5_3_2020_Covid_Prophet_Final.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.6.6"
    },
    "colab": {
      "name": " 5-3-2020 Covid_Prophet_Final.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "machine_shape": "hm",
      "include_colab_link": true
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/hussain0048/Machine-Learning-Driven-Approach-for-2019-nCoV-Warning-System/blob/master/Implementation/5_3_2020_Covid_Prophet_Final.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FfcjS4CKTyAP"
      },
      "source": [
        "# **1-Introducton** \n",
        "\n",
        "Predicting confirmed cases of Convid-19 with Prophet"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "r609ISbsmq2t"
      },
      "source": [
        "#**2-Importing necessary libraries**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FxdLWHIJTyAU"
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "from fbprophet import Prophet\n",
        "import matplotlib.pyplot as plt\n",
        " \n",
        "plt.rcParams['figure.figsize']=(20,10)\n",
        "plt.style.use('ggplot')"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5NeSSztObv-7"
      },
      "source": [
        "plt.rcParams['figure.figsize']=(20,10)\n",
        "plt.style.use('ggplot')"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XhadCtyRcFFY"
      },
      "source": [
        "pd.plotting.register_matplotlib_converters()"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-vcGxjHQGL1c"
      },
      "source": [
        "#[3]\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')\n",
        "\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from scipy import stats\n",
        "import statsmodels.api as sm\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "%matplotlib inline"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "d17uQ_kpKn5w"
      },
      "source": [
        "from fbprophet import Prophet\n",
        "\n",
        "import logging\n",
        "logging.getLogger().setLevel(logging.ERROR)"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "X5p8oL_STyAg"
      },
      "source": [
        "#**3-Import dataset**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4lsb_UGenA-Y",
        "outputId": "70bd3db4-797d-407c-88b1-0cf5182a48d2",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RDKPEa7keRWE"
      },
      "source": [
        "data = pd.read_csv(\"/content/drive/My Drive/Datasets/Covid-19/data_cases1.csv\", usecols=['Date','Confirmed'],index_col='Date', parse_dates=True )\n",
        "data.head()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_qRu1k2Dyt2O"
      },
      "source": [
        "data.head()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "t-E75RQRTyAx"
      },
      "source": [
        "data.shape"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "R8BthJfITyBS"
      },
      "source": [
        "#**4-Data cleaning and feature engineering**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ns2tk6CCGzDo"
      },
      "source": [
        "df = data.reset_index()"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WGxkU_bZ_eoq"
      },
      "source": [
        "df.head()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wiEfm8xA9Cb4"
      },
      "source": [
        "#we get rid of possible duplicates and missing values in the data [3]\n",
        "df = df[['Date', 'Confirmed']].dropna().drop_duplicates()"
      ],
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fqj24WXrFwbV",
        "outputId": "8bd45a73-a772-48f3-c96f-c05f27829a71",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "df.shape"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1157, 2)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8oKf0uV2GG02"
      },
      "source": [
        "#we need to convert Date to the datetime format because by default pandas treats this field as string-valued[3]\n",
        "df['Date'] = pd.to_datetime(df['Date'])"
      ],
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ijc8tkqbGqpE"
      },
      "source": [
        "df.head()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-cnFzmAAHIWq"
      },
      "source": [
        "#Let's sort the dataframe by time and take a look at what we've got [3]\n",
        "df.sort_values(by=['Date']).head(n=3)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vegTUwkq9kw8"
      },
      "source": [
        "#We will just trim our time series to keep only those rows that fall onto the period from March 10, 2020 to March 31, 2020[3]\n",
        "df = df[(df['Date'] > '2020-03-09') & (df['Date'] < '2020-04-01')].sort_values(by=['Date'])\n",
        "df.head(n=10)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rahdPkyK-81A"
      },
      "source": [
        "#As we are going to predict the number of confirm Cases of Covid-19, we will aggregate and count unique Cases at each given point in time. We will name the corresponding new column Confirmed_Cases:\n",
        "aggr_df = df.groupby('Date')[['Confirmed']].count()\n",
        "aggr_df.columns = ['Confirmed_Cases']"
      ],
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "K-av4_rg_pfa"
      },
      "source": [
        "aggr_df.head(n=3)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8WPfYYLU_6dD"
      },
      "source": [
        "#In this practice, we are interested in the number of confirmed cases a day. But at this moment all our data is divided into irregular time intervals that are less than a day. \n",
        "#This is called a sub-daily time series.To fix this, we need to aggregate the Cases counts by \"bins\" of a date size. \n",
        "#In time series analysis, this process is referred to as resampling. And if we reduce the sampling rate of data it is often called downsampling.\n",
        "#Luckily, pandas has a built-in functionality for this task. We will resample our time index down to 1-day bins [3]\n",
        "daily_df = aggr_df.resample('D').apply(sum)\n",
        "daily_df.head(n=3)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9n0GTxkvAeq5"
      },
      "source": [
        "weekly_df = daily_df.resample('W').apply(sum)"
      ],
      "execution_count": 22,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SUR6GRnOJp5T"
      },
      "source": [
        "weekly_df.head()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lVCbk-vsAz-n"
      },
      "source": [
        "daily_df = daily_df.loc[daily_df.index >= '2020-03-10']\n",
        "daily_df.head(n=3)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZQLlg3g4OvWh"
      },
      "source": [
        "plt.figure(figsize=(16,8))\n",
        "plt.title('Daily Confirmed Cases of COVID-19 (World)')\n",
        "plt.plot(daily_df['Confirmed_Cases'])\n",
        "plt.xlabel('Date', fontsize=18)\n",
        "plt.ylabel('Confirmed Cases', fontsize=18)\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "C8ivtMFNTyEs"
      },
      "source": [
        "#**5- Modelling**\n",
        "\n",
        "Let's focus on predictig the confirmed cases of COVI-ID.Prophet's API is very similar to the one you can find in sklearn. First we create a model, then call the method fit, and, finally, make a forecast. The input to the method fit is a DataFrame with two columns:\n",
        "\n",
        "ds (datestamp) must be of type date or datetime.\n",
        "y is a numeric value we want to predict."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IVFBqHaJBM2J"
      },
      "source": [
        "#[3]\n",
        "df = daily_df.reset_index()\n",
        "df.columns = ['ds', 'y']\n",
        "df.tail(n=3)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0_egvgIBTB2U"
      },
      "source": [
        "df.set_index('ds').y.plot().get_figure()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "m2sSEx8DLfGB"
      },
      "source": [
        "#To measure the quality of our forecast, we need to split our dataset into the historical part, which is the first and biggest slice of our data,\n",
        "# and the prediction part, which will be located at the end of the timeline. \n",
        "#We will remove the last 3 days from the dataset in order to use it later as a prediction target[3]\n",
        "prediction_size = 3\n",
        "train_df = df[:-prediction_size]\n",
        "train_df.tail(n=3)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Yi-9PTvvMaOg"
      },
      "source": [
        "# now we create model [3]\n",
        "m = Prophet()\n",
        "m.fit(train_df);"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "a_ki-3zQMxf9",
        "outputId": "c429c6f7-1b09-4972-9991-29d0041948ed",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 142
        }
      },
      "source": [
        "#[3]\n",
        "future = m.make_future_dataframe(periods=prediction_size)\n",
        "future.tail(n=3)"
      ],
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>ds</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>2020-03-24</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>2020-03-25</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>2020-03-26</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "           ds\n",
              "14 2020-03-24\n",
              "15 2020-03-25\n",
              "16 2020-03-26"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CTCGpSIMNM8X"
      },
      "source": [
        "#[3]\n",
        "forecast = m.predict(future)\n",
        "forecast.tail(n=3)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qJSMTYcAOl4M",
        "outputId": "cbdbfb3e-0699-469c-9f33-2938a4106d99",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "#[1]\n",
        "forecast[['ds', 'yhat', 'yhat_lower', 'yhat_upper']].tail()"
      ],
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>ds</th>\n",
              "      <th>yhat</th>\n",
              "      <th>yhat_lower</th>\n",
              "      <th>yhat_upper</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>12</th>\n",
              "      <td>2020-03-22</td>\n",
              "      <td>83.510020</td>\n",
              "      <td>56.517001</td>\n",
              "      <td>104.968454</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>13</th>\n",
              "      <td>2020-03-23</td>\n",
              "      <td>89.762552</td>\n",
              "      <td>66.425144</td>\n",
              "      <td>113.424199</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>14</th>\n",
              "      <td>2020-03-24</td>\n",
              "      <td>96.015084</td>\n",
              "      <td>74.384712</td>\n",
              "      <td>121.402555</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>15</th>\n",
              "      <td>2020-03-25</td>\n",
              "      <td>102.267617</td>\n",
              "      <td>76.591055</td>\n",
              "      <td>125.693865</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16</th>\n",
              "      <td>2020-03-26</td>\n",
              "      <td>108.520149</td>\n",
              "      <td>83.544659</td>\n",
              "      <td>133.043532</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "           ds        yhat  yhat_lower  yhat_upper\n",
              "12 2020-03-22   83.510020   56.517001  104.968454\n",
              "13 2020-03-23   89.762552   66.425144  113.424199\n",
              "14 2020-03-24   96.015084   74.384712  121.402555\n",
              "15 2020-03-25  102.267617   76.591055  125.693865\n",
              "16 2020-03-26  108.520149   83.544659  133.043532"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 33
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DPnDCm1jeFIp"
      },
      "source": [
        "#**6- Result Visualization**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BB5Kty-_N7k1"
      },
      "source": [
        "#[3]forecasting\n",
        "m.plot(forecast);\n",
        "plt.title(\"Forecasting Covid-19 confirmed Cases(world) \")\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Auutf8VWdchF"
      },
      "source": [
        "#[3]\n",
        "m.plot_components(forecast);\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Jirxk0dDTs-n"
      },
      "source": [
        "# [1]\n",
        "df.set_index('ds', inplace=True)\n",
        "forecast.set_index('ds', inplace=True)"
      ],
      "execution_count": 36,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fnUxJ25aTztl"
      },
      "source": [
        "viz_df = daily_df.join(forecast[['yhat', 'yhat_lower','yhat_upper']], how = 'outer')"
      ],
      "execution_count": 37,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pgD7XbqwT_LV",
        "outputId": "0db294b6-f0e2-436a-e438-07f711201add",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 235
        }
      },
      "source": [
        "viz_df.head()"
      ],
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Confirmed_Cases</th>\n",
              "      <th>yhat</th>\n",
              "      <th>yhat_lower</th>\n",
              "      <th>yhat_upper</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Date</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>2020-03-10</th>\n",
              "      <td>17</td>\n",
              "      <td>8.479636</td>\n",
              "      <td>-15.735117</td>\n",
              "      <td>31.833502</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-03-11</th>\n",
              "      <td>24</td>\n",
              "      <td>14.732168</td>\n",
              "      <td>-11.257902</td>\n",
              "      <td>38.105296</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-03-12</th>\n",
              "      <td>28</td>\n",
              "      <td>20.984699</td>\n",
              "      <td>-2.669664</td>\n",
              "      <td>46.315372</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-03-13</th>\n",
              "      <td>30</td>\n",
              "      <td>27.237231</td>\n",
              "      <td>4.537930</td>\n",
              "      <td>51.516875</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2020-03-14</th>\n",
              "      <td>37</td>\n",
              "      <td>33.489763</td>\n",
              "      <td>9.121017</td>\n",
              "      <td>57.230878</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "            Confirmed_Cases       yhat  yhat_lower  yhat_upper\n",
              "Date                                                          \n",
              "2020-03-10               17   8.479636  -15.735117   31.833502\n",
              "2020-03-11               24  14.732168  -11.257902   38.105296\n",
              "2020-03-12               28  20.984699   -2.669664   46.315372\n",
              "2020-03-13               30  27.237231    4.537930   51.516875\n",
              "2020-03-14               37  33.489763    9.121017   57.230878"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 38
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TacZpRviVQlQ",
        "outputId": "bcbf0e0f-e8af-4977-f6e7-7104a1a079f0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 595
        }
      },
      "source": [
        "viz_df[['Confirmed_Cases', 'yhat']].plot().get_figure()"
      ],
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 39
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEhCAYAAACTNXDdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deXxU5b348c+ZbGQnkLAviSi7gCIuoAT3ulRc6lOXW3G5Uq1eW6+3rv0Jbq2tW2mttth61dte5HGr+1WqElAEUQGBsAgkYQkkJIFM9mTmPL8/zkkIEEIyM8kkme/79eKVmXPOPPMNyfnmmef7nOdYxhiEEEJEBk+4AxBCCNF5JOkLIUQEkaQvhBARRJK+EEJEEEn6QggRQSTpCyFEBIkOdwAumTcqhBCBsdpzcFdJ+hQWFoY7BCGE6FYGDRrU7tfI8I4QQkQQSfpCCBFBJOkLIUQEOeqYvlLqReBioFhrPd7dthAY5R7SG9ivtZ6klMoENgCb3H3Ltda3BBKYMYba2lps28ay2lWnEN2EMQaPx0OvXr3kZyxEJ2lLIfcl4FnglcYNWusfNz5WSj0FlDc7fqvWelKwgdXW1hITE0N0dJepNYsO4PP5qK2tJT4+PtyhCBERjjq8o7VeApS1tE8pZQEKWBDiuLBtWxJ+BIiOjsa27XCHIUTECDarngEUaa2/b7YtSym1CvACv9JaLw2kYfm4HznkZy1E5wk26V/Nwb383cAwrXWpUmoy8E+l1DittffQFyqlZgOzAbTWpKenH7S/qKhIevoRIi4u7rCfvxCiddX/9xZc8uOjH3iIgLOqUioauByY3LhNa10H1LmPv1FKbQVGAl8f+nqt9XxgvvvUlJSUHLS/rq6OqKioQMMLmeLiYubMmcOaNWtISUkhIyODuXPnMmLEiHa1s2LFCu69916io6N55ZVXePDBB3nhhRc6KOoDjjvuOL7//vsj7t+6dStz585l27ZtJCUlkZmZyaOPPkpGRkaHx9aorq6OQ3/+QoiWGWMw776KeXcBvTsz6QPnABu11jsbNyilMoAyrbVfKXUMcBywLYj3CCtjDDfddBNXXnklzz//PADr16+npKSk3Un/zTff5Pbbb+eKK64AaDHh+3y+Tv10U1tby3XXXcecOXM477zzAFi2bBmlpaWdmvSFEG1jbD9mwQuYxR9gnXZWQG20ZcrmAmAGkK6U2gnM0Vr/DbiKwwu404GHlVINgA3corVusQjcHXzxxRfExMRw3XXXNW0bN24cxhgeeeQRPvvsMyzL4o477mDmzJksW7aMp59+mrS0NDZt2sSECRP44x//yIIFC3jvvffIycnhs88+45577mHWrFl8+umnLFy4kA8//JCqqips20YpxUcffUR1dTV5eXnccsst1NfX88YbbxAbG8v//M//kJaWRn5+Pg888AClpaXEx8fzxBNPcOyxx7J9+3Zuu+02qqurmxL5kfzzn/9k8uTJBx03depUAHbs2MEdd9xBdXU1AI8++ihTpkyhqKiIW2+9lYqKCvx+P7/5zW845ZRTyMnJ4cknn6S+vp7hw4fzzDPPkJiYyK9//Ws+/vhjoqOjmT59Og8++GAH/KSE6PlMQwPmxWcwX3+Odf5lWFdcH1A7R036Wuurj7D9sHfUWr8BvBFQJK2wX30BsyMvpG1aQ7PwXHVzq8ds2rSJ448//rDtH3zwAevXr2fRokWUlZVx4YUXcuqppwKwbt06Pv30UwYMGMDMmTNZuXIl11xzDV999RXnnHMOF198MTt27DiovbVr1/Kvf/2LtLQ0Fi5cyKZNm/joo4+oq6tj2rRp3H///Xz88cfMmTOH119/nZtvvpm7776bxx9/nGOOOYZvv/2W++67j9dee40HH3yQ6667jiuvvJKXXnqp1e9v48aNTJgwocV96enpLFiwgF69erFt2zZuu+02PvzwQ9566y2ys7P5+c9/jt/vp6amhrKyMubNm8fChQtJSEjgT3/6E/Pnz2fWrFl8+OGHLFmyBMuyKC8vb/G9hBCtM7XV2M/9BjaswfrR9XjOvzzgtqRSGoCvvvqKSy+9lKioKDIyMjj11FNZs2YNSUlJTJo0qWkRpHHjxrFjxw5OPvnkVtubPn06aWlpTc+nTp1KUlISSUlJJCcnc+655wIwZswYcnNzqaqq4ptvvuGnP/1p02vq6+sBWLlyZdPQ0RVXXMFjjz0W0PfY0NDAAw88QG5uLh6Ph23bnFG6SZMmcdddd+Hz+Tj//PMZP348ixYtYvPmzcycObPptZMnTyYlJYW4uDjuuusuzjnnHM4555yAYhEikpmKcux5D8GObVg3/BzP1LODaq9bJP2j9cg7ysiRI3n//ffb9ZrY2Nimx1FRUfh8vqO+JiEh4YhteDwe4uLiAGdqo9/vx7ZtUlJSWLRoUYvttXUK5KhRo/jyyy9b3PfCCy+QkZHBokWLsG2bY445BoBTTz2VN954g08++YQ777yT2bNnk5qayvTp03nuuecOa+f999/n888/5/333+e///u/ee2119oUmxACTGkx9jNzoGwvnp89gDVxStBtyto7rTj99NOpr6/n73//e9O23NxcUlNTeeedd/D7/ZSWlrJixQomTQr6IuQ2S05OZujQobz77ruAU3Bev349AFOmTOHtt98GnOJxay699FK++eYb/vWvfzVtW758ORs3bsTr9dKvXz88Hg9vvPEGfr8fgJ07d5KRkcG1117LNddcw9q1a5k8eTIrV64kL88Zgquurmbr1q1UVVVRUVHB2Wefzdy5c8nNzQ35/4UQPZXZtR378buhYj+eOx8OScKHbtLTDxfLsvjrX//KnDlzeO6554iLi2PIkCE89NBDVFVVce6552JZFg888AD9+vVjy5YtnRbbs88+y3333ce8efPw+XzMnDmTcePG8fDDD3Pbbbfx3HPPHbWQGx8fz8svv8ycOXOYM2cOMTExjBkzhocffphZs2Yxe/ZsXn/9dc4888ymTyPLli3jz3/+M9HR0SQmJjJv3jz69u3LM888w2233dY0zHT33XeTlJTEjTfeSF1dHcYY5syZ0+H/L0L0BGbLBuw/PgIxsXh++RusIZkha9sypkvctMocehOV6urqw4Y9RM8kP2shDjDfrcT+y2+hdzqeOx/CSu9/xGPd+mH3vHOWEEJEOnvZp5iX/wBDj8Fzx4NYKb1D/h6S9CPAhg0buOOOOw7aFhcXx3vvvRemiIQQh7I/fgvz2n/DmIl4fnYfVq+O+fQrwzsi7ORnLSKZMQbzxsuYj97EmjwN66b/xIqJadNrZXhHCCG6EeP3Y/7nWcwXn2DNuADr6tlYno5dc0ySvhBChIGpr8Oe/wSs+Qrrh1dh/fDqTllmXJK+EEJ0MlNdif3so7BlA9Y1t+A588JOe2+5OKudli1bdtACbG2xcOFC9uzZ00ERCSG6E7O/DPt398G2zVg3/7JTEz5I0u8Ur732GkVFReEOQwgRZqao0LnKtqQIzx0P4plyeqfHIEm/FU888cRB694//vjj5ObmUl1dzc0338z06dO5/fbbaZwB9cwzz3DhhRdy1llncffdd2OM4b333mPNmjXcfvvtnHvuudTU1ITr2xFChJEp2Ir923ugrhbPXY9hje28pVuak6TfiquuuorXX38dcG7U/s477zBw4EDWrVvHQw89xOLFiykoKGDlypUAXH/99XzwwQd8+umn1NTUsGjRIi6++GImTpzIs88+y6JFi4iPjw/ntySECAOzaR32k/dDbByeex7HyjoubLF0i0LuX78uIm9fbUjbzErrxb+fdOTLmwGGDh1KWloa69atY+/evYwbN460tLQjLp+8bNkynn/+eWpqati/fz+jRo066vo3QoiezXyfi/2Hh6BvP2fhtLS+YY2nWyT9cLr66qvRWlNcXMxVV10FtLx8cm1tLffffz8ffPABgwcP5qmnnqKuri5cYQshugCzbZOT8Puk47nrUazUtKO/qIN1i6R/tB55R7rgggt48skn8fl8/OlPf2LFihUtHteY4Pv06UNVVRXvv/8+F110EQCJiYlUVlZ2WsxCiPAzBVuwfz8XklPx/GfXSPjQTZJ+OMXGxjJ16lRSU1OJijrylXKpqalcc801nH322WRkZDBx4sSmfUop7r33Xnr16sU777wj4/pC9HBmRx720w9CQqJTtA3zkE5zsvbOUdi2zfnnn89f/vKXprtHidDqKj9rIULB7NruFG1jYvH88tdYGQM67L0CWXtHZu+0YvPmzUybNo3TTz9dEr4Q4qjMnp3YT/8KoqKdMfwOTPiBOurwjlLqReBioFhrPd7dNhe4GdjrHna/1voDd999wE2AH7hDa/1RB8TdKUaOHHnEe8gKIURzprgQ+6lfgTF4/usRrP6Dwh1Si9oypv8S8CzwyiHbn9FaP9l8g1JqLHAVMA4YBPxLKTVSa+0PQaxCCNElmZIiJ+H7GvD816+xBg4Nd0hHdNThHa31EqCsje3NBF7VWtdprfOALcDJgQTWRWoNohPIz1p0Z6Zsr5Pwa2ucefiDh4c7pFYFM3vndqXUdcDXwF1a633AYGB5s2N2utvazePx4PP5iI6WCUY9mc/nw+OR0pLonsz+Muyn/h9UVeC58xGsYSPCHdJRBZpRnwceAYz79SngxvY0oJSaDcwG0FqTnp5+0H5jDGVlZfh8vgBDFN1BTEwM/fv375R1xIUIJXt/GWXz5mJ599F7zu+JHX18uENqk4CSvta6aclIpdQLQOPNVncBzQezhrjbWmpjPjDffWpKSkpafK/W5saL7s8YQ2lpabjDEKJdTIUX+6kHYO9uPD+fizd9IBwhh3WkxuVg2iOgz9VKqYHNnl4GrHMfvwNcpZSKU0plAccBXwXyHkII0RWZqkrsZ/4fFO/Gc/v/wxo5PtwhtctRL85SSi0AZgDpQBEwx30+CWd4Jx/4qdZ6t3v8AzhDPT7gF1rrD9sQx2EXZwkhRFdjqquwn3kQdubhue0BrPGTwxpPIBdnddkrcoUQoisxtdXOWjr53+O59T6siQFNTAwpuSJXCCE6gKmrxf7jI5C3Gc/Nv+wSCT9QkvSFEKIVpr4O+0+PwfcbsG68E2vy1HCHFBRJ+kIIcQSmoQH7+d/Axu+wrr8DzynZ4Q4paJL0hRCiBcbXgP2X38K6b7F+chueqWeFO6SQkKQvhBCHMH4/9gtPwZqvsK75KZ4zes5tTyXpCyFEM8b2Y/72NHy7DEvdhOfMi8IdUkjJwjZCCIG78N+6b7Df/l8o2IJ1+Sw8584Md1ghJ0lfCBHxzIY12G//A7ZuhL79sG76Tzynzgh3WB1Ckr4QImKZ73OdZL9pLaSlY/3bz7CmnY0VHRPu0DqMJH0hRMQxed9jv/13WL8KUnpjXXUz1vTzsWJiwx1ah5OkL4SIGGZHntOzX/MVJCVj/eh6rBkXYcXFhTu0TiNJXwjR45nC7Zh3FmC++QLiE7FmXot1zg+xeiWEO7ROJ0lfCNFjmeJCzLuvYlbkQGwvrIsU1rmXYiUmhTu0sJGkL4TocUxpMea9hZhln0B0NNZ5l2KdfwVWckq4Qws7SfpCiB7D7C/FvP8aZunHYIF15kVYF/wIKzUt3KF1GZL0hRDdnvHux3z4BibnQ7D9WNPOxbroSqw+GeEOrcuRpC+E6LZMbQ3mg9cwn7wLDQ1Yp52JdfGPsTIGhDu0LkuSvhCiWzLe/dh/eBi2b8WacgbWD6/CGjAk3GF1eZL0hRDdjikpwn5mDuwvwXPbr7AmTgl3SN2GJH0hRLdiduY796ptqMNz5yNYx44Jd0jdiiR9IUS3Yb7PxX72EYiNw3P341iDh4c7pG7nqElfKfUicDFQrLUe7257AvghUA9sBW7QWu9XSmUCG4BN7suXa61v6YjAhRCRxaz5Cvsvv4M+GXjufAirb79wh9QttaWn/xLwLPBKs22LgPu01j6l1G+B+4B73H1btdaTQhqlECKi2V98gnnljzD0GDw/n4OVnBrukLqto945S2u9BCg7ZNvHWmuf+3Q5ICVzIUSHsD96E/PSPBh1PJ7/elQSfpBCMaZ/I7Cw2fMspdQqwAv8Smu9NATvIYSIMMYYzOsvYT5+y5mSecMvsGJ67jr3nSWopK+UegDwAf9wN+0GhmmtS5VSk4F/KqXGaa29Lbx2NjAbQGtNenp6MKEIIXoQ4/Phfe5xaj/7gPgLriD5pl9gRUWFO6weIeCkr5S6HqfAe7bW2gBoreuAOvfxN0qprcBI4OtDX6+1ng/Md5+akpKSQEMRQvQgpq4Oe/7v4LuVWJdcQ93FP6Z+375wh9UlDRo0qN2vCSjpK6V+ANwNZGutq5ttzwDKtNZ+pdQxwHHAtkDeQwgReUxVpTMlc+tGrGtvxTPjgnCH1ONYxphWD1BKLQBmAOlAETAHZ7ZOHFDqHrZca32LUuoK4GGgAbCBOVrrd9sQhyksLAzoGxBC9AxmXyn2vLlQtAvPv9+FNXlauEPq8tyevtWe1xw16XcSSfpCRDCzZxf27+dAZQWe2+7HGjMx3CF1C4EkfbkiVwgRVib/e+x5D4Fl4fnlY1jDjw13SD2aJH0hRNiY3NXYz/0GkpLx3PkwVv/2FyZF+0jSF0KEhb3yc8zfnoYBg/H8Yi5W777hDikiSNIXQnQ6+7MPMAv+AiPG4PmPX2ElRO6NyjubJH0hRKcxxmDeXYB591WYeDKe2b/Eio0Ld1gRRZK+EKJTGNuPWTAfs/hDrKlnY113u1xlGwaS9IUQncJ886WT8M+/DOuK67Gsds00FCFy1FU2hRAiJNZ+DUnJWJfPkoQfRpL0hRAdzhiDyV2NNWYSlkfSTjjJ/74QouPtKoDyMhh3QrgjiXiS9IUQHc7krgLAGiM31Qs3SfpCiA5n1q+GgUOx+sh9M8JNkr4QokOZ+jr4fj2WDO10CZL0hRAda0suNNRjjZWkHyp1PpsvCg67IWGbyDx9IUSHMutXQ3Q0jBwX7lC6Nb9tWFdczeI8L19ur6DGZ3PlaaPb3Y4kfSFEhzK5q+DYsVhxvcIdSrdjjCFvXx2L88pZWlBBWY2PhBgPU4clk52VElCbkvSFEB3G7C+DnflYV8wKdyjdSlFlPTn5XnLyvOz01hPtgcmDksjOSuGkQUnERQc+Mi9JXwjRYcyGNQAynt8G3jo/XxR4ycn3smFvDQBjM+K59eT+TBuWQnJcaNYpkqQvhOg4uasgORWGZIY7ki6pzmezclcli/O8rNpdic+Goamx/GRiBtMzU+iXFBPy95SkL4ToEMa2naUXxsrSC835bcPaompy8sv5cnslNT6bPvHRXDyqD9mZKWSlxXXo2kRtSvpKqReBi4FirfV4d1sfYCGQCeQDSmu9TyllAfOAC4Fq4Hqt9behD10I0aXtzAfvfpChHYwxbGtWkN3nFmSnDU8mOzOFcf0SiPJ0ziJ0be3pvwQ8C7zSbNu9wCda68eVUve6z+8BLgCOc/+dAjzvfhVCRJCmpRfGRu7SC0WV9eTkOeP0hxZkpwxOIjaq8z8BtSnpa62XKKUyD9k8E5jhPn4ZWIyT9GcCr2itDbBcKdVbKTVQa707JBELIboFk7saBg/H6t0n3KF0Km+tjy+2V7A4z8vGEqcgO65fPJeMHsDUYckhK8gGKpgx/f7NEvkeoL/7eDCwo9lxO91tkvSFiBCmzl164ayLwx1Kp6jz2Xy1s5Kc/HK+LazCb2BYaiw/mZTB9OEdU5ANVEgKuVpro5Qy7XmNUmo2MNt9PenpshCTED1F3bfL2e/zkXpaNnE99Nz224Zvduzno017ydlSSk2Dn4ykWH584mDOG5XBsemJXfJmMcEk/aLGYRul1ECg2N2+Cxja7Lgh7raDaK3nA/Pdp6akpCSIUIQQXYn9ZQ5Ex+DtNwSrB53bxhi2ltWxOL+cz/O97Kv1kxjjYdqwQwuytZSW1nZ4PIMGDWr3a4JJ+u8As4DH3a9vN9t+u1LqVZwCbrmM5wsRWUzuKhg5Dis2LtyhhMSeinqW5HtZnO9ll7eeaI/FSYMTyc5M4aQwFWQD1dYpmwtwirbpSqmdwBycZK+VUjcBBYByD/8AZ7rmFpwpmzeEOGYhRBdm9pVC4XasqWeHO5SgeGt9fO4WZDe5Bdnx/eK5dMwApg5NJinMBdlAWca0ayi+o5jCwsJwxyCECAH7i08wL83DM+cPWN3sStw6n82KnZUsaVaQHd47juzMFKZnppCR2HUKstA0vNOuwoFckSuECK3cVZCaBoOHhzuSNvHbhu+KqlmcV87yHZXU+mz6JkQzc4xzhWxmWs9aHVSSvhAiZJqWXjh+cpecudLIGMOWslpy8rwsLfCy3y3InjE8mRlZqYztF4+nC8cfDEn6QojQ2bENKr1ddumF3RUHliwurHAKslMGJ5KdmcrkwYndqiAbKEn6QoiQMesbl16YGOZIDiiv9fF5QQU5+eVsKnGmUY7vn8BlY/t064JsoCTpCyFCxuSuhqFZWClpYY2j1mezYkcFOfleVu2uwjaQ2TuOWZMyOKMLFmQ7kyR9IURImNoa2LIB65xLwvL+ftuwZk8VOXlelu+soNZnSE+I5tIeWpANlCR9IURobF4Hfh/WuM4bzzfG8H1pLTn5TkG2vNZPYqyH6ZkpZGf27IJsoCTpCyFCwuSuhthYOHZMh7/X7orGJYvLKaxocAuyjfeQTSQmAgqygZKkL4QICbN+FYwcjxUT2yHt76/18XmBM/Nmc2ktFk5B9vKxfTltWDJJsZFVkA2UJH0hRNBM6V7YsxNr+vkhbbfWZ7N8RwVLmhVks9LimHWCcw/Z9ITILcgGSpK+ECJoB+6SFfx4vt82rN5dRU6+lxVuQTYjIZrLxvQhOyuV4b17xiJu4SJJXwgRvNzV0LsvDBp69GNbYIxhs1uQ/dwtyCbFesjOTCU7M4UxUpANGUn6QoigGNuP2bAGa9LJ7V56odBbT05+OTn5XnZXNBDjsZgyJIkZmSmcKAXZDiFJXwgRnIKtUFXR5qUX9tf4WFrg3Cz8e7cge3z/BH40ri+nDU0mUQqyHUqSvhAiKGb9KrAsrDGTjnhMTYPNip0V5OR5Wb3nQEH2hhMzOGN4Cn2lINtpJOkLIYJiclfBsBFYySkHbfc1L8juqKDOb+iXGM3lY/uSnZXCsFQpyIaDJH0hRMBMTTVs24R13mXO88aCbF45nxdUUF7nJznWw5nHOAXZ0RlSkA03SfpCiMBtWgt+P4UjTmDJd3vJyfOyp7KB2KgDV8ieODCJmChJ9F2FJH0hRED21/jIWV/MkpPuYMu6GCxKOX5AAmq8c4VsQowUZLsiSfpCiDaraXCukM3J97JmTxU2x5HVax83ntiP04cnS0G2G5CkL4RoVVNBNs+5QtYpyMZwRWYcZyx8jKGXzMQz5rRwhynaKOCkr5QaBSxstukY4EGgN3AzsNfdfr/W+oOAIxRCdLrGguxityDrdQuyZzUryJqlH2Gqi0Oy9ILoPAEnfa31JmASgFIqCtgFvAXcADyjtX4yJBEKITrNTm8dOXleluQfKMiePCSJ7MwUTjikIGuvXw190mHA4DBGLNorVMM7ZwNbtdYFSqkQNSmE6Az7Gq+QzfOypawWj+VcIdtaQdb4/bBxDdbkae1eekGEV6iS/lXAgmbPb1dKXQd8Ddyltd4XovcRQoRAdYOf5Tsqycn38p17heyIPnHceGI/zshMoU/8UVJD/vdQXYU19shX4YquKeikr5SKBS4B7nM3PQ88Ahj361PAjS28bjYwG0BrTXp6erChCCFa4fPbrCjYz8ebilm6rYw6n82glDh+MmUo54/KYHifhDa3VfnJJqosi77TzsJzyJW4omsLRU//AuBbrXURQONXAKXUC8B7Lb1Iaz0fmO8+NSUlJSEIRQjRnDGGjSU15OR5+WK7W5CNi+KsrBSys1IYnR7vDM/Y1ZSUVLe5Xf/KLyDzOMrq6qFOzt1wGTRoULtfE4qkfzXNhnaUUgO11rvdp5cB60LwHkKIdthZXkdO/sEF2VOGJJGdmcqkgYlBXSFrqishbzPWBT8KYcSiswSV9JVSicC5wE+bbf6dUmoSzvBO/iH7hBAdpKzGx9J8Z8nirW5BdsKARH58fDqnDk0K3RWyG9eCbctUzW7KMsaEOwYAU1hYGO4YhOh2mgqyeeV8V1TtFmR7MSMrhdOHt6EgGwD7789hlufg+f0/sKLl+s5wcod32vWxTX5iQnQzDX7Dqt3OzJuvdlZS7zf0T4rhR+P6kp2ZwpAOXrLY5K6G0cdLwu+m5KcmRDdgjGHj3hrnHrLbK6hwC7LnjEglOzOVUem9OmW+vCneDXv3YJ07s8PfS3QMSfpCdGE7yp0rZHPyvRRXHSjIzshyCrLRns69MMrkrgKQ8fxuTJK+EF1MaXUDnxdUsDivnG376poKstdMSOeUUBZkA2DWr4a+/aDfwLDFIIIjSV+ILqC6wc+X2ytYnO9l7Z5qDHBsn178++R+nD48hbQOKMi2l/H5YNN3WFOmy9IL3Vj4f5OEiFANfsO3uyvJyfOycpdTkB2QFIM6vi/TM1MYktLF7iGbtxlqqrHGydIL3ZkkfSE6kd2sIPtFgZeKepsUtyA7IyuVkX07pyAbCJO7CiwPjJ4Q7lBEECTpC9EJtpc3LllcTnGVj9goi1OHJJOdlRKWgmwgzPpVcMxIrISkcIcigiBJX4gOUlrdwNICL4vzvOS5BdmJAxK5ZkJG2Auy7WWqKiB/C9bFsnR6dydJX4gQqqr38+WOCnLyvKwtcgqyx/V1CrJnDE+hdxcoyAZk43dgZOmFnqCb/gYK0XU0+A3fFlayON/Lyp2VNNgHCrLZmakMTokNd4hBM+tXQXwCZI0MdygiSJL0hQiAbQwb9jYuWeylst4mNS6K845NJbuLF2TbyxjjLr0wASuq+wxJiZZJ0heiHbbvd5YszskrZ2+1j7goi1OGJjMjM4WJ3aQg225FhVBajPWDK8IdiQgBSfpCHEVpdQNL3CWLGwuykwYk8m+TMjhlSDLxMZ5wh9ihmpZeGCfj+T2BJH0hWtBjC7IBMLmrIWMAVsaAcIciQiByfnOFOIoGv803hVXkNCvIDsCZIk8AABsNSURBVEyO4cduQXZQDyjItpfxNcDGtVinzQh3KCJEJOmLiGYbw4Zi9wrZ5gXZ43ozIzOF43pQQTYgWzdBXY0M7fQgkvRFRCrYX0dOXjlL8r1NBdlThyaT3ZMLsgEwuavA44FRsvRCTyFJX0SMErcgu6RZQfaEgYn8ZFIGpwxNpld0zy7IBsJZemE0VnxCuEMRISJJX/RolfXOksU5+V7WuQXZkX17Mfuk/kwbnkzvXnIKHImp8ML2rViXXB3uUEQIyW+86HEa/DZfF1aRk+fl611OQXZQcgxXTUgnOzOFgcmRV5ANhNm4BoyRpRd6mKCTvlIqH6gA/IBPa32SUqoPsBDIBPIBpbXeF+x7CXEktjHkFteQk1/OF9srqKq3Se0VxQ+O6012VgrH9onwgmwg1q+ChETIPDbckYgQClVP/0ytdUmz5/cCn2itH1dK3es+vydE7yVEk/x9teS44/Ql1T56RTcryA5IJEoKsgFpWnphzEQsjyy90JN01PDOTGCG+/hlYDGS9EWI7K1qYKl7hWz+/gMF2Vkn9OPkIUlSkA2QMQaKdmG2boRN62BfCdbYH4c7LBFioUj6BvhYKWWAv2it5wP9tda73f17gP4heB8RwSrr/SxzC7Lr3YLsqPR4Zp/Un9OHJ5MqBdl2M/V1kL8Fs3WDk+i3boDKCmdnQhLW5GlYJ50e3iBFyIXiTDlda71LKdUPWKSU2th8p9bauH8QDqKUmg3Mdo8hPT09BKGInqTeZ7Msv4yPN+5lWX4ZDX7D0N7x3HTqMM4dlcGQ3vHhDrFb8ZeV0LBpLQ0bvqNh0zoatm0Cnw+AqEHDiDl5OjGjjyd29ASiBg/D8sgnpp7IMuawfBwwpdRcoBK4GZihtd6tlBoILNZaj2rlpaawsDBkcYjuyzaG9cXV5OR5Wba9gqoGm969ojgjM4XsTCnItpWx/bBrO2bLBmjsyZcUOTtjYiHzWKwRY7BGjIYRY7CSU8IbsAjIoEGDANp1QgTV01dKJQIerXWF+/g84GHgHWAW8Lj79e1g3kf0fI0F2Zx8L6XVPnpFezhtaBLZWalM6J8gBdmjMDXVkLcJs8VN8Ns2QW2NszM1zUnsZ16EdewYGHYMVnRMeAMWYRPs8E5/4C2lVGNb/6u1/j+l1EpAK6VuAgoAubGmOMzeKnfJ4jwvBeV1RFlw4qBErj+hH6cMSSJOCrKHMQ0NULQTs2s77CrAFDpfm3rxlgWDM7FOneEk+hGjIb2/fDoSTUI6vBMEGd6JEJV1fpbtqCAnr5x1xU5PdFR6PDOyUjh9WDIpUpAFwPj9sHe3M0SzqwBTWAC7tkNxIdi2c1BUFPQfjDV4OAwahnXMSMgaJUsmRJBOH94Roi3q/TZf76okJ9/L17uq8NmGwSmxXONeITsggq+QNbYNZXud5F5Y4PTed22HPTvB1+AcZFmQMQAGDceaPBUGD8caNAz6D5JhGtFukvRFh7CNYV1RNTn5Xr50C7JpvaK4cGRvsjNTGdEnLmKGHIxtg3cflJXAvhJMaTEU7nCGZgp3QF3NgYP7pDvJfdwk5+vgYTBgKFZcXPi+AdGjSNIXIWOMIX9/HTl5zhWypTVOQXbqsCSyM1M5vgcWZI0xUFXhJPSyvZh9Je7jEsy+vc7j/aXg9x/8wuRUp8d++jnO0MygYc7XhMTwfCMiYkjSF0HbW9XgLIVwUEE2iRsyUzj5CAVZ4/OBsSE6psv2+I1tOzNgysugtFlC37cXU3bgMfX1B78wKhrS+kKfdGe2TJ8M53Ga85U+6ViJyeH5pkTEk6QvAlJR13iFbDnr3YLs6PR4bpnSn2lHKMgaXwOsX41ZuQSz+itnWCMqGuITmv1LhF7xWPGJB2/v5Xy1EhKbHh/YF3/Y+jCmoQFqqqCmGmqroboKamswjdua/h1he221k/APnehgWc4UyLR0GDIca8JJhyT0DEhOlQubRJclSV+0Wb3fZuWuSnLyvHxTWInPhiEpsVw70SnI9k86vCBr/H7Y9B3mq6WYVV86yTchCevkM5ziZLNkaxqTcOleTE2+k3Rrqg7MVsFZ86NFcfEQH+8Mo9RUNV1p2qqY2IP+0JCQCKlpWIf+UUntg5Xm9NDp3RcrWk4b0X3Jb69old92r5DNd66QrW6wSYuP5qKRaWRnpXJM2uEFWWPbsCUXs3Ip5ptlUFHu9MYnneok+zET2zzrxBgD9XWH9MzdPxJu7935w+F+jY4+8EnA/VRgHfZJIgHi42Xmi4hIkvTFYYwx5O2rIyffy1K3IBsf7eG0Yc6SxS0VZI0xkLfZSfRff+EUL2NjsSae4izadfxkrJj2T820LAviejn/evc5eF9Q36UQkUmSvmhSXOlcIbs4v5wd5fVEWTB5cBI3ZqYwZfDhBVljDOzIcxL9yqVQWuz0tMefhDXlBqwJU7B6yaJoQnQlkvR7GOPzOQXSXglYUUe/+UVFnZ8vtjtLIeTudQqyYzPcguzwFFLiDm/D7N7hjNF/vRT27HKuDB0zCeuSq50hHJl2KESXJUm/hzB7dmGWfoxZ9glUep2NsXGHzYohPpH6+CS+jhvMEs9AvvWl4sNiSKyPawcapg+IoX9aFMTbQAPGeLAsC1O82x26+Rx25juzWEYdj3XuTKwTpsoqjUJ0E5L0uzHT0IBZ9SVmyUewaa3T4554ijM3vLamqeBJTTW+mmrW2yksMZks94yg2o4jrcbLhcVLmF60iqzKwqYxcrv5m3g8zh+L6irn+bFjsK6ajTV5KtYhY+xCiK5Pkn43ZPbsbNarr4CMAViXX4c19Wys1LQDxzUryC7J91JW4yMhxsNpQ5PJzkphXHocUfVjoebyg2fFNJ8p0zgrJr0f1uTTsfpmhPE7F0IES5J+N2EaGjDfLnN69ZvXOb36SafgmX4+jJ540MVARZX1TkE2z8tObz3RHpg8KInszBROOrQgG50ECUnAgWQus2KE6Lkk6XdxZvdOzNKPMF9+2qxXPwtr2llYKQd69d46P18UODch2dCsIHvryf2ZNiyF5BYKskKIyCNJvwsyDfWYb79soVf/Axg9oalXX+dzrpBdnOdl1W7nCtmhqbH8ZGIG0zNT6JckFx8JIQ4mSb8LMbt3Ypa4vfqqlnv1ftuwdncVOfnlfLm9khqfTd/4aC4e1YfszBSyWrhCVgghGknSDzPTUI/5Zhlm6UeweT1ERTlz3aef39SrN8awtayWnLxylhRUsM8tyE4b7lwhO65fz1uyWAjRMSTpdwJj+6F8v3OHpH0l7rK87lK9m9Yd6NVfMQtr6oFefVFlPTl5zjj9QQXZLOcK2dgoWclRCNE+kvSDdOAmGnvdG2eUHLihhnunpBZvohHXC9LSscZOwjrjPOdCJ48Hb62PLzbvY3Gel40lTkF2XL94Lhk9gKnDkqUgK4QIiiT9NjC+Bti+DbOroJ030cjAOm6ss/Z645rrfdMhLQMSEpvG3ut8Nl9tryQnv5xvC6vwGxiWGstPJmUwfbgUZIUQoRNw0ldKDQVeAfrjLHM+X2s9Tyk1F7gZ2Oseer/W+oNgA+1MpsIL2zZitmzAbN0A+VugwU3sjTfR6JOBNSQTDrqJhnsjjTbcRMNvG9bucQqyy7ZXUusWZC8Z3YfsrBQye0tBVggResH09H3AXVrrb5VSycA3SqlF7r5ntNZPBh9exzO2DUW7MFs2wNYNmC0boWiXszMqGoYdg5V9gbO0wbBjnCGZAG+i4RRk61icX87n+V721fpJjPFw+vBkZmQ5BVmPJHohRAcKOOlrrXcDu93HFUqpDcDgUAXWUUxdHeR/j9mSi9m6EbZtcsbkAZKSYcQYrGnnYI0YDZnHYsXGBf2eeyrqycl3CrK7vPVEeyxOGpzYdIWsFGSFEJ3FMofeAzQASqlMYAkwHvhP4HrAC3yN82lg31GaMIWFhUHH0WLD+0oP9OK3boQd2w4UVQcOdZL7sWOcr/0Hh2xIpbzWx+cFFeTke9nkFmTH94snOyuVqUOTSZKCrBAiSIMGDYJ2rpwSdNJXSiUBOcBjWus3lVL9gRKccf5HgIFa6xtbeN1sYDaA1npy/aHF0ACZhgbqVi6lbnkO9Ru/w95b5OyIjSPmuLHEjj6emNHHEzNyPJ6U1JC8Z6PaBj9Lt5Xx8cZiVmzfj982jOibwHmj+3HOyHQGpPQK6fsJISJbbGwsdGbSV0rFAO8BH2mtn25hfybwntZ6/FGaCrqnb4oLMUsXYb74l3NP1tQ0rGPHur34MTA0q0NuaO23DWv2VJGT72X5DrcgmxBNdmYK2ZkpZKZJohdCdIxAevrBzN6xgL8BG5onfKXUQHe8H+AyYF2g73E0xtcAq1dgL/kINqxx1n6fcLKz8uS4SViejhlCMcawpayWnDwvSwu87HcLsmcMT2ZGVipj+8VLQVYI0SUF0/WdBvwEWKuUWu1uux+4Wik1CWd4Jx/4aVARtuCwXn2fDKyZ1zoF2LS+oX67JrsbC7J5XgornILslMGJZGemMnlwohRkhRBdXkgKuSFw1OEd42vArFrhrFHTvFeffT6M7bhe/YGCbDmbSmoBGN8/gezMFCnICiHCqlOHdzqLKS7ELHHvElVRDn37YV36b1jTzsbq3TG9+lqfzYodzsybVbursA1k9o5j1qQMzshMISNRrpAVQnRPXTLpN/Xql/wfbPzO6dVPdMfqO6hX31SQzfOyfGcFtT5DekI0l47pIwVZIUSP0aWSvikqdO4StezTQ3r153TITbiNMXxfWktOvlOQLa/1kxjrYXpmCtmZUpAVQvQ8XSbp+5/61YFe/aRT8JzR2KsPfXF0d0XjksXlFFY0uAVZZ8nikwYlEiMFWSFED9Vlkj5793Ror35/rY/PC5yZN5tLa7FwCrKXj+3LacOSSYqVgqwQoufrMrN3du3cGfJefa3PZvmOCpYcUpDNzkphemYK6QlSkBVCdF/devZOqBK+3zas3u1cIbuiWUH2sjF9yM5KZXjv4BdQE0KI7qrLJP1gGGPY7BZkPz+kIDsjM5UxUpAVQgigmyf9Qm89Ofnl5OR72V3RQIzH4qTBSczISmGyFGSFEOIw3S7p76/xsbTAWZv++2YF2R+N68upQ6UgK4QQrekWSb+mwWbFzgpy8rys3uMUZLPS4rj+BOcKWSnICiFE23TZpO9rXpDdUUGd35CREM3lY/uSnZnCMCnICiFEu3WppN9UkM0r5/OCCsrr/CTFepiRlUp2VgpjMqQgK4QQwegySf9/v9tLTp6XPZVOQfbkIUlkZ6ZwohRkhRAiZLpM0tdrSzl+QAJXju/LaUOTSZSCrBBChFyXuSJ37ZYC+kpBVggh2iyQK3K7zLiJJHwhhOh4XSbpCyGE6HiS9IUQIoJI0hdCiAgiSV8IISKIJH0hhIggkvSFECKCSNIXQogIIklfCCEiSJe5IjfcAQghRDfV/a7IVUp9gxN4yP4ppV6QNqVNabNrtdkdYuxmbX5DO3WJpN9B3pU2pU1ps8u12R1i7E5ttluXGN5RSn2ttT4p3HEIIUR3Ekju7Co9/fnhDkAIIbqhdufOLtHTD5ZS6kXgYqBYaz3e3XYlMBcYA5ystf46BG0+AfwQqAe2AjdorfcH2eYjwEzABoqB67XWhcG02WzfXcCTQIbWuiTIOOcCNwN73cPu11p/EGycSqn/AG4D/MD7Wuu7g4xzITDKPaQ3sF9rPSnINicBfwZ6AT7gZ1rrr4Job6LbXhKQD1yrtfa2I8ahwCtAf5xJEPO11vOUUn2AhUCm267SWu8Lss2Az6NW2gz4PGqlzYDPoyO12Wx/u8+jVuKcSxDnUSh0lZ5+sF4CfnDItnXA5cCSELa5CBivtZ4AbAbuC0GbT2itJ7iJ6T3gwRC02fhLdx6wvZ3tHbFN4Bmt9ST3X3t/UQ9rUyl1Js6JOlFrPQ7nxAqqTa31jxtjBN4A3gy2TeB3wENumw+6z4Np76/AvVrr44G3gF+2M0YfcJfWeixwKnCbUmoscC/widb6OOAT93mwbQZzHh2pzWDOoyO1Gcx5dKQ2gzmPjtgmwZ1HQev0pK+UelEpVayUWtdsWx+l1CKl1Pfu17T2tKm1XgKUHbJtg9Z6U6BxHqHNj7XWPvfpcmBICNps3sNLpJ3TV1tq0/UMcHd72ztKmwE7Qpu3Ao9rrevcY4pD0CYASikLUMCCELRpgBT3cSrQ5k9iR2hvJAeS6CLginbGuFtr/a37uALYAAzG+QP6snvYy8ClwbYZzHnUSpsBn0ettBnwedTK/ycEeB4dpc2AKKWGKqU+U0rlKqXWK6V+3mzffyilNrrbW+2UhKOn/xKH93yC6aGEy43Ah6FoSCn1mFJqB3At7e/pt9TeTGCX1npN0MEd7Hal1HfuH+52/WE+gpHAGUqpFUqpHKXUlBC02egMoEhr/X0I2voF8IT7M3qS9n/CO9R6nAQNcCUwNNCGlFKZwAnACqC/1nq3u2sPztBCsG2GRCttBnweHdpmKM6j5m2G6jxq4XsP9Dxq8dNDez8xd3rSP0LPJ+AeSjgopR7A+QH8IxTtaa0f0FoPddu7PcjYEoD7CcEfj0M8D4wAJgG7gadC0GY00AfnF/iXgHZ76KFwNe3s5bfiVuBO92d0J/C3INu7EfiZO8c6GWdsu92UUkk4Q1i/OLQmoLU2BPApr7U2A3WkNoM5j1pqM9jzqHmbblxBn0ctxBnwedTKp4d2fWLuKmP6IemhdAal1PU4hblr3RMrlP5BOz/qt2AEkAWsUUrl43x0/lYpNSCYRrXWRVprv9baBl4ATg4yToCdwJtaa+MWRm0gPdhGlVLROOPQC4NtyzWLA7WB1wjye9dab9Ran6e1nozzh2lre9tQSsXgJJN/aK0bYytSSg109w/EKWgG22ZQjtRmMOdRG+Js93nUQptBn0ctxRmq8+iQTw/t+sTcVZJ+k0B7KJ1BKfUDnPG9S7TW1SFq87hmT2cCG4NpT2u9VmvdT2udqbXOxEmsJ2qt9wTTbmMycV2GU+AL1j+BM932RwKxQJtnGbXiHGCj1npnCNoCZww/2318FhDUkJFSqp/71QP8CmcmT3teb+F82tigtX662a53cP5A4X59OwRtBuxIbQZzHrXSZsDnUUttBnsetRJn0OdRC58e2vWJOSxTNt2/Uu/pA1PYNgEztNa73f+UxVrrUa21cUh7C4AZOL3EImAOzhDSH4EMYD+wWmt9fpBt3gfEAaXuYcu11rcE2eaFOFMMbaAAuEVrvSuYNrXWf2u2Px84SbdvymZLcc7A+UhqcKYD/rTZp7NA2/wf4EW33Xrgv7TWnwbTptb6b0qpl3B+Nu1Kpq3EuQmYh3Ny1eJM2WzT5e9HaC8JZ5oqOJ8g7mtPb1cpdTqwFFiL83sDzlDECkADw3B+l5TWuk0F+VbajCPA86iVNv9AgOdRK23eRIDn0ZHabD6zpr3nUStxXk1w51EMzuykjxr/mCil/g/4rdb6M/f5VuBUrfXeltroKkn/CaBUa/24UupeoI9ux3xtIYTo6dze+8tAmdb6F8223wIM0lo/6H5i/gQYdqSORKcn/SP0fP5JgD0UIYSIBK18evgX7fjE3COuyBVCCNE2Xa6QK4QQouNI0hdCiAgS3RlvopQyOHNV/819Ho1zYcIKrfXFnRGDEEKIzuvpVwHjlVLx7vNzgTZPS4SmPxRCCCGC0JmJ9APgIuB1DlwifwaAUupknHnQvYAanKVWN7lX7V2OM685igMXyAghhAhAZ47pvwpcpZTqBUzg4IWXNgJnaK1PwFnr4tfN9p0I/EhrLQlfCCGC1GlJX2v9Hc7NHa7G6fU3lwq85i63/Awwrtm+RTJnXwghQqOzZ++8g7Ps56GrHz4CfOZeoftDnGGeRlWdFJsQQvR4nZ30X8S5A9HaQ7ancqCwe32nRiSEEBGkU2fEuKse/qGFXb8DXlZK/Qp4vzNjEkKISCLLMAghRASRK3KFECKCSNIXQogIIklfCCEiSIcUcpVSQ4FXcO51a4D5Wut5Sqk+OPctzcS5a4zSWu9TSl0L3ANYQAVwa+Md6N1bq83DuSL3r1rrxzsiZiGEiAQd1dP3AXdprcfi3LfxNqXUWOBe4BOt9XE4d3e51z0+D8jWWh+PM2d/PoBSKgr4E3ABMBa42m1HCCFEADqkp+/e83G3+7hCKbUBGIxzw+IZ7mEvA4uBe7TWy5q9fDnOnefBuVP8Fq31NgCl1KtuG7kdEbcQQvR0HT6m794P9wSctXb6N7sJ8B6c4Z9D3QR86D4eDOxotm+nu00IIUQAOjTpK6WSgDeAX2itvc33uTftNYccfyZO0r+nI+MSQohI1WFJXykVg5Pw/6G1ftPdXKSUGujuHwgUNzt+AvBXYKbWutTdvAsY2qzZIbRzHX4hhBAHdNTsHQv4G7BBa/10s13vALOAx92vb7vHDwPeBH6itd7c7PiVwHFKqSycZH8VcE1HxCyEEJGgQ5ZhUEqdDiwF1gK2u/l+nHF9DQwDCnCmbJYppf4KXOFuA/BprU9y27oQ+D3OlM0XtdaPhTxgIYSIELL2jhBCRBC5IlcIISKIJH0hhIggkvSFECKCSNIXQogIIklfCCEiiCR9IYSIIJ16j1whwkkplY+z3pMP8OMs3PcKztLfdisvbVxDKg+I0Vr7OjZSITqO9PRFpPmh1joZGI5zZfg9OFePCxERpKcvIpLWuhx4Rym1B1iulHoK5w/Bo8AIoBz4m9Z6rvuSJe7X/UopgHO11l8qpW4EfgkMAL4CZmutCxCii5KevohoWuuvcJbsPgOoAq4DegMXAbcqpS51D53ufu2ttU5yE/5MnOVFLgcycJYeWdCZ8QvRXtLTFwIKgT5a68XNtn2nlFoAZAP/PMLrbgF+o7XeAKCU+jVwv1JquPT2RVclSV8I58Y8ZUqpU3DG+ccDsUAc8ForrxsOzHOHhhpZbnuS9EWXJElfRDSl1BScJP05To/+WeACrXWtUur3QLp7aEsrE+4AHtNa/6NTghUiBGRMX0QkpVSKUupi4FXg71rrtUAyUOYm/JM5+N4Ne3GWCT+m2bY/A/cppca5baYqpa7snO9AiMBI0heR5l2lVAVOL/0B4GngBnffz4CH3f0P4tz7AQCtdTXwGPCFUmq/UupUrfVbwG+BV5VSXmAdcEHnfStCtJ+spy+EEBFEevpCCBFBJOkLIUQEkaQvhBARRJK+EEJEEEn6QggRQSTpCyFEBJGkL4QQEUSSvhBCRBBJ+kIIEUH+P57ziIDeavtqAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_3WYU3uraZOQ"
      },
      "source": [
        "#**7- Forecast quality evaluation** \n",
        "Let's evaluate the quality of the algorithm by calculating the error metrics for the last 30 days that we predicted. For this, we will need the observations  yi  and the corresponding predicted values  y^i ."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "V294mk5-PUcV"
      },
      "source": [
        "def make_comparison_dataframe(historical, forecast):\n",
        "    \"\"\"Join the history with the forecast.\n",
        "    \n",
        "       The resulting dataset will contain columns 'yhat', 'yhat_lower', 'yhat_upper' and 'y'.\n",
        "    \"\"\"\n",
        "    return forecast.set_index('ds')[['yhat', 'yhat_lower', 'yhat_upper']].join(historical.set_index('ds'))"
      ],
      "execution_count": 40,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7rkQB64bPdTp"
      },
      "source": [
        "cmp_df = make_comparison_dataframe(df, forecast)\n",
        "cmp_df.tail(n=3)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hHZkxBo2PvyI"
      },
      "source": [
        "def calculate_forecast_errors(df, prediction_size):\n",
        "    \"\"\"Calculate MAPE and MAE of the forecast.\n",
        "    \n",
        "       Args:\n",
        "           df: joined dataset with 'y' and 'yhat' columns.\n",
        "           prediction_size: number of days at the end to predict.\n",
        "    \"\"\"\n",
        "    \n",
        "    # Make a copy\n",
        "    df = df.copy()\n",
        "    \n",
        "    # Now we calculate the values of e_i and p_i according to the formulas given in the article above.\n",
        "    df['e'] = df['y'] - df['yhat']\n",
        "    df['p'] = 100 * df['e'] / df['y']\n",
        "    \n",
        "    # Recall that we held out the values of the last `prediction_size` days\n",
        "    # in order to predict them and measure the quality of the model. \n",
        "    \n",
        "    # Now cut out the part of the data which we made our prediction for.\n",
        "    predicted_part = df[-prediction_size:]\n",
        "    \n",
        "    # Define the function that averages absolute error values over the predicted part.\n",
        "    error_mean = lambda error_name: np.mean(np.abs(predicted_part[error_name]))\n",
        "    \n",
        "    # Now we can calculate MAPE and MAE and return the resulting dictionary of errors.\n",
        "    return {'MAPE': error_mean('p'), 'MAE': error_mean('e')}\n"
      ],
      "execution_count": 42,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6lpRKrLhP4NQ",
        "outputId": "ac59aaa7-0ac1-41d7-9c17-d19c296134f7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "for err_name, err_value in calculate_forecast_errors(cmp_df, prediction_size).items():\n",
        "    print(err_name, err_value)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "MAPE 34.774518380154696\n",
            "MAE 55.0657164657149\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HLConDmcvWW-"
      },
      "source": [
        "#**8- RMSE**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3mn74pRGqITN"
      },
      "source": [
        "se = np.square(forecast.loc[:, 'yhat'] - df['y'])\n",
        "mse = np.mean(se)\n",
        "rmse = np.sqrt(mse)"
      ],
      "execution_count": 46,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "P3nGmOylu3nv",
        "outputId": "d4b1e353-79a6-4435-99a0-c41b86a7ddbb",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "print(rmse)"
      ],
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "29.077235826002735\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QmrbiUI8sjn8"
      },
      "source": [
        "# **References** "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KZy09JqTtWPt"
      },
      "source": [
        "#[1]Forecasting Time Series Data With Prophet I & II\n",
        "# https://nextjournal.com/eric-brown/forecasting-with-prophet \n",
        "#[2]Forecasting Time Series Data With Prophet IV\n",
        "#https://nextjournal.com/eric-brown/forecasting-with-prophet-part-4\n",
        "#[3]Time series analysis in Python\n",
        "#https://mlcourse.ai/articles/topic9-part2-prophet/\n",
        "#https://www.kaggle.com/kashnitsky/topic-9-part-2-time-series-with-facebook-prophet\n",
        "#[4] Covid-19 Deaths Predictions (Time Series - Prophet) - DIY\n",
        "#https://www.youtube.com/watch?v=zlFKgMnaDXk&feature=youtu.be&fbclid=IwAR3vSH3rUeeQRvhxSzADMlqK4CY06GaSbJSaeq-mAYfHXZb9E0M8Dlor4wU\n"
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}