AR_and_ARIMA_Models.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "coursera": {
      "course_slug": "neural-networks-deep-learning",
      "graded_item_id": "XaIWT",
      "launcher_item_id": "zAgPl"
    },
    "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": "3.6.2"
    },
    "colab": {
      "name": "AR_and_ARIMA_Models.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "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/AR_and_ARIMA_Models.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2WZFQQORCAcS"
      },
      "source": [
        "#**1- Autoregressive (AR) and Auto Regressive Integrated Moving Average (ARIMA) Models**\n",
        "\n",
        "**Introduction:**\n",
        "\n",
        "A statistical model is autoregressive if it predicts future values based on past values. For example, an autoregressive model might seek to predict a stock's future prices based on its past performance [1].An AR(1) autoregressive process is one in which the current value is based on the immediately preceding value, while an AR(2) process is one in which the current value is based on the previous two values. An AR(0) process is used for white noise and has no dependence between the terms [1]"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Mn9QTzzT_0ty"
      },
      "source": [
        "!git clone https://github.com/hussain0048/Machine-Learning-Driven-Approach-for-2019-nCoV-Warning-System.git"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "tsY-ZNwbPThQ",
        "outputId": "cfe8b99f-cf5e-4b7a-c98e-0c51203a6980",
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/files.js": {
              "data": "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",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "application/javascript"
                ]
              ],
              "status": 200,
              "status_text": ""
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 72
        }
      },
      "source": [
        "# this code is used to upload dataset from Pc to colab\n",
        "from google.colab import files # Please First run this cod in chrom \n",
        "def getLocalFiles():\n",
        "    _files = files.upload() # upload StudentNextSessionf.csv datase\n",
        "    if len(_files) >0: # Then run above  libray \n",
        "       for k,v in _files.items():\n",
        "         open(k,'wb').write(v)\n",
        "getLocalFiles()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "\n",
              "     <input type=\"file\" id=\"files-e739dc9d-b01f-478f-8ad4-555f6d6d8b06\" name=\"files[]\" multiple disabled\n",
              "        style=\"border:none\" />\n",
              "     <output id=\"result-e739dc9d-b01f-478f-8ad4-555f6d6d8b06\">\n",
              "      Upload widget is only available when the cell has been executed in the\n",
              "      current browser session. Please rerun this cell to enable.\n",
              "      </output>\n",
              "      <script src=\"/nbextensions/google.colab/files.js\"></script> "
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Saving data_cases1.csv to data_cases1.csv\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lVaYgRILECmh",
        "outputId": "f11c866b-2705-40ac-af4c-a585bffcaa83",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "beVu-1UqCAcU"
      },
      "source": [
        "## **1 - Importing necessary libraries** ##\n",
        "we import the following python packges for this tutorial \n",
        " - pandas: importing index data\n",
        " - matplotlib: it is used for ploting data\n",
        " - statsmodels build time series models\n",
        " - statistic to help find statistic of models"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZF03EKpuCAcV"
      },
      "source": [
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7c3251v0N3z1",
        "outputId": "09b41ae2-2c0f-4c16-da00-07f4ec8d918e",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "from statsmodels.tsa.ar_model import AR\n",
        "from sklearn.metrics import mean_squared_error"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/statsmodels/tools/_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
            "  import pandas.util.testing as tm\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vlav6g_TvrD9"
      },
      "source": [
        "import statsmodels as sm\n",
        "from statsmodels.tsa.stattools import adfuller \n",
        "from statsmodels.graphics.tsaplots import plot_acf, plot_pacf  \n",
        "from statsmodels.tsa.arima_model import ARIMA \n",
        "import statistics"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "euL8yNRc7yE_"
      },
      "source": [
        "plt.rcParams.update({'figure.figsize':(9,7), 'figure.dpi':120})"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dZCZulRu3fib"
      },
      "source": [
        "from pandas import Series\n",
        "from pandas import DataFrame\n",
        "from pandas import concat\n",
        "from sklearn.metrics import mean_squared_error"
      ],
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "collapsed": true,
        "id": "nhYP15aoCAcf"
      },
      "source": [
        "## **2 - Load Datasets** ##"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "y1wsGi52CAch"
      },
      "source": [
        "daily_confirm = pd.read_csv(\"/content/drive/My Drive/Datasets/Covid-19/data_cases1.csv\", usecols=['Date','Confirmed'],index_col='Date', parse_dates=True )\n",
        "daily_confirm.head()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UKI68e66OsBL"
      },
      "source": [
        "## **3- Data Description** # "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TmxB0x3VOyIh"
      },
      "source": [
        "daily_confirm.index"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "p_Ttzuj2O6TB"
      },
      "source": [
        "daily_confirm.head()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "poqouIc3PBaB"
      },
      "source": [
        "daily_confirm['2020-10-03':'2020-15-03']"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vSN1E21RIYXh"
      },
      "source": [
        "## **3- Check and Fill the missing value** \n",
        "\n",
        "finding all columns that have nan"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KCtUtRplIrYB",
        "outputId": "fbba2288-5c70-4d39-b211-147e86510996",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "daily_confirm.isna().sum()"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Confirmed    11\n",
              "dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MvFSO1eVY4Ul"
      },
      "source": [
        "for j in range(0,1):       \n",
        "        daily_confirm.iloc[:,j]=daily_confirm.iloc[:,j].fillna(daily_confirm.iloc[:,j].mean())  "
      ],
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "P4QYI2IoHWpL"
      },
      "source": [
        "## **4 - Ploting** ##\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "p9JGZRKCCAcq"
      },
      "source": [
        " daily_confirm.plot(daily_confirm['Confirmed'])"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nR-PYX-bQsYr"
      },
      "source": [
        "%matplotlib inline\n",
        "daily_confirm.Confirmed.resample(\"M\").mean().plot()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A4baoniBQ5qv"
      },
      "source": [
        "daily_confirm.Confirmed.resample(\"W\").mean().plot()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L4jg1ktARE9o"
      },
      "source": [
        "daily_confirm.Confirmed.resample(\"D\").mean().plot()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "soIDaghNRORG"
      },
      "source": [
        "Quarterly_resampled_data = daily_confirm.Confirmed.resample('Q').mean().plot()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OXCttrtfR43-"
      },
      "source": [
        "daily_confirm.Confirmed.resample('QS').mean().plot()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jBsedv0KSDIf"
      },
      "source": [
        "daily_confirm.Confirmed.resample('M').mean().plot(kind=\"line\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ehTTQE9YJF1Q"
      },
      "source": [
        "##**5- Correlation** ##\n",
        "Visual check of correlation by plotting the scatter plot against the previous and next time step Pandas built in plot - Lagplot - to do visual check\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": true,
        "id": "TvqFi8-XCAcy"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Y5ay6wqxYGx4"
      },
      "source": [
        "##**6-Stationary**\n",
        "\n",
        "To check wither data is stationaly or not. stationary mean that is data is constant in mean and variance . we cehck this becasue many time sieries model required that data should be stationary in order to model it. It should be concentatn movment up and down momement"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xa_nh0I_YFQL"
      },
      "source": [
        "plt.plot(daily_confirm)\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "U_AmXR-xaMdo"
      },
      "source": [
        "now we make data stationary if data is not stationary \n",
        "  - Difference the data and make it satationary . difference mean substract the next value \n",
        "  - fill the miss value data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6fP07c68bmeo"
      },
      "source": [
        "# take serveral diff untill your data become stationary \n",
        "daily_confirm_diff1=daily_confirm.diff().fillna(daily_confirm)"
      ],
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9T4FkhX7sygg"
      },
      "source": [
        "daily_confirm_diff2=daily_confirm_diff1.diff().fillna(daily_confirm_diff1)"
      ],
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4TtGJAkCtBjT"
      },
      "source": [
        "daily_confirm_diff3=daily_confirm_diff2.diff().fillna(daily_confirm_diff2)"
      ],
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "keqNEyPpv1x9"
      },
      "source": [
        "daily_confirm_diff4=daily_confirm_diff3.diff().fillna(daily_confirm_diff3)"
      ],
      "execution_count": 20,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4mRDOrcvwYo7"
      },
      "source": [
        "daily_confirm_diff5=daily_confirm_diff4.diff().fillna(daily_confirm_diff4)"
      ],
      "execution_count": 21,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2ZhajJjfbyEU"
      },
      "source": [
        "plt.plot(daily_confirm_diff5)\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "X7TpYJAKuQS9"
      },
      "source": [
        "## 7-**Autocorrelation Function (ACF) plot**\n",
        "\n",
        "**Partial Autocorrelation Function(PACF) Plot ** \n",
        "\n",
        "Helps determine number of autoregressive AR term and moving average term\n",
        "\n",
        "can also be used to help see seasonality or periodic trends\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PV824D_Fyl64",
        "outputId": "fa531142-f17f-40d2-cc15-f97bebdd8086",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        }
      },
      "source": [
        "plot_acf(daily_confirm_diff5)\n",
        "plt.show()"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "59xk22YYyxFw"
      },
      "source": [
        "plot_pacf(daily_confirm_diff5)\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "izbM6eEUKAV-"
      },
      "source": [
        "## **7-Data Spliting** ##"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zt-vctO1Kb3o"
      },
      "source": [
        " X = daily_confirm_diff5.values\n",
        "train, test = X[1:len(X)-7], X[len(X)-7:]"
      ],
      "execution_count": 25,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bVnQBEQv7fqd"
      },
      "source": [
        "##**8 -Autoregressive model**##\n",
        "Forecast the next timestamp value by regressing over previous vaues"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1x8QNZN3CAdF"
      },
      "source": [
        "### **8.1 - Models training on training Data** ###\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zNBdXxiHLfLW"
      },
      "source": [
        "from statsmodels.tsa.ar_model import AR"
      ],
      "execution_count": 26,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ulKq5MEC4QTC"
      },
      "source": [
        "model = AR(train)\n",
        "model_fit = model.fit()"
      ],
      "execution_count": 27,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5zfaTqgI4Vis"
      },
      "source": [
        "print('Lag: %s' % model_fit.k_ar)\n",
        "print('Coefficients: %s' % model_fit.params)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "g78HCuLVCAdG"
      },
      "source": [
        "### **8.2 - Model prediction on test data**### "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JwOLIUoIPjY1"
      },
      "source": [
        "predictions = model_fit.predict(start=len(train), end=len(train)+len(test)-1, dynamic=False)"
      ],
      "execution_count": 29,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YhWBShUX5JbT"
      },
      "source": [
        "for i in range(len(predictions)):\n",
        "    print('predicted=%f, expected=%f' % (predictions[i], test[i]))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_JBFgfGjdddE"
      },
      "source": [
        "### **8.3- Model Evaulation**###"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j9xnyArC5dAD"
      },
      "source": [
        "error = mean_squared_error(test, predictions)"
      ],
      "execution_count": 44,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o8744XRH5lhP"
      },
      "source": [
        "print('Test MSE: %.3f' % error)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "u2LsL4c6xkx9"
      },
      "source": [
        "from math import sqrt\n",
        "rmse = sqrt(error)"
      ],
      "execution_count": 46,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-42m-k0nynr0",
        "outputId": "41433ff7-a5a4-43b5-f426-fbe5873188d1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "print(rmse)"
      ],
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "10.495376345684212\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "y5RGFYis5t2y",
        "outputId": "be0b9ecf-4f80-4c51-9c51-cd3d5990e08a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 546
        }
      },
      "source": [
        "plt.plot(test)\n",
        "plt.plot(predictions, color='red')\n",
        "plt.show()\n",
        "plt.title ('Comparsion between prediction and actual ')"
      ],
      "execution_count": 50,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Comparsion between prediction and actual ')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 50
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "i_GstlsrCAdc"
      },
      "source": [
        "### **8.4 -Model tunning**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "e1DF698ECAdd"
      },
      "source": [
        "window = model_fit.k_ar\n",
        "coef = model_fit.params"
      ],
      "execution_count": 51,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wd_svSDs6mQ-"
      },
      "source": [
        "history = train[len(train)-window:]\n",
        "history = [history[i] for i in range(len(history))]\n",
        "predictions = list()"
      ],
      "execution_count": 52,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SqVFvGGY6qsn",
        "outputId": "120a4b58-bb38-4bd9-eaa0-bb8d05a735a0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "for t in range(len(test)):\n",
        "    length = len(history)\n",
        "    lag = [history[i] for i in range(length-window,length)]\n",
        "    yhat = coef[0]\n",
        "    for d in range(window):\n",
        "        yhat += coef[d+1] * lag[window-d-1]\n",
        "    obs = test[t]\n",
        "    predictions.append(yhat)\n",
        "    history.append(obs)\n",
        "    print('predicted=%f, expected=%f' % (yhat, obs))"
      ],
      "execution_count": 53,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "predicted=23.529196, expected=13.000000\n",
            "predicted=-9.773814, expected=-26.000000\n",
            "predicted=43.635559, expected=51.000000\n",
            "predicted=-57.484915, expected=-52.000000\n",
            "predicted=37.988487, expected=44.000000\n",
            "predicted=-33.853246, expected=-23.000000\n",
            "predicted=-12.596304, expected=0.000000\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "lTe0Qqi3CAeA"
      },
      "source": [
        "### **8.5 - Model Evaulation** ###"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fakWwJ-h6-tm"
      },
      "source": [
        "error = mean_squared_error(test, predictions)\n",
        "print('Test MSE: %.3f' % error)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "92DpeEyZzyuS"
      },
      "source": [
        "from math import sqrt\n",
        "rmse = sqrt(error)"
      ],
      "execution_count": 55,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EciwUx62z06U"
      },
      "source": [
        "print(rmse)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XOG_8GjO7Dya",
        "outputId": "6e1e5fef-3462-4265-b246-5d773f817ecb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 666
        }
      },
      "source": [
        "plt.plot(test)\n",
        "plt.plot(predictions, color='red')\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x840 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "05Gq2A7BCAeX"
      },
      "source": [
        "## **9 - Auto Regressive Integrated Moving Average (ARIMA)**\n",
        "\n",
        " Forecast the next timestamps value by averaging over the previous values \n",
        "  - useful for non-stationary data\n",
        "  - allows us to difference the data\n",
        "  - also includes a seasona differencing parameter for seasonal non-stationary data "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rWCpBVwViJPO"
      },
      "source": [
        "### **9.1 Models training on training Data**"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ld1_nNE_mDcW"
      },
      "source": [
        "#5,2,0\n",
        "model_ARIMA = ARIMA(train,order=(1,1,0))\n",
        "model_fit_ARIMA = model_ARIMA.fit(disp=0)"
      ],
      "execution_count": 57,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o_rzpp-3b1ak",
        "outputId": "72468b6b-949a-46f3-8041-af2bcd2955f4",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "print('Lag: %s' % model_fit_ARIMA.k_ar)\n",
        "print('Coefficients: %s' % model_fit_ARIMA.params)"
      ],
      "execution_count": 58,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Lag: 1\n",
            "Coefficients: [ 0.06451809 -0.85719485]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "collapsed": true,
        "id": "yN3QOkmDCAed"
      },
      "source": [
        "### **9.2 - Model prediction on test data.** ###\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "f9Mfj47gcrlu"
      },
      "source": [
        "predictions_ARIMA = model_fit_ARIMA.predict(start=len(train), end=len(train)+len(test)-1, dynamic=False)"
      ],
      "execution_count": 59,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NaIYvkLwc97g"
      },
      "source": [
        "for i in range(len(predictions_ARIMA)):\n",
        "    print('predicted=%f, expected=%f' % (predictions_ARIMA[i], test[i]))"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MHVC12Sl0Xw1",
        "outputId": "c3fde3fc-5b34-4a28-a651-2e5fd24a425f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 384
        }
      },
      "source": [
        "# Another Mothod \n",
        "from statsmodels.tsa.arima_model import ARMA\n",
        "mod = ARMA(train, order=(0,1))\n",
        "res = mod.fit()\n",
        "res.plot_predict(start=1990, end=1991)\n",
        "plt.show()\n",
        "predictions_ARIMA = res.predict(start=len(train), end=len(train)+len(test)-1, dynamic=False)\n",
        "for i in range(len(predictions_ARIMA)):\n",
        "    print('predicted=%f, expected=%f' % (predictions_ARIMA[i], test[i]))\n"
      ],
      "execution_count": 61,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "predicted=11.348462, expected=13.000000\n",
            "predicted=0.000044, expected=-26.000000\n",
            "predicted=0.000044, expected=51.000000\n",
            "predicted=0.000044, expected=-52.000000\n",
            "predicted=0.000044, expected=44.000000\n",
            "predicted=0.000044, expected=-23.000000\n",
            "predicted=0.000044, expected=0.000000\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9q5Vg5WldwA0"
      },
      "source": [
        "###7.3 Model Evaulation##"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qNfne-7Ud2gl"
      },
      "source": [
        "error_ARIMA = mean_squared_error(test, predictions_ARIMA)"
      ],
      "execution_count": 62,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ofv08pr_eKUP",
        "outputId": "60cbb863-b8be-427c-a0ab-4c32d5bbaec4",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "print('Test MSE: %.3f' % error_ARIMA)"
      ],
      "execution_count": 63,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Test MSE: 1206.961\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zFm8Zkiu0DK0"
      },
      "source": [
        "from math import sqrt\n",
        "rmse_Fin = sqrt(error_ARIMA)"
      ],
      "execution_count": 64,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MpFVFOwF0N5r",
        "outputId": "b002e67d-3f7a-430a-cd6f-914d9ce0eaad",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "source": [
        "print(rmse_Fin)"
      ],
      "execution_count": 65,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "34.74134652195682\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vm7uTbEWCAei"
      },
      "source": [
        "References:\n",
        "\n",
        "Autoregressive¶\n",
        "https://www.investopedia.com/terms/a/autoregressive.asp\n",
        "\n",
        "Autoregressive Models\n",
        "https://online.stat.psu.edu/stat501/lesson/14/14.1#:~:text=An%20autoregressive%20model%20is%20when,from%20that%20same%20time%20series.&text=The%20order%20of%20an%20autoregression,value%20at%20the%20present%20time.\n",
        "\n",
        "ARIMA modeling and forecasting: Time Series in Python Part 2\n",
        "https://tutorials.datasciencedojo.com/arima-model-time-series-python/"
      ]
    }
  ]
}