timeseries_forecasting.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "\n",
    "# Laboratory work. Timeseries forecasting.\n",
    "\n",
    "*Maslov Aleksandr* <br/>*Group 5197*\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Task 1\n",
    "Make a primary analysis of the data _\"E_RRC_SETUP_FR_LTE\"_  and to visualize the time series with help of _matplotlib_ or _seaborn_. Visualize a slice of data over a short period of time. Make a histogram of considered data. Build a box-plot of the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "KPI = \"E_RRC_SETUP_FR\"\n",
    "data = pd.read_excel('E_RRC_SETUP_FR_LTE.xlsx',\n",
    "                     skiprows=0,\n",
    "                     usecols='A:B')\n",
    "data['DT'] = data['DT'].ffill()\n",
    "data['E_RRC_SETUP_FR'] = data['E_RRC_SETUP_FR'].ffill()\n",
    "data[:10]\n",
    "data['month'] = pd.DatetimeIndex(data['DT']).month"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax = fig.add_subplot(111)\n",
    "ax.boxplot([(data.loc[data['month'] == 12])['E_RRC_SETUP_FR'],\n",
    "            (data.loc[data['month'] == 11])['E_RRC_SETUP_FR'],\n",
    "            (data.loc[data['month'] == 10])['E_RRC_SETUP_FR'],\n",
    "            (data.loc[data['month'] == 9])['E_RRC_SETUP_FR'],\n",
    "            (data.loc[data['month'] == 8])['E_RRC_SETUP_FR'],\n",
    "            (data.loc[data['month'] == 7])['E_RRC_SETUP_FR'],\n",
    "            (data.loc[data['month'] == 6])['E_RRC_SETUP_FR'],\n",
    "            (data.loc[data['month'] == 5])['E_RRC_SETUP_FR'],\n",
    "            (data.loc[data['month'] == 4])['E_RRC_SETUP_FR'],\n",
    "            (data.loc[data['month'] == 3])['E_RRC_SETUP_FR'],\n",
    "            (data.loc[data['month'] == 2])['E_RRC_SETUP_FR'],\n",
    "            (data.loc[data['month'] == 1])['E_RRC_SETUP_FR']],\n",
    "            labels=['Dec', 'Nov', 'Oct', 'Sep', 'Aug', 'Jul','Jun', 'May', 'Apr', 'Mar', 'Feb', 'Jan'], vert=True)\n",
    "fig.set_figwidth(13)    \n",
    "fig.set_figheight(10)\n",
    "plt.xticks(np.linspace(1,13,13))\n",
    "plt.xlim([0,13])\n",
    "plt.xlabel(\"Month\")\n",
    "plt.ylabel(\"E_RRC_SETUP_FR\")\n",
    "plt.title(\"Month-wise Box plot\")\n",
    "plt.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "ax.hist(data[\"E_RRC_SETUP_FR\"], 2000, density=True, facecolor='b', alpha=0.7)\n",
    "plt.grid()\n",
    "plt.xlabel('E_RRC_SETUP_FR')\n",
    "plt.ylabel('Probability (%)')\n",
    "plt.xlim([0,0.4])\n",
    "fig.set_figwidth(20)\n",
    "fig.set_figheight(10)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    8761.000000\n",
       "mean        0.239066\n",
       "std         0.047830\n",
       "min         0.107856\n",
       "25%         0.204716\n",
       "50%         0.247050\n",
       "75%         0.276292\n",
       "max         0.484687\n",
       "Name: E_RRC_SETUP_FR, dtype: float64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data['E_RRC_SETUP_FR'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "ax = fig.add_subplot(1, 1, 1)\n",
    "ax.plot(data[\"DT\"][0:97], data[\"E_RRC_SETUP_FR\"][0:97])\n",
    "plt.grid()\n",
    "plt.xlabel('Date')\n",
    "plt.ylabel('E_RRC_SETUP_FR (%)')\n",
    "plt.title('E_RRC_SETUP_FR (from 01-01-2016 00:00 till 01-05-2016 00:00)')\n",
    "fig.set_figwidth(20)    \n",
    "fig.set_figheight(10)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Task 2\n",
    "For second task it is needs to be used Holt-Winter's method and XGboost for time series forecasting.\n",
    "So, we will do several preparations for this, exactly, define features and special class for models.\n",
    "We using sklearn library calculate time series cross-validation: calculate mean squared error (MSE), root mean squared error (RMSE), mean absolute error (MAE), median absolute error (Median AE) for HW and XGboost."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: xgboost==1.6.1 in ./anaconda3/lib/python3.7/site-packages (1.6.1)\n",
      "Requirement already satisfied: numpy in ./anaconda3/lib/python3.7/site-packages (from xgboost==1.6.1) (1.18.1)\n",
      "Requirement already satisfied: scipy in ./anaconda3/lib/python3.7/site-packages (from xgboost==1.6.1) (1.4.1)\n",
      "\n",
      "HW\n",
      "Median AE:  0.08378666566276978\n",
      "MAE:  0.0888880125309446\n",
      "MSE:  0.011920239280743676\n",
      "RMSE:  0.10321263782310791\n",
      "\n",
      "XGBoost\n",
      "Median AE:  0.015156277398268381\n",
      "MAE:  0.017786512285846837\n",
      "MSE:  0.0005470473658980282\n",
      "RMSE:  0.023008872781773798\n"
     ]
    }
   ],
   "source": [
    "\n",
    "!pip install xgboost==1.6.1\n",
    "from statsmodels.tsa.holtwinters import ExponentialSmoothing\n",
    "from xgboost import XGBRegressor, plot_importance\n",
    "def create_features(data, label=None):\n",
    "    \n",
    "    \"\"\"\n",
    "    Creates time series features from datetime index.\n",
    "    \n",
    "    Parameters\n",
    "    ----------\n",
    "    df: pandas.DataFrame\n",
    "        Input data frame.\n",
    "    label: str (default None)\n",
    "        Name of the considered KPI.\n",
    "\n",
    "    Returns\n",
    "    -------\n",
    "    pandas.DataFrame\n",
    "        Output data frame.\n",
    "    \"\"\"\n",
    "    \n",
    "    new_df = data.copy()\n",
    "    new_df.index = data[\"DT\"]\n",
    "    new_df['date'] = new_df.index\n",
    "    new_df['hour'] = new_df['date'].dt.hour\n",
    "    new_df['day_of_year'] = new_df['date'].dt.dayofyear\n",
    "\n",
    "    X = new_df[['hour', 'day_of_year']]\n",
    "\n",
    "    #new_df['day_of_week'] = new_df['date'].dt.dayofweek\n",
    "    #new_df['quarter'] = new_df['date'].dt.quarter\n",
    "    #new_df['month'] = new_df['date'].dt.month\n",
    "    #new_df['day_of_month'] = new_df['date'].dt.day\n",
    "    #new_df['week_of_year'] = new_df['date'].dt.isocalendar().week\n",
    "    \n",
    "    #X = new_df[['hour','day_of_week','quarter','month',\n",
    "    #       'day_of_year','day_of_month','week_of_year']]\n",
    "        \n",
    "    if label:\n",
    "        y = data[label]\n",
    "        return X, y\n",
    "    return X\n",
    "class ModelTrainer:\n",
    "    \"\"\"\n",
    "    Class for models training automatization.\n",
    "\n",
    "    Attributes\n",
    "    ----------\n",
    "    model_type: str\n",
    "    One of the available models name or abrivation:\n",
    "      - 'HW' (Holt-Winters model)\n",
    "      - 'SARIMA'\n",
    "      - 'SARIMAX'\n",
    "      - 'SARIMA (Kalman)'\n",
    "      - 'SARIMAX (Kalman)'\n",
    "      - 'DLM' (Dynamical Linear Models)\n",
    "      - 'SVR'\n",
    "      - 'XGboost'\n",
    "    \"\"\"\n",
    "    def __init__(self, \n",
    "                model_type, \n",
    "                scale_target=False):\n",
    "      \n",
    "      self.model_type = model_type\n",
    "\n",
    "    @staticmethod\n",
    "    def __holt_winters_fp(y_train, X_test, **model_kwargs):\n",
    "        \"\"\" Holt-Winter's method (triple exponential smoothing) \"\"\"\n",
    "\n",
    "        old_index = X_test.index\n",
    "        X_test.reset_index(inplace=True)\n",
    "\n",
    "        seasonal_periods = model_kwargs.get(\"seasonal_periods\", 24)\n",
    "        model = ExponentialSmoothing(y_train,\n",
    "                                     trend=\"add\",\n",
    "                                     seasonal=\"add\", \n",
    "                                     seasonal_periods=seasonal_periods).fit()\n",
    "\n",
    "        predictions = model.predict(start=X_test.index[0], end=X_test.index[-1])\n",
    "        X_test.set_index(old_index, inplace=True)\n",
    "        return predictions\n",
    "\n",
    "    \n",
    "    def fit_predict(self, \n",
    "                  y_train,\n",
    "                  X_test, \n",
    "                  X_train=None,\n",
    "                  **model_kwargs):\n",
    "\n",
    "        \"\"\" \n",
    "        Fits the model and predicts the values \n",
    "\n",
    "        Parameters\n",
    "        ----------\n",
    "\n",
    "        y_train: 1-D array\n",
    "          Training target vector.\n",
    "\n",
    "        X_test: 2-D array\n",
    "          Testing features matrix.\n",
    "\n",
    "        X_test: 2-D array\n",
    "          Training features matrix.\n",
    "\n",
    "        Returns\n",
    "        -------\n",
    "\n",
    "        predictions: 1-D array\n",
    "          Predicted values.\n",
    "\n",
    "        \"\"\"\n",
    "\n",
    "\n",
    "        if self.model_type == \"HW\":\n",
    "            predictions = self.__holt_winters_fp(y_train, \n",
    "                                               X_test, \n",
    "                                               **model_kwargs)\n",
    "\n",
    "       \n",
    "        elif self.model_type == \"XGBoost\":\n",
    "            model = XGBRegressor(n_estimators=1000).fit(X_train, y_train)\n",
    "            predictions = model.predict(X_test)\n",
    "\n",
    "        return predictions\n",
    "\n",
    "    \n",
    "\n",
    "from sklearn.metrics import (mean_absolute_error, \n",
    "                             median_absolute_error, \n",
    "                             mean_squared_error)\n",
    "from sklearn.model_selection import TimeSeriesSplit\n",
    "\n",
    "\n",
    "def timeseries_cv_score(series, \n",
    "                        KPI,\n",
    "                        model_name,\n",
    "                        n_splits,\n",
    "                        sc_X = None,\n",
    "                        sc_y = None,\n",
    "                        on_index=True,\n",
    "                        **model_kwargs):\n",
    "\n",
    "    \"\"\" Time series cross validation. \n",
    "        This function prints out the following scores:\n",
    "          - Median Absolute Error\n",
    "          - Mean Absolute Error (MAE)\n",
    "          - Mean Squared Error (MSE)\n",
    "          - Root Mean Squared Error (RMSE)\n",
    "    \n",
    "    Parameters\n",
    "    __________\n",
    "    \n",
    "    series: pandas.DataFrame\n",
    "      Input data frame.\n",
    "    \n",
    "    KPI: str\n",
    "      Name of the column to use (considered Key Performance Identifier).\n",
    "    \n",
    "    model_name: object\n",
    "      inititalizes the instance of the explored model class\n",
    "    \n",
    "    n_splits: int\n",
    "      How many splits should be done for the cross-validation procedure.\n",
    "    \n",
    "    \"\"\"\n",
    "\n",
    "    print(f\"\\n{model_name}\")\n",
    "    \n",
    "    # errors array\n",
    "    median_aes = []\n",
    "    maes = []\n",
    "    mses = []\n",
    "    rmses = []    \n",
    "    \n",
    "    # set the number of folds for cross-validation\n",
    "    tscv = TimeSeriesSplit(n_splits=n_splits) \n",
    "    \n",
    "    # iterating over folds, train model on each, forecast and calculate error\n",
    "    for train_indexes, test_indexes in tscv.split(series):\n",
    "        \n",
    "        X_train, y_train = create_features(series.iloc[train_indexes], \n",
    "                                           label=KPI) # training sets\n",
    "        X_test, y_test = create_features(series.iloc[test_indexes], \n",
    "                                         label=KPI) # testing sets\n",
    "        \n",
    "        \n",
    "        # Fit the model and predict the values \n",
    "        predictions = ModelTrainer(model_name).fit_predict(y_train, \n",
    "                                                           X_test, \n",
    "                                                           X_train,\n",
    "                                                           **model_kwargs)\n",
    "\n",
    "        # Errors counting\n",
    "        \n",
    "        median_aes.append(median_absolute_error(predictions, y_test))\n",
    "        maes.append(mean_absolute_error(predictions, y_test))\n",
    "        mses.append(mean_squared_error(predictions, y_test))\n",
    "        rmses.append(mean_squared_error(predictions, y_test, squared=False))\n",
    "    \n",
    "    print(\"Median AE: \", np.mean(np.array(median_aes)))\n",
    "    print(\"MAE: \", np.mean(np.array(maes)))\n",
    "    print(\"MSE: \", np.mean(np.array(mses)))\n",
    "    print(\"RMSE: \", np.mean(np.array(rmses)))\n",
    "    \n",
    "    \n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "# ARIMA takes couple of hours\n",
    "\n",
    "#cases = [\"SARIMAX\", \"SARIMAX (Kalman)\", \"SARIMA\", \"SARIMA (Kalman)\", \n",
    "#         \"SVR\", \"HW\", \"XGBoost\", \"DLM\"]\n",
    "\n",
    "cases = [\"HW\", \"XGBoost\"]\n",
    "\n",
    "for model_name in cases:\n",
    "    timeseries_cv_score(data, KPI, model_name, n_splits=12)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}