SVR - dataset creation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dataset creation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Train: 12-2017 - 11-2019  \n",
    "Test: 12-2019 - 05-2020"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "DatetimeIndex: 21889 entries, 2017-12-01 00:00:00+00:00 to 2020-05-31 00:00:00+00:00\n",
      "Data columns (total 34 columns):\n",
      " #   Column              Non-Null Count  Dtype  \n",
      "---  ------              --------------  -----  \n",
      " 0   demand              21889 non-null  float64\n",
      " 1   grid1-load          21889 non-null  float64\n",
      " 2   grid1-loss          21889 non-null  float64\n",
      " 3   grid1-temp          21889 non-null  float64\n",
      " 4   season_x            21889 non-null  float64\n",
      " 5   season_y            21889 non-null  float64\n",
      " 6   month_x             21889 non-null  float64\n",
      " 7   month_y             21889 non-null  float64\n",
      " 8   week_x              21889 non-null  float64\n",
      " 9   week_y              21889 non-null  float64\n",
      " 10  weekday_x           21889 non-null  float64\n",
      " 11  weekday_y           21889 non-null  float64\n",
      " 12  holiday             21889 non-null  bool   \n",
      " 13  hour_x              21889 non-null  float64\n",
      " 14  hour_y              21889 non-null  float64\n",
      " 15  has incorrect data  21889 non-null  bool   \n",
      " 16  temperature         21889 non-null  float64\n",
      " 17  dew_point           21889 non-null  float64\n",
      " 18  pressure            21889 non-null  float64\n",
      " 19  ground_pressure     21889 non-null  float64\n",
      " 20  humidity            21889 non-null  float64\n",
      " 21  clouds              21889 non-null  float64\n",
      " 22  wind_speed          21889 non-null  float64\n",
      " 23  wind_deg            21889 non-null  float64\n",
      " 24  rain                21889 non-null  float64\n",
      " 25  snow                21889 non-null  float64\n",
      " 26  ice                 21889 non-null  float64\n",
      " 27  fr_rain             21889 non-null  float64\n",
      " 28  convective          21889 non-null  float64\n",
      " 29  snow_depth          21889 non-null  float64\n",
      " 30  accumulated         21889 non-null  float64\n",
      " 31  hours               21889 non-null  float64\n",
      " 32  rate                21889 non-null  float64\n",
      " 33  probability         21889 non-null  float64\n",
      "dtypes: bool(2), float64(32)\n",
      "memory usage: 5.6 MB\n"
     ]
    }
   ],
   "source": [
    "test = pd.read_csv('../data/test_ow.csv', parse_dates=['datetime'], index_col='datetime')\n",
    "train = pd.read_csv('../data/train_ow.csv', parse_dates=['datetime'], index_col='datetime')\n",
    "dataset = pd.concat([train, test])\n",
    "dataset.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset['grid1-load'] = dataset['grid1-load'].shift(168)\n",
    "dataset = dataset.iloc[168:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset.to_csv('../data/dataset_all.csv') # save dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create datasets for individual hours"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = pd.read_csv('../data/dataset_all.csv',parse_dates=['datetime'],index_col='datetime')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "DatetimeIndex: 21721 entries, 2017-12-08 00:00:00+00:00 to 2020-05-31 00:00:00+00:00\n",
      "Data columns (total 34 columns):\n",
      " #   Column              Non-Null Count  Dtype  \n",
      "---  ------              --------------  -----  \n",
      " 0   demand              21721 non-null  float64\n",
      " 1   grid1-load          21721 non-null  float64\n",
      " 2   grid1-loss          21721 non-null  float64\n",
      " 3   grid1-temp          21721 non-null  float64\n",
      " 4   season_x            21721 non-null  float64\n",
      " 5   season_y            21721 non-null  float64\n",
      " 6   month_x             21721 non-null  float64\n",
      " 7   month_y             21721 non-null  float64\n",
      " 8   week_x              21721 non-null  float64\n",
      " 9   week_y              21721 non-null  float64\n",
      " 10  weekday_x           21721 non-null  float64\n",
      " 11  weekday_y           21721 non-null  float64\n",
      " 12  holiday             21721 non-null  bool   \n",
      " 13  hour_x              21721 non-null  float64\n",
      " 14  hour_y              21721 non-null  float64\n",
      " 15  has incorrect data  21721 non-null  bool   \n",
      " 16  temperature         21721 non-null  float64\n",
      " 17  dew_point           21721 non-null  float64\n",
      " 18  pressure            21721 non-null  float64\n",
      " 19  ground_pressure     21721 non-null  float64\n",
      " 20  humidity            21721 non-null  float64\n",
      " 21  clouds              21721 non-null  float64\n",
      " 22  wind_speed          21721 non-null  float64\n",
      " 23  wind_deg            21721 non-null  float64\n",
      " 24  rain                21721 non-null  float64\n",
      " 25  snow                21721 non-null  float64\n",
      " 26  ice                 21721 non-null  float64\n",
      " 27  fr_rain             21721 non-null  float64\n",
      " 28  convective          21721 non-null  float64\n",
      " 29  snow_depth          21721 non-null  float64\n",
      " 30  accumulated         21721 non-null  float64\n",
      " 31  hours               21721 non-null  float64\n",
      " 32  rate                21721 non-null  float64\n",
      " 33  probability         21721 non-null  float64\n",
      "dtypes: bool(2), float64(32)\n",
      "memory usage: 5.5 MB\n"
     ]
    }
   ],
   "source": [
    "dataset.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset.drop('has incorrect data', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "FOLDER_NAME = 'datasets'\n",
    "\n",
    "import os\n",
    "if not os.path.exists(f'../data/{FOLDER_NAME}'):\n",
    "    os.mkdir(f'../data/{FOLDER_NAME}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "00:00 - season_y (0.555), snow_depth (0.63), season_x (0.743), week_x (0.759), month_x (0.792), grid1-temp (0.868), grid1-load (0.873), dew_point (0.88), temperature (0.886), demand (0.947)\n",
      "01:00 - season_y (0.544), snow_depth (0.601), season_x (0.737), week_x (0.752), month_x (0.785), grid1-load (0.869), grid1-temp (0.874), dew_point (0.882), temperature (0.886), demand (0.947)\n",
      "02:00 - season_y (0.537), snow_depth (0.576), season_x (0.733), week_x (0.747), month_x (0.781), grid1-load (0.866), grid1-temp (0.878), dew_point (0.882), temperature (0.886), demand (0.947)\n",
      "03:00 - snow_depth (0.52), season_y (0.535), season_x (0.728), week_x (0.741), month_x (0.773), grid1-load (0.861), grid1-temp (0.88), dew_point (0.882), temperature (0.888), demand (0.945)\n",
      "04:00 - season_y (0.524), season_x (0.711), week_x (0.718), month_x (0.748), grid1-load (0.848), dew_point (0.874), grid1-temp (0.879), temperature (0.883), demand (0.935)\n",
      "05:00 - season_x (0.689), week_x (0.702), month_x (0.725), dew_point (0.846), grid1-load (0.852), grid1-temp (0.855), temperature (0.858), demand (0.929)\n",
      "06:00 - season_x (0.69), week_x (0.727), month_x (0.746), dew_point (0.829), grid1-temp (0.842), temperature (0.844), grid1-load (0.87), demand (0.945)\n",
      "07:00 - season_x (0.706), week_x (0.751), month_x (0.768), dew_point (0.834), grid1-temp (0.852), temperature (0.854), grid1-load (0.872), demand (0.95)\n",
      "08:00 - season_x (0.74), week_x (0.777), month_x (0.792), dew_point (0.843), grid1-temp (0.864), temperature (0.868), grid1-load (0.873), demand (0.951)\n",
      "09:00 - snow_depth (0.501), season_x (0.769), week_x (0.797), month_x (0.81), dew_point (0.846), grid1-temp (0.87), temperature (0.875), grid1-load (0.876), demand (0.952)\n",
      "10:00 - snow_depth (0.543), season_x (0.783), week_x (0.811), month_x (0.822), dew_point (0.847), grid1-temp (0.874), grid1-load (0.875), temperature (0.877), demand (0.953)\n",
      "11:00 - snow_depth (0.552), season_x (0.791), week_x (0.819), month_x (0.828), dew_point (0.841), grid1-temp (0.87), temperature (0.873), grid1-load (0.877), demand (0.954)\n",
      "12:00 - snow_depth (0.636), season_x (0.794), week_x (0.824), month_x (0.832), dew_point (0.837), grid1-temp (0.871), temperature (0.871), grid1-load (0.877), demand (0.955)\n",
      "13:00 - snow_depth (0.628), season_x (0.799), week_x (0.833), dew_point (0.836), month_x (0.838), temperature (0.873), grid1-temp (0.878), grid1-load (0.878), demand (0.955)\n",
      "14:00 - snow_depth (0.648), season_x (0.806), dew_point (0.835), week_x (0.842), month_x (0.843), temperature (0.876), grid1-load (0.881), grid1-temp (0.882), demand (0.957)\n",
      "15:00 - season_x (0.818), dew_point (0.837), week_x (0.847), month_x (0.847), temperature (0.876), grid1-temp (0.881), grid1-load (0.891), demand (0.961)\n",
      "16:00 - season_x (0.821), dew_point (0.844), week_x (0.845), month_x (0.847), temperature (0.878), grid1-temp (0.882), grid1-load (0.895), demand (0.962)\n",
      "17:00 - season_x (0.814), week_x (0.832), month_x (0.839), dew_point (0.852), temperature (0.878), grid1-temp (0.881), grid1-load (0.893), demand (0.961)\n",
      "18:00 - season_y (0.503), season_x (0.792), week_x (0.82), month_x (0.832), dew_point (0.859), temperature (0.879), grid1-temp (0.88), grid1-load (0.894), demand (0.96)\n",
      "19:00 - month_y (0.502), season_y (0.52), season_x (0.774), week_x (0.81), month_x (0.827), dew_point (0.863), temperature (0.879), grid1-temp (0.88), grid1-load (0.894), demand (0.959)\n",
      "20:00 - season_y (0.54), season_x (0.768), week_x (0.801), month_x (0.822), dew_point (0.863), temperature (0.877), grid1-temp (0.879), grid1-load (0.893), demand (0.958)\n",
      "21:00 - season_y (0.56), season_x (0.772), week_x (0.794), month_x (0.819), dew_point (0.862), temperature (0.875), grid1-temp (0.877), grid1-load (0.892), demand (0.956)\n",
      "22:00 - season_y (0.562), season_x (0.756), week_x (0.786), month_x (0.813), dew_point (0.867), temperature (0.878), grid1-temp (0.878), grid1-load (0.884), demand (0.951)\n",
      "23:00 - season_y (0.553), season_x (0.752), week_x (0.78), month_x (0.81), dew_point (0.876), grid1-temp (0.881), grid1-load (0.883), temperature (0.885), demand (0.952)\n"
     ]
    }
   ],
   "source": [
    "# calculate linear and nonlinear correlations\n",
    "def correls(dataset):\n",
    "    corrs = dataset.loc[:'11-2019'].corr().sort_values('grid1-loss')[['grid1-loss']].rename({'grid1-loss':'linear'},axis=1)\n",
    "    corrs['log'] = dataset.loc[:'11-2019'].apply(np.log).corr().sort_values('grid1-loss')[['grid1-loss']]\n",
    "    corrs['corr'] = corrs.apply(lambda x: max(abs(x)),axis=1)\n",
    "    return corrs.sort_values('corr',ascending=False)[['corr']]\n",
    "\n",
    "corrrs = pd.DataFrame()\n",
    "for hour in range(24):\n",
    "    corrrs = pd.concat([corrrs,correls(dataset[(dataset.index.hour == hour)].loc[:'11-2019']).rename({'corr':f'{str(hour).zfill(2)}:00'},axis=1)],axis=1)\n",
    "        \n",
    "# create datasets for individual minutes from correlations > +-0.5\n",
    "for hour in range(24):\n",
    "    c = corrrs[f'{str(hour).zfill(2)}:00'].apply(abs).sort_values().dropna().reset_index()\n",
    "    k = c[(c[f'{str(hour).zfill(2)}:00']>.5) & (c[f'{str(hour).zfill(2)}:00']<1)] if len(c[c[f'{str(hour).zfill(2)}:00']>.5]) > 2 else c[-3:]\n",
    "    print(f'{str(hour).zfill(2)}:00 - {\", \".join([f\"{i[1][0]} ({i[1][1]:.3})\" for i in k.iterrows()])}')\n",
    "    d = dataset[(dataset.index.hour == hour)][['grid1-loss',*[f\"{i[1][0]}\" for i in k.iterrows()]]]\n",
    "    d.to_csv(f'../data/{FOLDER_NAME}/{str(hour).zfill(2)}_00.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Add EWMA feature"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "FOLDER_NAME = 'datasets'\n",
    "FOLDER_NAME2 = 'datasets_ewma'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "for hour in range(24):\n",
    "    data = pd.read_csv(f'../data/{FOLDER_NAME}/{hour:02}_00.csv',parse_dates=['datetime'],index_col='datetime')\n",
    "    data.index.freq = 'D'\n",
    "    data['ewma'] = SimpleExpSmoothing(data['grid1-loss']).fit(\n",
    "        smoothing_level=SimpleExpSmoothing(data['grid1-loss'].loc[:'11-2019']).fit().params['smoothing_level'], \n",
    "        optimized=False).fittedvalues\n",
    "    data.to_csv(f'../data/{FOLDER_NAME2}/{hour:02}_00.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:xlabel='datetime'>"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data[['grid1-loss','ewma']].plot()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.10",
   "language": "python",
   "name": "py310"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}