week5_PracticalML_assignment_최혜빈.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Indian Liver Patient Records"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "우선 제가 관심있는 분야가 의료분야이고 관련전공이기도해서, 환자 진료기록 데이터를 다루고 싶었습니다\n",
    "\n",
    "관련데이터들을 쉽게 구하기 어려웠던 점에서, 제가 원했던 방향의 데이터는 찾지 못했지만!\n",
    "\n",
    "제가 관심있는 분야에서 나름 유명한 데이터를 통해 데이터 eda, 전처리, 모델 적용까지의 프로젝트를 완성하고자 하였습니다!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "이 과제에서 사용한 데이터는 Kaggle에서 다운받았으며 환자데이터들 중에서는 유명한 데이터입니다\n",
    "\n",
    "정확히는 liver disease. 간질환 환자에 대한 여러 수치들을 통해 질병유무를 예측하는 데이터셋 입니다\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "그래서 저는 데이터에 대한 전처리, EDA를 먼저 진행한 뒤 여러 모델을 사용해보고 적합한 파라미터를 찾아보는 방식으로 진행하였습니다!\n",
    "\n",
    "\n",
    "### ***\n",
    "-결측치 제거 등 전처리와 동시에 data관련 EDA를 통해 insight를 얻으려고 했습니다   \n",
    "-classification을 위한 적합한 모델을 여러개 사용해보려 하였고,   \n",
    "-적합한 parameter을 찾기위해 grid serach방식을 사용하였습니다   \n",
    "-scaling, model적용, grid search 사용 은 pipeline을 이용하여 한번에 이루어지도록 해보았습니다"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#사용할 모듈 불러오기\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "%matplotlib inline\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## <데이터 불러오기 + 살펴보기>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('indian_liver_patient.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Age</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Total_Bilirubin</th>\n",
       "      <th>Direct_Bilirubin</th>\n",
       "      <th>Alkaline_Phosphotase</th>\n",
       "      <th>Alamine_Aminotransferase</th>\n",
       "      <th>Aspartate_Aminotransferase</th>\n",
       "      <th>Total_Protiens</th>\n",
       "      <th>Albumin</th>\n",
       "      <th>Albumin_and_Globulin_Ratio</th>\n",
       "      <th>Dataset</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>65</td>\n",
       "      <td>Female</td>\n",
       "      <td>0.7</td>\n",
       "      <td>0.1</td>\n",
       "      <td>187</td>\n",
       "      <td>16</td>\n",
       "      <td>18</td>\n",
       "      <td>6.8</td>\n",
       "      <td>3.3</td>\n",
       "      <td>0.90</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>Male</td>\n",
       "      <td>10.9</td>\n",
       "      <td>5.5</td>\n",
       "      <td>699</td>\n",
       "      <td>64</td>\n",
       "      <td>100</td>\n",
       "      <td>7.5</td>\n",
       "      <td>3.2</td>\n",
       "      <td>0.74</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>62</td>\n",
       "      <td>Male</td>\n",
       "      <td>7.3</td>\n",
       "      <td>4.1</td>\n",
       "      <td>490</td>\n",
       "      <td>60</td>\n",
       "      <td>68</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.3</td>\n",
       "      <td>0.89</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>58</td>\n",
       "      <td>Male</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.4</td>\n",
       "      <td>182</td>\n",
       "      <td>14</td>\n",
       "      <td>20</td>\n",
       "      <td>6.8</td>\n",
       "      <td>3.4</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>72</td>\n",
       "      <td>Male</td>\n",
       "      <td>3.9</td>\n",
       "      <td>2.0</td>\n",
       "      <td>195</td>\n",
       "      <td>27</td>\n",
       "      <td>59</td>\n",
       "      <td>7.3</td>\n",
       "      <td>2.4</td>\n",
       "      <td>0.40</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Age  Gender  Total_Bilirubin  Direct_Bilirubin  Alkaline_Phosphotase  \\\n",
       "0   65  Female              0.7               0.1                   187   \n",
       "1   62    Male             10.9               5.5                   699   \n",
       "2   62    Male              7.3               4.1                   490   \n",
       "3   58    Male              1.0               0.4                   182   \n",
       "4   72    Male              3.9               2.0                   195   \n",
       "\n",
       "   Alamine_Aminotransferase  Aspartate_Aminotransferase  Total_Protiens  \\\n",
       "0                        16                          18             6.8   \n",
       "1                        64                         100             7.5   \n",
       "2                        60                          68             7.0   \n",
       "3                        14                          20             6.8   \n",
       "4                        27                          59             7.3   \n",
       "\n",
       "   Albumin  Albumin_and_Globulin_Ratio  Dataset  \n",
       "0      3.3                        0.90        1  \n",
       "1      3.2                        0.74        1  \n",
       "2      3.3                        0.89        1  \n",
       "3      3.4                        1.00        1  \n",
       "4      2.4                        0.40        1  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 583 entries, 0 to 582\n",
      "Data columns (total 11 columns):\n",
      "Age                           583 non-null int64\n",
      "Gender                        583 non-null object\n",
      "Total_Bilirubin               583 non-null float64\n",
      "Direct_Bilirubin              583 non-null float64\n",
      "Alkaline_Phosphotase          583 non-null int64\n",
      "Alamine_Aminotransferase      583 non-null int64\n",
      "Aspartate_Aminotransferase    583 non-null int64\n",
      "Total_Protiens                583 non-null float64\n",
      "Albumin                       583 non-null float64\n",
      "Albumin_and_Globulin_Ratio    579 non-null float64\n",
      "Dataset                       583 non-null int64\n",
      "dtypes: float64(5), int64(5), object(1)\n",
      "memory usage: 50.2+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>Age</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Total_Bilirubin</th>\n",
       "      <th>Direct_Bilirubin</th>\n",
       "      <th>Alkaline_Phosphotase</th>\n",
       "      <th>Alamine_Aminotransferase</th>\n",
       "      <th>Aspartate_Aminotransferase</th>\n",
       "      <th>Total_Protiens</th>\n",
       "      <th>Albumin</th>\n",
       "      <th>Albumin_and_Globulin_Ratio</th>\n",
       "      <th>Dataset</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>count</td>\n",
       "      <td>583.000000</td>\n",
       "      <td>583</td>\n",
       "      <td>583.000000</td>\n",
       "      <td>583.000000</td>\n",
       "      <td>583.000000</td>\n",
       "      <td>583.000000</td>\n",
       "      <td>583.000000</td>\n",
       "      <td>583.000000</td>\n",
       "      <td>583.000000</td>\n",
       "      <td>579.000000</td>\n",
       "      <td>583.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>unique</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>top</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>freq</td>\n",
       "      <td>NaN</td>\n",
       "      <td>441</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>mean</td>\n",
       "      <td>44.746141</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.298799</td>\n",
       "      <td>1.486106</td>\n",
       "      <td>290.576329</td>\n",
       "      <td>80.713551</td>\n",
       "      <td>109.910806</td>\n",
       "      <td>6.483190</td>\n",
       "      <td>3.141852</td>\n",
       "      <td>0.947064</td>\n",
       "      <td>1.286449</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>std</td>\n",
       "      <td>16.189833</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.209522</td>\n",
       "      <td>2.808498</td>\n",
       "      <td>242.937989</td>\n",
       "      <td>182.620356</td>\n",
       "      <td>288.918529</td>\n",
       "      <td>1.085451</td>\n",
       "      <td>0.795519</td>\n",
       "      <td>0.319592</td>\n",
       "      <td>0.452490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>min</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.400000</td>\n",
       "      <td>0.100000</td>\n",
       "      <td>63.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>2.700000</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.300000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>25%</td>\n",
       "      <td>33.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>175.500000</td>\n",
       "      <td>23.000000</td>\n",
       "      <td>25.000000</td>\n",
       "      <td>5.800000</td>\n",
       "      <td>2.600000</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>50%</td>\n",
       "      <td>45.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.300000</td>\n",
       "      <td>208.000000</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>42.000000</td>\n",
       "      <td>6.600000</td>\n",
       "      <td>3.100000</td>\n",
       "      <td>0.930000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>75%</td>\n",
       "      <td>58.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.600000</td>\n",
       "      <td>1.300000</td>\n",
       "      <td>298.000000</td>\n",
       "      <td>60.500000</td>\n",
       "      <td>87.000000</td>\n",
       "      <td>7.200000</td>\n",
       "      <td>3.800000</td>\n",
       "      <td>1.100000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>max</td>\n",
       "      <td>90.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>75.000000</td>\n",
       "      <td>19.700000</td>\n",
       "      <td>2110.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>4929.000000</td>\n",
       "      <td>9.600000</td>\n",
       "      <td>5.500000</td>\n",
       "      <td>2.800000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               Age Gender  Total_Bilirubin  Direct_Bilirubin  \\\n",
       "count   583.000000    583       583.000000        583.000000   \n",
       "unique         NaN      2              NaN               NaN   \n",
       "top            NaN   Male              NaN               NaN   \n",
       "freq           NaN    441              NaN               NaN   \n",
       "mean     44.746141    NaN         3.298799          1.486106   \n",
       "std      16.189833    NaN         6.209522          2.808498   \n",
       "min       4.000000    NaN         0.400000          0.100000   \n",
       "25%      33.000000    NaN         0.800000          0.200000   \n",
       "50%      45.000000    NaN         1.000000          0.300000   \n",
       "75%      58.000000    NaN         2.600000          1.300000   \n",
       "max      90.000000    NaN        75.000000         19.700000   \n",
       "\n",
       "        Alkaline_Phosphotase  Alamine_Aminotransferase  \\\n",
       "count             583.000000                583.000000   \n",
       "unique                   NaN                       NaN   \n",
       "top                      NaN                       NaN   \n",
       "freq                     NaN                       NaN   \n",
       "mean              290.576329                 80.713551   \n",
       "std               242.937989                182.620356   \n",
       "min                63.000000                 10.000000   \n",
       "25%               175.500000                 23.000000   \n",
       "50%               208.000000                 35.000000   \n",
       "75%               298.000000                 60.500000   \n",
       "max              2110.000000               2000.000000   \n",
       "\n",
       "        Aspartate_Aminotransferase  Total_Protiens     Albumin  \\\n",
       "count                   583.000000      583.000000  583.000000   \n",
       "unique                         NaN             NaN         NaN   \n",
       "top                            NaN             NaN         NaN   \n",
       "freq                           NaN             NaN         NaN   \n",
       "mean                    109.910806        6.483190    3.141852   \n",
       "std                     288.918529        1.085451    0.795519   \n",
       "min                      10.000000        2.700000    0.900000   \n",
       "25%                      25.000000        5.800000    2.600000   \n",
       "50%                      42.000000        6.600000    3.100000   \n",
       "75%                      87.000000        7.200000    3.800000   \n",
       "max                    4929.000000        9.600000    5.500000   \n",
       "\n",
       "        Albumin_and_Globulin_Ratio     Dataset  \n",
       "count                   579.000000  583.000000  \n",
       "unique                         NaN         NaN  \n",
       "top                            NaN         NaN  \n",
       "freq                           NaN         NaN  \n",
       "mean                      0.947064    1.286449  \n",
       "std                       0.319592    0.452490  \n",
       "min                       0.300000    1.000000  \n",
       "25%                       0.700000    1.000000  \n",
       "50%                       0.930000    1.000000  \n",
       "75%                       1.100000    2.000000  \n",
       "max                       2.800000    2.000000  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe(include='all')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Age', 'Gender', 'Total_Bilirubin', 'Direct_Bilirubin',\n",
       "       'Alkaline_Phosphotase', 'Alamine_Aminotransferase',\n",
       "       'Aspartate_Aminotransferase', 'Total_Protiens', 'Albumin',\n",
       "       'Albumin_and_Globulin_Ratio', 'Dataset'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## <데이터 전처리>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Age                           0\n",
       "Gender                        0\n",
       "Total_Bilirubin               0\n",
       "Direct_Bilirubin              0\n",
       "Alkaline_Phosphotase          0\n",
       "Alamine_Aminotransferase      0\n",
       "Aspartate_Aminotransferase    0\n",
       "Total_Protiens                0\n",
       "Albumin                       0\n",
       "Albumin_and_Globulin_Ratio    4\n",
       "Dataset                       0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum() #결측치확인"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "#비어있는 결측치는 중앙값으로 결측치 채움\n",
    "df.Albumin_and_Globulin_Ratio.fillna(df['Albumin_and_Globulin_Ratio'].median(), inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Age                           0\n",
       "Gender                        0\n",
       "Total_Bilirubin               0\n",
       "Direct_Bilirubin              0\n",
       "Alkaline_Phosphotase          0\n",
       "Alamine_Aminotransferase      0\n",
       "Aspartate_Aminotransferase    0\n",
       "Total_Protiens                0\n",
       "Albumin                       0\n",
       "Albumin_and_Globulin_Ratio    0\n",
       "Dataset                       0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum() #결측치 사라진 거 다시 확인! 이제 결측치 없음"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      Female\n",
       "1        Male\n",
       "2        Male\n",
       "3        Male\n",
       "4        Male\n",
       "        ...  \n",
       "578      Male\n",
       "579      Male\n",
       "580      Male\n",
       "581      Male\n",
       "582      Male\n",
       "Name: Gender, Length: 583, dtype: object"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['Gender']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "#유일하게 범주형 변수였던 gender변수는 labelencoder을 통해 인코딩을 진행하였습니다\n",
    "\n",
    "from sklearn import preprocessing\n",
    "\n",
    "le = preprocessing.LabelEncoder()\n",
    "le.fit(df.Gender.unique())\n",
    "df['Gender_Encoded'] = le.transform(df.Gender)\n",
    "df.drop(['Gender'], axis=1, inplace = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    441\n",
       "0    142\n",
       "Name: Gender_Encoded, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['Gender_Encoded'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## < EDA >"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "간질환으로 진단받은 환자 수 :  416\n",
      "간질환으로 진단 받지 않은 환자 수:  167\n"
     ]
    },
    {
     "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": [
    "#data 속  타겟변수인 환자 비율 확인\n",
    "\n",
    "sns.countplot(data=df, x = 'Dataset', label='Count')\n",
    "\n",
    "yes, no =df['Dataset'].value_counts()\n",
    "print('간질환으로 진단받은 환자 수 : ',yes)\n",
    "print('간질환으로 진단 받지 않은 환자 수: ',no)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "남자 환자 수 :  441\n",
      "여자 환자 수 :  142\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEHCAYAAABBW1qbAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAPrUlEQVR4nO3de7BdZXnH8e+PBMQboiYiTcBQTK04XtAIVHSKoe2grQQRHKyW1GaM41ArY22ldabaWmfqaEvVsbapqMFaFfECWqeUgnhp5ZLIRS46RCyQASWWi6IVCn36x37PyyY5STaQffY5nO9nZs9e61lrr/PsmTPnd9537b1WqgpJkgB2m3QDkqTZw1CQJHWGgiSpMxQkSZ2hIEnqFk66gYdi0aJFtWzZskm3IUlzysaNG39UVYun2zanQ2HZsmVs2LBh0m1I0pyS5PrtbXP6SJLUGQqSpM5QkCR1hoIkqTMUJEmdoSBJ6gwFSVJnKEiSOkNBktTN6W80Sw9nN/zFMyfdgmah/f/s22M9viMFSVJnKEiSOkNBktQZCpKkzlCQJHWGgiSpMxQkSZ2hIEnqDAVJUmcoSJI6Q0GS1BkKkqTOUJAkdYaCJKkzFCRJ3dhDIcmCJJcm+VJbPyDJRUmuTfLpJHu0+iPa+qa2fdm4e5Mk3d9MjBTeBFwztP5u4NSqWg7cBqxp9TXAbVX1VODUtp8kaQaNNRSSLAV+E/hwWw+wEjiz7bIeOKYtr2rrtO1Htv0lSTNk3COFvwX+GPi/tv5E4PaquqetbwaWtOUlwI0Abfsdbf/7SbI2yYYkG7Zs2TLO3iVp3hlbKCT5LeCWqto4XJ5m1xph232FqnVVtaKqVixevHgXdCpJmrJwjMc+HDg6yUuBPYG9GIwc9k6ysI0GlgI3tf03A/sBm5MsBB4H3DrG/iRJWxnbSKGq/qSqllbVMuAE4PyqejXwFeC4tttq4Ky2fHZbp20/v6q2GSlIksZnEt9TeCvw5iSbGJwzOK3VTwOe2OpvBk6ZQG+SNK+Nc/qoq6oLgAva8nXAIdPs83Pg+JnoR5I0Pb/RLEnqDAVJUmcoSJI6Q0GS1BkKkqTOUJAkdYaCJKkzFCRJnaEgSeoMBUlSZyhIkjpDQZLUGQqSpM5QkCR1hoIkqTMUJEmdoSBJ6gwFSVJnKEiSOkNBktQZCpKkzlCQJHWGgiSpMxQkSZ2hIEnqDAVJUmcoSJI6Q0GS1BkKkqTOUJAkdYaCJKkzFCRJnaEgSeoMBUlSZyhIkjpDQZLUGQqSpM5QkCR1YwuFJHsmuTjJ5UmuSvLnrX5AkouSXJvk00n2aPVHtPVNbfuycfUmSZreOEcKdwErq+rZwHOAo5IcBrwbOLWqlgO3AWva/muA26rqqcCpbT9J0gwaWyjUwJ1tdff2KGAlcGarrweOacur2jpt+5FJMq7+JEnbGus5hSQLklwG3AKcC3wPuL2q7mm7bAaWtOUlwI0AbfsdwBOnOebaJBuSbNiyZcs425ekeWesoVBV91bVc4ClwCHA06fbrT1PNyqobQpV66pqRVWtWLx48a5rVpI0M58+qqrbgQuAw4C9kyxsm5YCN7XlzcB+AG3744BbZ6I/SdLAOD99tDjJ3m35kcCvAdcAXwGOa7utBs5qy2e3ddr286tqm5GCJGl8Fu58lwdtX2B9kgUMwueMqvpSkquBTyX5S+BS4LS2/2nAx5NsYjBCOGGMvUmSpjG2UKiqK4CDp6lfx+D8wtb1nwPHj6sfSdLO+Y1mSVJnKEiSOkNBktQZCpKkzlCQJHWGgiSpMxQkSZ2hIEnqDAVJUmcoSJI6Q0GS1BkKkqTOUJAkdYaCJKkzFCRJ3UihkOS8UWqSpLlthzfZSbIn8ChgUZLHA2mb9gJ+Ycy9SZJm2M7uvPZ64GQGAbCR+0Lhx8AHx9iXJGkCdhgKVfU+4H1J3lhVH5ihniRJEzLSPZqr6gNJXgAsG35NVZ0+pr4kSRMwUigk+ThwIHAZcG8rF2AoSNLDyEihAKwADqqqGmczkqTJGvV7ClcCTx5nI5KkyRt1pLAIuDrJxcBdU8WqOnosXUmSJmLUUHjHOJuQJM0Oo3766KvjbkSSNHmjfvroJww+bQSwB7A78NOq2mtcjUmSZt6oI4XHDq8nOQY4ZCwdSZIm5kFdJbWqvgCs3MW9SJImbNTpo2OHVndj8L0Fv7MgSQ8zo3766GVDy/cA/wWs2uXdSJImatRzCq8ddyOSpMkb9SY7S5N8PsktSX6Y5LNJlo67OUnSzBr1RPNHgbMZ3FdhCfDFVpMkPYyMGgqLq+qjVXVPe3wMWDzGviRJEzBqKPwoyWuSLGiP1wD/Pc7GJEkzb9RQ+D3glcAPgJuB4wBPPkvSw8yoH0l9J7C6qm4DSPIE4L0MwkKS9DAx6kjhWVOBAFBVtwIH7+gFSfZL8pUk1yS5KsmbWv0JSc5Ncm17fnyrJ8n7k2xKckWS5z7YNyVJenBGDYXdpv54Qx8p7GyUcQ/wh1X1dOAw4KQkBwGnAOdV1XLgvLYO8BJgeXusBT408ruQJO0So04f/TXwn0nOZHB5i1cC79rRC6rqZgbnH6iqnyS5hsHHWVcBR7Td1gMXAG9t9dPbLT8vTLJ3kn3bcSRJM2DUbzSfnmQDg4vgBTi2qq4e9YckWcZguukiYJ+pP/RVdXOSJ7XdlgA3Dr1sc6vdLxSSrGUwkmD//fcftQVJ0ghGHSnQQmDkIJiS5DHAZ4GTq+rHSba763Q/dpo+1gHrAFasWOFF+SRpF3pQl84eVZLdGQTCJ6rqc638wyT7tu37Are0+mZgv6GXLwVuGmd/kqT7G1soZDAkOA24pqr+ZmjT2cDqtrwaOGuofmL7FNJhwB2eT5CkmTXy9NGDcDjwO8C3k1zWan8K/BVwRpI1wA3A8W3bl4GXApuAn+GX4yRpxo0tFKrqG0x/ngDgyGn2L+CkcfUjSdq5sZ5TkCTNLYaCJKkzFCRJnaEgSeoMBUlSZyhIkjpDQZLUGQqSpM5QkCR1hoIkqTMUJEmdoSBJ6gwFSVJnKEiSOkNBktQZCpKkzlCQJHWGgiSpMxQkSZ2hIEnqDAVJUrdw0g1M2vP+6PRJt6BZaON7Tpx0C9JEOFKQJHWGgiSpMxQkSZ2hIEnqDAVJUmcoSJI6Q0GS1BkKkqTOUJAkdYaCJKkzFCRJnaEgSeoMBUlSZyhIkjpDQZLUGQqSpG5soZDkI0luSXLlUO0JSc5Ncm17fnyrJ8n7k2xKckWS546rL0nS9o1zpPAx4KitaqcA51XVcuC8tg7wEmB5e6wFPjTGviRJ2zG2UKiqrwG3blVeBaxvy+uBY4bqp9fAhcDeSfYdV2+SpOnN9DmFfarqZoD2/KRWXwLcOLTf5lbbRpK1STYk2bBly5axNitJ881sOdGcaWo13Y5Vta6qVlTVisWLF4+5LUmaX2Y6FH44NS3Unm9p9c3AfkP7LQVumuHeJGnem+lQOBtY3ZZXA2cN1U9sn0I6DLhjappJkjRzFo7rwEk+CRwBLEqyGXg78FfAGUnWADcAx7fdvwy8FNgE/Ax47bj6kiRt39hCoapetZ1NR06zbwEnjasXSdJoZsuJZknSLGAoSJI6Q0GS1BkKkqTOUJAkdYaCJKkzFCRJnaEgSeoMBUlSZyhIkjpDQZLUGQqSpM5QkCR1hoIkqTMUJEmdoSBJ6gwFSVJnKEiSOkNBktQZCpKkzlCQJHWGgiSpMxQkSZ2hIEnqDAVJUmcoSJI6Q0GS1BkKkqTOUJAkdYaCJKkzFCRJnaEgSeoMBUlSZyhIkjpDQZLUGQqSpM5QkCR1hoIkqZtVoZDkqCTfTbIpySmT7keS5ptZEwpJFgAfBF4CHAS8KslBk+1KkuaXWRMKwCHApqq6rqruBj4FrJpwT5I0ryycdANDlgA3Dq1vBg7deqcka4G1bfXOJN+dgd7mi0XAjybdxGyQ966edAu6P383p7w9u+IoT9nehtkUCtO909qmULUOWDf+duafJBuqasWk+5C25u/mzJlN00ebgf2G1pcCN02oF0mal2ZTKFwCLE9yQJI9gBOAsyfckyTNK7Nm+qiq7kny+8A5wALgI1V11YTbmm+cltNs5e/mDEnVNtP2kqR5ajZNH0mSJsxQkCR1hoK8vIhmrSQfSXJLkisn3ct8YSjMc15eRLPcx4CjJt3EfGIoyMuLaNaqqq8Bt066j/nEUNB0lxdZMqFeJE2YoaCRLi8iaX4wFOTlRSR1hoK8vIikzlCY56rqHmDq8iLXAGd4eRHNFkk+CXwTeFqSzUnWTLqnhzsvcyFJ6hwpSJI6Q0GS1BkKkqTOUJAkdYaCJKkzFCRJnaGgOSnJPkn+Ocl1STYm+WaSl++C4x6R5Eu74DjLkvxPksuGHic+1OPu4Oe9I8lbHuBr7hxXP5q7Zs09mqVRJQnwBWB9Vf12qz0FOHoCvSxsXwCczveq6jkz2pD0EDlS0Fy0Eri7qv5+qlBV11fVB5IsSPKeJJckuSLJ66GPAC5IcmaS7yT5RAuXqZsMfSfJN4Bjp46Z5NHtJi+XJLk0yapW/90kn0nyReDfHmjzSe5M8q4klye5MMk+rb5Pks+3+uVJXtDqb05yZXucPHSct7WbI/078LSh+oFJ/rWNoL6e5Jdb/YA2orokyTsfaN+aHwwFzUXPAL61nW1rgDuq6vnA84HXJTmgbTsYOJnBzYR+ETg8yZ7APwIvA14EPHnoWG8Dzm/HejHwniSPbtt+BVhdVSt30OeBW00fvajVHw1cWFXPBr4GvK7V3w98tdWfC1yV5HnAa4FDgcPa+zm41U9o7+nY9l6nrAPeWFXPA94C/F2rvw/4UHs/P9hB35rHnD7SnJfkg8ALgbuB64FnJTmubX4csLxtu7iqNrfXXAYsA+4Evl9V17b6PwFr22t/Azh6aK5+T2D/tnxuVe3s5i/bmz66G5g6b7ER+PW2vBI4EaCq7gXuSPJC4PNV9dPW3+cYhNdurf6zVj+7PT8GeAHwmTYQAnhEez4ceEVb/jjw7p30r3nIUNBcdBX3/XGjqk5KsgjYANzA4L/kc4ZfkOQI4K6h0r3c9/u/vQuABXhFVX13q2MdCvz0IfT/v3XfRceG+9heD9szXd+7Abfv4FyGFzvTDjl9pLnofGDPJG8Yqj2qPZ8DvCHJ7gBJfmloymc63wEOSHJgW3/V0LZzgDcOnXs4eJd0v33nAW9oP2tBkr0YTC8dk+RR7X28HPh6q788ySOTPJbB9BdV9WPg+0mOb8dJkme34/8HgykngFeP+b1ojjIUNOe0/7KPAX41yfeTXAysB94KfBi4GvhWkiuBf2AH/4lX1c8ZTBf9SzvRfP3Q5ncCuwNXtGM90JOzW59T+IOd7P8m4MVJvs1gWukZVfUtBjevvxi4CPhwVV3a6p8GLgM+yyAoprwaWJPkcgajqlVDxz8pySUMptWkbXjpbElS50hBktR5oll6CJI8k8EneYbdVVWHTqIf6aFy+kiS1Dl9JEnqDAVJUmcoSJI6Q0GS1P0/YJqgErMb1CkAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#남녀 비율 확인 \n",
    "#남자 환자가 훨씬 많음을 확인 할 수 있다\n",
    "sns.countplot(data=df, x = 'Gender_Encoded', label='Count')\n",
    "\n",
    "male, female = df['Gender_Encoded'].value_counts()\n",
    "print('남자 환자 수 : ',male)\n",
    "print('여자 환자 수 : ',female)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1ef47e67348>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "sns.countplot(data=df, x = 'Dataset', hue = 'Gender_Encoded')\n",
    "#성별과 타겟변수관계를 확인\n",
    "#데이터자체에 남자가 많은 경향이고, 역시나 타겟변수인 간질환을 진단받은 환자의 성별 역시\n",
    "#남자가 훨씬 많습니다"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#간질환 진단 받은 환자와 아닌 환자의 나이별 분포 확인\n",
    "g = sns.FacetGrid(df, col=\"Dataset\", row=\"Gender_Encoded\", margin_titles=True)\n",
    "g.map(plt.hist, \"Age\")\n",
    "plt.subplots_adjust(top=0.9)\n",
    "g.fig.suptitle('Disease by Gender and Age');\n",
    "#간질환으로 진단받은 환자(dataset=1) 환자의 분포를 나이 분포를 보면 40-60사이에 많이 분포되어있음으로 확인할 수 있음"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#변수별 상관관계 확인\n",
    "\n",
    "correlations = df.corr()\n",
    "#상관계수 구하고 heatmap으로 시각화\n",
    "\n",
    "plt.figure(figsize=(10, 10))\n",
    "g = sns.heatmap(correlations, cbar = True, square = True, annot=True, fmt= '.2f', annot_kws={'size': 10},cmap= 'coolwarm')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#앞서 heatmap에서 상관관계가 높다고 확인된 변수들간의 선형관계를 확인"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Total_Bilirubin, Direct_bilirubin\n",
    "g = sns.FacetGrid(df, col=\"Gender_Encoded\", row=\"Dataset\", margin_titles=True)\n",
    "g.map(plt.scatter,\"Direct_Bilirubin\", \"Total_Bilirubin\", edgecolor=\"w\")\n",
    "plt.subplots_adjust(top=0.9)\n",
    "#이 두 feature은 전체적으로 모두 선형적인 관계를 보임\n",
    "#두 feature 중 하나는 없애도 될 것으로 판단!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.JointGrid at 0x1ef47fe2248>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.jointplot(\"Total_Bilirubin\", \"Direct_Bilirubin\", data=df, kind=\"reg\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Aspartate_Aminotransferase & Alamine_Aminotransferase \n",
    "g = sns.FacetGrid(df, col=\"Gender_Encoded\", row=\"Dataset\", margin_titles=True)\n",
    "g.map(plt.scatter,\"Aspartate_Aminotransferase\", \"Alamine_Aminotransferase\",  edgecolor=\"w\")\n",
    "plt.subplots_adjust(top=0.9)\n",
    "#이 두 feature도 선형적인 관계로 보임\n",
    "#두 feature 중 하나는 제거해도 괜찮을 것 같다고 판단"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.JointGrid at 0x1ef488ada88>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.jointplot(\"Aspartate_Aminotransferase\", \"Alamine_Aminotransferase\", data=df, kind=\"reg\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Total_Protiens & Albumin\n",
    "g = sns.FacetGrid(df, col=\"Gender_Encoded\", row=\"Dataset\", margin_titles=True)\n",
    "g.map(plt.scatter,\"Total_Protiens\", \"Albumin\",  edgecolor=\"w\")\n",
    "plt.subplots_adjust(top=0.9)\n",
    "#이 두 feature 역시나 선형적 관계 보임"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.JointGrid at 0x1ef47734548>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.jointplot(\"Total_Protiens\", \"Albumin\", data=df, kind=\"reg\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Albumin, Albumin_and_Globulin_Ratio\n",
    "g = sns.FacetGrid(df, col=\"Gender_Encoded\", row=\"Dataset\", margin_titles=True)\n",
    "g.map(plt.scatter,\"Albumin\", \"Albumin_and_Globulin_Ratio\",  edgecolor=\"w\")\n",
    "plt.subplots_adjust(top=0.9)\n",
    "#이 두 feature도 선형적 관게 보임"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.JointGrid at 0x1ef48202bc8>"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.jointplot(\"Albumin_and_Globulin_Ratio\", \"Albumin\", data=df, kind=\"reg\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "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>Age</th>\n",
       "      <th>Total_Bilirubin</th>\n",
       "      <th>Direct_Bilirubin</th>\n",
       "      <th>Alkaline_Phosphotase</th>\n",
       "      <th>Alamine_Aminotransferase</th>\n",
       "      <th>Aspartate_Aminotransferase</th>\n",
       "      <th>Total_Protiens</th>\n",
       "      <th>Albumin</th>\n",
       "      <th>Albumin_and_Globulin_Ratio</th>\n",
       "      <th>Dataset</th>\n",
       "      <th>Gender_Encoded</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>65</td>\n",
       "      <td>0.7</td>\n",
       "      <td>0.1</td>\n",
       "      <td>187</td>\n",
       "      <td>16</td>\n",
       "      <td>18</td>\n",
       "      <td>6.8</td>\n",
       "      <td>3.3</td>\n",
       "      <td>0.90</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>62</td>\n",
       "      <td>10.9</td>\n",
       "      <td>5.5</td>\n",
       "      <td>699</td>\n",
       "      <td>64</td>\n",
       "      <td>100</td>\n",
       "      <td>7.5</td>\n",
       "      <td>3.2</td>\n",
       "      <td>0.74</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>62</td>\n",
       "      <td>7.3</td>\n",
       "      <td>4.1</td>\n",
       "      <td>490</td>\n",
       "      <td>60</td>\n",
       "      <td>68</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.3</td>\n",
       "      <td>0.89</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>58</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.4</td>\n",
       "      <td>182</td>\n",
       "      <td>14</td>\n",
       "      <td>20</td>\n",
       "      <td>6.8</td>\n",
       "      <td>3.4</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>72</td>\n",
       "      <td>3.9</td>\n",
       "      <td>2.0</td>\n",
       "      <td>195</td>\n",
       "      <td>27</td>\n",
       "      <td>59</td>\n",
       "      <td>7.3</td>\n",
       "      <td>2.4</td>\n",
       "      <td>0.40</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\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",
       "      <td>578</td>\n",
       "      <td>60</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.1</td>\n",
       "      <td>500</td>\n",
       "      <td>20</td>\n",
       "      <td>34</td>\n",
       "      <td>5.9</td>\n",
       "      <td>1.6</td>\n",
       "      <td>0.37</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>579</td>\n",
       "      <td>40</td>\n",
       "      <td>0.6</td>\n",
       "      <td>0.1</td>\n",
       "      <td>98</td>\n",
       "      <td>35</td>\n",
       "      <td>31</td>\n",
       "      <td>6.0</td>\n",
       "      <td>3.2</td>\n",
       "      <td>1.10</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>580</td>\n",
       "      <td>52</td>\n",
       "      <td>0.8</td>\n",
       "      <td>0.2</td>\n",
       "      <td>245</td>\n",
       "      <td>48</td>\n",
       "      <td>49</td>\n",
       "      <td>6.4</td>\n",
       "      <td>3.2</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>581</td>\n",
       "      <td>31</td>\n",
       "      <td>1.3</td>\n",
       "      <td>0.5</td>\n",
       "      <td>184</td>\n",
       "      <td>29</td>\n",
       "      <td>32</td>\n",
       "      <td>6.8</td>\n",
       "      <td>3.4</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>582</td>\n",
       "      <td>38</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.3</td>\n",
       "      <td>216</td>\n",
       "      <td>21</td>\n",
       "      <td>24</td>\n",
       "      <td>7.3</td>\n",
       "      <td>4.4</td>\n",
       "      <td>1.50</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>583 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Age  Total_Bilirubin  Direct_Bilirubin  Alkaline_Phosphotase  \\\n",
       "0     65              0.7               0.1                   187   \n",
       "1     62             10.9               5.5                   699   \n",
       "2     62              7.3               4.1                   490   \n",
       "3     58              1.0               0.4                   182   \n",
       "4     72              3.9               2.0                   195   \n",
       "..   ...              ...               ...                   ...   \n",
       "578   60              0.5               0.1                   500   \n",
       "579   40              0.6               0.1                    98   \n",
       "580   52              0.8               0.2                   245   \n",
       "581   31              1.3               0.5                   184   \n",
       "582   38              1.0               0.3                   216   \n",
       "\n",
       "     Alamine_Aminotransferase  Aspartate_Aminotransferase  Total_Protiens  \\\n",
       "0                          16                          18             6.8   \n",
       "1                          64                         100             7.5   \n",
       "2                          60                          68             7.0   \n",
       "3                          14                          20             6.8   \n",
       "4                          27                          59             7.3   \n",
       "..                        ...                         ...             ...   \n",
       "578                        20                          34             5.9   \n",
       "579                        35                          31             6.0   \n",
       "580                        48                          49             6.4   \n",
       "581                        29                          32             6.8   \n",
       "582                        21                          24             7.3   \n",
       "\n",
       "     Albumin  Albumin_and_Globulin_Ratio  Dataset  Gender_Encoded  \n",
       "0        3.3                        0.90        1               0  \n",
       "1        3.2                        0.74        1               1  \n",
       "2        3.3                        0.89        1               1  \n",
       "3        3.4                        1.00        1               1  \n",
       "4        2.4                        0.40        1               1  \n",
       "..       ...                         ...      ...             ...  \n",
       "578      1.6                        0.37        2               1  \n",
       "579      3.2                        1.10        1               1  \n",
       "580      3.2                        1.00        1               1  \n",
       "581      3.4                        1.00        1               1  \n",
       "582      4.4                        1.50        2               1  \n",
       "\n",
       "[583 rows x 11 columns]"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "#앞서 얻은 insight를 기반으로 \n",
    "# Total_Bilirubin, Direct_Bilirubin 중 하나만 사용하기로 결정하였고, 타겟변수인 dataset과 더 상관계수가 작게 나온 direct_Bilirubin을 제외시키는 것으로 판단하였습니다\n",
    "#또한 Aspartate_Aminotransferase,Alamine_Aminotransferase 중 하나만 사용하기로 결정하였고, 역시나 상관계수가 더 작게 나온 Alamine_Aminotransferase을 제외시켰습니다\n",
    "#또 다른 feautre를 제거할까 고민했지만, 다른 feature에 비해 타겟변수간와의 상관계수가 높다는 점을 감안하여 최종 feature로 사용하기로 결정하였습니다\n",
    "\n",
    "X=df[['Age', 'Gender_Encoded',\n",
    "      'Total_Bilirubin','Alkaline_Phosphotase',\n",
    "       'Aspartate_Aminotransferase', 'Total_Protiens', 'Albumin',\n",
    "       'Albumin_and_Globulin_Ratio']]\n",
    "y=df['Dataset']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## <모델 적용>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train,X_test,y_train,y_test=train_test_split(X,y,test_size=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(466, 8) (466,) (117, 8)\n"
     ]
    }
   ],
   "source": [
    "print(X_train.shape, y_train.shape, X_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "#모델은 logistic regression, RandomForest, SVM 세가지를 사용해보았습니다!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 필요한 모듈 import\n",
    "from sklearn import preprocessing\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.feature_selection import SelectKBest, f_classif\n",
    "from sklearn.pipeline import Pipeline\n",
    "\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    " "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Pipeline으로 구성하였습니다!\n",
    "def pipeline_LR(X_train, y_train):\n",
    "     \n",
    "    #scaling\n",
    "    scale = preprocessing.StandardScaler()\n",
    "     \n",
    "    # 높은 설명력 기준으로 변수의 수 설정!(앞서 변수를 미리 몇개 제거하기는 했지만 더 적은 변수가 더 높은 accuracy를 보일수도 있다고 판단해서 feature selection을 위한 부분을 넣었습니다)\n",
    "    selectK = SelectKBest(score_func=f_classif)\n",
    "     \n",
    "    #모델 fit\n",
    "    logistic_model = LogisticRegression()\n",
    "     \n",
    "    pipeline_object = Pipeline([('scaler', scale),  #파이프라인을 통해 스케일링\n",
    "                                ('feature_selection', selectK), # feature 수 선택\n",
    "                                ('model', logistic_model)]) #모델 선택을 한번에 진행하도록 함!\n",
    "     \n",
    "    # Parmeters 튜닝\n",
    "    params = [{'feature_selection__k': [5,6,7,8],\n",
    "               'model__C': [0.01, 0.1, 1, 10],\n",
    "               'model__penalty': ['l1', 'l2']}]\n",
    "     \n",
    "\n",
    "    grid_search = GridSearchCV(pipeline_object, param_grid = params, scoring = 'roc_auc', cv=10) #그를 기반으로 그리드서치 한 값을 리턴하도록 함\n",
    "     \n",
    "    # Model fit\n",
    "    grid_search.fit(X_train, y_train)\n",
    "         \n",
    "    print(grid_search.best_params_)\n",
    "    print(grid_search.best_score_)\n",
    "    return grid_search"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n",
      "C:\\Users\\user\\Anaconda3\\lib\\site-packages\\sklearn\\model_selection\\_validation.py:536: FitFailedWarning: Estimator fit failed. The score on this train-test partition for these parameters will be set to nan. Details: \n",
      "ValueError: Solver lbfgs supports only 'l2' or 'none' penalties, got l1 penalty.\n",
      "\n",
      "  FitFailedWarning)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'feature_selection__k': 8, 'model__C': 1, 'model__penalty': 'l2'}\n",
      "0.7333137450784509\n"
     ]
    }
   ],
   "source": [
    "model_LR = pipeline_LR(X_train, y_train)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "         steps=[('scale',\n",
       "                 StandardScaler(copy=True, with_mean=True, with_std=True)),\n",
       "                ('feature_select',\n",
       "                 SelectKBest(k=8,\n",
       "                             score_func=<function f_classif at 0x0000026A17BFBC18>)),\n",
       "                ('LR',\n",
       "                 LogisticRegression(C=1, class_weight=None, dual=False,\n",
       "                                    fit_intercept=True, intercept_scaling=1,\n",
       "                                    l1_ratio=None, max_iter=100,\n",
       "                                    multi_class='auto', n_jobs=None,\n",
       "                                    penalty='l2', random_state=None,\n",
       "                                    solver='lbfgs', tol=0.0001, verbose=0,\n",
       "                                    warm_start=False))],\n",
       "         verbose=False)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scale = preprocessing.StandardScaler()\n",
    "select_feature = SelectKBest(f_classif, k=8)\n",
    "model = LogisticRegression()\n",
    "\n",
    "pp_LR = Pipeline([('scale', scale),\n",
    "                  ('feature_select', select_feature), \n",
    "                  ('LR', model)])\n",
    "\n",
    "pp_LR.set_params(LR__C = 1).fit(X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           1       0.76      0.94      0.84        85\n",
      "           2       0.58      0.22      0.32        32\n",
      "\n",
      "    accuracy                           0.74       117\n",
      "   macro avg       0.67      0.58      0.58       117\n",
      "weighted avg       0.71      0.74      0.70       117\n",
      "\n",
      "Accuracy: 0.7435897435897436\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x26a183ca788>"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    " # 모델을 통해 Testset 예측\n",
    "preds = pp_LR.predict(X_test)\n",
    "\n",
    "print(metrics.classification_report(y_test, preds))\n",
    "print(\"Accuracy:\", metrics.accuracy_score(y_true = y_test, y_pred=preds))\n",
    "\n",
    "sns.heatmap(confusion_matrix(y_test,preds),annot=True)\n",
    "\n",
    "#accuracy: 0.74 정도 나오네요!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## RandomForest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Pipeline 구성 및 모델링\n",
    "\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "\n",
    "def pipeline_RFC(X_train, y_train):\n",
    "     \n",
    "    #scale\n",
    "    scale = preprocessing.StandardScaler()\n",
    "     \n",
    "    # FIXME: 높은 설명력 기준으로 변수의 수 설정\n",
    "    selectK = SelectKBest(score_func=f_classif)\n",
    "     \n",
    "    #모델 fit\n",
    "    rfc = RandomForestClassifier(random_state=42)\n",
    "     \n",
    "    pipeline_object = Pipeline([('scaler', scale), \n",
    "                                ('feature_selection', selectK),\n",
    "                                ('model', rfc)])\n",
    "     \n",
    "    # Parmeters 튜닝\n",
    "    params = [{'feature_selection__k': [5,6,7,8],\n",
    "               'model__n_estimators': [200, 500],\n",
    "                'model__max_features': ['auto', 'sqrt', 'log2'],\n",
    "                'model__max_depth' : [4,5,6,7,8],\n",
    "                'model__criterion' :['gini', 'entropy'] }]\n",
    "     \n",
    "\n",
    "    grid_search = GridSearchCV(pipeline_object, param_grid = params, cv=10)\n",
    "     \n",
    "    # Model fit\n",
    "    grid_search.fit(X_train, y_train)\n",
    "         \n",
    "    print(grid_search.best_params_)\n",
    "    print(grid_search.best_score_)\n",
    "    return grid_search"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'feature_selection__k': 7, 'model__criterion': 'entropy', 'model__max_depth': 4, 'model__max_features': 'auto', 'model__n_estimators': 500}\n",
      "0.714708603145236\n"
     ]
    }
   ],
   "source": [
    "model_RFC = pipeline_RFC(X_train, y_train)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "         steps=[('scale',\n",
       "                 StandardScaler(copy=True, with_mean=True, with_std=True)),\n",
       "                ('feature_select',\n",
       "                 SelectKBest(k=7,\n",
       "                             score_func=<function f_classif at 0x0000026A17BFBC18>)),\n",
       "                ('rfc',\n",
       "                 RandomForestClassifier(bootstrap=True, ccp_alpha=0.0,\n",
       "                                        class_weight=None, criterion='entropy',\n",
       "                                        max_depth=4, max_features='auto',\n",
       "                                        max_leaf_nodes=None, max_samples=None,\n",
       "                                        min_impurity_decrease=0.0,\n",
       "                                        min_impurity_split=None,\n",
       "                                        min_samples_leaf=1, min_samples_split=2,\n",
       "                                        min_weight_fraction_leaf=0.0,\n",
       "                                        n_estimators=500, n_jobs=None,\n",
       "                                        oob_score=False, random_state=42,\n",
       "                                        verbose=0, warm_start=False))],\n",
       "         verbose=False)"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scale = preprocessing.StandardScaler()\n",
    "select_feature = SelectKBest(f_classif, k=7)\n",
    "model = RandomForestClassifier(random_state=42)\n",
    "\n",
    "pp_rfc = Pipeline([('scale', scale),\n",
    "                  ('feature_select', select_feature), \n",
    "                  ('rfc', model)])\n",
    "\n",
    "pp_rfc.set_params(rfc__n_estimators = 500, rfc__max_features = 'auto', rfc__max_depth = 4, rfc__criterion = 'entropy').fit(X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           1       0.76      0.96      0.85        85\n",
      "           2       0.67      0.19      0.29        32\n",
      "\n",
      "    accuracy                           0.75       117\n",
      "   macro avg       0.71      0.58      0.57       117\n",
      "weighted avg       0.73      0.75      0.70       117\n",
      "\n",
      "Accuracy: 0.7521367521367521\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x26a185fea88>"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 모델을 통해 Testset 예측\n",
    "preds = pp_rfc.predict(X_test)\n",
    "\n",
    "print(metrics.classification_report(y_test, preds))\n",
    "print(\"Accuracy:\", metrics.accuracy_score(y_true = y_test, y_pred=preds))\n",
    "sns.heatmap(confusion_matrix(y_test,preds),annot=True)\n",
    "\n",
    "#accuracy: 0.75 나오네요!!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "#SVM\n",
    "\n",
    "# Pipeline 구성 및 모델링\n",
    "\n",
    "from sklearn.svm import SVC, LinearSVC\n",
    "\n",
    "def pipeline_SVC(X_train, y_train):\n",
    "     \n",
    "    #scaling\n",
    "    scale = preprocessing.StandardScaler()\n",
    "     \n",
    "    #높은 설명력 기준으로 변수의 수 설정\n",
    "    selectK = SelectKBest(score_func=f_classif)\n",
    "     \n",
    "    #모델 fit\n",
    "    svc = SVC(class_weight = 'balanced')\n",
    "     \n",
    "    pipeline_object = Pipeline([('scaler', scale), \n",
    "                                ('feature_selection', selectK),\n",
    "                                ('model', svc)])\n",
    "     \n",
    "    # Parmeters 튜닝\n",
    "    params = [{'feature_selection__k': [5,6,7,8],\n",
    "               'model__C'     : [0.0001, 0.001, 0.01, 0.1, 1, 10, 100, 1000],\n",
    "            'model__gamma' : ['auto']}]\n",
    "     \n",
    "\n",
    "    grid_search = GridSearchCV(pipeline_object, param_grid = params, verbose = 3, cv=10)\n",
    "     \n",
    "    # Model fit\n",
    "    grid_search.fit(X_train, y_train)\n",
    "         \n",
    "    print(grid_search.best_params_)\n",
    "    print(grid_search.best_score_)\n",
    "    return grid_search"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 10 folds for each of 32 candidates, totalling 320 fits\n",
      "[CV] feature_selection__k=5, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=5, model__C=0.0001, model__gamma=auto, score=0.277, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=5, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=5, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=5, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=5, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=5, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n",
      "[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:    0.0s remaining:    0.0s\n",
      "[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:    0.0s remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "[CV] feature_selection__k=5, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=5, model__C=0.0001, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=5, model__C=0.0001, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=5, model__C=0.0001, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=5, model__C=0.0001, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=5, model__C=0.001, model__gamma=auto, score=0.277, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=5, model__C=0.001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=5, model__C=0.001, model__gamma=auto, score=0.298, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=5, model__C=0.001, model__gamma=auto, score=0.298, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=5, model__C=0.001, model__gamma=auto, score=0.298, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=5, model__C=0.001, model__gamma=auto, score=0.298, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=5, model__C=0.001, model__gamma=auto, score=0.717, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=5, model__C=0.001, model__gamma=auto, score=0.717, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=5, model__C=0.001, model__gamma=auto, score=0.717, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=5, model__C=0.001, model__gamma=auto, score=0.717, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=0.01, model__gamma=auto, score=0.277, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=0.01, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=0.01, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=0.01, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=0.01, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=0.1, model__gamma=auto, score=0.596, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=0.1, model__gamma=auto, score=0.468, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=0.1, model__gamma=auto, score=0.617, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=0.1, model__gamma=auto, score=0.447, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=0.1, model__gamma=auto, score=0.681, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=0.1, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=0.1, model__gamma=auto, score=0.543, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=0.1, model__gamma=auto, score=0.413, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=0.1, model__gamma=auto, score=0.565, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=0.1, model__gamma=auto, score=0.630, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=5, model__C=1, model__gamma=auto, score=0.574, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=5, model__C=1, model__gamma=auto, score=0.511, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=5, model__C=1, model__gamma=auto, score=0.574, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=5, model__C=1, model__gamma=auto, score=0.489, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=5, model__C=1, model__gamma=auto, score=0.660, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=5, model__C=1, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=5, model__C=1, model__gamma=auto, score=0.543, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=5, model__C=1, model__gamma=auto, score=0.457, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=5, model__C=1, model__gamma=auto, score=0.565, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=5, model__C=1, model__gamma=auto, score=0.674, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=5, model__C=10, model__gamma=auto, score=0.596, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=5, model__C=10, model__gamma=auto, score=0.553, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=5, model__C=10, model__gamma=auto, score=0.532, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=5, model__C=10, model__gamma=auto, score=0.489, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=5, model__C=10, model__gamma=auto, score=0.681, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=5, model__C=10, model__gamma=auto, score=0.660, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=5, model__C=10, model__gamma=auto, score=0.543, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=5, model__C=10, model__gamma=auto, score=0.457, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=5, model__C=10, model__gamma=auto, score=0.543, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=5, model__C=10, model__gamma=auto, score=0.674, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=100, model__gamma=auto .........\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  feature_selection__k=5, model__C=100, model__gamma=auto, score=0.574, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=100, model__gamma=auto, score=0.511, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=100, model__gamma=auto, score=0.574, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=100, model__gamma=auto, score=0.511, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=100, model__gamma=auto, score=0.660, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=100, model__gamma=auto, score=0.681, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=100, model__gamma=auto, score=0.565, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=100, model__gamma=auto, score=0.457, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=100, model__gamma=auto, score=0.543, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=5, model__C=100, model__gamma=auto, score=0.674, total=   0.0s\n",
      "[CV] feature_selection__k=5, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=1000, model__gamma=auto, score=0.617, total=   0.2s\n",
      "[CV] feature_selection__k=5, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=1000, model__gamma=auto, score=0.553, total=   0.3s\n",
      "[CV] feature_selection__k=5, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=1000, model__gamma=auto, score=0.617, total=   0.3s\n",
      "[CV] feature_selection__k=5, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=1000, model__gamma=auto, score=0.511, total=   0.2s\n",
      "[CV] feature_selection__k=5, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=1000, model__gamma=auto, score=0.617, total=   0.2s\n",
      "[CV] feature_selection__k=5, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=1000, model__gamma=auto, score=0.638, total=   0.2s\n",
      "[CV] feature_selection__k=5, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=1000, model__gamma=auto, score=0.587, total=   0.3s\n",
      "[CV] feature_selection__k=5, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=1000, model__gamma=auto, score=0.565, total=   0.2s\n",
      "[CV] feature_selection__k=5, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=1000, model__gamma=auto, score=0.609, total=   0.3s\n",
      "[CV] feature_selection__k=5, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=5, model__C=1000, model__gamma=auto, score=0.674, total=   0.2s\n",
      "[CV] feature_selection__k=6, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=6, model__C=0.0001, model__gamma=auto, score=0.277, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=6, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=6, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=6, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=6, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=6, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=6, model__C=0.0001, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=6, model__C=0.0001, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=6, model__C=0.0001, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=6, model__C=0.0001, model__gamma=auto, score=0.283, total=   0.1s\n",
      "[CV] feature_selection__k=6, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=6, model__C=0.001, model__gamma=auto, score=0.277, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=6, model__C=0.001, model__gamma=auto, score=0.298, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=6, model__C=0.001, model__gamma=auto, score=0.298, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=6, model__C=0.001, model__gamma=auto, score=0.298, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=6, model__C=0.001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=6, model__C=0.001, model__gamma=auto, score=0.298, total=   0.1s\n",
      "[CV] feature_selection__k=6, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=6, model__C=0.001, model__gamma=auto, score=0.717, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=6, model__C=0.001, model__gamma=auto, score=0.717, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=6, model__C=0.001, model__gamma=auto, score=0.717, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=6, model__C=0.001, model__gamma=auto, score=0.717, total=   0.1s\n",
      "[CV] feature_selection__k=6, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=0.01, model__gamma=auto, score=0.277, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=0.01, model__gamma=auto, score=0.702, total=   0.1s\n",
      "[CV] feature_selection__k=6, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=0.01, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=0.01, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=0.01, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=0.01, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=0.1, model__gamma=auto, score=0.617, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=0.1, model__gamma=auto, score=0.511, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.1, model__gamma=auto .........\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  feature_selection__k=6, model__C=0.1, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=0.1, model__gamma=auto, score=0.532, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=0.1, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=0.1, model__gamma=auto, score=0.660, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=0.1, model__gamma=auto, score=0.543, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=0.1, model__gamma=auto, score=0.478, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=0.1, model__gamma=auto, score=0.565, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=0.1, model__gamma=auto, score=0.565, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=6, model__C=1, model__gamma=auto, score=0.660, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=6, model__C=1, model__gamma=auto, score=0.553, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=6, model__C=1, model__gamma=auto, score=0.660, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=6, model__C=1, model__gamma=auto, score=0.532, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=6, model__C=1, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=6, model__C=1, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=6, model__C=1, model__gamma=auto, score=0.587, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=6, model__C=1, model__gamma=auto, score=0.543, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=6, model__C=1, model__gamma=auto, score=0.565, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=6, model__C=1, model__gamma=auto, score=0.652, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=6, model__C=10, model__gamma=auto, score=0.638, total=   0.1s\n",
      "[CV] feature_selection__k=6, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=6, model__C=10, model__gamma=auto, score=0.511, total=   0.1s\n",
      "[CV] feature_selection__k=6, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=6, model__C=10, model__gamma=auto, score=0.723, total=   0.1s\n",
      "[CV] feature_selection__k=6, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=6, model__C=10, model__gamma=auto, score=0.489, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=6, model__C=10, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=6, model__C=10, model__gamma=auto, score=0.638, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=6, model__C=10, model__gamma=auto, score=0.543, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=6, model__C=10, model__gamma=auto, score=0.565, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=6, model__C=10, model__gamma=auto, score=0.587, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=6, model__C=10, model__gamma=auto, score=0.609, total=   0.1s\n",
      "[CV] feature_selection__k=6, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=100, model__gamma=auto, score=0.574, total=   0.1s\n",
      "[CV] feature_selection__k=6, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=100, model__gamma=auto, score=0.511, total=   0.1s\n",
      "[CV] feature_selection__k=6, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=100, model__gamma=auto, score=0.638, total=   0.1s\n",
      "[CV] feature_selection__k=6, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=100, model__gamma=auto, score=0.574, total=   0.1s\n",
      "[CV] feature_selection__k=6, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=100, model__gamma=auto, score=0.723, total=   0.1s\n",
      "[CV] feature_selection__k=6, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=100, model__gamma=auto, score=0.638, total=   0.1s\n",
      "[CV] feature_selection__k=6, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=100, model__gamma=auto, score=0.543, total=   0.1s\n",
      "[CV] feature_selection__k=6, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=100, model__gamma=auto, score=0.500, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=100, model__gamma=auto, score=0.587, total=   0.0s\n",
      "[CV] feature_selection__k=6, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=6, model__C=100, model__gamma=auto, score=0.696, total=   0.1s\n",
      "[CV] feature_selection__k=6, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=1000, model__gamma=auto, score=0.660, total=   0.2s\n",
      "[CV] feature_selection__k=6, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=1000, model__gamma=auto, score=0.532, total=   0.2s\n",
      "[CV] feature_selection__k=6, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=1000, model__gamma=auto, score=0.702, total=   0.2s\n",
      "[CV] feature_selection__k=6, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=1000, model__gamma=auto, score=0.617, total=   0.3s\n",
      "[CV] feature_selection__k=6, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=1000, model__gamma=auto, score=0.681, total=   0.3s\n",
      "[CV] feature_selection__k=6, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=1000, model__gamma=auto, score=0.702, total=   0.5s\n",
      "[CV] feature_selection__k=6, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=1000, model__gamma=auto, score=0.587, total=   0.3s\n",
      "[CV] feature_selection__k=6, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=1000, model__gamma=auto, score=0.500, total=   0.2s\n",
      "[CV] feature_selection__k=6, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=1000, model__gamma=auto, score=0.630, total=   0.2s\n",
      "[CV] feature_selection__k=6, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=6, model__C=1000, model__gamma=auto, score=0.696, total=   0.2s\n",
      "[CV] feature_selection__k=7, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=7, model__C=0.0001, model__gamma=auto, score=0.277, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=7, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=7, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=7, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.0001, model__gamma=auto ......\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  feature_selection__k=7, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.1s\n",
      "[CV] feature_selection__k=7, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=7, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=7, model__C=0.0001, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=7, model__C=0.0001, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=7, model__C=0.0001, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=7, model__C=0.0001, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=7, model__C=0.001, model__gamma=auto, score=0.277, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=7, model__C=0.001, model__gamma=auto, score=0.298, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=7, model__C=0.001, model__gamma=auto, score=0.298, total=   0.1s\n",
      "[CV] feature_selection__k=7, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=7, model__C=0.001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=7, model__C=0.001, model__gamma=auto, score=0.298, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=7, model__C=0.001, model__gamma=auto, score=0.298, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=7, model__C=0.001, model__gamma=auto, score=0.717, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=7, model__C=0.001, model__gamma=auto, score=0.717, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=7, model__C=0.001, model__gamma=auto, score=0.717, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=7, model__C=0.001, model__gamma=auto, score=0.717, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=0.01, model__gamma=auto, score=0.277, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=0.01, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=0.01, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=0.01, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=0.01, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=0.1, model__gamma=auto, score=0.660, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=0.1, model__gamma=auto, score=0.511, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=0.1, model__gamma=auto, score=0.681, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=0.1, model__gamma=auto, score=0.553, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=0.1, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=0.1, model__gamma=auto, score=0.660, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=0.1, model__gamma=auto, score=0.587, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=0.1, model__gamma=auto, score=0.500, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=0.1, model__gamma=auto, score=0.543, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=0.1, model__gamma=auto, score=0.543, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=7, model__C=1, model__gamma=auto, score=0.660, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=7, model__C=1, model__gamma=auto, score=0.574, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=7, model__C=1, model__gamma=auto, score=0.723, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=7, model__C=1, model__gamma=auto, score=0.511, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=7, model__C=1, model__gamma=auto, score=0.681, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=7, model__C=1, model__gamma=auto, score=0.723, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=7, model__C=1, model__gamma=auto, score=0.565, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=7, model__C=1, model__gamma=auto, score=0.565, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=7, model__C=1, model__gamma=auto, score=0.500, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=7, model__C=1, model__gamma=auto, score=0.630, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=7, model__C=10, model__gamma=auto, score=0.638, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=7, model__C=10, model__gamma=auto, score=0.511, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=7, model__C=10, model__gamma=auto, score=0.723, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=7, model__C=10, model__gamma=auto, score=0.511, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=7, model__C=10, model__gamma=auto, score=0.681, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=10, model__gamma=auto ..........\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  feature_selection__k=7, model__C=10, model__gamma=auto, score=0.681, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=7, model__C=10, model__gamma=auto, score=0.522, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=7, model__C=10, model__gamma=auto, score=0.587, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=7, model__C=10, model__gamma=auto, score=0.565, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=7, model__C=10, model__gamma=auto, score=0.587, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=100, model__gamma=auto, score=0.660, total=   0.1s\n",
      "[CV] feature_selection__k=7, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=100, model__gamma=auto, score=0.574, total=   0.1s\n",
      "[CV] feature_selection__k=7, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=100, model__gamma=auto, score=0.638, total=   0.1s\n",
      "[CV] feature_selection__k=7, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=100, model__gamma=auto, score=0.617, total=   0.1s\n",
      "[CV] feature_selection__k=7, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=100, model__gamma=auto, score=0.638, total=   0.1s\n",
      "[CV] feature_selection__k=7, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=100, model__gamma=auto, score=0.681, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=100, model__gamma=auto, score=0.609, total=   0.1s\n",
      "[CV] feature_selection__k=7, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=100, model__gamma=auto, score=0.522, total=   0.1s\n",
      "[CV] feature_selection__k=7, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=100, model__gamma=auto, score=0.652, total=   0.1s\n",
      "[CV] feature_selection__k=7, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=7, model__C=100, model__gamma=auto, score=0.609, total=   0.0s\n",
      "[CV] feature_selection__k=7, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=1000, model__gamma=auto, score=0.638, total=   0.1s\n",
      "[CV] feature_selection__k=7, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=1000, model__gamma=auto, score=0.532, total=   0.2s\n",
      "[CV] feature_selection__k=7, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=1000, model__gamma=auto, score=0.660, total=   0.1s\n",
      "[CV] feature_selection__k=7, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=1000, model__gamma=auto, score=0.596, total=   0.1s\n",
      "[CV] feature_selection__k=7, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=1000, model__gamma=auto, score=0.617, total=   0.3s\n",
      "[CV] feature_selection__k=7, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=1000, model__gamma=auto, score=0.723, total=   0.1s\n",
      "[CV] feature_selection__k=7, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=1000, model__gamma=auto, score=0.652, total=   0.2s\n",
      "[CV] feature_selection__k=7, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=1000, model__gamma=auto, score=0.478, total=   0.2s\n",
      "[CV] feature_selection__k=7, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=1000, model__gamma=auto, score=0.652, total=   0.3s\n",
      "[CV] feature_selection__k=7, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=7, model__C=1000, model__gamma=auto, score=0.652, total=   0.1s\n",
      "[CV] feature_selection__k=8, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=8, model__C=0.0001, model__gamma=auto, score=0.277, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=8, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=8, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=8, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=8, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=8, model__C=0.0001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=8, model__C=0.0001, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=8, model__C=0.0001, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=8, model__C=0.0001, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.0001, model__gamma=auto ......\n",
      "[CV]  feature_selection__k=8, model__C=0.0001, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=8, model__C=0.001, model__gamma=auto, score=0.277, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=8, model__C=0.001, model__gamma=auto, score=0.298, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=8, model__C=0.001, model__gamma=auto, score=0.298, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=8, model__C=0.001, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=8, model__C=0.001, model__gamma=auto, score=0.298, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=8, model__C=0.001, model__gamma=auto, score=0.298, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=8, model__C=0.001, model__gamma=auto, score=0.717, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=8, model__C=0.001, model__gamma=auto, score=0.717, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=8, model__C=0.001, model__gamma=auto, score=0.717, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.001, model__gamma=auto .......\n",
      "[CV]  feature_selection__k=8, model__C=0.001, model__gamma=auto, score=0.717, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=0.01, model__gamma=auto, score=0.277, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=0.01, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=0.01, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=0.01, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=0.01, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.01, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=0.01, model__gamma=auto, score=0.283, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=0.1, model__gamma=auto, score=0.660, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=0.1, model__gamma=auto, score=0.468, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=0.1, model__gamma=auto, score=0.660, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.1, model__gamma=auto .........\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  feature_selection__k=8, model__C=0.1, model__gamma=auto, score=0.553, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=0.1, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=0.1, model__gamma=auto, score=0.660, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=0.1, model__gamma=auto, score=0.587, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=0.1, model__gamma=auto, score=0.478, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=0.1, model__gamma=auto, score=0.543, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=0.1, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=0.1, model__gamma=auto, score=0.543, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=8, model__C=1, model__gamma=auto, score=0.681, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=8, model__C=1, model__gamma=auto, score=0.553, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=8, model__C=1, model__gamma=auto, score=0.723, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=8, model__C=1, model__gamma=auto, score=0.574, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=8, model__C=1, model__gamma=auto, score=0.681, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=8, model__C=1, model__gamma=auto, score=0.681, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=8, model__C=1, model__gamma=auto, score=0.522, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=8, model__C=1, model__gamma=auto, score=0.609, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=8, model__C=1, model__gamma=auto, score=0.543, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=1, model__gamma=auto ...........\n",
      "[CV]  feature_selection__k=8, model__C=1, model__gamma=auto, score=0.630, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=8, model__C=10, model__gamma=auto, score=0.638, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=8, model__C=10, model__gamma=auto, score=0.511, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=8, model__C=10, model__gamma=auto, score=0.766, total=   0.1s\n",
      "[CV] feature_selection__k=8, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=8, model__C=10, model__gamma=auto, score=0.511, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=8, model__C=10, model__gamma=auto, score=0.702, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=8, model__C=10, model__gamma=auto, score=0.596, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=8, model__C=10, model__gamma=auto, score=0.609, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=8, model__C=10, model__gamma=auto, score=0.609, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=8, model__C=10, model__gamma=auto, score=0.565, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=10, model__gamma=auto ..........\n",
      "[CV]  feature_selection__k=8, model__C=10, model__gamma=auto, score=0.674, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=100, model__gamma=auto, score=0.617, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=100, model__gamma=auto, score=0.553, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=100, model__gamma=auto, score=0.638, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=100, model__gamma=auto, score=0.617, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=100, model__gamma=auto, score=0.638, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=100, model__gamma=auto, score=0.723, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=100, model__gamma=auto, score=0.652, total=   0.1s\n",
      "[CV] feature_selection__k=8, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=100, model__gamma=auto, score=0.587, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=100, model__gamma=auto, score=0.609, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=100, model__gamma=auto .........\n",
      "[CV]  feature_selection__k=8, model__C=100, model__gamma=auto, score=0.717, total=   0.0s\n",
      "[CV] feature_selection__k=8, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=1000, model__gamma=auto, score=0.553, total=   0.2s\n",
      "[CV] feature_selection__k=8, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=1000, model__gamma=auto, score=0.511, total=   0.1s\n",
      "[CV] feature_selection__k=8, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=1000, model__gamma=auto, score=0.702, total=   0.2s\n",
      "[CV] feature_selection__k=8, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=1000, model__gamma=auto, score=0.617, total=   0.1s\n",
      "[CV] feature_selection__k=8, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=1000, model__gamma=auto, score=0.596, total=   0.2s\n",
      "[CV] feature_selection__k=8, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=1000, model__gamma=auto, score=0.638, total=   0.2s\n",
      "[CV] feature_selection__k=8, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=1000, model__gamma=auto, score=0.587, total=   0.2s\n",
      "[CV] feature_selection__k=8, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=1000, model__gamma=auto, score=0.522, total=   0.1s\n",
      "[CV] feature_selection__k=8, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=1000, model__gamma=auto, score=0.674, total=   0.1s\n",
      "[CV] feature_selection__k=8, model__C=1000, model__gamma=auto ........\n",
      "[CV]  feature_selection__k=8, model__C=1000, model__gamma=auto, score=0.696, total=   0.1s\n",
      "{'feature_selection__k': 8, 'model__C': 100, 'model__gamma': 'auto'}\n",
      "0.635245143385754\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done 320 out of 320 | elapsed:   18.8s finished\n"
     ]
    }
   ],
   "source": [
    "model_SVC = pipeline_SVC(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "         steps=[('scale',\n",
       "                 StandardScaler(copy=True, with_mean=True, with_std=True)),\n",
       "                ('feature_select',\n",
       "                 SelectKBest(k=8,\n",
       "                             score_func=<function f_classif at 0x0000026A17BFBC18>)),\n",
       "                ('svm',\n",
       "                 SVC(C=100, break_ties=False, cache_size=200,\n",
       "                     class_weight='balanced', coef0=0.0,\n",
       "                     decision_function_shape='ovr', degree=3, gamma='scale',\n",
       "                     kernel='rbf', max_iter=-1, probability=False,\n",
       "                     random_state=None, shrinking=True, tol=0.001,\n",
       "                     verbose=False))],\n",
       "         verbose=False)"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scale = preprocessing.StandardScaler()\n",
    "select_feature = SelectKBest(f_classif, k=8)\n",
    "model = SVC(class_weight = 'balanced')\n",
    "\n",
    "pp_svc = Pipeline([('scale', scale),\n",
    "                  ('feature_select', select_feature), \n",
    "                  ('svm', model)])\n",
    "\n",
    "pp_svc.set_params(svm__C = 100).fit(X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           1       0.85      0.67      0.75        85\n",
      "           2       0.44      0.69      0.54        32\n",
      "\n",
      "    accuracy                           0.68       117\n",
      "   macro avg       0.65      0.68      0.64       117\n",
      "weighted avg       0.74      0.68      0.69       117\n",
      "\n",
      "Accuracy: 0.6752136752136753\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x26a1869fdc8>"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVoAAAD9CAYAAAD9P7+UAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAARKklEQVR4nO3de5RddXXA8e8OEPDBI82DBEKMVDIoPsJqoFDA8hIxILSCLYiIVpj6XKC2QdCW+uryiXZZtWsUUCyiUYO6EESUZ5YaCAHlEUcwxhiTEB4JWsHAzN39Y25wiJO5d8L9zZmcfD9Zv8Xcc+793c1amb129tn33MhMJEnljKs6AEmqOxOtJBVmopWkwky0klSYiVaSCjPRSlJh21cdQJ11dXUtB34P9AN9vb29c7q6ur4GdDWfshuwvre3d3Y1EaoCewGXAlOBBtAD/BcwG/gfYCegD3gLcEtFMarDTLTlHdHb2/vgxge9vb3/uPHnrq6uTwCPVBKVqtIHvAtYAuwM3AZcC3wUeB9wNTC3+fjwakJUp7VMtBGxL3AisCeQwCrgO5m5tHBstdbV1RXAPwBHVh2LRtXq5oKBf+0s5U+/W7s0j+/KwO+ZamLYHm1EnAt8FQgG/hlza/PnyyPi3eXD2+ol8P2urq7burq6ujc5dxhwf29v770VxKWxYSawP7AIOAf4GPAb4OPAedWFpU6L4T6CGxG/APbLzCc2OT4euDsz99nM67qBboDPfuKDf3Xm607tXMRbkbUPPMSUyRN5aN16zjrnfM5/x5uZM/tFALz/Y59mxvQ9eP2pJ1UcZTW+OPvfqw6hUts/c0eO/8Z7uePT32b51Ys5+P2ns/onP2f5Vbey9/F/zb6nHcFVp3646jBH3Vkr/zee7h5PPLis7fsK7DBp76f9fu1oNXXQAPYY4vi05rkhZWZPZs7JzDnbapIFmDJ5IgATJ+zGUS/9G+68pxeAvr5+fnDjjzj2qJdWGZ4qEttvx8t6zuaXV/yI5VcvBmDWyYex/KpbAVh25SImz/7LKkNUh7Xq0Z4D/DAi7mXgnzQAM4DnAW8rGdjW7tHH/kg2GjzrWc/k0cf+yI9uWcKb3/AaAH6y+Hb2fs50pk6ZXHGUqsLffvxM1t23ijs/f/WTx/5w/zqmHfx8Vv94KXscsh+P/GpNhRFu5Rr9HdsqIpYzaHIoM+c0j7+dgRzYB3w3M+cNt8+wiTYzvxcRs4ADGWjYB7ASuDUzO/d/U0MPPbyOs8//AAD9ff3MPeZwDj1oDgBX/+BGXnH04RVGp6rsfsAs9jn5MB5auoJXXfMhAG79yHxunncRB7/vdMZtP47+DU+w8NyLKo50K9bf1+kdj8jMJyeHIuIIBgYEXpyZGyJiSqsNhu3RdsJI+iXadmzrPVoNrRM92sdX3d12zhm/x37Dvl+zop2zSaKdD/Rk5g/afR8/GSapXhqNtldEdEfE4kFr0+mgBL4fEbcNOjcLOCwiFkXEjRFxQKuQ/MCCpHrJzV6n//OnZvYw8Om8zTkkM1c12wPXRsTPGcibE4CDgAOA+RGxdw7THjDRSqqXDl4My8xVzf+ujYgrGLhetRJY0Eyst0REA5gEPLC5fWwdSKqXbLS/hhERz4qInTf+DBwD3AV8i+YnOpvDAuOBBze3D1jRSqqZ7NzUwe7AFREBA7nyK81JrPHAxRFxF/A4cMZwbYONL5ak+mi036MdTmYuA14yxPHHgdeOZC8TraR6GcHFsNFiopVULx28GNYpJlpJ9WJFK0mFdf4juE+biVZSvXToYlgnmWgl1cpYvN+ViVZSvdijlaTCbB1IUmFWtJJUWP8TrZ8zyky0kurF1oEkFWbrQJIKs6KVpMJMtJJUVnoxTJIKs0crSYXZOpCkwqxoJakwK1pJKsyKVpIK6/PG35JUlhWtJBVmj1aSCrOilaTCrGglqTArWkkqzKkDSSoss+oI/sy4qgOQpI5qNNpfbYiI7SLi9oi4svn4qIhYEhF3RMTCiHheqz1MtJLqpcOJFjgbWDro8eeA0zJzNvAV4L2tNjDRSqqXbLS/WoiI6cBxwBcGvwOwS/PnXYFVrfaxRyupXvr7235qRHQD3YMO9WRmz6DHnwLmATsPOnYmcFVEPAb8Djio1ftY0UqqlxG0DjKzJzPnDFpPJtmIOB5Ym5m3bfIO7wDmZuZ04BLgwlYhWdFKqpfOfWDhEOCEiJgL7ATsEhHfBfbNzEXN53wN+F6rjaxoJdVLh3q0mXleZk7PzJnAKcB1wInArhExq/m0l/HUC2VDsqKVVCvZKDdHm5l9EXEW8M2IaADrgH9q9ToTraR6KXCvg8y8Abih+fMVwBUjeb2JVlK9jGDqYLSYaCXVi3fvkqTCTLSSVNgYvKmMiVZSvVjRSlJhBce7tpSJVlK9OHUgSWWlrQNJKszWgSQV5pczSlJhVrSSVFifF8MkqSxbB5JUmK0DSSrL8S5JKs2KVpIKM9FKUmF+BFeSyir5nWFbykQrqV5MtJJUmFMHklSYFa0kFWailaSyst/WgSSVZUUrSWU53iVJpZloJamwsdeiZVzVAUhSJ2Vfo+3VjojYLiJuj4grm4+fGxGLIuLeiPhaRIxvtYeJVlK9NEaw2nM2sHTQ448An8zMfYB1wBtbbWCilVQr2ci2VysRMR04DvhC83EARwLfaD7lS8DftdrHRCupXkZQ0UZEd0QsHrS6N9ntU8A8/lT/TgTWZ2Zf8/FKYM9WIXkxTFKtjGS8KzN7gJ6hzkXE8cDazLwtIg7feHiobVq9j4lWUr10burgEOCEiJgL7ATswkCFu1tEbN+saqcDq1ptZOtAUq1kX/tr2H0yz8vM6Zk5EzgFuC4zTwOuB05uPu0M4NutYjLRSqqVbLS/ttC5wDsj4j4GerYXtXqBrQNJ9VLgAwuZeQNwQ/PnZcCBI3m9iVZSrTyNSrUYE62kWjHRSlJh2T/UBFa1TLSSasWKVpIKy4YVrSQVZUUrSYVlWtFKUlFWtJJUWMOpA0kqy4thklSYiVaSCsux9yW4JlpJ9WJFK0mFOd4lSYX1O3UgSWVZ0UpSYfZoJakwpw4kqTArWkkqrL8x9r5z1kQrqVZsHUhSYQ2nDiSpLMe7JKmwbbJ18Iw9Div9FtoKvX6Pg6sOQWPQWR3Yw9aBJBXm1IEkFTYGOwcmWkn1YutAkgrr1NRBROwE3ATsyECu/EZmXhARlwFzgCeAW4B/zswnhttr7DUzJOlpaIxgtbABODIzXwLMBo6NiIOAy4B9gRcBzwDObLWRFa2kWkk6U9FmZgL/13y4Q3NlZl618TkRcQswvdVeVrSSaqUvo+0VEd0RsXjQ6h68V0RsFxF3AGuBazNz0aBzOwCnA99rFZMVraRaGUlFm5k9QM8w5/uB2RGxG3BFRLwwM+9qnv4scFNm3tzqfaxoJdVKB3u0T8rM9cANwLEAEXEBMBl4ZzuvN9FKqpUk2l7DiYjJzUqWiHgGcDTw84g4E3g5cGpmtpWvbR1IqpWRVKotTAO+FBHbMVCUzs/MKyOiD/g18OOIAFiQme8fbiMTraRa6e/c1MHPgP2HOD7ivGmilVQrY/CbbEy0kuql0aGKtpNMtJJqxZvKSFJhHbwY1jEmWkm10ghbB5JUVH/VAQzBRCupVpw6kKTCnDqQpMKcOpCkwmwdSFJhjndJUmH9VrSSVJYVrSQVZqKVpMI69G3jHWWilVQrVrSSVJgfwZWkwpyjlaTCbB1IUmEmWkkqzHsdSFJh9mglqTCnDiSpsMYYbB6YaCXVihfDJKmwsVfPmmgl1YwVrSQV1hdjr6YdV3UAktRJOYI1nIjYKyKuj4ilEXF3RJy9yfl/iYiMiEmtYrKilVQrHWwd9AHvyswlEbEzcFtEXJuZ90TEXsDLgBXtbGRFK6lWGmTbaziZuTozlzR//j2wFNizefqTwDzavPZmopVUKyNpHUREd0QsHrS6h9ozImYC+wOLIuIE4LeZ+dN2Y7J1IKlWRtI6yMweoGe450TEs4FvAucw0E54D3DMSGIy0Uqqlf4OTtJGxA4MJNnLMnNBRLwIeC7w04gAmA4siYgDM3PN5vYx0UqqlU5dDIuBTHoRsDQzLwTIzDuBKYOesxyYk5kPDreXPVpJtZIj+NPCIcDpwJERcUdzzd2SmKxoJdVKpyrazFwIDHvTxcyc2c5eJlpJteLduySpsLGXZk20kmqmbwymWhOtpFpp4yLXqDPRSqoVb5MoSYVZ0UpSYVa0klRYf1rRSlJRztFKUmH2aCWpMHu0klSYrQNJKszWgSQV5tSBJBVm60CSCvNimCQVZo9WkgqzdSBJhaUXwySprE5+3XinmGgl1YqtA0kqzNaBJBVmRStJhTneJUmF+RFcSSrM1oEkFWailaTCxuLUwbiqA5CkTmqQba9WIuLiiFgbEXdtcvztEdEbEXdHxEdb7WNFK6lWOjx18EXgv4FLNx6IiCOAE4EXZ+aGiJjSahMTraRa6c/O3SgxM2+KiJmbHH4z8OHM3NB8ztpW+9g6kFQrmdn2iojuiFg8aHW38RazgMMiYlFE3BgRB7R6gRWtpFoZydRBZvYAPSN8i+2BCcBBwAHA/IjYO4e5CmdFK6lWcgR/ttBKYEEOuIWBL3WYNNwLTLSSaqWR2fbaQt8CjgSIiFnAeODB4V5g60BSrXRy6iAiLgcOByZFxErgAuBi4OLmyNfjwBnDtQ3ARCupZjo8dXDqZk69diT7mGgl1crTaAkUY6KVVCveJlGSCrOilaTCrGglqbD+7K86hD9jopVUK2PxNokmWkm14o2/JakwK1pJKsypA0kqzKkDSSqskx/B7RQTraRasUcrSYXZo5WkwqxoJakw52glqTArWkkqzKkDSSrMi2HbkM/3fILj5h7N2gceZPb+RwEwYcJuXH7Z53jOc/bi17/+Dae85k2sX/9IxZFqtEyYNpE3Xvh2dp28G41GctPl1/LDS67i5PNO5yVHz6H/8T7WrljDJf/6GR773aNVh7vVGoutA78Ft5BLL53Pccef9pRj5857K9ddv5Dn73co112/kHPnvbWi6FSFRl8/8z/4Jf7t6HP4z78/jyNOP5Zpz5vOPQt/xgXHvIP/eMW7uP9Xq5n7lldVHepWbRS+bnzETLSF3LxwEQ+vW/+UY6985cu59MtfB+DSL3+dE044torQVJFHHljPirt/BcCGP/yR1b/8LROm/gX33PxTGv0DfcVlt/+CCVMnVhnmVi8z216jZYsTbUS8oZOBbAt2nzKJNWvWArBmzVqmTPYXals1cfpkZrxgJsvuuPcpxw999ZHcdcOSiqKqh0Zm22u0xJZm9YhYkZkzNnOuG+huPuzJzJ4tjG9rNxO4EnghwIYNGx7dcccdnzno/DpgQgVxqVrPBm4EPgQsiIju5u/Ie4A5wKtgDA6DaosNezEsIn62uVPA7pt7XfMvzbaaXDdr9erV42bOnDkNWA1MA9ZWHJJG3w7AN4HLgAXNY93ABuB44ChMsrXTqnWwO/A64JVDrIfKhlY/11xzzXrgjObDM4BvVxiORl8AFwFLgQs3HjzppJN2Ac4FTgAcN6ihYVsHEXERcElmLhzi3Fcy8zUlg9vKXQ4cDkwC7gcumDp16tvWrFnzMDADWAG8Gni4sgg12g4FbgbuBDZO1Z+/YsWKBTNmzHiAPxUvPwHeVEF8KmSLe7QauUG9OOlJ/r2oPxOtJBXmHK0kFWailaTCTLSjJCKOjYjeiLgvIt5ddTyqXkRcHBFrI+KuqmNRWSbaURAR2wGfAV4BvAA4NSJeUG1UGgO+CPg57G2AiXZ0HAjcl5nLMvNx4KvAiRXHpIpl5k043rdNMNGOjj2B3wx6vLJ5TNI2wEQ7OmKIY87VSdsIE+3oWAnsNejxdGBVRbFIGmUm2tFxK7BPRDw3IsYDpwDfqTgmSaPERDsKMrMPeBtwDQM3FJmfmXdXG5WqFhGXAz8GuiJiZUS8seqYVIYfwZWkwqxoJakwE60kFWailaTCTLSSVJiJVpIKM9FKUmEmWkkq7P8B2XqVda21CEgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 모델을 통해 Testset 예측\n",
    "preds = pp_svc.predict(X_test)\n",
    "\n",
    "print(metrics.classification_report(y_test, preds))\n",
    "print(\"Accuracy:\", metrics.accuracy_score(y_true = y_test, y_pred=preds))\n",
    "\n",
    "sns.heatmap(confusion_matrix(y_test,preds),annot=True,fmt=\"d\")\n",
    "#accuracy 0.67 나오네요"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "세 모델 비교 결과 random forest가 가장 accuracy가 높게 나왔고\n",
    "이 데이터를 classification를 하는 가장 좋은 모델은 randomforest로 0.75라는 그렇게 높은 accuracy를 얻진 못했지만 데이터 전처리, eda부터 모델 적용까지 많이 고민해볼수 있었던 과제였던 것 같습니당!"
   ]
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}