02_12_2023 Beginners' pandas for Time series forecasting.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dea37193",
   "metadata": {},
   "outputs": [],
   "source": [
    "#!/usr/bin/env python3\n",
    "# -*- coding: utf-8 -*-\n",
    "\"\"\"\n",
    "Created on Wed Jul 14 13:37:58 2021\n",
    "@author: sadrachpierre\n",
    "\"\"\"\n",
    "\n",
    "\n",
    "import pandas as pd \n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns \n",
    "from statsmodels.tsa.stattools import adfuller\n",
    "from statsmodels.tsa.seasonal import seasonal_decompose\n",
    "\n",
    "df = pd.read_csv(\"AirPassengers.csv\")\n",
    "\n",
    "print(df.head())\n",
    "\n",
    "print(df.tail())\n",
    "\n",
    "df['Month'] = pd.to_datetime(df['Month'], format='%Y-%m')\n",
    "\n",
    "\n",
    "\n",
    "df.index = df['Month']\n",
    "del df['Month']\n",
    "print(df.head())\n",
    "\n",
    "\n",
    "# sns.lineplot(data=df)\n",
    "# plt.ylabel(\"Number of Passengers\")\n",
    "# plt.show()\n",
    "\n",
    "rolling_mean = df.rolling(7).mean()\n",
    "rolling_std = df.rolling(7).std()\n",
    "\n",
    "#!/usr/bin/env python3\n",
    "# -*- coding: utf-8 -*-\n",
    "\"\"\"\n",
    "Created on Wed Jul 14 13:37:58 2021\n",
    "@author: sadrachpierre\n",
    "\"\"\"\n",
    "\n",
    "\n",
    "import pandas as pd \n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns \n",
    "from statsmodels.tsa.stattools import adfuller\n",
    "from statsmodels.tsa.seasonal import seasonal_decompose\n",
    "\n",
    "df = pd.read_csv(\"AirPassengers.csv\")\n",
    "\n",
    "print(df.head())\n",
    "\n",
    "print(df.tail())\n",
    "\n",
    "df['Month'] = pd.to_datetime(df['Month'], format='%Y-%m')\n",
    "\n",
    "\n",
    "\n",
    "df.index = df['Month']\n",
    "del df['Month']\n",
    "print(df.head())\n",
    "\n",
    "\n",
    "# sns.lineplot(data=df)\n",
    "# plt.ylabel(\"Number of Passengers\")\n",
    "# plt.show()\n",
    "\n",
    "rolling_mean = df.rolling(7).mean()\n",
    "rolling_std = df.rolling(7).std()\n",
    "\n",
    "\n",
    "plt.plot(df, color=\"blue\",label=\"Original Passenger Data\")\n",
    "plt.plot(rolling_mean, color=\"red\", \n",
    "         label=\"Rolling Mean #Passenger\")\n",
    "plt.plot(rolling_std, color=\"black\",\n",
    "         label = \"Rolling Standard Deviation in #Passenger\")\n",
    "plt.title(\"Passenger Time Series, \\\n",
    "         Rolling Mean, Standard Deviation\")\n",
    "plt.legend(loc=\"best\")\n",
    "plt.show()\n",
    "\n",
    "adft = adfuller(df,autolag=\"AIC\")\n",
    " \n",
    "output_df = pd.DataFrame({\n",
    "    \"Values\":[adft[0],adft[1],adft[2],adft[3],\n",
    "              adft[4]['1%'], adft[4]['5%'], adft[4]['10%']]\n",
    "    , \"Metric\":[\"Test Statistics\",\"p-value\",\n",
    "                \"No. of lags used\",\"Number of observations used\", \n",
    "                \"critical value (1%)\", \"critical value (5%)\",\n",
    "                \"critical value (10%)\"]})\n",
    "print(output_df)\n",
    "\n",
    "\n",
    "autocorrelation_lag1 = df['#Passengers'].autocorr(lag=1)\n",
    "print(\"One Month Lag: \", autocorrelation_lag1)\n",
    "\n",
    "autocorrelation_lag3 = df['#Passengers'].autocorr(lag=3)\n",
    "print(\"Three Month Lag: \", autocorrelation_lag3)\n",
    "\n",
    "autocorrelation_lag6 = df['#Passengers'].autocorr(lag=6)\n",
    "print(\"Six Month Lag: \", autocorrelation_lag6)\n",
    "\n",
    "autocorrelation_lag9 = df['#Passengers'].autocorr(lag=9)\n",
    "print(\"Nine Month Lag: \", autocorrelation_lag9)\n",
    "\n",
    "\n",
    "decompose = seasonal_decompose(df['#Passengers'],\n",
    "                               model='additive', period=7)\n",
    "decompose.plot()\n",
    "plt.show()\n",
    "df['Date'] = df.index\n",
    "\n",
    "train = df[df['Date'] < pd.to_datetime(\"1960-08\", format='%Y-%m')]\n",
    "train['train'] = train['#Passengers']\n",
    "del train['Date']\n",
    "del train['#Passengers']\n",
    "test = df[df['Date'] >= pd.to_datetime(\"1960-08\", format='%Y-%m')]\n",
    "del test['Date']\n",
    "test['test'] = test['#Passengers']\n",
    "del test['#Passengers']\n",
    "plt.plot(train, color = \"black\")\n",
    "plt.plot(test, color = \"red\")\n",
    "plt.title(\"Train/Test split for Passenger Data\")\n",
    "plt.ylabel(\"Passenger Number\")\n",
    "plt.xlabel('Year-Month')\n",
    "sns.set()\n",
    "plt.show()\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "from pmdarima.arima import auto_arima\n",
    "model = auto_arima(train, trace=True, \n",
    "                   error_action='ignore', suppress_warnings=True)\n",
    "model.fit(train)\n",
    "forecast = model.predict(n_periods=len(test))\n",
    "forecast = pd.DataFrame(forecast,\n",
    "                        index = test.index,columns=['Prediction'])\n",
    "\n",
    "\n",
    "\n",
    "plt.plot(train, label='Train')\n",
    "plt.plot(test, label='Test')\n",
    "plt.plot(forecast, label='Prediction')\n",
    "plt.title('#Passenger Prediction')\n",
    "plt.xlabel('Date')\n",
    "plt.ylabel('Actual #Passenger')\n",
    "plt.legend(loc='upper left', fontsize=8)\n",
    "plt.show()\n",
    "\n",
    "\n",
    "\n",
    "from math import sqrt\n",
    "from sklearn.metrics import mean_squared_error\n",
    "print(\"RMSE: \", rms)\n",
    "plt.plot(df, color=\"blue\",label=\"Original Passenger Data\")\n",
    "plt.plot(rolling_mean, color=\"red\",\n",
    "         label=\"Rolling Mean #Passenger\")\n",
    "plt.plot(rolling_std, color=\"black\", label = \n",
    "         \"Rolling Standard Deviation in #Passenger\")\n",
    "plt.title(\"Passenger Time Series, Rolling Mean, \\\n",
    "          Standard Deviation\")\n",
    "plt.legend(loc=\"best\")\n",
    "plt.show()\n",
    "\n",
    "adft = adfuller(df,autolag=\"AIC\")\n",
    " \n",
    "output_df = pd.DataFrame({\"Values\":[\n",
    "    adft[0],adft[1],adft[2],adft[3], adft[4]['1%'], \n",
    "    adft[4]['5%'], adft[4]['10%']]  , \n",
    "    \"Metric\":[\"Test Statistics\",\"p-value\",\n",
    "    \"No. of lags used\",\"Number of observations used\", \n",
    "    \"critical value (1%)\", \"critical value (5%)\", \n",
    "              \"critical value (10%)\"]})\n",
    "print(output_df)\n",
    "\n",
    "\n",
    "autocorrelation_lag1 = df['#Passengers'].autocorr(lag=1)\n",
    "print(\"One Month Lag: \", autocorrelation_lag1)\n",
    "\n",
    "autocorrelation_lag3 = df['#Passengers'].autocorr(lag=3)\n",
    "print(\"Three Month Lag: \", autocorrelation_lag3)\n",
    "\n",
    "autocorrelation_lag6 = df['#Passengers'].autocorr(lag=6)\n",
    "print(\"Six Month Lag: \", autocorrelation_lag6)\n",
    "\n",
    "autocorrelation_lag9 = df['#Passengers'].autocorr(lag=9)\n",
    "print(\"Nine Month Lag: \", autocorrelation_lag9)\n",
    "\n",
    "\n",
    "decompose = seasonal_decompose(df['#Passengers'],\n",
    "                               model='additive', period=7)\n",
    "decompose.plot()\n",
    "plt.show()\n",
    "df['Date'] = df.index\n",
    "\n",
    "train = df[df['Date'] < pd.to_datetime(\"1960-08\", format='%Y-%m')]\n",
    "train['train'] = train['#Passengers']\n",
    "del train['Date']\n",
    "del train['#Passengers']\n",
    "test = df[df['Date'] >= pd.to_datetime(\"1960-08\", format='%Y-%m')]\n",
    "del test['Date']\n",
    "test['test'] = test['#Passengers']\n",
    "del test['#Passengers']\n",
    "plt.plot(train, color = \"black\")\n",
    "plt.plot(test, color = \"red\")\n",
    "plt.title(\"Train/Test split for Passenger Data\")\n",
    "plt.ylabel(\"Passenger Number\")\n",
    "plt.xlabel('Year-Month')\n",
    "sns.set()\n",
    "plt.show()\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "from pmdarima.arima import auto_arima\n",
    "model = auto_arima(train, trace=True, error_action='ignore',\n",
    "                   suppress_warnings=True)\n",
    "model.fit(train)\n",
    "forecast = model.predict(n_periods=len(test))\n",
    "forecast = pd.DataFrame(forecast,index = test.index,\n",
    "                        columns=['Prediction'])\n",
    "\n",
    "\n",
    "\n",
    "plt.plot(train, label='Train')\n",
    "plt.plot(test, label='Test')\n",
    "plt.plot(forecast, label='Prediction')\n",
    "plt.title('#Passenger Prediction')\n",
    "plt.xlabel('Date')\n",
    "plt.ylabel('Actual #Passenger')\n",
    "plt.legend(loc='upper left', fontsize=8)\n",
    "plt.show()\n",
    "\n",
    "\n",
    "\n",
    "from math import sqrt\n",
    "from sklearn.metrics import mean_squared_error\n",
    "print(\"RMSE: \", rms)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}