rdu-weather-history-SVR-Linear.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Índice\n",
    "1. [Introdução](#introduction)\n",
    "2. [Lendo os dados](#reading)   \n",
    "3. [Visão Geral](#visao)    \n",
    " 3.1. [Mapas de Calor](#seaborn)\n",
    "4. [Pré Processamento](#processing)   \n",
    " 4.1. [Agrupamento dos dados para teste](#grouping)   \n",
    " 4.1.1. [Criando nova coluna \"Season\"](#column-season)   \n",
    " 4.2. [Criando DataFrames separados para cada estação](#dataframe-season)   \n",
    "5. [SVR (Support Vector Regression)](#svr)    \n",
    "    5.1. [SVR com kernel linear](#svrLinear)  \n",
    "    5.2. [SVR kernel RBF](#svrRbf)   \n",
    "6. [Referências](#references)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Introdução <a name=\"introduction\"></a>\n",
    "<p>Este <i>dataset</i> contém dados meteorológicos do Aeroporto Internacional de Raleigh Durham retirados do serviço da Web do NOAA.</p>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Lendo os Dados <a name=\"reading\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "# As colunas estão separadas pelo símbolo ';' #. \n",
    "weather_history = pd.read_csv('rdu-weather-history.csv', parse_dates=['date'], sep=';')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Visão geral <a name=\"visao\"></a>\n",
    "<p>O <i>dataset</i> contém informação sobre dados meteorológicos do Aeroporto Internacional de Raleigh Durham, desde 2007. Possui dados como temperaturas mínimas e máximas, volume de chuva, velocidade do vento e etc.</p>   \n",
    "</p>Visualizando as primeiras cinco entradas do DataFrame:</p>    "
   ]
  },
  {
   "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>date</th>\n",
       "      <th>temperaturemin</th>\n",
       "      <th>temperaturemax</th>\n",
       "      <th>precipitation</th>\n",
       "      <th>snowfall</th>\n",
       "      <th>snowdepth</th>\n",
       "      <th>avgwindspeed</th>\n",
       "      <th>fastest2minwinddir</th>\n",
       "      <th>fastest2minwindspeed</th>\n",
       "      <th>fastest5secwinddir</th>\n",
       "      <th>...</th>\n",
       "      <th>drizzle</th>\n",
       "      <th>snow</th>\n",
       "      <th>freezingrain</th>\n",
       "      <th>smokehaze</th>\n",
       "      <th>thunder</th>\n",
       "      <th>highwind</th>\n",
       "      <th>hail</th>\n",
       "      <th>blowingsnow</th>\n",
       "      <th>dust</th>\n",
       "      <th>freezingfog</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2009-10-03</td>\n",
       "      <td>55.0</td>\n",
       "      <td>82.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.91</td>\n",
       "      <td>240.0</td>\n",
       "      <td>16.11</td>\n",
       "      <td>230.0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2009-10-10</td>\n",
       "      <td>59.0</td>\n",
       "      <td>79.0</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.83</td>\n",
       "      <td>220.0</td>\n",
       "      <td>17.00</td>\n",
       "      <td>220.0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2009-10-14</td>\n",
       "      <td>46.9</td>\n",
       "      <td>61.0</td>\n",
       "      <td>0.14</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8.72</td>\n",
       "      <td>40.0</td>\n",
       "      <td>14.99</td>\n",
       "      <td>50.0</td>\n",
       "      <td>...</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>Yes</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2009-10-17</td>\n",
       "      <td>45.0</td>\n",
       "      <td>57.9</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6.26</td>\n",
       "      <td>30.0</td>\n",
       "      <td>14.09</td>\n",
       "      <td>40.0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2009-10-29</td>\n",
       "      <td>48.0</td>\n",
       "      <td>68.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.82</td>\n",
       "      <td>80.0</td>\n",
       "      <td>14.99</td>\n",
       "      <td>70.0</td>\n",
       "      <td>...</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "      <td>No</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 28 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        date  temperaturemin  temperaturemax  precipitation  snowfall  \\\n",
       "0 2009-10-03            55.0            82.0           0.00       0.0   \n",
       "1 2009-10-10            59.0            79.0           0.02       0.0   \n",
       "2 2009-10-14            46.9            61.0           0.14       0.0   \n",
       "3 2009-10-17            45.0            57.9           0.00       0.0   \n",
       "4 2009-10-29            48.0            68.0           0.00       0.0   \n",
       "\n",
       "   snowdepth  avgwindspeed  fastest2minwinddir  fastest2minwindspeed  \\\n",
       "0        0.0          2.91               240.0                 16.11   \n",
       "1        0.0          7.83               220.0                 17.00   \n",
       "2        0.0          8.72                40.0                 14.99   \n",
       "3        0.0          6.26                30.0                 14.09   \n",
       "4        0.0          5.82                80.0                 14.99   \n",
       "\n",
       "   fastest5secwinddir     ...      drizzle snow freezingrain smokehaze  \\\n",
       "0               230.0     ...           No   No           No        No   \n",
       "1               220.0     ...           No   No           No        No   \n",
       "2                50.0     ...          Yes   No           No        No   \n",
       "3                40.0     ...           No   No           No        No   \n",
       "4                70.0     ...           No   No           No        No   \n",
       "\n",
       "  thunder highwind hail blowingsnow dust freezingfog  \n",
       "0      No       No   No          No   No          No  \n",
       "1      No       No  Yes          No   No          No  \n",
       "2      No       No  Yes          No   No          No  \n",
       "3      No       No   No          No   No          No  \n",
       "4      No       No   No          No   No          No  \n",
       "\n",
       "[5 rows x 28 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weather_history.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('N\\xc3\\xbamero de linhas do dataset:', 4137)\n",
      "\n",
      "\n",
      "Resumo estatístico do DataFrame:\n"
     ]
    },
    {
     "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>temperaturemin</th>\n",
       "      <th>temperaturemax</th>\n",
       "      <th>precipitation</th>\n",
       "      <th>snowfall</th>\n",
       "      <th>snowdepth</th>\n",
       "      <th>avgwindspeed</th>\n",
       "      <th>fastest2minwinddir</th>\n",
       "      <th>fastest2minwindspeed</th>\n",
       "      <th>fastest5secwinddir</th>\n",
       "      <th>fastest5secwindspeed</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>4136.000000</td>\n",
       "      <td>4136.000000</td>\n",
       "      <td>4136.000000</td>\n",
       "      <td>4135.000000</td>\n",
       "      <td>4136.000000</td>\n",
       "      <td>4134.000000</td>\n",
       "      <td>4135.000000</td>\n",
       "      <td>4135.000000</td>\n",
       "      <td>4118.000000</td>\n",
       "      <td>4118.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>50.540063</td>\n",
       "      <td>72.017021</td>\n",
       "      <td>0.126630</td>\n",
       "      <td>0.012965</td>\n",
       "      <td>0.017384</td>\n",
       "      <td>5.860614</td>\n",
       "      <td>172.541717</td>\n",
       "      <td>15.957151</td>\n",
       "      <td>177.056824</td>\n",
       "      <td>21.801610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>16.229527</td>\n",
       "      <td>16.530515</td>\n",
       "      <td>0.371318</td>\n",
       "      <td>0.195214</td>\n",
       "      <td>0.213953</td>\n",
       "      <td>2.958446</td>\n",
       "      <td>94.603272</td>\n",
       "      <td>5.270319</td>\n",
       "      <td>96.850988</td>\n",
       "      <td>7.096004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>4.100000</td>\n",
       "      <td>23.200000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>4.920000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>6.930000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>37.000000</td>\n",
       "      <td>60.100000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.580000</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>12.970000</td>\n",
       "      <td>90.000000</td>\n",
       "      <td>17.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>52.000000</td>\n",
       "      <td>73.900000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.370000</td>\n",
       "      <td>210.000000</td>\n",
       "      <td>14.990000</td>\n",
       "      <td>210.000000</td>\n",
       "      <td>21.030000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>64.900000</td>\n",
       "      <td>86.000000</td>\n",
       "      <td>0.040000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7.610000</td>\n",
       "      <td>240.000000</td>\n",
       "      <td>18.120000</td>\n",
       "      <td>240.000000</td>\n",
       "      <td>25.050000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>80.100000</td>\n",
       "      <td>105.100000</td>\n",
       "      <td>6.450000</td>\n",
       "      <td>6.690000</td>\n",
       "      <td>5.910000</td>\n",
       "      <td>19.010000</td>\n",
       "      <td>360.000000</td>\n",
       "      <td>59.950000</td>\n",
       "      <td>360.000000</td>\n",
       "      <td>86.120000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       temperaturemin  temperaturemax  precipitation     snowfall  \\\n",
       "count     4136.000000     4136.000000    4136.000000  4135.000000   \n",
       "mean        50.540063       72.017021       0.126630     0.012965   \n",
       "std         16.229527       16.530515       0.371318     0.195214   \n",
       "min          4.100000       23.200000       0.000000     0.000000   \n",
       "25%         37.000000       60.100000       0.000000     0.000000   \n",
       "50%         52.000000       73.900000       0.000000     0.000000   \n",
       "75%         64.900000       86.000000       0.040000     0.000000   \n",
       "max         80.100000      105.100000       6.450000     6.690000   \n",
       "\n",
       "         snowdepth  avgwindspeed  fastest2minwinddir  fastest2minwindspeed  \\\n",
       "count  4136.000000   4134.000000         4135.000000           4135.000000   \n",
       "mean      0.017384      5.860614          172.541717             15.957151   \n",
       "std       0.213953      2.958446           94.603272              5.270319   \n",
       "min       0.000000      0.000000           10.000000              4.920000   \n",
       "25%       0.000000      3.580000           80.000000             12.970000   \n",
       "50%       0.000000      5.370000          210.000000             14.990000   \n",
       "75%       0.000000      7.610000          240.000000             18.120000   \n",
       "max       5.910000     19.010000          360.000000             59.950000   \n",
       "\n",
       "       fastest5secwinddir  fastest5secwindspeed  \n",
       "count         4118.000000           4118.000000  \n",
       "mean           177.056824             21.801610  \n",
       "std             96.850988              7.096004  \n",
       "min             10.000000              6.930000  \n",
       "25%             90.000000             17.000000  \n",
       "50%            210.000000             21.030000  \n",
       "75%            240.000000             25.050000  \n",
       "max            360.000000             86.120000  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Número de entradas\n",
    "print(\"Número de linhas do dataset:\", weather_history.shape[0])\n",
    "\n",
    "print(\"\\n\\nResumo estatístico do DataFrame:\")\n",
    "weather_history.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 4137 entries, 0 to 4136\n",
      "Data columns (total 28 columns):\n",
      "date                    4137 non-null datetime64[ns]\n",
      "temperaturemin          4136 non-null float64\n",
      "temperaturemax          4136 non-null float64\n",
      "precipitation           4136 non-null float64\n",
      "snowfall                4135 non-null float64\n",
      "snowdepth               4136 non-null float64\n",
      "avgwindspeed            4134 non-null float64\n",
      "fastest2minwinddir      4135 non-null float64\n",
      "fastest2minwindspeed    4135 non-null float64\n",
      "fastest5secwinddir      4118 non-null float64\n",
      "fastest5secwindspeed    4118 non-null float64\n",
      "fog                     4137 non-null object\n",
      "fogheavy                4137 non-null object\n",
      "mist                    4137 non-null object\n",
      "rain                    4137 non-null object\n",
      "fogground               4137 non-null object\n",
      "ice                     4137 non-null object\n",
      "glaze                   4137 non-null object\n",
      "drizzle                 4137 non-null object\n",
      "snow                    4137 non-null object\n",
      "freezingrain            4137 non-null object\n",
      "smokehaze               4137 non-null object\n",
      "thunder                 4137 non-null object\n",
      "highwind                4137 non-null object\n",
      "hail                    4137 non-null object\n",
      "blowingsnow             4137 non-null object\n",
      "dust                    4137 non-null object\n",
      "freezingfog             4137 non-null object\n",
      "dtypes: datetime64[ns](1), float64(10), object(17)\n",
      "memory usage: 905.0+ KB\n"
     ]
    }
   ],
   "source": [
    "weather_history.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3.1 Utilizando a função Seaborn de mapas de calor <a name=\"seaborn\"></a>\n",
    "<p>Seaborn é uma biblioteca em python para criação de visualizações estatísticas.</p>\n",
    "<p>O método ```.corr() ```, calcula o <b>coeficiente de correlação de Pearson</b> entre todos os pares de colunas numéricas do DataFrame.</p>\n",
    "<p>Interpretando o coeficiente:\n",
    "- 0.9 para mais ou para menos indica uma correlação muito forte.\n",
    "- 0.7 a 0.9 positivo ou negativo indica uma correlação forte.\n",
    "- 0.5 a 0.7 positivo ou negativo indica uma correlação moderada.\n",
    "- 0.3 a 0.5 positivo ou negativo indica uma correlação fraca.\n",
    "- 0 a 0.3 positivo ou negativo indica uma correlação desprezível."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns; sns.set()\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "plt.figure(figsize=(18,10))\n",
    "weather_map = sns.heatmap(weather_history.corr(), annot=True, square=True, cmap=\"YlGnBu\", linewidths=.3)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Pré-processamento dos dados <a name=\"processing\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Substituindo as string por números binários -- No = 0.0 e Yes = 1.0\n",
    "\n",
    "weather_history = weather_history.replace('No', 0.0)\n",
    "weather_history = weather_history.replace('Yes', 1.0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "date                     0\n",
       "temperaturemin           1\n",
       "temperaturemax           1\n",
       "precipitation            1\n",
       "snowfall                 2\n",
       "snowdepth                1\n",
       "avgwindspeed             3\n",
       "fastest2minwinddir       2\n",
       "fastest2minwindspeed     2\n",
       "fastest5secwinddir      19\n",
       "fastest5secwindspeed    19\n",
       "fog                      0\n",
       "fogheavy                 0\n",
       "mist                     0\n",
       "rain                     0\n",
       "fogground                0\n",
       "ice                      0\n",
       "glaze                    0\n",
       "drizzle                  0\n",
       "snow                     0\n",
       "freezingrain             0\n",
       "smokehaze                0\n",
       "thunder                  0\n",
       "highwind                 0\n",
       "hail                     0\n",
       "blowingsnow              0\n",
       "dust                     0\n",
       "freezingfog              0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Verificando a soma dos valores nulos em cada feature\n",
    "\n",
    "weather_history.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "date                    0\n",
       "temperaturemin          0\n",
       "temperaturemax          0\n",
       "precipitation           0\n",
       "snowfall                0\n",
       "snowdepth               0\n",
       "avgwindspeed            0\n",
       "fastest2minwinddir      0\n",
       "fastest2minwindspeed    0\n",
       "fastest5secwinddir      0\n",
       "fastest5secwindspeed    0\n",
       "fog                     0\n",
       "fogheavy                0\n",
       "mist                    0\n",
       "rain                    0\n",
       "fogground               0\n",
       "ice                     0\n",
       "glaze                   0\n",
       "drizzle                 0\n",
       "snow                    0\n",
       "freezingrain            0\n",
       "smokehaze               0\n",
       "thunder                 0\n",
       "highwind                0\n",
       "hail                    0\n",
       "blowingsnow             0\n",
       "dust                    0\n",
       "freezingfog             0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.preprocessing import Imputer\n",
    "import numpy as np\n",
    "\n",
    "#Substituindo os valores nulos pela média dos valores referentes a feature\n",
    "imputer = Imputer(missing_values=np.nan, strategy='median', axis=0)\n",
    "weather_history[['temperaturemin']] = imputer.fit_transform(weather_history[['temperaturemin']])\n",
    "weather_history[['temperaturemax']] = imputer.fit_transform(weather_history[['temperaturemax']])\n",
    "weather_history[['precipitation']] = imputer.fit_transform(weather_history[['precipitation']])\n",
    "weather_history[['snowfall']] = imputer.fit_transform(weather_history[['snowfall']])\n",
    "weather_history[['snowdepth']] = imputer.fit_transform(weather_history[['snowdepth']])\n",
    "weather_history[['avgwindspeed']] = imputer.fit_transform(weather_history[['avgwindspeed']])\n",
    "weather_history[['fastest2minwinddir']] = imputer.fit_transform(weather_history[['fastest2minwinddir']])\n",
    "weather_history[['fastest2minwindspeed']] = imputer.fit_transform(weather_history[['fastest2minwindspeed']])\n",
    "weather_history[['fastest5secwinddir']] = imputer.fit_transform(weather_history[['fastest5secwinddir']])\n",
    "weather_history[['fastest5secwindspeed']] = imputer.fit_transform(weather_history[['fastest5secwindspeed']])\n",
    "\n",
    "weather_history.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.1. Agrupamento dos dados para teste    <a name=\"grouping\"></a>\n",
    "\n",
    "Como o clima de uma região possui uma grande relação com as estações do ano, os dados para teste foram agrupados de acordo com a estação que occorre em uma data específica.\n",
    "\n",
    "- Primavera: 20 de março a 20 de junho \n",
    "- Verão: 21 de junho a 21 de setembro\n",
    "- Outono: 22 de setembro a 20 de dezembro\n",
    "- Inverno: 21 de dezembro a 19 de março"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.1.1. Criando nova coluna \"Season\"    <a name=\"column-season\"></a>\n",
    "\n",
    "A coluna \"season\" foi criada com base nos periodos de inicio e de fim das estações do ano no local em que os dados foram colhidos."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Antes de utilizar o algoritimo para definir as estações a partir das datas é feita uma pequena alteração no formato da data.  \n",
    "Antes: AAAA-MM-DD  \n",
    "Depois: AAAAMMDD"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "season_array = []\n",
    "\n",
    "for date in weather_history['date']:\n",
    "    if int(date.strftime('%m%d')) >= 1221 or int(date.strftime('%m%d')) <= 319:\n",
    "        season = 0 #'winter'\n",
    "    elif int(date.strftime('%m%d')) >= 320 and int(date.strftime('%m%d')) <= 620:\n",
    "        season = 1 #'spring'\n",
    "    elif int(date.strftime('%m%d')) >= 621 and int(date.strftime('%m%d')) <= 921:\n",
    "        season = 2 #'summer'\n",
    "    elif int(date.strftime('%m%d')) >= 922 and int(date.strftime('%m%d')) <= 1220:\n",
    "        season = 3 #'autumn'    \n",
    "    \n",
    "    season_array.append(season)\n",
    "\n",
    "weather_history['season'] = season_array"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>date</th>\n",
       "      <th>temperaturemin</th>\n",
       "      <th>temperaturemax</th>\n",
       "      <th>precipitation</th>\n",
       "      <th>snowfall</th>\n",
       "      <th>snowdepth</th>\n",
       "      <th>avgwindspeed</th>\n",
       "      <th>fastest2minwinddir</th>\n",
       "      <th>fastest2minwindspeed</th>\n",
       "      <th>fastest5secwinddir</th>\n",
       "      <th>...</th>\n",
       "      <th>snow</th>\n",
       "      <th>freezingrain</th>\n",
       "      <th>smokehaze</th>\n",
       "      <th>thunder</th>\n",
       "      <th>highwind</th>\n",
       "      <th>hail</th>\n",
       "      <th>blowingsnow</th>\n",
       "      <th>dust</th>\n",
       "      <th>freezingfog</th>\n",
       "      <th>season</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2070</th>\n",
       "      <td>2007-01-01</td>\n",
       "      <td>48.9</td>\n",
       "      <td>68.0</td>\n",
       "      <td>0.45</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.75</td>\n",
       "      <td>190.0</td>\n",
       "      <td>25.05</td>\n",
       "      <td>180.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>813</th>\n",
       "      <td>2007-01-02</td>\n",
       "      <td>32.0</td>\n",
       "      <td>55.9</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.13</td>\n",
       "      <td>320.0</td>\n",
       "      <td>12.97</td>\n",
       "      <td>330.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>814</th>\n",
       "      <td>2007-01-03</td>\n",
       "      <td>28.9</td>\n",
       "      <td>62.1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.24</td>\n",
       "      <td>220.0</td>\n",
       "      <td>14.09</td>\n",
       "      <td>220.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>815</th>\n",
       "      <td>2007-01-04</td>\n",
       "      <td>46.0</td>\n",
       "      <td>69.1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.47</td>\n",
       "      <td>220.0</td>\n",
       "      <td>14.09</td>\n",
       "      <td>230.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2449</th>\n",
       "      <td>2007-01-05</td>\n",
       "      <td>57.0</td>\n",
       "      <td>72.0</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8.05</td>\n",
       "      <td>190.0</td>\n",
       "      <td>21.03</td>\n",
       "      <td>190.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 29 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           date  temperaturemin  temperaturemax  precipitation  snowfall  \\\n",
       "2070 2007-01-01            48.9            68.0           0.45       0.0   \n",
       "813  2007-01-02            32.0            55.9           0.00       0.0   \n",
       "814  2007-01-03            28.9            62.1           0.00       0.0   \n",
       "815  2007-01-04            46.0            69.1           0.00       0.0   \n",
       "2449 2007-01-05            57.0            72.0           0.86       0.0   \n",
       "\n",
       "      snowdepth  avgwindspeed  fastest2minwinddir  fastest2minwindspeed  \\\n",
       "2070        0.0         12.75               190.0                 25.05   \n",
       "813         0.0          3.13               320.0                 12.97   \n",
       "814         0.0          2.24               220.0                 14.09   \n",
       "815         0.0          4.47               220.0                 14.09   \n",
       "2449        0.0          8.05               190.0                 21.03   \n",
       "\n",
       "      fastest5secwinddir   ...    snow  freezingrain  smokehaze  thunder  \\\n",
       "2070               180.0   ...     0.0           0.0        0.0      0.0   \n",
       "813                330.0   ...     0.0           0.0        0.0      0.0   \n",
       "814                220.0   ...     0.0           0.0        0.0      0.0   \n",
       "815                230.0   ...     0.0           0.0        0.0      0.0   \n",
       "2449               190.0   ...     0.0           0.0        0.0      0.0   \n",
       "\n",
       "      highwind  hail  blowingsnow  dust  freezingfog  season  \n",
       "2070       0.0   0.0          0.0   0.0          0.0       0  \n",
       "813        0.0   0.0          0.0   0.0          0.0       0  \n",
       "814        0.0   0.0          0.0   0.0          0.0       0  \n",
       "815        0.0   0.0          0.0   0.0          0.0       0  \n",
       "2449       0.0   0.0          0.0   0.0          0.0       0  \n",
       "\n",
       "[5 rows x 29 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Ordenando o dataset pela coluna data\n",
    "\n",
    "weather_history = weather_history.sort_values('date', ascending=True)\n",
    "weather_history.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4.2. Criando DataFrames separados para cada estação. <a name=\"dataframe-season\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "winter_weather_history = weather_history[weather_history['season'] == 0]\n",
    "spring_weather_history = weather_history[weather_history['season'] == 1]\n",
    "summer_weather_history = weather_history[weather_history['season'] == 2]\n",
    "autumn_weather_history = weather_history[weather_history['season'] == 3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1ed0e13f90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "len_winter = len(winter_weather_history['date'])\n",
    "len_summer = len(summer_weather_history['date'])\n",
    "len_spring = len(spring_weather_history['date'])\n",
    "len_autumn = len(autumn_weather_history['date'])\n",
    "\n",
    "X = ['winter', 'summer', 'spring', 'autumn']\n",
    "Y = [len_winter, len_summer, len_spring, len_autumn]\n",
    "\n",
    "plt.bar(X, Y)\n",
    "plt.xlabel('Seasons')\n",
    "plt.ylabel('Amount of data for each season')\n",
    "plt.title('Season per amount')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f1f08595710>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1ed0e17350>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Exemplo para mostrar que a correlação com a divisão do dataframe em estações aumentou. Nesse exemplo\n",
    "# tem os dados do verão de 2007\n",
    "\n",
    "plt.figure(figsize=(10,8))\n",
    "\n",
    "summer_weather_history = summer_weather_history[summer_weather_history['date'] <= '2007-12-31']                                        \n",
    "summer_weather_history = summer_weather_history.drop(columns=['snowfall', 'snowdepth', 'ice', 'glaze', 'freezingrain', 'highwind', 'hail', 'blowingsnow', 'dust', 'freezingfog', 'season', 'snow'])\n",
    "sns.heatmap(summer_weather_history.corr(), square=True, cmap=\"YlGnBu\", linewidths=.2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. SVR (Support Vector Regression)  <a name=\"svr\"></a>  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Support Vector Regression é uma extensão do Support Vector Classification que é capaz de realizar regressões. Esse método realiza o mapeamento dos dados em um espaço multidimensional através de um mapeamento não-linear (geralmente utilizando kernels) e então realiza uma regressão linear neste espaço multidimensional transformado.\n",
    "\n",
    "A ideia básico dos SV é a criação de hiperplanos para separar os dados.\n",
    "\n",
    "O SVR depende apenas de um subconjunto dos dados de treinamento, porque a função custo para a construção do modelo ignora os dados de treinamento perto da previsão do modelo.\n",
    "\n",
    "Devido ao uso dos kernel tricks, o SVR pode tratar dados pouco linearmente separáveis como linearmente sepaŕaveis através do aumento do espaço dimensional onde os dados estão inseridos.\n",
    "\n",
    "O uso de kernel pode ser entendido através dos seguintes gráficos:\n",
    "\n",
    "![grafico1](https://cdn-images-1.medium.com/max/1600/0*ngkO1BblQXnOTcmr.png)\n",
    "\n",
    "Os principais parâmetros do SVR aqui utilizados são:\n",
    "\n",
    "* Kernel: define o tipo de kernel que será utilizado (aqui serão utilizados os kernels linear e RBF)\n",
    "\n",
    "\n",
    "* C (cost): define o peso de uma classificação errônea de exemplos de treinamento com relação a simplicidade da superfície de decisão. Um C baixo torna a superfície de decisão suave, enquanto um C alto tem como objetivo classificar todos os exemplos de treinamento corretamente, dando ao modelo a liberdade de selecionar mais amostras como vetores de suporte. Ou seja, o parâmetro faz o peso entre acurácia e eficácia, onde um C pequeno tende a aumentar a acurácia e um C grande tende a executar mais rapidamente.\n",
    "\n",
    "\n",
    "* gamma (utilizado somente no RBF): define até onde chega a influência de um único exemplo de treinamento, com valores baixos significando \"longe\" e valores altos significando \"perto\". Basicamente um gamma muito grande causa overfitting e um gamma muito pequeno faria com que o RBF se tornasse muito parecido com um linear, já que a influência de um vetor seria muito grande.\n",
    "\n",
    "\n",
    "* epsilon: é, resumidamente, a margem de erro que não é penalizado na regressão. Assim sendo, um epsilon muito grande permite que os erros não sejam penalizados, e um epsilon muito pequeno penaliza praticamente todos os erros.\n",
    "\n",
    "Os demais ou não se adequam aos tipos de kernel (como degree e coef0) ou a variação de seus valores dentro do intervalo testado não surtiram diferenças significativas.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "   \n",
    "import matplotlib.pyplot as plt\n",
    "X = weather_history.drop('temperaturemin', axis=1).values\n",
    "y = weather_history['temperaturemin'].values\n",
    "\n",
    "weather_history_no_min = weather_history.drop('temperaturemin', axis=1)\n",
    "\n",
    "X_new = weather_history_no_min\n",
    "for i in enumerate(weather_history_no_min.columns):\n",
    "    related_value = weather_history_no_min[i[1]].values\n",
    "    if np.array_equal(np.unique(related_value), [0, 1.]):\n",
    "        X_new = X_new.drop(i[1], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5.1. SVR com kernel linear <a name=\"svrLinear\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVR(C=2.0, cache_size=200, coef0=0.0, degree=3, epsilon=0.1, gamma='auto',\n",
       "  kernel='linear', max_iter=-1, shrinking=True, tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.svm import SVR\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X_new_new = X_new.drop('date', axis=1)\n",
    "X_new_new = X_new_new.values\n",
    "\n",
    "regr = SVR(kernel='linear', C=2.0)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_new_new, y, test_size=0.25, random_state=42)\n",
    "regr.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Training set score: 0.853\n",
      "\n",
      "Testing set score: 0.855\n",
      "\n",
      "Features: 11\n"
     ]
    }
   ],
   "source": [
    "print(\"\\nTraining set score: {:.3f}\".format(regr.score(X_train, y_train)))\n",
    "print(\"\\nTesting set score: {:.3f}\".format(regr.score(X_test, y_test)))\n",
    "print(\"\\nFeatures: {}\".format(np.sum(regr.coef_.shape)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1ed0196fd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "predict = regr.predict(X_test)\n",
    "figure, axes = plt.subplots(figsize=(9, 9))\n",
    "ax = sns.regplot(x = y_test, y = predict, scatter_kws={\"color\": \"orange\"}, line_kws={\"color\": \"blue\"})\n",
    "ax = ax.set(xlabel='Actual Temperaturemin: Test Set', ylabel = 'Predicted Temperaturemin: Test Set')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Winter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Training set score: 0.726\n",
      "Test set score: 0.716\n"
     ]
    }
   ],
   "source": [
    "X_winter = winter_weather_history.drop('temperaturemin', axis=1)\n",
    "X_winter = X_winter.drop('date', axis=1)\n",
    "X_winter = X_winter.drop('season', axis=1)\n",
    "X_winter.shape\n",
    "\n",
    "y_winter = (winter_weather_history['temperaturemin'])\n",
    "y_winter.shape\n",
    "\n",
    "# Separando o dataset em treino e teste. 30% para teste, e 70% para treinamento.\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_winter, y_winter, test_size = 0.3, random_state=42)\n",
    "\n",
    "regr = SVR(kernel='linear', C=2.0)\n",
    "regr.fit(X_train, y_train) \n",
    "\n",
    "print(\"\\nTraining set score: {:.3f}\".format(regr.score(X_train, y_train)))\n",
    "print(\"Test set score: {:.3f}\".format(regr.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Summer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Training set score: 0.567\n",
      "Test set score: 0.606\n"
     ]
    }
   ],
   "source": [
    "X_summer = summer_weather_history.drop('temperaturemin', axis=1)\n",
    "X_summer = X_summer.drop('date', axis=1)\n",
    "X_summer.shape\n",
    "\n",
    "y_summer = (summer_weather_history['temperaturemin'])\n",
    "y_summer.shape\n",
    "\n",
    "# Separando o dataset em treino e teste. 30% para teste, e 70% para treinamento.\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_summer, y_summer, test_size = 0.3, random_state=42)\n",
    "\n",
    "regr = SVR(kernel='linear', C=2.0)\n",
    "regr.fit(X_train, y_train) \n",
    "\n",
    "print(\"\\nTraining set score: {:.3f}\".format(regr.score(X_train, y_train)))\n",
    "print(\"Test set score: {:.3f}\".format(regr.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Spring"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Training set score: 0.746\n",
      "Test set score: 0.727\n"
     ]
    }
   ],
   "source": [
    "X_spring = spring_weather_history.drop('temperaturemin', axis=1)\n",
    "X_spring = X_spring.drop('date', axis=1)\n",
    "X_spring = X_spring.drop('season', axis=1)\n",
    "X_spring.shape\n",
    "\n",
    "y_spring = (spring_weather_history['temperaturemin'])\n",
    "y_spring.shape\n",
    "\n",
    "# Separando o dataset em treino e teste. 30% para teste, e 70% para treinamento.\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_spring, y_spring, test_size = 0.3, random_state=42)\n",
    "\n",
    "regr = SVR(kernel='linear', C=2.0)\n",
    "regr.fit(X_train, y_train) \n",
    "\n",
    "print(\"\\nTraining set score: {:.3f}\".format(regr.score(X_train, y_train)))\n",
    "print(\"Test set score: {:.3f}\".format(regr.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Autumn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Training set score: 0.786\n",
      "Test set score: 0.768\n"
     ]
    }
   ],
   "source": [
    "X_autumn = autumn_weather_history.drop('temperaturemin', axis=1)\n",
    "X_autumn = X_autumn.drop('date', axis=1)\n",
    "X_autumn = X_autumn.drop('season', axis=1)\n",
    "X_autumn.shape\n",
    "\n",
    "y_autumn = (autumn_weather_history['temperaturemin'])\n",
    "y_autumn.shape\n",
    "\n",
    "# Separando o dataset em treino e teste. 30% para teste, e 70% para treinamento.\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_autumn, y_autumn, test_size = 0.3, random_state=42)\n",
    "\n",
    "regr = SVR(kernel='linear', C=2.0)\n",
    "regr.fit(X_train, y_train) \n",
    "\n",
    "print(\"\\nTraining set score: {:.3f}\".format(regr.score(X_train, y_train)))\n",
    "print(\"Test set score: {:.3f}\".format(regr.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5.2. SVR kernel RBF <a name=\"svrRbf\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Training set score: 0.627\n",
      "\n",
      "Testing set score: 0.539\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1ed00a2a10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "regrRbf = SVR(kernel='rbf', gamma=0.001, C=2.0)\n",
    "regrRbf.fit(X_train, y_train)\n",
    "regrRbf.score(X_train, y_train)\n",
    "predictRbf = regrRbf.predict(X_test)\n",
    "\n",
    "figure, axes = plt.subplots(figsize=(9, 9))\n",
    "ax = sns.regplot(x = y_test, y = predictRbf, scatter_kws={\"color\": \"orange\"}, line_kws={\"color\": \"blue\"})\n",
    "ax = ax.set(xlabel='Actual Temperaturemin: Test Set', ylabel = 'Predicted Temperaturemin: Test Set')\n",
    "\n",
    "print(\"\\nTraining set score: {:.3f}\".format(regrRbf.score(X_train, y_train)))\n",
    "print(\"\\nTesting set score: {:.3f}\".format(regrRbf.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Winter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Training set score: 0.588\n",
      "Test set score: 0.519\n"
     ]
    }
   ],
   "source": [
    "X_winter = winter_weather_history.drop('temperaturemin', axis=1)\n",
    "X_winter = X_winter.drop('date', axis=1)\n",
    "X_winter = X_winter.drop('season', axis=1)\n",
    "X_winter.shape\n",
    "\n",
    "y_winter = (winter_weather_history['temperaturemin'])\n",
    "y_winter.shape\n",
    "\n",
    "# Separando o dataset em treino e teste. 30% para teste, e 70% para treinamento.\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_winter, y_winter, test_size = 0.3, random_state=42)\n",
    "\n",
    "regrRbf = SVR(kernel='rbf', gamma=0.001, C=2.0)\n",
    "regrRbf.fit(X_train, y_train) \n",
    "\n",
    "print(\"\\nTraining set score: {:.3f}\".format(regrRbf.score(X_train, y_train)))\n",
    "print(\"Test set score: {:.3f}\".format(regrRbf.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Summer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Training set score: 0.350\n",
      "Test set score: 0.160\n"
     ]
    }
   ],
   "source": [
    "X_summer = summer_weather_history.drop('temperaturemin', axis=1)\n",
    "X_summer = X_summer.drop('date', axis=1)\n",
    "X_summer.shape\n",
    "\n",
    "y_summer = (summer_weather_history['temperaturemin'])\n",
    "y_summer.shape\n",
    "\n",
    "# Separando o dataset em treino e teste. 30% para teste, e 70% para treinamento.\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_summer, y_summer, test_size = 0.3, random_state=42)\n",
    "\n",
    "regrRbf = SVR(kernel='rbf', gamma=0.001, C=2.0)\n",
    "regrRbf.fit(X_train, y_train) \n",
    "\n",
    "print(\"\\nTraining set score: {:.3f}\".format(regrRbf.score(X_train, y_train)))\n",
    "print(\"Test set score: {:.3f}\".format(regrRbf.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Spring"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Training set score: 0.614\n",
      "Test set score: 0.566\n"
     ]
    }
   ],
   "source": [
    "X_spring = spring_weather_history.drop('temperaturemin', axis=1)\n",
    "X_spring = X_spring.drop('date', axis=1)\n",
    "X_spring = X_spring.drop('season', axis=1)\n",
    "X_spring.shape\n",
    "\n",
    "y_spring = (spring_weather_history['temperaturemin'])\n",
    "y_spring.shape\n",
    "\n",
    "# Separando o dataset em treino e teste. 30% para teste, e 70% para treinamento.\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_spring, y_spring, test_size = 0.3, random_state=42)\n",
    "\n",
    "regrRbf = SVR(kernel='rbf', gamma=0.001, C=2.0)\n",
    "regrRbf.fit(X_train, y_train) \n",
    "\n",
    "print(\"\\nTraining set score: {:.3f}\".format(regrRbf.score(X_train, y_train)))\n",
    "print(\"Test set score: {:.3f}\".format(regrRbf.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Autumn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Training set score: 0.627\n",
      "Test set score: 0.539\n"
     ]
    }
   ],
   "source": [
    "X_autumn = autumn_weather_history.drop('temperaturemin', axis=1)\n",
    "X_autumn = X_autumn.drop('date', axis=1)\n",
    "X_autumn = X_autumn.drop('season', axis=1)\n",
    "X_autumn.shape\n",
    "\n",
    "y_autumn = (autumn_weather_history['temperaturemin'])\n",
    "y_autumn.shape\n",
    "\n",
    "# Separando o dataset em treino e teste. 30% para teste, e 70% para treinamento.\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_autumn, y_autumn, test_size = 0.3, random_state=42)\n",
    "\n",
    "regrRbf = SVR(kernel='rbf', gamma=0.001, C=2.0)\n",
    "regrRbf.fit(X_train, y_train) \n",
    "\n",
    "print(\"\\nTraining set score: {:.3f}\".format(regrRbf.score(X_train, y_train)))\n",
    "print(\"Test set score: {:.3f}\".format(regrRbf.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6. Referências  <a name=\"references\"></a>  \n",
    "\n",
    "> [Coeficiente de correlação de Pearson](https://pt.wikipedia.org/wiki/Coeficiente_de_correla%C3%A7%C3%A3o_de_Pearson#Refer%C3%AAncias)\n",
    "\n",
    "> [Kernel Trick](http://crsouza.com/2010/02/02/the-kernel-trick/)\n",
    "\n",
    "> [Parâmetros do SVR](http://scikit-learn.org/stable/auto_examples/svm/plot_rbf_parameters.html)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}