LSTM_TensorFlow_WeeklyPrecipitation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "8da86ec5",
   "metadata": {
    "executionInfo": {
     "elapsed": 2734,
     "status": "ok",
     "timestamp": 1625317288370,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "8da86ec5"
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from sklearn import preprocessing\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras import Sequential, layers, callbacks\n",
    "from tensorflow.keras.layers import Dense, LSTM, Dropout, GRU, Bidirectional\n",
    "import numpy as np\n",
    "\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "xL6Y-IS3ksaQ",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 23123,
     "status": "ok",
     "timestamp": 1625317311488,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "xL6Y-IS3ksaQ",
    "outputId": "e3563886-d5db-4bcd-a12d-3e2895486e1e"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"\\n#connect with drive\\nfrom google.colab import drive\\ndrive.mount('/content/drive')\\n\""
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"\n",
    "#connect with drive\n",
    "from google.colab import drive\n",
    "drive.mount('/content/drive')\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "p7MwJk2JksoL",
   "metadata": {
    "executionInfo": {
     "elapsed": 4,
     "status": "ok",
     "timestamp": 1625317311488,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "p7MwJk2JksoL"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"\\n# Path to data:\\n# David's path\\npath = '/content/drive/MyDrive/Dengue_GIS_Visualization/DengueData/'\\n# Dana's path\\n#path='/content/drive/MyDrive/Dengue_GIS Visualization/Dengue_GIS_Visualization/DengueData/'\\n\""
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"\n",
    "# Path to data:\n",
    "# David's path\n",
    "path = '/content/drive/MyDrive/Dengue_GIS_Visualization/DengueData/'\n",
    "# Dana's path\n",
    "#path='/content/drive/MyDrive/Dengue_GIS Visualization/Dengue_GIS_Visualization/DengueData/'\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "lxt7_uXJKXrK",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 609
    },
    "executionInfo": {
     "elapsed": 790,
     "status": "ok",
     "timestamp": 1625317313463,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "lxt7_uXJKXrK",
    "outputId": "1e119a71-292e-4244-c08b-ca89bd72a557"
   },
   "outputs": [
    {
     "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>cases_medellin</th>\n",
       "      <th>DATA</th>\n",
       "      <th>YEAR</th>\n",
       "      <th>YearWeek</th>\n",
       "      <th>LastDayWeek</th>\n",
       "      <th>MONTH</th>\n",
       "      <th>temperature_medellin</th>\n",
       "      <th>percipitation_medellin</th>\n",
       "      <th>percipitation_medellin_rural</th>\n",
       "      <th>percipitation_medellin_urbano</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>['2007', 'w01']</td>\n",
       "      <td>2007</td>\n",
       "      <td>200701</td>\n",
       "      <td>2007-01-06</td>\n",
       "      <td>1</td>\n",
       "      <td>21.009283</td>\n",
       "      <td>13.096688</td>\n",
       "      <td>14.097693</td>\n",
       "      <td>12.628933</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>['2007', 'w02']</td>\n",
       "      <td>2007</td>\n",
       "      <td>200702</td>\n",
       "      <td>2007-01-13</td>\n",
       "      <td>1</td>\n",
       "      <td>21.009283</td>\n",
       "      <td>8.437044</td>\n",
       "      <td>6.910390</td>\n",
       "      <td>9.067214</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>['2007', 'w03']</td>\n",
       "      <td>2007</td>\n",
       "      <td>200703</td>\n",
       "      <td>2007-01-20</td>\n",
       "      <td>1</td>\n",
       "      <td>21.009283</td>\n",
       "      <td>19.373165</td>\n",
       "      <td>21.616968</td>\n",
       "      <td>18.406665</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>['2007', 'w04']</td>\n",
       "      <td>2007</td>\n",
       "      <td>200704</td>\n",
       "      <td>2007-01-27</td>\n",
       "      <td>1</td>\n",
       "      <td>21.009283</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>['2007', 'w05']</td>\n",
       "      <td>2007</td>\n",
       "      <td>200705</td>\n",
       "      <td>2007-02-03</td>\n",
       "      <td>2</td>\n",
       "      <td>20.982925</td>\n",
       "      <td>3.725206</td>\n",
       "      <td>3.144225</td>\n",
       "      <td>3.901021</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>619</th>\n",
       "      <td>28</td>\n",
       "      <td>['2018', 'w48']</td>\n",
       "      <td>2018</td>\n",
       "      <td>201848</td>\n",
       "      <td>2018-12-01</td>\n",
       "      <td>12</td>\n",
       "      <td>20.936353</td>\n",
       "      <td>8.916642</td>\n",
       "      <td>8.656351</td>\n",
       "      <td>9.114419</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>620</th>\n",
       "      <td>12</td>\n",
       "      <td>['2018', 'w49']</td>\n",
       "      <td>2018</td>\n",
       "      <td>201849</td>\n",
       "      <td>2018-12-08</td>\n",
       "      <td>12</td>\n",
       "      <td>20.936353</td>\n",
       "      <td>17.006086</td>\n",
       "      <td>16.076869</td>\n",
       "      <td>17.334870</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>621</th>\n",
       "      <td>27</td>\n",
       "      <td>['2018', 'w50']</td>\n",
       "      <td>2018</td>\n",
       "      <td>201850</td>\n",
       "      <td>2018-12-15</td>\n",
       "      <td>12</td>\n",
       "      <td>20.936353</td>\n",
       "      <td>5.673248</td>\n",
       "      <td>5.388778</td>\n",
       "      <td>5.794555</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>622</th>\n",
       "      <td>17</td>\n",
       "      <td>['2018', 'w51']</td>\n",
       "      <td>2018</td>\n",
       "      <td>201851</td>\n",
       "      <td>2018-12-22</td>\n",
       "      <td>12</td>\n",
       "      <td>20.936353</td>\n",
       "      <td>2.785103</td>\n",
       "      <td>2.708024</td>\n",
       "      <td>2.810241</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>623</th>\n",
       "      <td>17</td>\n",
       "      <td>['2018', 'w52']</td>\n",
       "      <td>2018</td>\n",
       "      <td>201852</td>\n",
       "      <td>2018-12-29</td>\n",
       "      <td>12</td>\n",
       "      <td>20.936353</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>624 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     cases_medellin             DATA  YEAR  YearWeek LastDayWeek  MONTH  \\\n",
       "0                 1  ['2007', 'w01']  2007    200701  2007-01-06      1   \n",
       "1                 0  ['2007', 'w02']  2007    200702  2007-01-13      1   \n",
       "2                 0  ['2007', 'w03']  2007    200703  2007-01-20      1   \n",
       "3                 0  ['2007', 'w04']  2007    200704  2007-01-27      1   \n",
       "4                 0  ['2007', 'w05']  2007    200705  2007-02-03      2   \n",
       "..              ...              ...   ...       ...         ...    ...   \n",
       "619              28  ['2018', 'w48']  2018    201848  2018-12-01     12   \n",
       "620              12  ['2018', 'w49']  2018    201849  2018-12-08     12   \n",
       "621              27  ['2018', 'w50']  2018    201850  2018-12-15     12   \n",
       "622              17  ['2018', 'w51']  2018    201851  2018-12-22     12   \n",
       "623              17  ['2018', 'w52']  2018    201852  2018-12-29     12   \n",
       "\n",
       "     temperature_medellin  percipitation_medellin  \\\n",
       "0               21.009283               13.096688   \n",
       "1               21.009283                8.437044   \n",
       "2               21.009283               19.373165   \n",
       "3               21.009283                0.000000   \n",
       "4               20.982925                3.725206   \n",
       "..                    ...                     ...   \n",
       "619             20.936353                8.916642   \n",
       "620             20.936353               17.006086   \n",
       "621             20.936353                5.673248   \n",
       "622             20.936353                2.785103   \n",
       "623             20.936353                0.000000   \n",
       "\n",
       "     percipitation_medellin_rural  percipitation_medellin_urbano  \n",
       "0                       14.097693                      12.628933  \n",
       "1                        6.910390                       9.067214  \n",
       "2                       21.616968                      18.406665  \n",
       "3                        0.000000                       0.000000  \n",
       "4                        3.144225                       3.901021  \n",
       "..                            ...                            ...  \n",
       "619                      8.656351                       9.114419  \n",
       "620                     16.076869                      17.334870  \n",
       "621                      5.388778                       5.794555  \n",
       "622                      2.708024                       2.810241  \n",
       "623                      0.000000                       0.000000  \n",
       "\n",
       "[624 rows x 10 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Read Data\n",
    "merge_cases_temp_precip = pd.read_csv('Data/merge_cases_temperature_WeeklyPrecipitation_timeseries.csv')\n",
    "# Remove extra column\n",
    "merge_cases_temp_precip = merge_cases_temp_precip.drop('Unnamed: 0', 1)\n",
    "merge_cases_temp_precip.LastDayWeek = pd.to_datetime(merge_cases_temp_precip.LastDayWeek)\n",
    "merge_cases_temp_precip"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "JDVHbhn9zGhb",
   "metadata": {
    "id": "JDVHbhn9zGhb"
   },
   "source": [
    "# Time Series"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "zxNDtkPwhKJt",
   "metadata": {
    "id": "zxNDtkPwhKJt"
   },
   "source": [
    "## Data visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "6d55f16d",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 497
    },
    "executionInfo": {
     "elapsed": 12,
     "status": "ok",
     "timestamp": 1625317313464,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "6d55f16d",
    "outputId": "89ed7899-90bf-4dff-a41b-9e2f2d2aa910"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Dengue cases in time\n",
    "def timeseries (x_axis, y_axis, x_label):\n",
    "    plt.figure(figsize = (12, 8))\n",
    "    plt.plot(x_axis, y_axis, color ='black')\n",
    "    plt.xlabel(x_label) \n",
    "    plt.ylabel('Dengue Cases')\n",
    "\n",
    "timeseries(merge_cases_temp_precip['LastDayWeek'], merge_cases_temp_precip['cases_medellin'], 'cases by Week')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3be7476a",
   "metadata": {
    "id": "3be7476a"
   },
   "source": [
    "## DataSet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "b3934515",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 450
    },
    "executionInfo": {
     "elapsed": 11,
     "status": "ok",
     "timestamp": 1625317313465,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "b3934515",
    "outputId": "6da015f2-7c83-4c67-b931-30866f6f91a9"
   },
   "outputs": [
    {
     "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>temperature_medellin</th>\n",
       "      <th>percipitation_medellin</th>\n",
       "      <th>cases_medellin</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LastDayWeek</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2007-01-06</th>\n",
       "      <td>21.009283</td>\n",
       "      <td>13.096688</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2007-01-13</th>\n",
       "      <td>21.009283</td>\n",
       "      <td>8.437044</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2007-01-20</th>\n",
       "      <td>21.009283</td>\n",
       "      <td>19.373165</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2007-01-27</th>\n",
       "      <td>21.009283</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2007-02-03</th>\n",
       "      <td>20.982925</td>\n",
       "      <td>3.725206</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-12-01</th>\n",
       "      <td>20.936353</td>\n",
       "      <td>8.916642</td>\n",
       "      <td>28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-12-08</th>\n",
       "      <td>20.936353</td>\n",
       "      <td>17.006086</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-12-15</th>\n",
       "      <td>20.936353</td>\n",
       "      <td>5.673248</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-12-22</th>\n",
       "      <td>20.936353</td>\n",
       "      <td>2.785103</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-12-29</th>\n",
       "      <td>20.936353</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>624 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             temperature_medellin  percipitation_medellin  cases_medellin\n",
       "LastDayWeek                                                              \n",
       "2007-01-06              21.009283               13.096688               1\n",
       "2007-01-13              21.009283                8.437044               0\n",
       "2007-01-20              21.009283               19.373165               0\n",
       "2007-01-27              21.009283                0.000000               0\n",
       "2007-02-03              20.982925                3.725206               0\n",
       "...                           ...                     ...             ...\n",
       "2018-12-01              20.936353                8.916642              28\n",
       "2018-12-08              20.936353               17.006086              12\n",
       "2018-12-15              20.936353                5.673248              27\n",
       "2018-12-22              20.936353                2.785103              17\n",
       "2018-12-29              20.936353                0.000000              17\n",
       "\n",
       "[624 rows x 3 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset = merge_cases_temp_precip[['temperature_medellin','percipitation_medellin','cases_medellin']]\n",
    "dataset.index = merge_cases_temp_precip.LastDayWeek\n",
    "dataset  #DF"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "JeddvRFhkTnY",
   "metadata": {
    "id": "JeddvRFhkTnY"
   },
   "source": [
    "# Prepare data to supervised learning time series"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "LFsBjJYWhZaQ",
   "metadata": {
    "id": "LFsBjJYWhZaQ"
   },
   "source": [
    "we will use:\n",
    "* data: is the dataframe in our case (Dengue Cases, Precipitation and Temperature)\n",
    "* n_in: is the number of lag weeks in the past (length of window)\n",
    "\n",
    "The heart of this \"series_to_supervised\" function is the <a href=\"https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.shift.html\">shift</a> fuction of pandas\n",
    "\n",
    "This function gets as input the number of periods(in this case the number of weeks represented as rows up or down in the dataframe) to move the columns of a dataframe.\n",
    "E.g. \n",
    "* If we have merge_cases_temp_precip['cases_medellin'].shift(1) all the rows of column cases_medellin will move one row down\n",
    "* If we have merge_cases_temp_precip['cases_medellin'].shift(-1) all the rows of column cases_medellin will move one row up\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "BH5JesTNkW1V",
   "metadata": {
    "executionInfo": {
     "elapsed": 386,
     "status": "ok",
     "timestamp": 1625317313842,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "BH5JesTNkW1V"
   },
   "outputs": [],
   "source": [
    "# prepare data for lstm\n",
    "from pandas import read_csv\n",
    "from pandas import DataFrame\n",
    "from pandas import concat\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "# 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\n",
    " "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4hM9aJnekTOF",
   "metadata": {
    "id": "4hM9aJnekTOF"
   },
   "source": [
    "### normalize features\n",
    "As we are working with a Neural Network the data values ​​must be normalized to help backpropagation algorithm\n",
    "So we will use the <a href=\"https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.MinMaxScaler.html\">MinMaxScaler</a> from sklearn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "gq1acAcggdew",
   "metadata": {
    "executionInfo": {
     "elapsed": 32,
     "status": "ok",
     "timestamp": 1625317313842,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "gq1acAcggdew"
   },
   "outputs": [],
   "source": [
    "scaler = MinMaxScaler(feature_range=(0, 1)) # Scaler between 0 and 1\n",
    "scaled = scaler.fit_transform(dataset) # As we can see data set has 3 Columns (This shape is also important for inverse scaler as we will see in future)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "HhBFNwjskQ0u",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 241
    },
    "executionInfo": {
     "elapsed": 32,
     "status": "ok",
     "timestamp": 1625317313843,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "HhBFNwjskQ0u",
    "outputId": "51d0b826-fa10-4193-8cbc-c222813928e9"
   },
   "outputs": [
    {
     "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>var1(t-10)</th>\n",
       "      <th>var2(t-10)</th>\n",
       "      <th>var3(t-10)</th>\n",
       "      <th>var1(t-9)</th>\n",
       "      <th>var2(t-9)</th>\n",
       "      <th>var3(t-9)</th>\n",
       "      <th>var1(t-8)</th>\n",
       "      <th>var2(t-8)</th>\n",
       "      <th>var3(t-8)</th>\n",
       "      <th>var1(t-7)</th>\n",
       "      <th>...</th>\n",
       "      <th>var3(t-3)</th>\n",
       "      <th>var1(t-2)</th>\n",
       "      <th>var2(t-2)</th>\n",
       "      <th>var3(t-2)</th>\n",
       "      <th>var1(t-1)</th>\n",
       "      <th>var2(t-1)</th>\n",
       "      <th>var3(t-1)</th>\n",
       "      <th>var1(t)</th>\n",
       "      <th>var2(t)</th>\n",
       "      <th>var3(t)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.364241</td>\n",
       "      <td>0.103033</td>\n",
       "      <td>0.00122</td>\n",
       "      <td>0.364241</td>\n",
       "      <td>0.066375</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.364241</td>\n",
       "      <td>0.152411</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.364241</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.493659</td>\n",
       "      <td>0.251632</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.493659</td>\n",
       "      <td>0.194133</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.493659</td>\n",
       "      <td>0.339599</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.364241</td>\n",
       "      <td>0.066375</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.364241</td>\n",
       "      <td>0.152411</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.364241</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.356053</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.493659</td>\n",
       "      <td>0.194133</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.493659</td>\n",
       "      <td>0.339599</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.493659</td>\n",
       "      <td>0.287922</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.364241</td>\n",
       "      <td>0.152411</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.364241</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.356053</td>\n",
       "      <td>0.029307</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.356053</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.493659</td>\n",
       "      <td>0.339599</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.493659</td>\n",
       "      <td>0.287922</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.493659</td>\n",
       "      <td>0.718713</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.364241</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.356053</td>\n",
       "      <td>0.029307</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.356053</td>\n",
       "      <td>0.095100</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.356053</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.493659</td>\n",
       "      <td>0.287922</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.493659</td>\n",
       "      <td>0.718713</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.434987</td>\n",
       "      <td>0.456353</td>\n",
       "      <td>0.00122</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.356053</td>\n",
       "      <td>0.029307</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.356053</td>\n",
       "      <td>0.095100</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.356053</td>\n",
       "      <td>0.011560</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.356053</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.493659</td>\n",
       "      <td>0.718713</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.434987</td>\n",
       "      <td>0.456353</td>\n",
       "      <td>0.00122</td>\n",
       "      <td>0.434987</td>\n",
       "      <td>0.371302</td>\n",
       "      <td>0.00000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 33 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    var1(t-10)  var2(t-10)  var3(t-10)  var1(t-9)  var2(t-9)  var3(t-9)  \\\n",
       "10    0.364241    0.103033     0.00122   0.364241   0.066375        0.0   \n",
       "11    0.364241    0.066375     0.00000   0.364241   0.152411        0.0   \n",
       "12    0.364241    0.152411     0.00000   0.364241   0.000000        0.0   \n",
       "13    0.364241    0.000000     0.00000   0.356053   0.029307        0.0   \n",
       "14    0.356053    0.029307     0.00000   0.356053   0.095100        0.0   \n",
       "\n",
       "    var1(t-8)  var2(t-8)  var3(t-8)  var1(t-7)  ...  var3(t-3)  var1(t-2)  \\\n",
       "10   0.364241   0.152411        0.0   0.364241  ...        0.0   0.493659   \n",
       "11   0.364241   0.000000        0.0   0.356053  ...        0.0   0.493659   \n",
       "12   0.356053   0.029307        0.0   0.356053  ...        0.0   0.493659   \n",
       "13   0.356053   0.095100        0.0   0.356053  ...        0.0   0.493659   \n",
       "14   0.356053   0.011560        0.0   0.356053  ...        0.0   0.493659   \n",
       "\n",
       "    var2(t-2)  var3(t-2)  var1(t-1)  var2(t-1)  var3(t-1)   var1(t)   var2(t)  \\\n",
       "10   0.251632        0.0   0.493659   0.194133    0.00000  0.493659  0.339599   \n",
       "11   0.194133        0.0   0.493659   0.339599    0.00000  0.493659  0.287922   \n",
       "12   0.339599        0.0   0.493659   0.287922    0.00000  0.493659  0.718713   \n",
       "13   0.287922        0.0   0.493659   0.718713    0.00000  0.434987  0.456353   \n",
       "14   0.718713        0.0   0.434987   0.456353    0.00122  0.434987  0.371302   \n",
       "\n",
       "    var3(t)  \n",
       "10  0.00000  \n",
       "11  0.00000  \n",
       "12  0.00000  \n",
       "13  0.00122  \n",
       "14  0.00000  \n",
       "\n",
       "[5 rows x 33 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# length of window\n",
    "weeks = 10\n",
    "\n",
    "# frame as supervised learning\n",
    "data = series_to_supervised(scaled, n_in=weeks)\n",
    "DataFrame(data).head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "Uss0IPaQqnWB",
   "metadata": {
    "id": "Uss0IPaQqnWB"
   },
   "source": [
    "## Features Set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "eWFelunOgdml",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 31,
     "status": "ok",
     "timestamp": 1625317313843,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "eWFelunOgdml",
    "outputId": "b7aa5b76-6d78-406f-d713-fa066de429b5"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.36424094, 0.10303316, 0.00121951, ..., 0.49365894, 0.19413316,\n",
       "        0.        ],\n",
       "       [0.36424094, 0.0663752 , 0.        , ..., 0.49365894, 0.33959896,\n",
       "        0.        ],\n",
       "       [0.36424094, 0.15241093, 0.        , ..., 0.49365894, 0.28792167,\n",
       "        0.        ],\n",
       "       ...,\n",
       "       [0.28320738, 0.60463713, 0.03292683, ..., 0.34158643, 0.13378884,\n",
       "        0.01463415],\n",
       "       [0.28320738, 0.20007748, 0.02804878, ..., 0.34158643, 0.0446321 ,\n",
       "        0.03292683],\n",
       "       [0.28320738, 0.3982439 , 0.02926829, ..., 0.34158643, 0.02191073,\n",
       "        0.02073171]])"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# We define the number of features as 3 (Temperature, Precipitation and Dengue Cases)\n",
    "n_features = 3\n",
    "# The features to train the model will be all except the values of the actual week \n",
    "# We can't use the temperature and precipitation in week t because whe need to resample a a 3D Array\n",
    "features_set = DataFrame(data.values[:,:-n_features])\n",
    "# Convert pandas data frame to np.array to reshape as 3D Array\n",
    "features_set = features_set.to_numpy()\n",
    "features_set"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ojvAeBMPr3ol",
   "metadata": {
    "id": "ojvAeBMPr3ol"
   },
   "source": [
    "## Labels Set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "q4jhxnI5gdqz",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 30,
     "status": "ok",
     "timestamp": 1625317313844,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "q4jhxnI5gdqz",
    "outputId": "8575301d-eecd-441a-b190-f8f7df84e935"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.00121951],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.00121951],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.00121951],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.        ],\n",
       "       [0.00121951],\n",
       "       [0.03170732],\n",
       "       [0.02926829],\n",
       "       [0.02560976],\n",
       "       [0.02195122],\n",
       "       [0.01829268],\n",
       "       [0.02073171],\n",
       "       [0.02073171],\n",
       "       [0.02195122],\n",
       "       [0.02682927],\n",
       "       [0.0304878 ],\n",
       "       [0.0195122 ],\n",
       "       [0.01829268],\n",
       "       [0.02439024],\n",
       "       [0.0195122 ],\n",
       "       [0.02073171],\n",
       "       [0.02195122],\n",
       "       [0.01463415],\n",
       "       [0.0195122 ],\n",
       "       [0.01463415],\n",
       "       [0.02317073],\n",
       "       [0.01463415],\n",
       "       [0.02439024],\n",
       "       [0.02560976],\n",
       "       [0.02804878],\n",
       "       [0.01829268],\n",
       "       [0.01585366],\n",
       "       [0.00853659],\n",
       "       [0.01829268],\n",
       "       [0.01829268],\n",
       "       [0.02073171],\n",
       "       [0.01219512],\n",
       "       [0.01829268],\n",
       "       [0.02560976],\n",
       "       [0.01463415],\n",
       "       [0.02804878],\n",
       "       [0.0195122 ],\n",
       "       [0.02682927],\n",
       "       [0.01707317],\n",
       "       [0.0195122 ],\n",
       "       [0.02195122],\n",
       "       [0.02682927],\n",
       "       [0.02317073],\n",
       "       [0.01829268],\n",
       "       [0.0304878 ],\n",
       "       [0.02804878],\n",
       "       [0.01219512],\n",
       "       [0.02317073],\n",
       "       [0.01829268],\n",
       "       [0.02073171],\n",
       "       [0.01585366],\n",
       "       [0.02317073],\n",
       "       [0.0195122 ],\n",
       "       [0.00487805],\n",
       "       [0.00487805],\n",
       "       [0.00365854],\n",
       "       [0.00243902],\n",
       "       [0.00365854],\n",
       "       [0.00121951],\n",
       "       [0.00121951],\n",
       "       [0.00365854],\n",
       "       [0.00121951],\n",
       "       [0.00121951],\n",
       "       [0.00243902],\n",
       "       [0.00487805],\n",
       "       [0.        ],\n",
       "       [0.00243902],\n",
       "       [0.00365854],\n",
       "       [0.00243902],\n",
       "       [0.        ],\n",
       "       [0.00121951],\n",
       "       [0.        ],\n",
       "       [0.00609756],\n",
       "       [0.00609756],\n",
       "       [0.00609756],\n",
       "       [0.00853659],\n",
       "       [0.00731707],\n",
       "       [0.00365854],\n",
       "       [0.00731707],\n",
       "       [0.00853659],\n",
       "       [0.0097561 ],\n",
       "       [0.00609756],\n",
       "       [0.00731707],\n",
       "       [0.00731707],\n",
       "       [0.00609756],\n",
       "       [0.01707317],\n",
       "       [0.01219512],\n",
       "       [0.01219512],\n",
       "       [0.00853659],\n",
       "       [0.00731707],\n",
       "       [0.01219512],\n",
       "       [0.0304878 ],\n",
       "       [0.02439024],\n",
       "       [0.02195122],\n",
       "       [0.01829268],\n",
       "       [0.01585366],\n",
       "       [0.01341463],\n",
       "       [0.01585366],\n",
       "       [0.01829268],\n",
       "       [0.01463415],\n",
       "       [0.03170732],\n",
       "       [0.0195122 ],\n",
       "       [0.02926829],\n",
       "       [0.02560976],\n",
       "       [0.0195122 ],\n",
       "       [0.03658537],\n",
       "       [0.04390244],\n",
       "       [0.04512195],\n",
       "       [0.05609756],\n",
       "       [0.10121951],\n",
       "       [0.10121951],\n",
       "       [0.11707317],\n",
       "       [0.11585366],\n",
       "       [0.1402439 ],\n",
       "       [0.13414634],\n",
       "       [0.16585366],\n",
       "       [0.16219512],\n",
       "       [0.20487805],\n",
       "       [0.28536585],\n",
       "       [0.38292683],\n",
       "       [0.54268293],\n",
       "       [0.62317073],\n",
       "       [0.68536585],\n",
       "       [0.78658537],\n",
       "       [0.89146341],\n",
       "       [0.85243902],\n",
       "       [0.89512195],\n",
       "       [1.        ],\n",
       "       [0.95365854],\n",
       "       [0.79878049],\n",
       "       [0.79390244],\n",
       "       [0.69756098],\n",
       "       [0.65487805],\n",
       "       [0.63658537],\n",
       "       [0.54512195],\n",
       "       [0.65365854],\n",
       "       [0.60121951],\n",
       "       [0.60121951],\n",
       "       [0.54634146],\n",
       "       [0.49878049],\n",
       "       [0.42317073],\n",
       "       [0.32073171],\n",
       "       [0.35365854],\n",
       "       [0.25487805],\n",
       "       [0.2       ],\n",
       "       [0.20243902],\n",
       "       [0.15487805],\n",
       "       [0.14756098],\n",
       "       [0.09756098],\n",
       "       [0.10853659],\n",
       "       [0.08170732],\n",
       "       [0.05121951],\n",
       "       [0.05365854],\n",
       "       [0.05609756],\n",
       "       [0.05731707],\n",
       "       [0.0402439 ],\n",
       "       [0.03292683],\n",
       "       [0.04634146],\n",
       "       [0.03536585],\n",
       "       [0.03536585],\n",
       "       [0.03170732],\n",
       "       [0.02317073],\n",
       "       [0.01707317],\n",
       "       [0.01585366],\n",
       "       [0.01707317],\n",
       "       [0.02317073],\n",
       "       [0.01707317],\n",
       "       [0.01097561],\n",
       "       [0.01707317],\n",
       "       [0.00853659],\n",
       "       [0.01829268],\n",
       "       [0.02195122],\n",
       "       [0.0195122 ],\n",
       "       [0.0195122 ],\n",
       "       [0.01097561],\n",
       "       [0.01463415],\n",
       "       [0.01829268],\n",
       "       [0.02682927],\n",
       "       [0.01707317],\n",
       "       [0.02682927],\n",
       "       [0.02195122],\n",
       "       [0.02195122],\n",
       "       [0.02560976],\n",
       "       [0.03170732],\n",
       "       [0.02317073],\n",
       "       [0.01341463],\n",
       "       [0.01097561],\n",
       "       [0.0195122 ],\n",
       "       [0.01463415],\n",
       "       [0.01463415],\n",
       "       [0.01829268],\n",
       "       [0.02439024],\n",
       "       [0.01097561],\n",
       "       [0.01219512],\n",
       "       [0.01707317],\n",
       "       [0.01707317],\n",
       "       [0.00853659],\n",
       "       [0.01707317],\n",
       "       [0.01463415],\n",
       "       [0.01097561],\n",
       "       [0.01463415],\n",
       "       [0.00731707],\n",
       "       [0.00853659],\n",
       "       [0.01341463],\n",
       "       [0.01097561],\n",
       "       [0.01707317],\n",
       "       [0.01707317],\n",
       "       [0.02073171],\n",
       "       [0.02682927],\n",
       "       [0.0195122 ],\n",
       "       [0.01707317],\n",
       "       [0.02195122],\n",
       "       [0.01463415],\n",
       "       [0.01585366],\n",
       "       [0.0097561 ],\n",
       "       [0.00487805],\n",
       "       [0.01463415],\n",
       "       [0.01829268],\n",
       "       [0.01707317],\n",
       "       [0.0195122 ],\n",
       "       [0.01585366],\n",
       "       [0.0097561 ],\n",
       "       [0.02195122],\n",
       "       [0.01219512],\n",
       "       [0.01097561],\n",
       "       [0.01219512],\n",
       "       [0.01219512],\n",
       "       [0.01829268],\n",
       "       [0.01585366],\n",
       "       [0.01097561],\n",
       "       [0.01097561],\n",
       "       [0.02317073],\n",
       "       [0.0097561 ],\n",
       "       [0.01829268],\n",
       "       [0.01585366],\n",
       "       [0.01585366],\n",
       "       [0.01585366],\n",
       "       [0.01585366],\n",
       "       [0.01829268],\n",
       "       [0.01463415],\n",
       "       [0.01585366],\n",
       "       [0.01585366],\n",
       "       [0.02073171],\n",
       "       [0.01219512],\n",
       "       [0.02926829],\n",
       "       [0.02195122],\n",
       "       [0.02195122],\n",
       "       [0.01829268],\n",
       "       [0.0195122 ],\n",
       "       [0.02560976],\n",
       "       [0.02682927],\n",
       "       [0.01829268],\n",
       "       [0.01707317],\n",
       "       [0.01707317],\n",
       "       [0.01463415],\n",
       "       [0.0195122 ],\n",
       "       [0.0304878 ],\n",
       "       [0.01219512],\n",
       "       [0.02073171],\n",
       "       [0.01341463],\n",
       "       [0.00609756],\n",
       "       [0.03536585],\n",
       "       [0.05609756],\n",
       "       [0.03780488],\n",
       "       [0.04512195],\n",
       "       [0.05      ],\n",
       "       [0.05      ],\n",
       "       [0.03902439],\n",
       "       [0.04756098],\n",
       "       [0.06097561],\n",
       "       [0.05243902],\n",
       "       [0.05609756],\n",
       "       [0.06341463],\n",
       "       [0.04634146],\n",
       "       [0.07073171],\n",
       "       [0.07195122],\n",
       "       [0.06463415],\n",
       "       [0.05853659],\n",
       "       [0.05487805],\n",
       "       [0.05      ],\n",
       "       [0.07560976],\n",
       "       [0.05609756],\n",
       "       [0.07317073],\n",
       "       [0.07439024],\n",
       "       [0.0804878 ],\n",
       "       [0.08292683],\n",
       "       [0.05609756],\n",
       "       [0.06219512],\n",
       "       [0.07682927],\n",
       "       [0.05487805],\n",
       "       [0.05609756],\n",
       "       [0.03658537],\n",
       "       [0.05      ],\n",
       "       [0.05365854],\n",
       "       [0.06585366],\n",
       "       [0.04390244],\n",
       "       [0.04878049],\n",
       "       [0.04146341],\n",
       "       [0.04756098],\n",
       "       [0.05      ],\n",
       "       [0.05243902],\n",
       "       [0.05121951],\n",
       "       [0.05853659],\n",
       "       [0.05853659],\n",
       "       [0.04878049],\n",
       "       [0.05853659],\n",
       "       [0.04512195],\n",
       "       [0.05609756],\n",
       "       [0.05      ],\n",
       "       [0.04634146],\n",
       "       [0.05853659],\n",
       "       [0.04634146],\n",
       "       [0.02804878],\n",
       "       [0.0597561 ],\n",
       "       [0.0695122 ],\n",
       "       [0.07439024],\n",
       "       [0.07195122],\n",
       "       [0.04756098],\n",
       "       [0.04634146],\n",
       "       [0.04878049],\n",
       "       [0.0402439 ],\n",
       "       [0.03658537],\n",
       "       [0.0304878 ],\n",
       "       [0.04512195],\n",
       "       [0.03902439],\n",
       "       [0.03536585],\n",
       "       [0.04146341],\n",
       "       [0.03902439],\n",
       "       [0.03292683],\n",
       "       [0.05243902],\n",
       "       [0.05365854],\n",
       "       [0.04268293],\n",
       "       [0.03536585],\n",
       "       [0.03536585],\n",
       "       [0.03414634],\n",
       "       [0.05487805],\n",
       "       [0.05365854],\n",
       "       [0.03780488],\n",
       "       [0.05365854],\n",
       "       [0.09390244],\n",
       "       [0.06829268],\n",
       "       [0.08414634],\n",
       "       [0.09268293],\n",
       "       [0.13658537],\n",
       "       [0.11829268],\n",
       "       [0.1195122 ],\n",
       "       [0.1       ],\n",
       "       [0.10731707],\n",
       "       [0.15365854],\n",
       "       [0.15      ],\n",
       "       [0.14268293],\n",
       "       [0.13658537],\n",
       "       [0.11707317],\n",
       "       [0.11585366],\n",
       "       [0.08292683],\n",
       "       [0.10853659],\n",
       "       [0.09268293],\n",
       "       [0.08292683],\n",
       "       [0.08292683],\n",
       "       [0.0695122 ],\n",
       "       [0.05365854],\n",
       "       [0.0902439 ],\n",
       "       [0.07439024],\n",
       "       [0.09146341],\n",
       "       [0.07682927],\n",
       "       [0.12682927],\n",
       "       [0.09146341],\n",
       "       [0.07073171],\n",
       "       [0.0597561 ],\n",
       "       [0.06707317],\n",
       "       [0.04512195],\n",
       "       [0.04878049],\n",
       "       [0.03780488],\n",
       "       [0.05487805],\n",
       "       [0.03292683],\n",
       "       [0.04146341],\n",
       "       [0.04390244],\n",
       "       [0.03780488],\n",
       "       [0.04146341],\n",
       "       [0.03780488],\n",
       "       [0.02317073],\n",
       "       [0.0195122 ],\n",
       "       [0.03658537],\n",
       "       [0.03414634],\n",
       "       [0.03658537],\n",
       "       [0.02926829],\n",
       "       [0.03170732],\n",
       "       [0.05487805],\n",
       "       [0.0695122 ],\n",
       "       [0.04634146],\n",
       "       [0.08414634],\n",
       "       [0.07926829],\n",
       "       [0.06341463],\n",
       "       [0.08536585],\n",
       "       [0.05609756],\n",
       "       [0.0597561 ],\n",
       "       [0.07073171],\n",
       "       [0.08902439],\n",
       "       [0.08780488],\n",
       "       [0.08170732],\n",
       "       [0.12682927],\n",
       "       [0.12804878],\n",
       "       [0.14512195],\n",
       "       [0.14390244],\n",
       "       [0.12560976],\n",
       "       [0.14146341],\n",
       "       [0.15731707],\n",
       "       [0.15      ],\n",
       "       [0.12560976],\n",
       "       [0.14756098],\n",
       "       [0.12804878],\n",
       "       [0.11829268],\n",
       "       [0.16463415],\n",
       "       [0.15853659],\n",
       "       [0.17926829],\n",
       "       [0.22926829],\n",
       "       [0.2695122 ],\n",
       "       [0.28658537],\n",
       "       [0.33414634],\n",
       "       [0.30731707],\n",
       "       [0.3195122 ],\n",
       "       [0.33414634],\n",
       "       [0.32439024],\n",
       "       [0.28536585],\n",
       "       [0.31463415],\n",
       "       [0.33170732],\n",
       "       [0.2304878 ],\n",
       "       [0.28902439],\n",
       "       [0.33170732],\n",
       "       [0.41707317],\n",
       "       [0.3695122 ],\n",
       "       [0.45365854],\n",
       "       [0.4304878 ],\n",
       "       [0.41829268],\n",
       "       [0.50853659],\n",
       "       [0.54512195],\n",
       "       [0.56585366],\n",
       "       [0.5695122 ],\n",
       "       [0.57804878],\n",
       "       [0.62804878],\n",
       "       [0.64268293],\n",
       "       [0.64268293],\n",
       "       [0.62682927],\n",
       "       [0.72195122],\n",
       "       [0.64146341],\n",
       "       [0.56219512],\n",
       "       [0.58292683],\n",
       "       [0.61097561],\n",
       "       [0.59512195],\n",
       "       [0.65731707],\n",
       "       [0.56829268],\n",
       "       [0.48780488],\n",
       "       [0.46097561],\n",
       "       [0.41707317],\n",
       "       [0.4       ],\n",
       "       [0.43170732],\n",
       "       [0.32073171],\n",
       "       [0.3195122 ],\n",
       "       [0.31097561],\n",
       "       [0.27560976],\n",
       "       [0.22560976],\n",
       "       [0.22439024],\n",
       "       [0.23414634],\n",
       "       [0.22073171],\n",
       "       [0.17439024],\n",
       "       [0.14146341],\n",
       "       [0.15121951],\n",
       "       [0.14878049],\n",
       "       [0.13414634],\n",
       "       [0.10853659],\n",
       "       [0.12073171],\n",
       "       [0.09512195],\n",
       "       [0.08414634],\n",
       "       [0.0695122 ],\n",
       "       [0.06341463],\n",
       "       [0.05731707],\n",
       "       [0.06219512],\n",
       "       [0.06097561],\n",
       "       [0.05853659],\n",
       "       [0.04878049],\n",
       "       [0.05121951],\n",
       "       [0.05      ],\n",
       "       [0.06829268],\n",
       "       [0.05243902],\n",
       "       [0.0695122 ],\n",
       "       [0.04634146],\n",
       "       [0.04512195],\n",
       "       [0.04878049],\n",
       "       [0.05609756],\n",
       "       [0.05731707],\n",
       "       [0.05121951],\n",
       "       [0.0402439 ],\n",
       "       [0.04756098],\n",
       "       [0.02926829],\n",
       "       [0.04634146],\n",
       "       [0.06219512],\n",
       "       [0.06097561],\n",
       "       [0.04634146],\n",
       "       [0.04390244],\n",
       "       [0.04634146],\n",
       "       [0.03292683],\n",
       "       [0.01707317],\n",
       "       [0.05365854],\n",
       "       [0.03536585],\n",
       "       [0.03780488],\n",
       "       [0.04756098],\n",
       "       [0.04512195],\n",
       "       [0.05121951],\n",
       "       [0.03780488],\n",
       "       [0.03536585],\n",
       "       [0.03292683],\n",
       "       [0.03292683],\n",
       "       [0.02926829],\n",
       "       [0.03292683],\n",
       "       [0.02682927],\n",
       "       [0.03414634],\n",
       "       [0.02560976],\n",
       "       [0.03658537],\n",
       "       [0.01707317],\n",
       "       [0.02682927],\n",
       "       [0.02317073],\n",
       "       [0.0195122 ],\n",
       "       [0.04268293],\n",
       "       [0.03170732],\n",
       "       [0.03170732],\n",
       "       [0.03658537],\n",
       "       [0.02682927],\n",
       "       [0.02804878],\n",
       "       [0.02560976],\n",
       "       [0.0304878 ],\n",
       "       [0.02682927],\n",
       "       [0.01707317],\n",
       "       [0.02926829],\n",
       "       [0.02439024],\n",
       "       [0.02926829],\n",
       "       [0.02682927],\n",
       "       [0.02682927],\n",
       "       [0.02439024],\n",
       "       [0.02926829],\n",
       "       [0.02439024],\n",
       "       [0.02560976],\n",
       "       [0.02073171],\n",
       "       [0.02317073],\n",
       "       [0.02804878],\n",
       "       [0.02682927],\n",
       "       [0.02804878],\n",
       "       [0.02804878],\n",
       "       [0.03292683],\n",
       "       [0.04268293],\n",
       "       [0.02682927],\n",
       "       [0.02317073],\n",
       "       [0.02439024],\n",
       "       [0.02682927],\n",
       "       [0.02195122],\n",
       "       [0.0304878 ],\n",
       "       [0.03292683],\n",
       "       [0.0304878 ],\n",
       "       [0.01463415],\n",
       "       [0.02682927],\n",
       "       [0.02682927],\n",
       "       [0.03292683],\n",
       "       [0.02804878],\n",
       "       [0.02926829],\n",
       "       [0.02926829],\n",
       "       [0.02073171],\n",
       "       [0.02439024],\n",
       "       [0.01829268],\n",
       "       [0.05      ],\n",
       "       [0.03414634],\n",
       "       [0.01463415],\n",
       "       [0.03292683],\n",
       "       [0.02073171],\n",
       "       [0.02073171]])"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# We will use Dengue cases in last week \n",
    "labels_set = DataFrame(data.values[:,-1])\n",
    "# Convert pandas data frame to np.array\n",
    "labels_set = labels_set.to_numpy()\n",
    "labels_set"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "RlwXfwaLsHza",
   "metadata": {
    "id": "RlwXfwaLsHza"
   },
   "source": [
    "## Train Test Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "-sb-9LCMgdwr",
   "metadata": {
    "executionInfo": {
     "elapsed": 28,
     "status": "ok",
     "timestamp": 1625317313844,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "-sb-9LCMgdwr"
   },
   "outputs": [],
   "source": [
    "# We need a sequence so we can't split randomly\n",
    "# To divide into Train (90%) and test (10%) to do that we need to know the 90% of the total dataframe\n",
    "size = features_set.shape[0]\n",
    "split = int(size*(9/10))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "_2LNWDMusZsl",
   "metadata": {
    "id": "_2LNWDMusZsl"
   },
   "source": [
    "### train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "FJan_zurgd1q",
   "metadata": {
    "executionInfo": {
     "elapsed": 28,
     "status": "ok",
     "timestamp": 1625317313845,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "FJan_zurgd1q"
   },
   "outputs": [],
   "source": [
    "# We will train with 1st 90% of data and test with last 10%\n",
    "train_X = features_set[:split] ##90% train\n",
    "train_y = labels_set[:split]  ##90% train"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "t17j8OtzemAG",
   "metadata": {
    "id": "t17j8OtzemAG"
   },
   "source": [
    "### test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "DS6rGhy5gd7B",
   "metadata": {
    "executionInfo": {
     "elapsed": 28,
     "status": "ok",
     "timestamp": 1625317313845,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "DS6rGhy5gd7B"
   },
   "outputs": [],
   "source": [
    "test_X = features_set[split:] ##10% test\n",
    "test_y = labels_set[split:] ##10% test"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "UExWrkq9Ib_J",
   "metadata": {
    "id": "UExWrkq9Ib_J"
   },
   "source": [
    "## Reshape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "G1xmvWGQgeDB",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 27,
     "status": "ok",
     "timestamp": 1625317313845,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "G1xmvWGQgeDB",
    "outputId": "9d0fddc4-3864-4047-cb25-0f87bb404b7c"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(552, 10, 3) (552, 1) (62, 10, 3) (62, 1)\n"
     ]
    }
   ],
   "source": [
    "# reshape input to be 3D [samples, timesteps, features]\n",
    "train_X = train_X.reshape((train_X.shape[0], weeks, n_features))\n",
    "test_X = test_X.reshape((test_X.shape[0], weeks, n_features))\n",
    "print(train_X.shape, train_y.shape, test_X.shape, test_y.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ac223af3",
   "metadata": {
    "id": "ac223af3"
   },
   "source": [
    "## Model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "h5ZtewNE5vXX",
   "metadata": {
    "id": "h5ZtewNE5vXX"
   },
   "source": [
    "### LSTM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "wPD_nuIfgeK9",
   "metadata": {
    "executionInfo": {
     "elapsed": 595,
     "status": "ok",
     "timestamp": 1625317314419,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "wPD_nuIfgeK9"
   },
   "outputs": [],
   "source": [
    "# Set Seed\n",
    "tf.random.set_seed(1)\n",
    "\n",
    "# design network\n",
    "model = Sequential()\n",
    "model.add(LSTM(60, input_shape=(train_X.shape[1], train_X.shape[2])))\n",
    "model.add(Dense(1))\n",
    "model.compile(loss='mae', optimizer='adam')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "CRKoVYEqmcav",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 5925,
     "status": "ok",
     "timestamp": 1625317320340,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "CRKoVYEqmcav",
    "outputId": "abca9a8a-e6a4-46ec-da6d-91605b9a8e9b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 552 samples, validate on 62 samples\n",
      "Epoch 1/50\n",
      "552/552 - 3s - loss: 0.1071 - val_loss: 0.0755\n",
      "Epoch 2/50\n",
      "552/552 - 0s - loss: 0.1112 - val_loss: 0.0313\n",
      "Epoch 3/50\n",
      "552/552 - 0s - loss: 0.0929 - val_loss: 0.0083\n",
      "Epoch 4/50\n",
      "552/552 - 0s - loss: 0.0695 - val_loss: 0.0286\n",
      "Epoch 5/50\n",
      "552/552 - 0s - loss: 0.0671 - val_loss: 0.0130\n",
      "Epoch 6/50\n",
      "552/552 - 0s - loss: 0.0546 - val_loss: 0.0165\n",
      "Epoch 7/50\n",
      "552/552 - 0s - loss: 0.0435 - val_loss: 0.0106\n",
      "Epoch 8/50\n",
      "552/552 - 0s - loss: 0.0347 - val_loss: 0.0079\n",
      "Epoch 9/50\n",
      "552/552 - 0s - loss: 0.0332 - val_loss: 0.0080\n",
      "Epoch 10/50\n",
      "552/552 - 0s - loss: 0.0341 - val_loss: 0.0185\n",
      "Epoch 11/50\n",
      "552/552 - 0s - loss: 0.0339 - val_loss: 0.0093\n",
      "Epoch 12/50\n",
      "552/552 - 0s - loss: 0.0320 - val_loss: 0.0107\n",
      "Epoch 13/50\n",
      "552/552 - 0s - loss: 0.0327 - val_loss: 0.0107\n",
      "Epoch 14/50\n",
      "552/552 - 0s - loss: 0.0307 - val_loss: 0.0120\n",
      "Epoch 15/50\n",
      "552/552 - 0s - loss: 0.0297 - val_loss: 0.0077\n",
      "Epoch 16/50\n",
      "552/552 - 0s - loss: 0.0287 - val_loss: 0.0079\n",
      "Epoch 17/50\n",
      "552/552 - 0s - loss: 0.0286 - val_loss: 0.0065\n",
      "Epoch 18/50\n",
      "552/552 - 0s - loss: 0.0283 - val_loss: 0.0137\n",
      "Epoch 19/50\n",
      "552/552 - 0s - loss: 0.0292 - val_loss: 0.0089\n",
      "Epoch 20/50\n",
      "552/552 - 0s - loss: 0.0268 - val_loss: 0.0102\n",
      "Epoch 21/50\n",
      "552/552 - 0s - loss: 0.0282 - val_loss: 0.0070\n",
      "Epoch 22/50\n",
      "552/552 - 0s - loss: 0.0267 - val_loss: 0.0125\n",
      "Epoch 23/50\n",
      "552/552 - 0s - loss: 0.0295 - val_loss: 0.0160\n",
      "Epoch 24/50\n",
      "552/552 - 0s - loss: 0.0271 - val_loss: 0.0161\n",
      "Epoch 25/50\n",
      "552/552 - 0s - loss: 0.0314 - val_loss: 0.0091\n",
      "Epoch 26/50\n",
      "552/552 - 0s - loss: 0.0273 - val_loss: 0.0193\n",
      "Epoch 27/50\n",
      "552/552 - 0s - loss: 0.0318 - val_loss: 0.0064\n",
      "Epoch 28/50\n",
      "552/552 - 0s - loss: 0.0247 - val_loss: 0.0138\n",
      "Epoch 29/50\n",
      "552/552 - 0s - loss: 0.0271 - val_loss: 0.0065\n",
      "Epoch 30/50\n",
      "552/552 - 0s - loss: 0.0244 - val_loss: 0.0147\n",
      "Epoch 31/50\n",
      "552/552 - 0s - loss: 0.0274 - val_loss: 0.0069\n",
      "Epoch 32/50\n",
      "552/552 - 0s - loss: 0.0237 - val_loss: 0.0137\n",
      "Epoch 33/50\n",
      "552/552 - 0s - loss: 0.0250 - val_loss: 0.0064\n",
      "Epoch 34/50\n",
      "552/552 - 0s - loss: 0.0244 - val_loss: 0.0184\n",
      "Epoch 35/50\n",
      "552/552 - 0s - loss: 0.0273 - val_loss: 0.0137\n",
      "Epoch 36/50\n",
      "552/552 - 0s - loss: 0.0260 - val_loss: 0.0085\n",
      "Epoch 37/50\n",
      "552/552 - 0s - loss: 0.0227 - val_loss: 0.0157\n",
      "Epoch 38/50\n",
      "552/552 - 0s - loss: 0.0264 - val_loss: 0.0064\n",
      "Epoch 39/50\n",
      "552/552 - 0s - loss: 0.0216 - val_loss: 0.0129\n",
      "Epoch 40/50\n",
      "552/552 - 0s - loss: 0.0222 - val_loss: 0.0082\n",
      "Epoch 41/50\n",
      "552/552 - 0s - loss: 0.0234 - val_loss: 0.0142\n",
      "Epoch 42/50\n",
      "552/552 - 0s - loss: 0.0235 - val_loss: 0.0160\n",
      "Epoch 43/50\n",
      "552/552 - 0s - loss: 0.0254 - val_loss: 0.0072\n",
      "Epoch 44/50\n",
      "552/552 - 0s - loss: 0.0206 - val_loss: 0.0143\n",
      "Epoch 45/50\n",
      "552/552 - 0s - loss: 0.0240 - val_loss: 0.0059\n",
      "Epoch 46/50\n",
      "552/552 - 0s - loss: 0.0193 - val_loss: 0.0120\n",
      "Epoch 47/50\n",
      "552/552 - 0s - loss: 0.0201 - val_loss: 0.0069\n",
      "Epoch 48/50\n",
      "552/552 - 0s - loss: 0.0213 - val_loss: 0.0145\n",
      "Epoch 49/50\n",
      "552/552 - 0s - loss: 0.0217 - val_loss: 0.0174\n",
      "Epoch 50/50\n",
      "552/552 - 0s - loss: 0.0249 - val_loss: 0.0061\n"
     ]
    }
   ],
   "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)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "9yqmdQXsmcfH",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 265
    },
    "executionInfo": {
     "elapsed": 428,
     "status": "ok",
     "timestamp": 1625317320749,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "9yqmdQXsmcfH",
    "outputId": "914fb0c4-eb4f-4dd6-9ab7-796cff5dc4a7"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot\n",
    "# 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()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "C21mhuRullNL",
   "metadata": {
    "executionInfo": {
     "elapsed": 7,
     "status": "ok",
     "timestamp": 1625317320750,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "C21mhuRullNL"
   },
   "outputs": [],
   "source": [
    "from math import sqrt\n",
    "from numpy import concatenate\n",
    "\n",
    "# make a prediction\n",
    "yhat = model.predict(test_X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "TtordubT422m",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 12,
     "status": "ok",
     "timestamp": 1625317321210,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "TtordubT422m",
    "outputId": "26ee2559-6d1b-445b-fa85-21b224c2ad43"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(62, 1)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yhat.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "5Ozj32Jwq4VL",
   "metadata": {
    "executionInfo": {
     "elapsed": 11,
     "status": "ok",
     "timestamp": 1625317321211,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "5Ozj32Jwq4VL"
   },
   "outputs": [],
   "source": [
    "# Convert test data to 2D \n",
    "test_X = test_X.reshape((test_X.shape[0], weeks*n_features))\n",
    "\n",
    "# invert scaling for forecast\n",
    "# As we said Scaler needs 3 columns so we can take those columns from test data and take again the predictions\n",
    "# Concatenate last 2 columns of test data with predicted data (yhat)\n",
    "inv_yhat = concatenate((test_X[:, -(n_features-1):], yhat), axis=1)\n",
    "# Inverse Scaler\n",
    "inv_yhat = scaler.inverse_transform(inv_yhat)\n",
    "# Take predicted data scaled to original Dengue cases\n",
    "inv_yhat = inv_yhat[:,-1]\n",
    "\n",
    "# invert scaling for actual\n",
    "# Same process than for predicted data (yhat)\n",
    "test_y = test_y.reshape((len(test_y), 1))\n",
    "inv_y = concatenate((test_X[:, -(n_features-1):], test_y), axis=1)\n",
    "inv_y = scaler.inverse_transform(inv_y)\n",
    "inv_y = inv_y[:,-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "TdDRHwvNllQR",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 11,
     "status": "ok",
     "timestamp": 1625317321211,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "TdDRHwvNllQR",
    "outputId": "a09cd4cf-adfd-4adf-e797-a22eaeca1782"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test MAE: 4.969\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import  mean_absolute_error\n",
    "\n",
    "# calculate MAE\n",
    "mae = mean_absolute_error(inv_y, inv_yhat)\n",
    "print('Test MAE: %.3f' % mae)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "PJzx3p9psAqW",
   "metadata": {
    "id": "PJzx3p9psAqW"
   },
   "source": [
    "#### Plot predicted vs actual dengue cases"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "nPOiCVoxsr-1",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 294
    },
    "executionInfo": {
     "elapsed": 9,
     "status": "ok",
     "timestamp": 1625317321212,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "nPOiCVoxsr-1",
    "outputId": "0f7a4019-9838-4a92-bdbd-ca2c3ae27858"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "data_predict = inv_yhat  ## predicted target  cases\n",
    "dataY_plot = inv_y  ##  real test-target cases\n",
    "\n",
    "data_predict = data_predict.reshape(len(data_predict), 1)\n",
    "dataY_plot = dataY_plot.reshape(len(dataY_plot), 1)\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.plot(dataY_plot, label = 'actual')\n",
    "plt.plot(data_predict, label = 'predicted')\n",
    "plt.legend(loc=\"upper left\")\n",
    "\n",
    "plt.suptitle('Time-Series Prediction')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "kMiD5Vp27n5W",
   "metadata": {
    "executionInfo": {
     "elapsed": 8,
     "status": "ok",
     "timestamp": 1625317321212,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "kMiD5Vp27n5W"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "9R-Zl5OgLo_h",
   "metadata": {
    "id": "9R-Zl5OgLo_h"
   },
   "source": [
    "<ol>\n",
    "  <li> <a href=\"https://towardsdatascience.com/predictive-analytics-time-series-forecasting-with-gru-and-bilstm-in-tensorflow-87588c852915\">Predictive Analytics: Time-Series Forecasting with GRU and BiLSTM in TensorFlow</a></li>\n",
    "  <li><a href=\"https://machinelearningmastery.com/multivariate-time-series-forecasting-lstms-keras/\">Multivariate Time Series Forecasting with LSTMs in Keras</a></li>\n",
    "</ol>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "LA6ELSVJLpp0",
   "metadata": {
    "executionInfo": {
     "elapsed": 7,
     "status": "ok",
     "timestamp": 1625317321212,
     "user": {
      "displayName": "DAVID RESTREPO",
      "photoUrl": "https://lh3.googleusercontent.com/a-/AOh14GgqxItBQRJ0dDzIdCZM0BMuA0KeTKTlKRNGMApY=s64",
      "userId": "09638303782015555303"
     },
     "user_tz": 300
    },
    "id": "LA6ELSVJLpp0"
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "LSTM_TensorFlow_WeeklyPrecipitation.ipynb",
   "provenance": [
    {
     "file_id": "1yhwDjCPk8baHY-Odebdo8C1jIJVI5Uny",
     "timestamp": 1625026832717
    }
   ]
  },
  "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": "3.6.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}