analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "LOADING DATA: 100%|██████████| 14/14 [03:30<00:00, 15.00s/it]\n"
     ]
    },
    {
     "data": {
      "text/plain": "                                        passengers     noise  weekend   \ndatetime            origin destination                                  \n2022-01-01 00:00:00 12     19                159.0  0.304939     True  \\\n                           LM                  6.0 -1.456441     True   \n                           OW                 25.0 -1.558501     True   \n                    16     24                 78.0  0.451318     True   \n                           CC                 82.0  0.865496     True   \n...                                            ...       ...      ...   \n2023-06-30 23:00:00 WD     ED                110.0 -1.519143    False   \n                    WP     NC                 10.0  0.890731    False   \n                           PC                 53.0  0.916043    False   \n                    WS     FM                 34.0  0.878724    False   \n                           ML                 91.0  0.328786    False   \n\n                                        hours_00_04  hours_04_08  hours_08_12   \ndatetime            origin destination                                          \n2022-01-01 00:00:00 12     19                  True        False        False  \\\n                           LM                  True        False        False   \n                           OW                  True        False        False   \n                    16     24                  True        False        False   \n                           CC                  True        False        False   \n...                                             ...          ...          ...   \n2023-06-30 23:00:00 WD     ED                 False        False        False   \n                    WP     NC                 False        False        False   \n                           PC                 False        False        False   \n                    WS     FM                 False        False        False   \n                           ML                 False        False        False   \n\n                                        hours_12_16  hours_16_20  hours_20_24   \ndatetime            origin destination                                          \n2022-01-01 00:00:00 12     19                 False        False        False  \\\n                           LM                 False        False        False   \n                           OW                 False        False        False   \n                    16     24                 False        False        False   \n                           CC                 False        False        False   \n...                                             ...          ...          ...   \n2023-06-30 23:00:00 WD     ED                 False        False         True   \n                    WP     NC                 False        False         True   \n                           PC                 False        False         True   \n                    WS     FM                 False        False         True   \n                           ML                 False        False         True   \n\n                                        event_capacity  ...   \ndatetime            origin destination                  ...   \n2022-01-01 00:00:00 12     19                      0.0  ...  \\\n                           LM                      0.0  ...   \n                           OW                      0.0  ...   \n                    16     24                      0.0  ...   \n                           CC                      0.0  ...   \n...                                                ...  ...   \n2023-06-30 23:00:00 WD     ED                      0.0  ...   \n                    WP     NC                      0.0  ...   \n                           PC                      0.0  ...   \n                    WS     FM                      0.0  ...   \n                           ML                      0.0  ...   \n\n                                        event_type_festival event_type_lgbtq   \ndatetime            origin destination                                         \n2022-01-01 00:00:00 12     19                             0                0  \\\n                           LM                             0                0   \n                           OW                             0                0   \n                    16     24                             0                0   \n                           CC                             0                0   \n...                                                     ...              ...   \n2023-06-30 23:00:00 WD     ED                             0                0   \n                    WP     NC                             0                0   \n                           PC                             0                0   \n                    WS     FM                             0                0   \n                           ML                             0                0   \n\n                                       event_type_experiences   \ndatetime            origin destination                          \n2022-01-01 00:00:00 12     19                               0  \\\n                           LM                               0   \n                           OW                               0   \n                    16     24                               0   \n                           CC                               0   \n...                                                       ...   \n2023-06-30 23:00:00 WD     ED                               0   \n                    WP     NC                               0   \n                           PC                               0   \n                    WS     FM                               0   \n                           ML                               0   \n\n                                       event_type_music event_type_party   \ndatetime            origin destination                                     \n2022-01-01 00:00:00 12     19                         0                0  \\\n                           LM                         0                0   \n                           OW                         0                0   \n                    16     24                         0                0   \n                           CC                         0                0   \n...                                                 ...              ...   \n2023-06-30 23:00:00 WD     ED                         0                0   \n                    WP     NC                         0                0   \n                           PC                         0                0   \n                    WS     FM                         0                0   \n                           ML                         0                0   \n\n                                       event_type_education   \ndatetime            origin destination                        \n2022-01-01 00:00:00 12     19                             0  \\\n                           LM                             0   \n                           OW                             0   \n                    16     24                             0   \n                           CC                             0   \n...                                                     ...   \n2023-06-30 23:00:00 WD     ED                             0   \n                    WP     NC                             0   \n                           PC                             0   \n                    WS     FM                             0   \n                           ML                             0   \n\n                                       event_type_food-drink event_type_arts   \ndatetime            origin destination                                         \n2022-01-01 00:00:00 12     19                              0               0  \\\n                           LM                              0               0   \n                           OW                              0               0   \n                    16     24                              0               0   \n                           CC                              0               0   \n...                                                      ...             ...   \n2023-06-30 23:00:00 WD     ED                              0               0   \n                    WP     NC                              0               0   \n                           PC                              0               0   \n                    WS     FM                              0               0   \n                           ML                              0               0   \n\n                                         time   weekday  \ndatetime            origin destination                   \n2022-01-01 00:00:00 12     19           00:00  Saturday  \n                           LM           00:00  Saturday  \n                           OW           00:00  Saturday  \n                    16     24           00:00  Saturday  \n                           CC           00:00  Saturday  \n...                                       ...       ...  \n2023-06-30 23:00:00 WD     ED           23:00    Friday  \n                    WP     NC           23:00    Friday  \n                           PC           23:00    Friday  \n                    WS     FM           23:00    Friday  \n                           ML           23:00    Friday  \n\n[1336608 rows x 25 columns]",
      "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></th>\n      <th></th>\n      <th>passengers</th>\n      <th>noise</th>\n      <th>weekend</th>\n      <th>hours_00_04</th>\n      <th>hours_04_08</th>\n      <th>hours_08_12</th>\n      <th>hours_12_16</th>\n      <th>hours_16_20</th>\n      <th>hours_20_24</th>\n      <th>event_capacity</th>\n      <th>...</th>\n      <th>event_type_festival</th>\n      <th>event_type_lgbtq</th>\n      <th>event_type_experiences</th>\n      <th>event_type_music</th>\n      <th>event_type_party</th>\n      <th>event_type_education</th>\n      <th>event_type_food-drink</th>\n      <th>event_type_arts</th>\n      <th>time</th>\n      <th>weekday</th>\n    </tr>\n    <tr>\n      <th>datetime</th>\n      <th>origin</th>\n      <th>destination</th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n      <th></th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th rowspan=\"5\" valign=\"top\">2022-01-01 00:00:00</th>\n      <th rowspan=\"3\" valign=\"top\">12</th>\n      <th>19</th>\n      <td>159.0</td>\n      <td>0.304939</td>\n      <td>True</td>\n      <td>True</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>00:00</td>\n      <td>Saturday</td>\n    </tr>\n    <tr>\n      <th>LM</th>\n      <td>6.0</td>\n      <td>-1.456441</td>\n      <td>True</td>\n      <td>True</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>00:00</td>\n      <td>Saturday</td>\n    </tr>\n    <tr>\n      <th>OW</th>\n      <td>25.0</td>\n      <td>-1.558501</td>\n      <td>True</td>\n      <td>True</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>00:00</td>\n      <td>Saturday</td>\n    </tr>\n    <tr>\n      <th rowspan=\"2\" valign=\"top\">16</th>\n      <th>24</th>\n      <td>78.0</td>\n      <td>0.451318</td>\n      <td>True</td>\n      <td>True</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>00:00</td>\n      <td>Saturday</td>\n    </tr>\n    <tr>\n      <th>CC</th>\n      <td>82.0</td>\n      <td>0.865496</td>\n      <td>True</td>\n      <td>True</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>00:00</td>\n      <td>Saturday</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <th>...</th>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th rowspan=\"5\" valign=\"top\">2023-06-30 23:00:00</th>\n      <th>WD</th>\n      <th>ED</th>\n      <td>110.0</td>\n      <td>-1.519143</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>True</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>23:00</td>\n      <td>Friday</td>\n    </tr>\n    <tr>\n      <th rowspan=\"2\" valign=\"top\">WP</th>\n      <th>NC</th>\n      <td>10.0</td>\n      <td>0.890731</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>True</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>23:00</td>\n      <td>Friday</td>\n    </tr>\n    <tr>\n      <th>PC</th>\n      <td>53.0</td>\n      <td>0.916043</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>True</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>23:00</td>\n      <td>Friday</td>\n    </tr>\n    <tr>\n      <th rowspan=\"2\" valign=\"top\">WS</th>\n      <th>FM</th>\n      <td>34.0</td>\n      <td>0.878724</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>True</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>23:00</td>\n      <td>Friday</td>\n    </tr>\n    <tr>\n      <th>ML</th>\n      <td>91.0</td>\n      <td>0.328786</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>False</td>\n      <td>True</td>\n      <td>0.0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>23:00</td>\n      <td>Friday</td>\n    </tr>\n  </tbody>\n</table>\n<p>1336608 rows × 25 columns</p>\n</div>"
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from datetime import date\n",
    "\n",
    "from util.dataset import FeaturePassengerFlowDataset\n",
    "\n",
    "# noinspection PyProtectedMember\n",
    "data = FeaturePassengerFlowDataset(\n",
    "    min_date=date(2022, 1, 1),\n",
    "    max_date=date(2023, 7, 1)\n",
    ")._data\n",
    "\n",
    "data['time'] = data.index.to_series().apply(lambda idx: f'{idx[0].hour:02d}:{idx[0].minute:02d}')\n",
    "data['weekday'] = data.index.to_series().apply(lambda idx: ['Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday'][idx[0].weekday()])\n",
    "\n",
    "data"
   ]
  },
  {
   "cell_type": "markdown",
   "source": [
    "### Regularity of flow"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "outputs": [
    {
     "data": {
      "text/plain": "REMOVE_TREND DONE:   0%|          | 0/17136 [00:00<?, ?it/s]",
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "10c881760157450b8c7565ef7f00738e"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": "                                              mean      lower       upper\ndatetime            origin destination                                   \n2022-01-01 00:00:00 12     19           150.237769  22.011172  278.745071\n                           LM            26.755525   5.500901   49.890638\n                           OW            57.749715   8.759476  132.351973\n                    16     24            44.516290 -19.917108  131.146463\n                           CC            95.415129  67.231212  128.688774\n...                                            ...        ...         ...\n2023-06-30 23:00:00 WD     ED           125.755604  90.775952  172.747879\n                    WP     NC            13.950183   9.941646   19.380242\n                           PC            42.070807  27.247870   58.695092\n                    WS     FM            31.055462  22.842618   38.381706\n                           ML            71.776997  53.442871   94.615959\n\n[1336608 rows x 3 columns]",
      "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></th>\n      <th></th>\n      <th>mean</th>\n      <th>lower</th>\n      <th>upper</th>\n    </tr>\n    <tr>\n      <th>datetime</th>\n      <th>origin</th>\n      <th>destination</th>\n      <th></th>\n      <th></th>\n      <th></th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th rowspan=\"5\" valign=\"top\">2022-01-01 00:00:00</th>\n      <th rowspan=\"3\" valign=\"top\">12</th>\n      <th>19</th>\n      <td>150.237769</td>\n      <td>22.011172</td>\n      <td>278.745071</td>\n    </tr>\n    <tr>\n      <th>LM</th>\n      <td>26.755525</td>\n      <td>5.500901</td>\n      <td>49.890638</td>\n    </tr>\n    <tr>\n      <th>OW</th>\n      <td>57.749715</td>\n      <td>8.759476</td>\n      <td>132.351973</td>\n    </tr>\n    <tr>\n      <th rowspan=\"2\" valign=\"top\">16</th>\n      <th>24</th>\n      <td>44.516290</td>\n      <td>-19.917108</td>\n      <td>131.146463</td>\n    </tr>\n    <tr>\n      <th>CC</th>\n      <td>95.415129</td>\n      <td>67.231212</td>\n      <td>128.688774</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <th>...</th>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th rowspan=\"5\" valign=\"top\">2023-06-30 23:00:00</th>\n      <th>WD</th>\n      <th>ED</th>\n      <td>125.755604</td>\n      <td>90.775952</td>\n      <td>172.747879</td>\n    </tr>\n    <tr>\n      <th rowspan=\"2\" valign=\"top\">WP</th>\n      <th>NC</th>\n      <td>13.950183</td>\n      <td>9.941646</td>\n      <td>19.380242</td>\n    </tr>\n    <tr>\n      <th>PC</th>\n      <td>42.070807</td>\n      <td>27.247870</td>\n      <td>58.695092</td>\n    </tr>\n    <tr>\n      <th rowspan=\"2\" valign=\"top\">WS</th>\n      <th>FM</th>\n      <td>31.055462</td>\n      <td>22.842618</td>\n      <td>38.381706</td>\n    </tr>\n    <tr>\n      <th>ML</th>\n      <td>71.776997</td>\n      <td>53.442871</td>\n      <td>94.615959</td>\n    </tr>\n  </tbody>\n</table>\n<p>1336608 rows × 3 columns</p>\n</div>"
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "from parallel_pandas import ParallelPandas\n",
    "from tqdm.notebook import tqdm\n",
    "\n",
    "tqdm.pandas()\n",
    "ParallelPandas.initialize()\n",
    "\n",
    "def remove_trend(df: pd.DataFrame) -> pd.Series:\n",
    "    from statsmodels.tsa.stl._stl import STL\n",
    "\n",
    "    stl_fit = STL(df.passengers,\n",
    "                  period=3,\n",
    "                  low_pass=5,\n",
    "                  robust=True).fit()\n",
    "\n",
    "    return stl_fit.trend\n",
    "\n",
    "# Detect trend\n",
    "normalized = data.assign(trend=data\n",
    "                         .groupby(['origin', 'destination', 'weekday', 'time'])\n",
    "                         .p_apply(remove_trend)\n",
    "                         .droplevel([0, 1, 2, 3]))\n",
    "normalized['normalized'] = normalized.passengers - normalized.trend\n",
    "normalized['mean'] = normalized.groupby(['origin', 'destination', 'weekday', 'time']).apply(lambda df: df.normalized.mean() + df.trend).droplevel([0, 1, 2, 3])\n",
    "normalized['lower'] = normalized.groupby(['origin', 'destination', 'weekday', 'time']).apply(lambda df: df.normalized.quantile(0.10) + df.trend).droplevel([0, 1, 2, 3])\n",
    "normalized['upper'] = normalized.groupby(['origin', 'destination', 'weekday', 'time']).apply(lambda df: df.normalized.quantile(0.90) + df.trend).droplevel([0, 1, 2, 3])\n",
    "\n",
    "normalized = normalized[['mean', 'lower', 'upper']]\n",
    "\n",
    "normalized"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 1400x200 with 1 Axes>",
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib.dates import DateFormatter\n",
    "from util.plotstyle import style_plot\n",
    "from datetime import datetime\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "regularity_example = normalized.loc[:, 'EM', 'OW'].loc[datetime(2022, 7, 4):datetime(2022, 7, 11)]\n",
    "\n",
    "fig, axes = plt.subplots(1, 1, figsize=(14, 2))\n",
    "\n",
    "axes.fill_between(\n",
    "    regularity_example.index.to_series(),\n",
    "    regularity_example['upper'],\n",
    "    regularity_example['lower'],\n",
    "    color='tab:blue',\n",
    "    alpha=0.1\n",
    ")\n",
    "sns.lineplot(regularity_example['mean'])\n",
    "\n",
    "axes.set_xlabel('Time [d]')\n",
    "axes.set_ylabel('Passenger mean')\n",
    "style_plot(axes)\n",
    "\n",
    "axes.xaxis.set_major_formatter(DateFormatter(\"%a\"))\n",
    "\n",
    "plt.savefig('figures/flow-regularity-EM-OW.png', dpi=300, bbox_inches='tight', transparent=True)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "### SARIMA Order Selection"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Regular: Non-stationary (0.1226)\n",
      "Seasonal (24): Stationary (0.0232)\n"
     ]
    }
   ],
   "source": [
    "from statsmodels.tsa.stattools import adfuller\n",
    "\n",
    "test = adfuller(data.loc[:, 'EM', 'OW'].passengers,\n",
    "                maxlag=168+4,\n",
    "                regression='c')\n",
    "print('Regular:', 'Stationary' if test[1] <= 0.05 else 'Non-stationary', f'({test[1]:.4f})')\n",
    "\n",
    "test = adfuller(data.loc[:, 'EM', 'OW'].passengers.iloc[::24],\n",
    "                regression='c')\n",
    "print('Seasonal (24):', 'Stationary' if test[1] <= 0.05 else 'Non-stationary', f'({test[1]:.4f})')"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 700x200 with 1 Axes>",
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib.ticker import MultipleLocator\n",
    "from statsmodels.graphics.tsaplots import plot_acf\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig, axes = plt.subplots(1, 1, figsize=(7, 2))\n",
    "axes.axhline(y=0,\n",
    "             alpha=0.7,\n",
    "             color='black',\n",
    "             linewidth=1)\n",
    "\n",
    "plot_acf(data.loc[:, 'EM', 'OW'].passengers,\n",
    "         lags=72+4,\n",
    "         title=None,\n",
    "         use_vlines=False,\n",
    "         ax=axes)\n",
    "\n",
    "style_plot(axes)\n",
    "axes.spines['bottom'].set_visible(False)\n",
    "\n",
    "axes.xaxis.set_major_locator(MultipleLocator(24))\n",
    "\n",
    "axes.set_xlim(0, 80)\n",
    "axes.set_ylim(-1.02, 1)\n",
    "\n",
    "axes.set_xlabel('Amount of lags')\n",
    "axes.set_ylabel('Autocorrelation')\n",
    "\n",
    "plt.savefig('figures/autocorrelation.png', dpi=300, bbox_inches='tight', transparent=True)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 700x200 with 1 Axes>",
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from statsmodels.graphics.tsaplots import plot_pacf\n",
    "\n",
    "fig, axes = plt.subplots(1, 1, figsize=(7, 2))\n",
    "axes.axhline(y=0,\n",
    "             alpha=0.7,\n",
    "             color='black',\n",
    "             linewidth=1)\n",
    "\n",
    "plot_pacf(data.loc[:, 'EM', 'OW'].passengers,\n",
    "          lags=72+4,\n",
    "          method=\"ols\",\n",
    "          title=None,\n",
    "          use_vlines=False,\n",
    "          ax=axes)\n",
    "\n",
    "style_plot(axes)\n",
    "axes.spines['bottom'].set_visible(False)\n",
    "\n",
    "axes.xaxis.set_major_locator(MultipleLocator(24))\n",
    "\n",
    "axes.set_xlim(0, 80)\n",
    "axes.set_ylim(-1.02, 1)\n",
    "\n",
    "axes.set_xlabel('Amount of lags')\n",
    "axes.set_ylabel('Partial autocorrelation')\n",
    "\n",
    "plt.savefig('figures/autocorrelation-partial.png', dpi=300, bbox_inches='tight', transparent=True)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 700x200 with 1 Axes>",
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(1, 1, figsize=(7, 2))\n",
    "axes.axhline(y=0,\n",
    "             alpha=0.7,\n",
    "             color='black',\n",
    "             linewidth=1)\n",
    "\n",
    "plot_acf(data.loc[:, 'EM', 'OW'].passengers.iloc[::24],\n",
    "         lags=29,\n",
    "         title=None,\n",
    "         use_vlines=False,\n",
    "         ax=axes)\n",
    "\n",
    "style_plot(axes)\n",
    "axes.spines['bottom'].set_visible(False)\n",
    "\n",
    "axes.xaxis.set_major_locator(MultipleLocator(7))\n",
    "\n",
    "axes.set_xlim(0, 30)\n",
    "axes.set_ylim(-1.02, 1)\n",
    "\n",
    "axes.set_xlabel('Amount of lags')\n",
    "axes.set_ylabel('Autocorrelation')\n",
    "\n",
    "plt.savefig('figures/seasonal-autocorrelation.png', dpi=300, bbox_inches='tight', transparent=True)"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 700x200 with 1 Axes>",
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from statsmodels.graphics.tsaplots import plot_pacf\n",
    "\n",
    "fig, axes = plt.subplots(1, 1, figsize=(7, 2))\n",
    "axes.axhline(y=0,\n",
    "             alpha=0.7,\n",
    "             color='black',\n",
    "             linewidth=1)\n",
    "\n",
    "plot_pacf(data.loc[:, 'EM', 'OW'].passengers.iloc[::24],\n",
    "          lags=29,\n",
    "          method=\"ols\",\n",
    "          title=None,\n",
    "          use_vlines=False,\n",
    "          ax=axes)\n",
    "\n",
    "style_plot(axes)\n",
    "axes.spines['bottom'].set_visible(False)\n",
    "\n",
    "axes.xaxis.set_major_locator(MultipleLocator(7))\n",
    "\n",
    "axes.set_xlim(0, 30)\n",
    "axes.set_ylim(-1.02, 1)\n",
    "\n",
    "axes.set_xlabel('Amount of lags')\n",
    "axes.set_ylabel('Partial autocorrelation')\n",
    "\n",
    "plt.savefig('figures/seasonal-autocorrelation-partial.png', dpi=300, bbox_inches='tight', transparent=True)"
   ],
   "metadata": {
    "collapsed": false
   }
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}