAPPylY3/c\nrqQCUs9VMDiqA+uP5XBZ/y507xDu5q78R6Bt+mOBJCllCoAQYhlwM+BzoZ+cX94gu+nKgxmkFzZO\nyGjklBrc71SHahepiD1xpnpDSZWied6l+hJsw0Rnf++fNLpJeeVWbaouthEtjbl3Lc48p85o7fb5\nexg7oCtxZ2pNI1JCss218hoQmdNQFtiEsdYVJi+vPmH3fWtifr17Plde7bId3LVRgk2nnJRXTmJO\nGR9uSbIKuKdXJnD/VQNc34QP0OYtaKbRi/t04v7xMVzcp7PPrnEiu9Rhe8z5ub5j+bL/bKxX5oxj\nbhaub6ijVnvuH/rfftqHBfPjo+PVLZKnVx7lxpG9uHJQpN0xwUGCAZHtG1zXhiICOVFBCHELME1K\n+YD6/W7gcinlo86OiY2NlfHxnqcpHfbcOqpcDOl1dHR0mjONGekJIfZLKR160Jtl9I4Q4kHgQYD+\n/b1bqeeNP45yOIyrS3FlDUUVRkwWC4UVNYQGBxEaLDCaJceyShgT05WEzBL6dG5LQmYJsdFd+P5g\nJpcP6D01mLUAAAdESURBVIrRLNmmLqIxdXgU3SLCyS6pwmA0syelkI5tQpjzmxGsSchBCEVzCw0O\n4qZRvflqbxoTB0cyOKoDb25I5JJ+nRkc1YF2YcEcSiumY9sQxg/271oCBqOZPckFXNizA5XVJnYk\nnePCHh04kVPKXZdHk5RfTlFFDaEhQaw/mqPYrq+IIau4is0n8xgcFUF6YZXVgX3jyF6EBAl+OJTF\n9It60rNTG9qEBvPz4Syiu7Vj2kW9rBE1xzJLCA4SxJ0ppE+XtuSUGPjT5f357kAm1w2Pchl54w6T\n2cKJ7FIu7tuZg2lFFFXUEBkRTrvwEGK6teNEdilVRjNnCyq5cWQv2oaFsCwujQujOnBZdBe+3Z9B\neEgQeWXVhAcHkVlcRe/ObejRsQ2rj2TzfxMGEBQk+HRbCreP6UdGURXl1SZio7vQJjSYwsoajmaW\ncP9VA0jOLyenxMCYAV1Jya/AYDRjslhoHx5Ct/ZhlFaZKDMYuaBHBJ3a1eZ3yiquIlgIojq1qXd/\nheXVlBlMRDvQAvNKDVTVmDFZJIN6RFif7wHd22O2WDiaWUrvzm05W1BB785t6d25LQXl1cSfLSK/\nrJq4M4XcMbYfUiomrrgzhdSYLdwyui81Jgv7UgvJLjEwOroLQULRYEdHdyG7xMCxrFIu7d+ZhIwS\nTBbJU9OGkl5USUR4CAfOFhHVqQ2X9utMfnk1F3SPIDw02Ovf2Jb0wkrahgYTaWMyKSivZl9qIZOH\n9OBEdimZRVVUGc1Ed2tnFzFXWF7DxX07seFYDtUmCwUVNbQNDebeK2PILTUwOKoDB9OK2H+2iMKK\nGqZd1JPgIEFxpZELozqwcNcZjGYLArh6SA/Kq010ahuKySKprDERER5K+/BgYmO6IqXkuwOZ/GZU\nb8JCat2piTml9O3Sjvbh9uI4PMQ/LtdAa/pXAHOklNer32cDSClfdXFMPnDWy0tGAoENiWl56G3k\nHr2NXKO3j3sC3UbRUkqHWmOghX4IcAq4FsgE9gF/klL6ZUaHECLe2RBHR0FvI/fobeQavX3c05za\nKKDmHSmlSQjxKLAeCAYW+kvg6+jo6OjUJ+A2fSnlGlpgXn0dHR2d1kBrn5E7v6kr0ALQ28g9ehu5\nRm8f9zSbNgqoTV9HR0dHp2lp7Zq+jo6Ojo4NutDX0dHROY9olUK/uSZ18xdCiIVCiDwhxFGbsq5C\niI1CiNPq/y5quRBCvKe2zREhxGU2x9yj7n9aCHGPTfloIUSCesx7whe5jwOMEKKfEGKLEOK4EOKY\nEGKmWq63k4oQoo0QIk4IcVhtoxfV8gFCiL3qfX0jhAhTy8PV70nq9hibc81WyxOFENfblLf4d1MI\nESyEOCiEWKV+b1ntI6VsVX8ooaDJwEAgDDgMDG/qevn5nicClwFHbcrmAbPUz7OA19TPM4C1KNmT\nxwF71fKuQIr6v4v6uYu6LU7dV6jHTm/qe/aijXoBl6mfO6DMFxmut5NdGwkgQv0cCuxV72c5cLta\n/gnwsPr5b8An6ufbgW/Uz8PV9y4cGKC+j8Gt5d0EngS+Alap31tU+7RGTd+a1E1KWQNoSd1aLVLK\n7UDdBOs3A1+on78AfmtTvkQq7AE6CyF6AdcDG6WUhVLKImAjME3d1lFKuUcqT+wSm3O1GKSU2VLK\nA+rnMuAE0Ae9nayo96plLQtV/yRwDfCtWl63jbS2+xa4Vh3d3Awsk1JWSynPAEko72WLfzeFEH2B\nG4DP1e+CFtY+rVHo9wFsc99mqGXnG1FSSm01ixwgSv3srH1clWc4KG+xqMPsS1E0Wb2dbFBNF4eA\nPJQOLRkollJqSelt78vaFur2EqAbnrddS+Id4N+Alh63Gy2sfVqj0Nepg6p56rG5gBAiAvgOeFxK\naZc3V28nkFKapZSXAH1RNM+hTVylZoMQ4kYgT0rpeMWkFkJrFPqZQD+b733VsvONXNXkgPpfS5bv\nrH1clfd1UN7iEEKEogj8L6WU36vFejs5QEpZDGwBrkAxbWmz923vy9oW6vZOQAGet11L4SrgN0KI\nVBTTyzXAu7S09mlqp4iv/1BSS6SgOEg0Z8iIpq5XAO47BntH7uvYOyjnqZ9vwN5BGaeWdwXOoDgn\nu6ifu6rb6jooZzT1/XrRPgLFzv5OnXK9nWrbojvQWf3cFtgB3AiswN5R+Tf18yPYOyqXq59HYO+o\nTEFxUraadxOYTK0jt0W1T5M3np9+kBko0RnJwDNNXZ8A3O/XQDZgRLED/gXFdvgLcBrYZCOYBMqS\nlclAAhBrc577UZxKScB9NuWxwFH1mA9QZ3K3pD9gPIrp5ghwSP2bobeTXRuNBA6qbXQUeF4tH4jS\noSWpAi5cLW+jfk9Stw+0OdczajskYhPF1FrezTpCv0W1j56GQUdHR+c8ojXa9HV0dHR0nKALfR0d\nHZ3zCF3o6+jo6JxH6EJfR0dH5zxCF/o6Ojo65xG60NfR0dE5j9CFvo6Ojs55xP8D8hfMZVap1O4A\nAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 7 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "38IPFb23sf6E",
        "colab_type": "text"
      },
      "source": [
        "# **2.** **Label** **Encoding**\n",
        "The Wind direction doesn’t contain numerical values so label encoding is done.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "p9hfAc0PsMMb",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# integer encode direction\n",
        "encoder = LabelEncoder()\n",
        "values[:,4] = encoder.fit_transform(values[:,4])\n",
        "# ensure all data is float\n",
        "values = values.astype('float32')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iubdsXNGtK13",
        "colab_type": "text"
      },
      "source": [
        "# **3. Data Normalization**\n",
        "Data Normalization is done using MinMaxScaler function of sklearn."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BTfuMJvOtRrp",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# normalize features\n",
        "scaler = MinMaxScaler(feature_range=(0, 1))\n",
        "scaled = scaler.fit_transform(values)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MSf8Y-G-t5Tj",
        "colab_type": "text"
      },
      "source": [
        "# **4. Converting to Time Series Data**\n",
        "Since we use the LSTM neural network, we must sort the data according to the time. The dataset is transformed into a supervised learning problem. The weather variables for the hour to be predicted (t) are then removed. So, we have features for previous timestep (t-1) and for prediction of pollution PM2.5 taking its current timestep (t) data."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DgeD-HP6vL7U",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# convert series to supervised learning\n",
        "def series_to_supervised(data, n_in=1, n_out=1, dropnan=True):\n",
        "\tn_vars = 1 if type(data) is list else data.shape[1]\n",
        "\tdf = DataFrame(data)\n",
        "\tcols, names = list(), list()\n",
        "\t# input sequence (t-n, ... t-1)\n",
        "\tfor i in range(n_in, 0, -1):\n",
        "\t\tcols.append(df.shift(i))\n",
        "\t\tnames += [('var%d(t-%d)' % (j+1, i)) for j in range(n_vars)]\n",
        "\t# forecast sequence (t, t+1, ... t+n)\n",
        "\tfor i in range(0, n_out):\n",
        "\t\tcols.append(df.shift(-i))\n",
        "\t\tif i == 0:\n",
        "\t\t\tnames += [('var%d(t)' % (j+1)) for j in range(n_vars)]\n",
        "\t\telse:\n",
        "\t\t\tnames += [('var%d(t+%d)' % (j+1, i)) for j in range(n_vars)]\n",
        "\t# put it all together\n",
        "\tagg = concat(cols, axis=1)\n",
        "\tagg.columns = names\n",
        "\t# drop rows with NaN values\n",
        "\tif dropnan:\n",
        "\t\tagg.dropna(inplace=True)\n",
        "\treturn agg"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9JbFtgb3vQJ1",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 153
        },
        "outputId": "e1316a15-1c87-4907-9402-4d495f071baa"
      },
      "source": [
        "# frame as supervised learning\n",
        "reframed = series_to_supervised(scaled, 1, 1)\n",
        "# drop columns we don't want to predict\n",
        "reframed.drop(reframed.columns[[9,10,11,12,13,14,15]], axis=1, inplace=True)\n",
        "print(reframed.head())"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "   var1(t-1)  var2(t-1)  var3(t-1)  ...  var7(t-1)  var8(t-1)   var1(t)\n",
            "1   0.129779   0.352941   0.245902  ...   0.000000        0.0  0.148893\n",
            "2   0.148893   0.367647   0.245902  ...   0.000000        0.0  0.159960\n",
            "3   0.159960   0.426471   0.229508  ...   0.000000        0.0  0.182093\n",
            "4   0.182093   0.485294   0.229508  ...   0.037037        0.0  0.138833\n",
            "5   0.138833   0.485294   0.229508  ...   0.074074        0.0  0.109658\n",
            "\n",
            "[5 rows x 9 columns]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0KS25njlvaX6",
        "colab_type": "text"
      },
      "source": [
        "# Splitting into train and test sets\n",
        "We will only fit the model on the first 2 years(365* 24 * 2 hours) of data, then evaluate it on the remaining 3 years of data. "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "law1jSvUwLTl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# split into train and test sets\n",
        "values = reframed.values\n",
        "n_train_hours = 365 * 24*2\n",
        "train = values[:n_train_hours, :]\n",
        "test = values[n_train_hours:, :]\n",
        "# split into input and outputs\n",
        "train_X, train_y = train[:, :8], train[:, 8]\n",
        "test_X, test_y = test[:, :8], test[:, 8]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vxlh_uxZwTuw",
        "colab_type": "text"
      },
      "source": [
        "# reshaping input to be 3D [samples, timesteps, features]\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8jxAhIVVwhcn",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "f3c8eaea-8b92-41b5-e156-22386677ca6e"
      },
      "source": [
        "train_X = train_X.reshape((train_X.shape[0], 1, train_X.shape[1]))\n",
        "test_X = test_X.reshape((test_X.shape[0], 1, test_X.shape[1]))\n",
        "print(train_X.shape, train_y.shape, test_X.shape, test_y.shape)"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(17520, 1, 8) (17520,) (24236, 1, 8) (24236,)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IkZb5nqRwi8Y",
        "colab_type": "text"
      },
      "source": [
        "# **Designing Network**\n",
        "We will define the LSTM with 100 neurons in the first hidden layer and a Dropout Layer of 0.3,Next there will be another hidden layer of 50 neurons and a Dropout of 0.2. Similarly, there will two more hidden layers with respective 0.2 Dropouts and 50 neurons. In Final Layer, 1 neuron in the output layer for predicting pollution. In the activation Function we used linear function, because of sequential dataset. In the batch size we used three days (24*3 Hours) data. Optimizer function we used Adam method. Loss function we used mean squared error. By monitoring the value of test data loss function, stop the training model when it is not decreasing, and save the current best model."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "saLSkgO1x_0T",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# design network\n",
        "model = Sequential()\n",
        "model.add(LSTM(100, return_sequences = True, input_shape=(train_X.shape[1], train_X.shape[2])))\n",
        "model.add(Dropout(0.3))\n",
        "\n",
        "model.add(LSTM(units = 50, return_sequences = True))\n",
        "model.add(Dropout(0.2))\n",
        "\n",
        "model.add(LSTM(units = 50, return_sequences = True))\n",
        "model.add(Dropout(0.2))\n",
        "\n",
        "model.add(LSTM(units = 50))\n",
        "model.add(Dropout(0.2))\n",
        "\n",
        "model.add(Dense(1,activation='linear'))\n",
        "\n",
        "model.compile(loss='mse', optimizer='adam')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0hozBMtLhORS",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 459
        },
        "outputId": "4ab3cfaf-7c5a-49b2-c6bc-dab1fc800804"
      },
      "source": [
        "model.summary()"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "lstm (LSTM)                  (None, 1, 100)            43600     \n",
            "_________________________________________________________________\n",
            "dropout (Dropout)            (None, 1, 100)            0         \n",
            "_________________________________________________________________\n",
            "lstm_1 (LSTM)                (None, 1, 50)             30200     \n",
            "_________________________________________________________________\n",
            "dropout_1 (Dropout)          (None, 1, 50)             0         \n",
            "_________________________________________________________________\n",
            "lstm_2 (LSTM)                (None, 1, 50)             20200     \n",
            "_________________________________________________________________\n",
            "dropout_2 (Dropout)          (None, 1, 50)             0         \n",
            "_________________________________________________________________\n",
            "lstm_3 (LSTM)                (None, 50)                20200     \n",
            "_________________________________________________________________\n",
            "dropout_3 (Dropout)          (None, 50)                0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 1)                 51        \n",
            "=================================================================\n",
            "Total params: 114,251\n",
            "Trainable params: 114,251\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EJ2iNauoWCsf",
        "colab_type": "text"
      },
      "source": [
        "# **Fitting the Network**\n",
        "Network is fit with epochs size of 50 , batch size of 72."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BFpzGnxzV6CF",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "bb898936-e74f-4154-bd17-e3b19cbab816"
      },
      "source": [
        "# fit network\n",
        "history = model.fit(train_X, train_y, epochs=50, batch_size=72, validation_data=(test_X, test_y), verbose=2, shuffle=False)"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/50\n",
            "244/244 - 4s - loss: 0.0074 - val_loss: 0.0045\n",
            "Epoch 2/50\n",
            "244/244 - 3s - loss: 0.0015 - val_loss: 0.0025\n",
            "Epoch 3/50\n",
            "244/244 - 3s - loss: 0.0011 - val_loss: 0.0010\n",
            "Epoch 4/50\n",
            "244/244 - 3s - loss: 0.0010 - val_loss: 6.0220e-04\n",
            "Epoch 5/50\n",
            "244/244 - 3s - loss: 9.9046e-04 - val_loss: 5.7982e-04\n",
            "Epoch 6/50\n",
            "244/244 - 3s - loss: 9.6692e-04 - val_loss: 5.6456e-04\n",
            "Epoch 7/50\n",
            "244/244 - 3s - loss: 9.5543e-04 - val_loss: 5.9899e-04\n",
            "Epoch 8/50\n",
            "244/244 - 3s - loss: 9.2703e-04 - val_loss: 5.7770e-04\n",
            "Epoch 9/50\n",
            "244/244 - 3s - loss: 9.2540e-04 - val_loss: 5.8980e-04\n",
            "Epoch 10/50\n",
            "244/244 - 3s - loss: 9.5063e-04 - val_loss: 6.0340e-04\n",
            "Epoch 11/50\n",
            "244/244 - 3s - loss: 9.2028e-04 - val_loss: 6.2078e-04\n",
            "Epoch 12/50\n",
            "244/244 - 3s - loss: 8.9136e-04 - val_loss: 6.0795e-04\n",
            "Epoch 13/50\n",
            "244/244 - 3s - loss: 8.9024e-04 - val_loss: 6.3853e-04\n",
            "Epoch 14/50\n",
            "244/244 - 3s - loss: 9.0460e-04 - val_loss: 6.6691e-04\n",
            "Epoch 15/50\n",
            "244/244 - 3s - loss: 9.1434e-04 - val_loss: 6.4268e-04\n",
            "Epoch 16/50\n",
            "244/244 - 3s - loss: 9.0130e-04 - val_loss: 7.5625e-04\n",
            "Epoch 17/50\n",
            "244/244 - 3s - loss: 8.9012e-04 - val_loss: 6.9763e-04\n",
            "Epoch 18/50\n",
            "244/244 - 3s - loss: 8.9030e-04 - val_loss: 7.6910e-04\n",
            "Epoch 19/50\n",
            "244/244 - 3s - loss: 8.9762e-04 - val_loss: 7.3056e-04\n",
            "Epoch 20/50\n",
            "244/244 - 3s - loss: 8.8961e-04 - val_loss: 7.4914e-04\n",
            "Epoch 21/50\n",
            "244/244 - 3s - loss: 9.0369e-04 - val_loss: 7.4577e-04\n",
            "Epoch 22/50\n",
            "244/244 - 3s - loss: 8.8337e-04 - val_loss: 7.5675e-04\n",
            "Epoch 23/50\n",
            "244/244 - 3s - loss: 8.9423e-04 - val_loss: 7.5954e-04\n",
            "Epoch 24/50\n",
            "244/244 - 3s - loss: 9.0866e-04 - val_loss: 7.0824e-04\n",
            "Epoch 25/50\n",
            "244/244 - 3s - loss: 8.8323e-04 - val_loss: 7.8057e-04\n",
            "Epoch 26/50\n",
            "244/244 - 3s - loss: 8.9432e-04 - val_loss: 7.5012e-04\n",
            "Epoch 27/50\n",
            "244/244 - 3s - loss: 8.7444e-04 - val_loss: 8.1429e-04\n",
            "Epoch 28/50\n",
            "244/244 - 3s - loss: 8.8533e-04 - val_loss: 7.6706e-04\n",
            "Epoch 29/50\n",
            "244/244 - 3s - loss: 8.7173e-04 - val_loss: 7.3549e-04\n",
            "Epoch 30/50\n",
            "244/244 - 3s - loss: 9.0746e-04 - val_loss: 7.5835e-04\n",
            "Epoch 31/50\n",
            "244/244 - 3s - loss: 9.0267e-04 - val_loss: 6.5081e-04\n",
            "Epoch 32/50\n",
            "244/244 - 3s - loss: 9.0185e-04 - val_loss: 6.9924e-04\n",
            "Epoch 33/50\n",
            "244/244 - 3s - loss: 8.8022e-04 - val_loss: 7.0197e-04\n",
            "Epoch 34/50\n",
            "244/244 - 3s - loss: 8.7723e-04 - val_loss: 7.0989e-04\n",
            "Epoch 35/50\n",
            "244/244 - 3s - loss: 8.9710e-04 - val_loss: 7.0431e-04\n",
            "Epoch 36/50\n",
            "244/244 - 3s - loss: 8.9159e-04 - val_loss: 7.6559e-04\n",
            "Epoch 37/50\n",
            "244/244 - 3s - loss: 8.7120e-04 - val_loss: 7.0626e-04\n",
            "Epoch 38/50\n",
            "244/244 - 3s - loss: 8.9914e-04 - val_loss: 7.2507e-04\n",
            "Epoch 39/50\n",
            "244/244 - 3s - loss: 8.7866e-04 - val_loss: 7.5507e-04\n",
            "Epoch 40/50\n",
            "244/244 - 3s - loss: 8.8854e-04 - val_loss: 7.0360e-04\n",
            "Epoch 41/50\n",
            "244/244 - 3s - loss: 8.8052e-04 - val_loss: 7.4718e-04\n",
            "Epoch 42/50\n",
            "244/244 - 3s - loss: 8.8404e-04 - val_loss: 7.6117e-04\n",
            "Epoch 43/50\n",
            "244/244 - 3s - loss: 8.4083e-04 - val_loss: 6.8127e-04\n",
            "Epoch 44/50\n",
            "244/244 - 3s - loss: 8.4414e-04 - val_loss: 7.2791e-04\n",
            "Epoch 45/50\n",
            "244/244 - 3s - loss: 8.5387e-04 - val_loss: 7.9109e-04\n",
            "Epoch 46/50\n",
            "244/244 - 3s - loss: 8.7035e-04 - val_loss: 6.6664e-04\n",
            "Epoch 47/50\n",
            "244/244 - 3s - loss: 8.7271e-04 - val_loss: 7.6874e-04\n",
            "Epoch 48/50\n",
            "244/244 - 3s - loss: 8.9019e-04 - val_loss: 7.4561e-04\n",
            "Epoch 49/50\n",
            "244/244 - 3s - loss: 8.7932e-04 - val_loss: 7.4655e-04\n",
            "Epoch 50/50\n",
            "244/244 - 3s - loss: 8.7452e-04 - val_loss: 7.3149e-04\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9T1ttuNCWbHx",
        "colab_type": "text"
      },
      "source": [
        "# **Plotting the graph of Train Loss and Test Loss**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "i8ZcS4QyWfq8",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "outputId": "c8eb60a4-82c2-436b-8817-c5934b2e2235"
      },
      "source": [
        "# plot history\n",
        "pyplot.plot(history.history['loss'], label='train')\n",
        "pyplot.plot(history.history['val_loss'], label='test')\n",
        "pyplot.legend()\n",
        "pyplot.show()"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAD4CAYAAADlwTGnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\njDHGp/4/mP9frvjZ/sYAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lVP-g5r3XMnZ",
        "colab_type": "text"
      },
      "source": [
        "# **Making Prediction/Forecasting**\n",
        "Forecasting the results and invert the scaling of the prediction and test data to check."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wEL6sJGrXcvt",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# make a prediction\n",
        "yhat = model.predict(test_X)\n",
        "test_X = test_X.reshape((test_X.shape[0], test_X.shape[2]))\n",
        "\n",
        "# invert scaling for forecast\n",
        "inv_yhat = concatenate((yhat, test_X[:, 1:]), axis=1)\n",
        "inv_yhat = scaler.inverse_transform(inv_yhat)\n",
        "inv_yhat = inv_yhat[:,0]\n",
        "\n",
        "# invert scaling for actual\n",
        "test_y = test_y.reshape((len(test_y), 1))\n",
        "inv_y = concatenate((test_y, test_X[:, 1:]), axis=1)\n",
        "inv_y = scaler.inverse_transform(inv_y)\n",
        "inv_y = inv_y[:,0]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CQPmkiDZYaah",
        "colab_type": "text"
      },
      "source": [
        "# **Calculating the RMSE andd MAE values**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vFUpVIiiYh3u",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "54122cd3-ab1b-421e-834f-c16f25cfe964"
      },
      "source": [
        "# calculate RMSE and MAE\n",
        "rmse = sqrt(mean_squared_error(inv_y, inv_yhat))\n",
        "print('Test RMSE: %.3f' % rmse)\n",
        "mae = (mean_absolute_error(inv_y, inv_yhat))\n",
        "print('Test MAE: %.3f' % mae)"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Test RMSE: 26.897\n",
            "Test MAE: 17.021\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "G6jEnuZbcLpB",
        "colab_type": "text"
      },
      "source": [
        "# **Plotting the Graph of Actual vs Predicted**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r_YnmnUoa6W8",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        },
        "outputId": "169186e7-3600-4f2d-e07e-1d03563f55f2"
      },
      "source": [
        "print('Actual :', inv_y)\n",
        "print('Predicted:', inv_yhat)"
      ],
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Actual : [110. 113. 110. ...  10.   8.  12.]\n",
            "Predicted: [113.372086 111.72432  115.032684 ...  11.640506  11.598233  10.386721]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bhakZROQb9Cf",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "outputId": "0b0d3323-973a-4149-da30-1c5447632ef4"
      },
      "source": [
        "# plot history\n",
        "pyplot.plot(inv_y, label='Actual')\n",
        "pyplot.plot(inv_yhat, label='Predicted')\n",
        "pyplot.legend()\n",
        "pyplot.show()"
      ],
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAAD4CAYAAAAKA1qZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/wfFiXrh1ga5wAAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}