Notebooks/AQUDA-Novelty detection.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Important libraries\n",
    "from dateutil.parser import parse \n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn import preprocessing\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.font_manager\n",
    "import seaborn as sns\n",
    "from sklearn.svm import OneClassSVM\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "#We load the data\n",
    "data=pd.read_csv(\"PM2.5_dataset.csv\", parse_dates=['stime'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>Unnamed: 0</th>\n",
       "      <th>station_id</th>\n",
       "      <th>stime</th>\n",
       "      <th>air_data_value</th>\n",
       "      <th>RH</th>\n",
       "      <th>UGRD</th>\n",
       "      <th>VGRD</th>\n",
       "      <th>HPBL</th>\n",
       "      <th>TMP</th>\n",
       "      <th>goes_measurement</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>06-011-0007</td>\n",
       "      <td>2019-01-02 20:00:00</td>\n",
       "      <td>17.0</td>\n",
       "      <td>31.600000</td>\n",
       "      <td>-2.106623</td>\n",
       "      <td>-1.797583</td>\n",
       "      <td>256.61905</td>\n",
       "      <td>282.81880</td>\n",
       "      <td>-0.005922</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>06-019-0500</td>\n",
       "      <td>2019-01-02 20:00:00</td>\n",
       "      <td>13.0</td>\n",
       "      <td>62.200000</td>\n",
       "      <td>1.205877</td>\n",
       "      <td>1.764917</td>\n",
       "      <td>337.49405</td>\n",
       "      <td>281.63130</td>\n",
       "      <td>0.087090</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>06-061-0003</td>\n",
       "      <td>2019-01-02 20:00:00</td>\n",
       "      <td>21.0</td>\n",
       "      <td>61.500000</td>\n",
       "      <td>1.518377</td>\n",
       "      <td>1.014917</td>\n",
       "      <td>270.61905</td>\n",
       "      <td>280.13130</td>\n",
       "      <td>0.094333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>06-073-1201</td>\n",
       "      <td>2019-01-02 20:00:00</td>\n",
       "      <td>6.0</td>\n",
       "      <td>15.400001</td>\n",
       "      <td>2.080877</td>\n",
       "      <td>-1.610083</td>\n",
       "      <td>1009.30660</td>\n",
       "      <td>288.19380</td>\n",
       "      <td>-0.024185</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>06-079-2004</td>\n",
       "      <td>2019-01-02 20:00:00</td>\n",
       "      <td>7.0</td>\n",
       "      <td>50.700000</td>\n",
       "      <td>2.393377</td>\n",
       "      <td>-1.172583</td>\n",
       "      <td>460.43155</td>\n",
       "      <td>285.19380</td>\n",
       "      <td>-0.014013</td>\n",
       "    </tr>\n",
       "    <tr>\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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31565</th>\n",
       "      <td>103560</td>\n",
       "      <td>49-035-4002</td>\n",
       "      <td>2019-10-30 23:00:00</td>\n",
       "      <td>6.3</td>\n",
       "      <td>25.500000</td>\n",
       "      <td>0.723850</td>\n",
       "      <td>-2.034275</td>\n",
       "      <td>892.42080</td>\n",
       "      <td>273.49664</td>\n",
       "      <td>0.054725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31566</th>\n",
       "      <td>103591</td>\n",
       "      <td>49-021-0005</td>\n",
       "      <td>2019-10-16 20:00:00</td>\n",
       "      <td>6.4</td>\n",
       "      <td>8.100000</td>\n",
       "      <td>6.964932</td>\n",
       "      <td>4.869552</td>\n",
       "      <td>1664.20080</td>\n",
       "      <td>297.60638</td>\n",
       "      <td>0.010030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31567</th>\n",
       "      <td>103592</td>\n",
       "      <td>49-035-4002</td>\n",
       "      <td>2019-10-16 20:00:00</td>\n",
       "      <td>8.5</td>\n",
       "      <td>8.900001</td>\n",
       "      <td>2.214932</td>\n",
       "      <td>4.682052</td>\n",
       "      <td>842.38837</td>\n",
       "      <td>297.60638</td>\n",
       "      <td>0.282129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31568</th>\n",
       "      <td>103594</td>\n",
       "      <td>49-035-4002</td>\n",
       "      <td>2019-10-16 21:00:00</td>\n",
       "      <td>8.1</td>\n",
       "      <td>8.300000</td>\n",
       "      <td>1.386543</td>\n",
       "      <td>5.465472</td>\n",
       "      <td>1292.07180</td>\n",
       "      <td>298.64655</td>\n",
       "      <td>0.233581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31569</th>\n",
       "      <td>103595</td>\n",
       "      <td>49-021-0005</td>\n",
       "      <td>2019-10-18 18:00:00</td>\n",
       "      <td>2.1</td>\n",
       "      <td>44.100002</td>\n",
       "      <td>-3.219561</td>\n",
       "      <td>-4.265393</td>\n",
       "      <td>980.89050</td>\n",
       "      <td>283.69052</td>\n",
       "      <td>0.200753</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>31570 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Unnamed: 0   station_id               stime  air_data_value         RH  \\\n",
       "0               0  06-011-0007 2019-01-02 20:00:00            17.0  31.600000   \n",
       "1               1  06-019-0500 2019-01-02 20:00:00            13.0  62.200000   \n",
       "2               2  06-061-0003 2019-01-02 20:00:00            21.0  61.500000   \n",
       "3               3  06-073-1201 2019-01-02 20:00:00             6.0  15.400001   \n",
       "4               4  06-079-2004 2019-01-02 20:00:00             7.0  50.700000   \n",
       "...           ...          ...                 ...             ...        ...   \n",
       "31565      103560  49-035-4002 2019-10-30 23:00:00             6.3  25.500000   \n",
       "31566      103591  49-021-0005 2019-10-16 20:00:00             6.4   8.100000   \n",
       "31567      103592  49-035-4002 2019-10-16 20:00:00             8.5   8.900001   \n",
       "31568      103594  49-035-4002 2019-10-16 21:00:00             8.1   8.300000   \n",
       "31569      103595  49-021-0005 2019-10-18 18:00:00             2.1  44.100002   \n",
       "\n",
       "           UGRD      VGRD        HPBL        TMP  goes_measurement  \n",
       "0     -2.106623 -1.797583   256.61905  282.81880         -0.005922  \n",
       "1      1.205877  1.764917   337.49405  281.63130          0.087090  \n",
       "2      1.518377  1.014917   270.61905  280.13130          0.094333  \n",
       "3      2.080877 -1.610083  1009.30660  288.19380         -0.024185  \n",
       "4      2.393377 -1.172583   460.43155  285.19380         -0.014013  \n",
       "...         ...       ...         ...        ...               ...  \n",
       "31565  0.723850 -2.034275   892.42080  273.49664          0.054725  \n",
       "31566  6.964932  4.869552  1664.20080  297.60638          0.010030  \n",
       "31567  2.214932  4.682052   842.38837  297.60638          0.282129  \n",
       "31568  1.386543  5.465472  1292.07180  298.64655          0.233581  \n",
       "31569 -3.219561 -4.265393   980.89050  283.69052          0.200753  \n",
       "\n",
       "[31570 rows x 10 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "#We remove a redundant column\n",
    "data=data.iloc[:,1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>station_id</th>\n",
       "      <th>stime</th>\n",
       "      <th>air_data_value</th>\n",
       "      <th>RH</th>\n",
       "      <th>UGRD</th>\n",
       "      <th>VGRD</th>\n",
       "      <th>HPBL</th>\n",
       "      <th>TMP</th>\n",
       "      <th>goes_measurement</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>06-011-0007</td>\n",
       "      <td>2019-01-02 20:00:00</td>\n",
       "      <td>17.0</td>\n",
       "      <td>31.600000</td>\n",
       "      <td>-2.106623</td>\n",
       "      <td>-1.797583</td>\n",
       "      <td>256.61905</td>\n",
       "      <td>282.81880</td>\n",
       "      <td>-0.005922</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>06-019-0500</td>\n",
       "      <td>2019-01-02 20:00:00</td>\n",
       "      <td>13.0</td>\n",
       "      <td>62.200000</td>\n",
       "      <td>1.205877</td>\n",
       "      <td>1.764917</td>\n",
       "      <td>337.49405</td>\n",
       "      <td>281.63130</td>\n",
       "      <td>0.087090</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>06-061-0003</td>\n",
       "      <td>2019-01-02 20:00:00</td>\n",
       "      <td>21.0</td>\n",
       "      <td>61.500000</td>\n",
       "      <td>1.518377</td>\n",
       "      <td>1.014917</td>\n",
       "      <td>270.61905</td>\n",
       "      <td>280.13130</td>\n",
       "      <td>0.094333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>06-073-1201</td>\n",
       "      <td>2019-01-02 20:00:00</td>\n",
       "      <td>6.0</td>\n",
       "      <td>15.400001</td>\n",
       "      <td>2.080877</td>\n",
       "      <td>-1.610083</td>\n",
       "      <td>1009.30660</td>\n",
       "      <td>288.19380</td>\n",
       "      <td>-0.024185</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>06-079-2004</td>\n",
       "      <td>2019-01-02 20:00:00</td>\n",
       "      <td>7.0</td>\n",
       "      <td>50.700000</td>\n",
       "      <td>2.393377</td>\n",
       "      <td>-1.172583</td>\n",
       "      <td>460.43155</td>\n",
       "      <td>285.19380</td>\n",
       "      <td>-0.014013</td>\n",
       "    </tr>\n",
       "    <tr>\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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31565</th>\n",
       "      <td>49-035-4002</td>\n",
       "      <td>2019-10-30 23:00:00</td>\n",
       "      <td>6.3</td>\n",
       "      <td>25.500000</td>\n",
       "      <td>0.723850</td>\n",
       "      <td>-2.034275</td>\n",
       "      <td>892.42080</td>\n",
       "      <td>273.49664</td>\n",
       "      <td>0.054725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31566</th>\n",
       "      <td>49-021-0005</td>\n",
       "      <td>2019-10-16 20:00:00</td>\n",
       "      <td>6.4</td>\n",
       "      <td>8.100000</td>\n",
       "      <td>6.964932</td>\n",
       "      <td>4.869552</td>\n",
       "      <td>1664.20080</td>\n",
       "      <td>297.60638</td>\n",
       "      <td>0.010030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31567</th>\n",
       "      <td>49-035-4002</td>\n",
       "      <td>2019-10-16 20:00:00</td>\n",
       "      <td>8.5</td>\n",
       "      <td>8.900001</td>\n",
       "      <td>2.214932</td>\n",
       "      <td>4.682052</td>\n",
       "      <td>842.38837</td>\n",
       "      <td>297.60638</td>\n",
       "      <td>0.282129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31568</th>\n",
       "      <td>49-035-4002</td>\n",
       "      <td>2019-10-16 21:00:00</td>\n",
       "      <td>8.1</td>\n",
       "      <td>8.300000</td>\n",
       "      <td>1.386543</td>\n",
       "      <td>5.465472</td>\n",
       "      <td>1292.07180</td>\n",
       "      <td>298.64655</td>\n",
       "      <td>0.233581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31569</th>\n",
       "      <td>49-021-0005</td>\n",
       "      <td>2019-10-18 18:00:00</td>\n",
       "      <td>2.1</td>\n",
       "      <td>44.100002</td>\n",
       "      <td>-3.219561</td>\n",
       "      <td>-4.265393</td>\n",
       "      <td>980.89050</td>\n",
       "      <td>283.69052</td>\n",
       "      <td>0.200753</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>31570 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        station_id               stime  air_data_value         RH      UGRD  \\\n",
       "0      06-011-0007 2019-01-02 20:00:00            17.0  31.600000 -2.106623   \n",
       "1      06-019-0500 2019-01-02 20:00:00            13.0  62.200000  1.205877   \n",
       "2      06-061-0003 2019-01-02 20:00:00            21.0  61.500000  1.518377   \n",
       "3      06-073-1201 2019-01-02 20:00:00             6.0  15.400001  2.080877   \n",
       "4      06-079-2004 2019-01-02 20:00:00             7.0  50.700000  2.393377   \n",
       "...            ...                 ...             ...        ...       ...   \n",
       "31565  49-035-4002 2019-10-30 23:00:00             6.3  25.500000  0.723850   \n",
       "31566  49-021-0005 2019-10-16 20:00:00             6.4   8.100000  6.964932   \n",
       "31567  49-035-4002 2019-10-16 20:00:00             8.5   8.900001  2.214932   \n",
       "31568  49-035-4002 2019-10-16 21:00:00             8.1   8.300000  1.386543   \n",
       "31569  49-021-0005 2019-10-18 18:00:00             2.1  44.100002 -3.219561   \n",
       "\n",
       "           VGRD        HPBL        TMP  goes_measurement  \n",
       "0     -1.797583   256.61905  282.81880         -0.005922  \n",
       "1      1.764917   337.49405  281.63130          0.087090  \n",
       "2      1.014917   270.61905  280.13130          0.094333  \n",
       "3     -1.610083  1009.30660  288.19380         -0.024185  \n",
       "4     -1.172583   460.43155  285.19380         -0.014013  \n",
       "...         ...         ...        ...               ...  \n",
       "31565 -2.034275   892.42080  273.49664          0.054725  \n",
       "31566  4.869552  1664.20080  297.60638          0.010030  \n",
       "31567  4.682052   842.38837  297.60638          0.282129  \n",
       "31568  5.465472  1292.07180  298.64655          0.233581  \n",
       "31569 -4.265393   980.89050  283.69052          0.200753  \n",
       "\n",
       "[31570 rows x 9 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>air_data_value</th>\n",
       "      <th>RH</th>\n",
       "      <th>UGRD</th>\n",
       "      <th>VGRD</th>\n",
       "      <th>HPBL</th>\n",
       "      <th>TMP</th>\n",
       "      <th>goes_measurement</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>31570.000000</td>\n",
       "      <td>31570.000000</td>\n",
       "      <td>31570.000000</td>\n",
       "      <td>31570.000000</td>\n",
       "      <td>31570.000000</td>\n",
       "      <td>31570.000000</td>\n",
       "      <td>31570.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>7.948898</td>\n",
       "      <td>42.447127</td>\n",
       "      <td>1.077064</td>\n",
       "      <td>0.459147</td>\n",
       "      <td>1109.665863</td>\n",
       "      <td>294.993541</td>\n",
       "      <td>0.360805</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>5.912331</td>\n",
       "      <td>21.691148</td>\n",
       "      <td>2.895799</td>\n",
       "      <td>2.927115</td>\n",
       "      <td>845.487755</td>\n",
       "      <td>9.300556</td>\n",
       "      <td>0.574477</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>2.000000</td>\n",
       "      <td>2.700000</td>\n",
       "      <td>-12.932089</td>\n",
       "      <td>-16.077074</td>\n",
       "      <td>19.441486</td>\n",
       "      <td>252.974610</td>\n",
       "      <td>-0.050000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>4.000000</td>\n",
       "      <td>24.800001</td>\n",
       "      <td>-0.811573</td>\n",
       "      <td>-1.375287</td>\n",
       "      <td>404.699665</td>\n",
       "      <td>289.247043</td>\n",
       "      <td>0.083083</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>6.500000</td>\n",
       "      <td>40.400000</td>\n",
       "      <td>0.945436</td>\n",
       "      <td>0.298247</td>\n",
       "      <td>978.844550</td>\n",
       "      <td>295.964700</td>\n",
       "      <td>0.197401</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>10.000000</td>\n",
       "      <td>58.100002</td>\n",
       "      <td>2.859779</td>\n",
       "      <td>2.253714</td>\n",
       "      <td>1607.010550</td>\n",
       "      <td>301.890297</td>\n",
       "      <td>0.408507</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>91.000000</td>\n",
       "      <td>100.000000</td>\n",
       "      <td>15.319673</td>\n",
       "      <td>13.251497</td>\n",
       "      <td>4843.839400</td>\n",
       "      <td>321.825350</td>\n",
       "      <td>4.999973</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       air_data_value            RH          UGRD          VGRD          HPBL  \\\n",
       "count    31570.000000  31570.000000  31570.000000  31570.000000  31570.000000   \n",
       "mean         7.948898     42.447127      1.077064      0.459147   1109.665863   \n",
       "std          5.912331     21.691148      2.895799      2.927115    845.487755   \n",
       "min          2.000000      2.700000    -12.932089    -16.077074     19.441486   \n",
       "25%          4.000000     24.800001     -0.811573     -1.375287    404.699665   \n",
       "50%          6.500000     40.400000      0.945436      0.298247    978.844550   \n",
       "75%         10.000000     58.100002      2.859779      2.253714   1607.010550   \n",
       "max         91.000000    100.000000     15.319673     13.251497   4843.839400   \n",
       "\n",
       "                TMP  goes_measurement  \n",
       "count  31570.000000      31570.000000  \n",
       "mean     294.993541          0.360805  \n",
       "std        9.300556          0.574477  \n",
       "min      252.974610         -0.050000  \n",
       "25%      289.247043          0.083083  \n",
       "50%      295.964700          0.197401  \n",
       "75%      301.890297          0.408507  \n",
       "max      321.825350          4.999973  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Some basic statistics\n",
    "data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 31570 entries, 0 to 31569\n",
      "Data columns (total 9 columns):\n",
      " #   Column            Non-Null Count  Dtype         \n",
      "---  ------            --------------  -----         \n",
      " 0   station_id        31570 non-null  object        \n",
      " 1   stime             31570 non-null  datetime64[ns]\n",
      " 2   air_data_value    31570 non-null  float64       \n",
      " 3   RH                31570 non-null  float64       \n",
      " 4   UGRD              31570 non-null  float64       \n",
      " 5   VGRD              31570 non-null  float64       \n",
      " 6   HPBL              31570 non-null  float64       \n",
      " 7   TMP               31570 non-null  float64       \n",
      " 8   goes_measurement  31570 non-null  float64       \n",
      "dtypes: datetime64[ns](1), float64(7), object(1)\n",
      "memory usage: 2.2+ MB\n"
     ]
    }
   ],
   "source": [
    "#Type of data of each column\n",
    "data.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Let's do the SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 255,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Split the data into training and test set\n",
    "train, test = train_test_split(data, test_size=.2) \n",
    "\n",
    "#We only focus on RH and goes_measurement\n",
    "X_train_raw = train.loc[train['goes_measurement']<2.5,['RH','goes_measurement']].to_numpy() \n",
    "X_test_raw = test.loc[test['goes_measurement']<2.5,['RH','goes_measurement']].to_numpy() \n",
    "X_outliers_raw=data.loc[data['goes_measurement']>2.5,['RH','goes_measurement']].to_numpy()\n",
    "outliers_prop = len(X_outliers) / len(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "metadata": {},
   "outputs": [],
   "source": [
    "#We scale the data\n",
    "X_train=preprocessing.scale(X_train_raw)\n",
    "X_test=preprocessing.scale(X_test_raw)     \n",
    "X_outliers=preprocessing.scale(X_outliers_raw)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 304,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#We make plots of the training, test and novelty points\n",
    "plt.scatter([a[0] for a in X_train],[a[1] for a in X_train], c=\"g\", label=\"Training\")\n",
    "plt.scatter([a[0] for a in X_test],[a[1] for a in X_test], c=\"y\", label=\"Test\")\n",
    "plt.scatter([a[0] for a in X_outliers],[a[1] for a in X_outliers], c=\"r\", label=\"Novelty\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 299,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "OneClassSVM(cache_size=200, coef0=0.0, degree=3, gamma='auto', kernel='rbf',\n",
       "            max_iter=-1, nu=0.32, shrinking=True, tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 299,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Train the model\n",
    "clf = OneClassSVM(kernel='rbf',gamma='auto',nu=.32)\n",
    "clf.fit(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 302,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_oneclass_svm(svm):\n",
    "    # Definimos una grilla de puntos sobre la cual vamos a determinar la frontera de detección de anomalías:\n",
    "    xx, yy = np.meshgrid(np.linspace(-3, 3, 500), np.linspace(-2, 6, 500))\n",
    "\n",
    "    # Obtenemos la distancia con la frontera de decisión para cada punto\n",
    "    Z = svm.decision_function(np.c_[xx.ravel(), yy.ravel()])\n",
    "    Z = Z.reshape(xx.shape)\n",
    "    plt.title(\"Fronteras de detección de anomalías (en rojo)\")\n",
    "    \n",
    "    # Ploteamos fronteras y pintamos regiones interna y externa a la frontera\n",
    "    plt.contourf(xx, yy, Z, levels=[Z.min(), 0], colors=\"gray\") # Región anómala\n",
    "    a = plt.contour(xx, yy, Z, levels=[0], linewidths=4, colors='red') # Fronteras de decisión\n",
    "    plt.contourf(xx, yy, Z, levels=[0, Z.max()], colors='palevioletred') # Región de tipicidad\n",
    "    \n",
    "    # Ploteamos los puntos de entrenamiento, test y anomalías\n",
    "    s = 40\n",
    "    b1 = plt.scatter(X_train[:, 0], X_train[:, 1], s=s, edgecolors='k', c=\"g\") # Puntos de entrenamiento\n",
    "    b2 = plt.scatter(X_test[:, 0], X_test[:, 1], s=s, edgecolors='k', c=\"y\") # Puntos de Test\n",
    "    c = plt.scatter(X_outliers[:, 0], X_outliers[:, 1], s=s, edgecolors='k', c=\"r\") # Puntos excepcionales\n",
    "    \n",
    "    #Leyenda\n",
    "    plt.axis('tight') # Solo el espacio necesario\n",
    "    plt.xlim((-3, 3))\n",
    "    plt.ylim((-2, 6))\n",
    "    plt.legend([a.collections[0], b1, b2, c],\n",
    "               [\"Novelty frontier\", \"Training\", \"Test normal\", \"Test novelty\"],\n",
    "               loc=\"upper left\",\n",
    "               prop=matplotlib.font_manager.FontProperties(size=11))\n",
    "    plt.show()\n",
    "    \n",
    "    # Calculamos accuracy del training, test positivos y negativos\n",
    "    y_pred_train = clf.predict(X_train)\n",
    "    y_pred_test = clf.predict(X_test)\n",
    "    y_pred_outliers = clf.predict(X_outliers)\n",
    "    n_error_train = y_pred_train[y_pred_train == -1].size\n",
    "    n_error_test = y_pred_test[y_pred_test == -1].size\n",
    "    n_error_outliers = y_pred_outliers[y_pred_outliers == 1].size\n",
    "    \n",
    "    print(\"Accuracy training set: \"+str(1-n_error_train/len(X_train)))\n",
    "    print(\"Recall (normal) test set: \"+str(1-n_error_test/len(X_test)))\n",
    "    print(\"Specificity (novelty) test set: \"+str(1-n_error_outliers/len(X_outliers)))\n",
    "    print(\"Accuracy all test set: \"+ str(1-(n_error_test+n_error_outliers)/(len(X_test)+len(X_outliers))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 303,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy training set: 0.6799807507218479\n",
      "Recall (normal) test set: 0.68854033290653\n",
      "Specificity (novelty) test set: 0.8134715025906736\n",
      "Accuracy all test set: 0.6958094663852878\n"
     ]
    }
   ],
   "source": [
    "plot_oneclass_svm(clf)"
   ]
  }
 ],
 "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
}