trianta2/sarimax-debugging-daily-predictor.ipynb Secret

## sarimax-debugging-daily-predictor.ipynb
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/trianta/.pyenv/versions/miniconda3-latest/envs/py36/lib/python3.6/site-packages/statsmodels/compat/pandas.py:56: FutureWarning: The pandas.core.datetools module is deprecated and will be removed in a future version. Please use the pandas.tseries module instead.\n",
      "  from pandas.core import datetools\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import statsmodels.tsa.api as smt\n",
    "import statsmodels.api as sm\n",
    "from statsmodels.tsa.stattools import adfuller\n",
    "from statsmodels.tsa.statespace.sarimax import SARIMAX\n",
    "\n",
    "\n",
    "def ts_diagnostics(y, lags=None, title=''):\n",
    "    '''\n",
    "    Calculate acf, pacf, qq plot and Augmented Dickey Fuller test for a given time series\n",
    "\n",
    "    Modified from http://dacatay.com/data-science/part-3-time-series-stationarity-python/\n",
    "    '''\n",
    "    if not isinstance(y, pd.Series):\n",
    "        y = pd.Series(y)\n",
    "\n",
    "    # weekly moving averages (5 day window because of workdays)\n",
    "    rolling_mean = y.rolling(window=12).mean()\n",
    "    rolling_std = y.rolling(window=12).std()\n",
    "\n",
    "    plt.figure(figsize=(14, 12))\n",
    "    layout = (3, 2)\n",
    "    ts_ax = plt.subplot2grid(layout, (0, 0), colspan=2)\n",
    "    acf_ax = plt.subplot2grid(layout, (1, 0))\n",
    "    pacf_ax = plt.subplot2grid(layout, (1, 1))\n",
    "    qq_ax = plt.subplot2grid(layout, (2, 0))\n",
    "    hist_ax = plt.subplot2grid(layout, (2, 1))\n",
    "\n",
    "    # time series plot\n",
    "    y.plot(ax=ts_ax)\n",
    "    rolling_mean.plot(ax=ts_ax, color='crimson', label='rolling-mean');\n",
    "    rolling_std.plot(ax=ts_ax, color='darkslateblue', label='rolling-std');\n",
    "    ts_ax.legend(loc='best')\n",
    "    ts_ax.set_title(title, fontsize=24);\n",
    "\n",
    "    # acf and pacf\n",
    "    smt.graphics.plot_acf(y, lags=lags, ax=acf_ax, alpha=0.5)\n",
    "    smt.graphics.plot_pacf(y, lags=lags, ax=pacf_ax, alpha=0.5)\n",
    "\n",
    "    # qq plot\n",
    "    sm.qqplot(y, line='s', ax=qq_ax)\n",
    "    qq_ax.set_title('QQ Plot')\n",
    "\n",
    "    # hist plot\n",
    "    y.plot(ax=hist_ax, kind='hist', bins=25);\n",
    "    hist_ax.set_title('Histogram');\n",
    "    plt.tight_layout();\n",
    "    plt.show()\n",
    "\n",
    "    # perform Augmented Dickey Fuller test\n",
    "    print('Results of Dickey-Fuller test:')\n",
    "    dftest = adfuller(y, autolag='AIC')\n",
    "    dfoutput = pd.Series(dftest[0:4], index=['test statistic', 'p-value', '# of lags', '# of observations'])\n",
    "    for key, value in dftest[4].items():\n",
    "        dfoutput['Critical Value (%s)'%key] = value\n",
    "    print(dfoutput)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "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>endog</th>\n",
       "      <th>exog</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dt_utc</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2017-01-01 05:00:00</th>\n",
       "      <td>3.514526</td>\n",
       "      <td>37.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-01 06:00:00</th>\n",
       "      <td>3.467297</td>\n",
       "      <td>37.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-01 07:00:00</th>\n",
       "      <td>3.375196</td>\n",
       "      <td>37.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-01 08:00:00</th>\n",
       "      <td>3.219676</td>\n",
       "      <td>37.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-01 09:00:00</th>\n",
       "      <td>3.293612</td>\n",
       "      <td>37.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        endog  exog\n",
       "dt_utc                             \n",
       "2017-01-01 05:00:00  3.514526  37.0\n",
       "2017-01-01 06:00:00  3.467297  37.0\n",
       "2017-01-01 07:00:00  3.375196  37.0\n",
       "2017-01-01 08:00:00  3.219676  37.0\n",
       "2017-01-01 09:00:00  3.293612  37.0"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "csv = pd.read_csv('sample.csv')\n",
    "data = csv.set_index(pd.DatetimeIndex(csv['dt_utc'], freq='H')).drop('dt_utc', axis=1)\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "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>endog_daily</th>\n",
       "      <th>exog_daily</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dt_utc</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2017-01-01 00:00:00</th>\n",
       "      <td>3.432436</td>\n",
       "      <td>41.789474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-01 01:00:00</th>\n",
       "      <td>3.432436</td>\n",
       "      <td>41.789474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-01 02:00:00</th>\n",
       "      <td>3.432436</td>\n",
       "      <td>41.789474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-01 03:00:00</th>\n",
       "      <td>3.432436</td>\n",
       "      <td>41.789474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017-01-01 04:00:00</th>\n",
       "      <td>3.432436</td>\n",
       "      <td>41.789474</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     endog_daily  exog_daily\n",
       "dt_utc                                      \n",
       "2017-01-01 00:00:00     3.432436   41.789474\n",
       "2017-01-01 01:00:00     3.432436   41.789474\n",
       "2017-01-01 02:00:00     3.432436   41.789474\n",
       "2017-01-01 03:00:00     3.432436   41.789474\n",
       "2017-01-01 04:00:00     3.432436   41.789474"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# create daily \"prediction\" from hourly data\n",
    "\n",
    "daily = data.resample('D').mean()\n",
    "daily_resamp = daily.resample('H').ffill()\n",
    "daily_resamp.columns = [\"{}_daily\".format(c) for c in daily_resamp.columns]\n",
    "\n",
    "daily_resamp.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dt_utc\n",
      "2017-01-01 05:00:00    3.514526\n",
      "2017-01-01 06:00:00    3.467297\n",
      "2017-01-01 07:00:00    3.375196\n",
      "2017-01-01 08:00:00    3.219676\n",
      "2017-01-01 09:00:00    3.293612\n",
      "Freq: H, Name: endog, dtype: float64\n",
      "dt_utc\n",
      "2017-10-01 00:00:00    2.928524\n",
      "2017-10-01 01:00:00    2.814810\n",
      "2017-10-01 02:00:00    2.631889\n",
      "2017-10-01 03:00:00    2.613740\n",
      "2017-10-01 04:00:00    2.576422\n",
      "Freq: H, Name: endog, dtype: float64\n",
      "                     endog_daily\n",
      "dt_utc                          \n",
      "2017-01-01 05:00:00     3.432436\n",
      "2017-01-01 06:00:00     3.432436\n",
      "2017-01-01 07:00:00     3.432436\n",
      "2017-01-01 08:00:00     3.432436\n",
      "2017-01-01 09:00:00     3.432436\n",
      "                     endog_daily\n",
      "dt_utc                          \n",
      "2017-10-01 00:00:00     2.645857\n",
      "2017-10-01 01:00:00     2.645857\n",
      "2017-10-01 02:00:00     2.645857\n",
      "2017-10-01 03:00:00     2.645857\n",
      "2017-10-01 04:00:00     2.645857\n"
     ]
    }
   ],
   "source": [
    "data_wdaily = pd.concat((data, daily_resamp), axis=1).ffill().dropna()\n",
    "\n",
    "endog_train = data_wdaily['endog'].loc[:'2017-09']\n",
    "endog_test = data_wdaily['endog'].loc['2017-10':]\n",
    "\n",
    "exog_train = data_wdaily[['endog_daily']].loc[:'2017-09']\n",
    "exog_test = data_wdaily[['endog_daily']].loc['2017-10':]\n",
    "\n",
    "for df in (endog_train, endog_test, exog_train, exog_test):\n",
    "    print(df.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table class=\"simpletable\">\n",
       "<caption>Statespace Model Results</caption>\n",
       "<tr>\n",
       "  <th>Dep. Variable:</th>                <td>endog</td>             <th>  No. Observations:  </th>    <td>6547</td>   \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Model:</th>           <td>SARIMAX(0, 1, 0)x(1, 1, 1, 24)</td> <th>  Log Likelihood     </th>  <td>8438.200</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Date:</th>                   <td>Thu, 05 Apr 2018</td>        <th>  AIC                </th> <td>-16868.399</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Time:</th>                       <td>17:26:32</td>            <th>  BIC                </th> <td>-16841.252</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Sample:</th>                    <td>01-01-2017</td>           <th>  HQIC               </th> <td>-16859.013</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th></th>                          <td>- 09-30-2017</td>          <th>                     </th>      <td> </td>    \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Covariance Type:</th>               <td>opg</td>              <th>                     </th>      <td> </td>    \n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "       <td></td>          <th>coef</th>     <th>std err</th>      <th>z</th>      <th>P>|z|</th>  <th>[0.025</th>    <th>0.975]</th>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>endog_daily</th> <td>    0.0444</td> <td>    0.041</td> <td>    1.085</td> <td> 0.278</td> <td>   -0.036</td> <td>    0.124</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ar.S.L24</th>    <td>    0.2419</td> <td>    0.009</td> <td>   26.280</td> <td> 0.000</td> <td>    0.224</td> <td>    0.260</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>ma.S.L24</th>    <td>   -0.9139</td> <td>    0.005</td> <td> -189.224</td> <td> 0.000</td> <td>   -0.923</td> <td>   -0.904</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>sigma2</th>      <td>    0.0044</td> <td> 4.26e-05</td> <td>  102.852</td> <td> 0.000</td> <td>    0.004</td> <td>    0.004</td>\n",
       "</tr>\n",
       "</table>\n",
       "<table class=\"simpletable\">\n",
       "<tr>\n",
       "  <th>Ljung-Box (Q):</th>          <td>293.89</td> <th>  Jarque-Bera (JB):  </th> <td>5694.27</td>\n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(Q):</th>                 <td>0.00</td>  <th>  Prob(JB):          </th>  <td>0.00</td>  \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Heteroskedasticity (H):</th>  <td>1.51</td>  <th>  Skew:              </th>  <td>-0.11</td> \n",
       "</tr>\n",
       "<tr>\n",
       "  <th>Prob(H) (two-sided):</th>     <td>0.00</td>  <th>  Kurtosis:          </th>  <td>7.57</td>  \n",
       "</tr>\n",
       "</table>"
      ],
      "text/plain": [
       "<class 'statsmodels.iolib.summary.Summary'>\n",
       "\"\"\"\n",
       "                                 Statespace Model Results                                 \n",
       "==========================================================================================\n",
       "Dep. Variable:                              endog   No. Observations:                 6547\n",
       "Model:             SARIMAX(0, 1, 0)x(1, 1, 1, 24)   Log Likelihood                8438.200\n",
       "Date:                            Thu, 05 Apr 2018   AIC                         -16868.399\n",
       "Time:                                    17:26:32   BIC                         -16841.252\n",
       "Sample:                                01-01-2017   HQIC                        -16859.013\n",
       "                                     - 09-30-2017                                         \n",
       "Covariance Type:                              opg                                         \n",
       "===============================================================================\n",
       "                  coef    std err          z      P>|z|      [0.025      0.975]\n",
       "-------------------------------------------------------------------------------\n",
       "endog_daily     0.0444      0.041      1.085      0.278      -0.036       0.124\n",
       "ar.S.L24        0.2419      0.009     26.280      0.000       0.224       0.260\n",
       "ma.S.L24       -0.9139      0.005   -189.224      0.000      -0.923      -0.904\n",
       "sigma2          0.0044   4.26e-05    102.852      0.000       0.004       0.004\n",
       "===================================================================================\n",
       "Ljung-Box (Q):                      293.89   Jarque-Bera (JB):              5694.27\n",
       "Prob(Q):                              0.00   Prob(JB):                         0.00\n",
       "Heteroskedasticity (H):               1.51   Skew:                            -0.11\n",
       "Prob(H) (two-sided):                  0.00   Kurtosis:                         7.57\n",
       "===================================================================================\n",
       "\n",
       "Warnings:\n",
       "[1] Covariance matrix calculated using the outer product of gradients (complex-step).\n",
       "\"\"\""
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sarimax = SARIMAX(endog=endog_train, exog=exog_train,\n",
    "                  order=(0,1,0), seasonal_order=(1,1,1,24),\n",
    "                  trend='n')\n",
    "res = sarimax.fit()\n",
    "res.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x11386beb8>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11386b8d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# in-sample prediction\n",
    "\n",
    "pred_in = res.predict()\n",
    "pred_in.name = 'pred_in'\n",
    "\n",
    "comp_in = pd.concat((endog_train, pred_in), axis=1)\n",
    "comp_in.loc['2017-02-01':'2017-02-07'].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x11381ada0>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x113876dd8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# include tail in-sample data to highlight how forecast goes wrong\n",
    "\n",
    "_exog = exog_test.values.reshape((-1, 1)) if len(exog_test.shape) == 1 else exog_test.values\n",
    "\n",
    "pred_both = res.predict('2017-09-25', '2017-10-31 23:00:00', exog=_exog)\n",
    "pred_both.name = 'pred_both'\n",
    "\n",
    "comp_both = pd.concat((data['endog'].loc['2017-09-25':], pred_both), axis=1)\n",
    "comp_both.plot()"
   ]
  }
 ],
 "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}