feature_selection_methods.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "\n",
    "sns.set()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data = pd.read_csv('data/modile_prices_train.csv')\n",
    "# test_data = pd.read_csv('data/datasets_11167_15520_test.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Basic EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2000, 21)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "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>battery_power</th>\n",
       "      <th>blue</th>\n",
       "      <th>clock_speed</th>\n",
       "      <th>dual_sim</th>\n",
       "      <th>fc</th>\n",
       "      <th>four_g</th>\n",
       "      <th>int_memory</th>\n",
       "      <th>m_dep</th>\n",
       "      <th>mobile_wt</th>\n",
       "      <th>n_cores</th>\n",
       "      <th>...</th>\n",
       "      <th>px_height</th>\n",
       "      <th>px_width</th>\n",
       "      <th>ram</th>\n",
       "      <th>sc_h</th>\n",
       "      <th>sc_w</th>\n",
       "      <th>talk_time</th>\n",
       "      <th>three_g</th>\n",
       "      <th>touch_screen</th>\n",
       "      <th>wifi</th>\n",
       "      <th>price_range</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1217</th>\n",
       "      <td>1451</td>\n",
       "      <td>1</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>57</td>\n",
       "      <td>0.6</td>\n",
       "      <td>181</td>\n",
       "      <td>8</td>\n",
       "      <td>...</td>\n",
       "      <td>812</td>\n",
       "      <td>1456</td>\n",
       "      <td>3038</td>\n",
       "      <td>15</td>\n",
       "      <td>7</td>\n",
       "      <td>18</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      battery_power  blue  clock_speed  dual_sim  fc  four_g  int_memory  \\\n",
       "1217           1451     1          0.5         0   5       1          57   \n",
       "\n",
       "      m_dep  mobile_wt  n_cores  ...  px_height  px_width   ram  sc_h  sc_w  \\\n",
       "1217    0.6        181        8  ...        812      1456  3038    15     7   \n",
       "\n",
       "      talk_time  three_g  touch_screen  wifi  price_range  \n",
       "1217         18        1             0     1            3  \n",
       "\n",
       "[1 rows x 21 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.sample()"
   ]
  },
  {
   "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>battery_power</th>\n",
       "      <th>blue</th>\n",
       "      <th>clock_speed</th>\n",
       "      <th>dual_sim</th>\n",
       "      <th>fc</th>\n",
       "      <th>four_g</th>\n",
       "      <th>int_memory</th>\n",
       "      <th>m_dep</th>\n",
       "      <th>mobile_wt</th>\n",
       "      <th>n_cores</th>\n",
       "      <th>...</th>\n",
       "      <th>px_height</th>\n",
       "      <th>px_width</th>\n",
       "      <th>ram</th>\n",
       "      <th>sc_h</th>\n",
       "      <th>sc_w</th>\n",
       "      <th>talk_time</th>\n",
       "      <th>three_g</th>\n",
       "      <th>touch_screen</th>\n",
       "      <th>wifi</th>\n",
       "      <th>price_range</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.0000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "      <td>2000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>1238.518500</td>\n",
       "      <td>0.4950</td>\n",
       "      <td>1.522250</td>\n",
       "      <td>0.509500</td>\n",
       "      <td>4.309500</td>\n",
       "      <td>0.521500</td>\n",
       "      <td>32.046500</td>\n",
       "      <td>0.501750</td>\n",
       "      <td>140.249000</td>\n",
       "      <td>4.520500</td>\n",
       "      <td>...</td>\n",
       "      <td>645.108000</td>\n",
       "      <td>1251.515500</td>\n",
       "      <td>2124.213000</td>\n",
       "      <td>12.306500</td>\n",
       "      <td>5.767000</td>\n",
       "      <td>11.011000</td>\n",
       "      <td>0.761500</td>\n",
       "      <td>0.503000</td>\n",
       "      <td>0.507000</td>\n",
       "      <td>1.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>439.418206</td>\n",
       "      <td>0.5001</td>\n",
       "      <td>0.816004</td>\n",
       "      <td>0.500035</td>\n",
       "      <td>4.341444</td>\n",
       "      <td>0.499662</td>\n",
       "      <td>18.145715</td>\n",
       "      <td>0.288416</td>\n",
       "      <td>35.399655</td>\n",
       "      <td>2.287837</td>\n",
       "      <td>...</td>\n",
       "      <td>443.780811</td>\n",
       "      <td>432.199447</td>\n",
       "      <td>1084.732044</td>\n",
       "      <td>4.213245</td>\n",
       "      <td>4.356398</td>\n",
       "      <td>5.463955</td>\n",
       "      <td>0.426273</td>\n",
       "      <td>0.500116</td>\n",
       "      <td>0.500076</td>\n",
       "      <td>1.118314</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>501.000000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.100000</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>500.000000</td>\n",
       "      <td>256.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>851.750000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>0.200000</td>\n",
       "      <td>109.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>282.750000</td>\n",
       "      <td>874.750000</td>\n",
       "      <td>1207.500000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1226.000000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>1.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>32.000000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>141.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>564.000000</td>\n",
       "      <td>1247.000000</td>\n",
       "      <td>2146.500000</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1615.250000</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>2.200000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>48.000000</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>170.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>947.250000</td>\n",
       "      <td>1633.000000</td>\n",
       "      <td>3064.500000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.250000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1998.000000</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>64.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>200.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>1960.000000</td>\n",
       "      <td>1998.000000</td>\n",
       "      <td>3998.000000</td>\n",
       "      <td>19.000000</td>\n",
       "      <td>18.000000</td>\n",
       "      <td>20.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       battery_power       blue  clock_speed     dual_sim           fc  \\\n",
       "count    2000.000000  2000.0000  2000.000000  2000.000000  2000.000000   \n",
       "mean     1238.518500     0.4950     1.522250     0.509500     4.309500   \n",
       "std       439.418206     0.5001     0.816004     0.500035     4.341444   \n",
       "min       501.000000     0.0000     0.500000     0.000000     0.000000   \n",
       "25%       851.750000     0.0000     0.700000     0.000000     1.000000   \n",
       "50%      1226.000000     0.0000     1.500000     1.000000     3.000000   \n",
       "75%      1615.250000     1.0000     2.200000     1.000000     7.000000   \n",
       "max      1998.000000     1.0000     3.000000     1.000000    19.000000   \n",
       "\n",
       "            four_g   int_memory        m_dep    mobile_wt      n_cores  ...  \\\n",
       "count  2000.000000  2000.000000  2000.000000  2000.000000  2000.000000  ...   \n",
       "mean      0.521500    32.046500     0.501750   140.249000     4.520500  ...   \n",
       "std       0.499662    18.145715     0.288416    35.399655     2.287837  ...   \n",
       "min       0.000000     2.000000     0.100000    80.000000     1.000000  ...   \n",
       "25%       0.000000    16.000000     0.200000   109.000000     3.000000  ...   \n",
       "50%       1.000000    32.000000     0.500000   141.000000     4.000000  ...   \n",
       "75%       1.000000    48.000000     0.800000   170.000000     7.000000  ...   \n",
       "max       1.000000    64.000000     1.000000   200.000000     8.000000  ...   \n",
       "\n",
       "         px_height     px_width          ram         sc_h         sc_w  \\\n",
       "count  2000.000000  2000.000000  2000.000000  2000.000000  2000.000000   \n",
       "mean    645.108000  1251.515500  2124.213000    12.306500     5.767000   \n",
       "std     443.780811   432.199447  1084.732044     4.213245     4.356398   \n",
       "min       0.000000   500.000000   256.000000     5.000000     0.000000   \n",
       "25%     282.750000   874.750000  1207.500000     9.000000     2.000000   \n",
       "50%     564.000000  1247.000000  2146.500000    12.000000     5.000000   \n",
       "75%     947.250000  1633.000000  3064.500000    16.000000     9.000000   \n",
       "max    1960.000000  1998.000000  3998.000000    19.000000    18.000000   \n",
       "\n",
       "         talk_time      three_g  touch_screen         wifi  price_range  \n",
       "count  2000.000000  2000.000000   2000.000000  2000.000000  2000.000000  \n",
       "mean     11.011000     0.761500      0.503000     0.507000     1.500000  \n",
       "std       5.463955     0.426273      0.500116     0.500076     1.118314  \n",
       "min       2.000000     0.000000      0.000000     0.000000     0.000000  \n",
       "25%       6.000000     1.000000      0.000000     0.000000     0.750000  \n",
       "50%      11.000000     1.000000      1.000000     1.000000     1.500000  \n",
       "75%      16.000000     1.000000      1.000000     1.000000     2.250000  \n",
       "max      20.000000     1.000000      1.000000     1.000000     3.000000  \n",
       "\n",
       "[8 rows x 21 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_data.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 3600x2160 with 25 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_data.hist(train_data.columns, figsize=(50, 30), xlabelsize=8, ylabelsize=8)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "В общем то с этим и картинками с кэгла видно, что все признаки или числовые или уже бинарные, но надо отскейлить:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "numeric_columns = train_data.loc[:, (train_data.nunique() > 2).values].columns.values\n",
    "numeric_columns = numeric_columns[numeric_columns != 'price_range']\n",
    "\n",
    "categ_columns = list(set(train_data.columns.values) - set(numeric_columns) - set(['price_range']))\n",
    "\n",
    "target_column = 'price_range'\n",
    "\n",
    "len(numeric_columns) + len(categ_columns) + 1 == len(train_data.columns.values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "scaler = StandardScaler()\n",
    "data_scaled = pd.DataFrame(scaler.fit_transform(train_data[numeric_columns]),\n",
    "                           columns=numeric_columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_data = pd.concat([data_scaled, train_data[categ_columns]], axis=1)\n",
    "target_data = train_data[target_column]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_data, target_data, test_size=0.3, random_state=42)\n",
    "\n",
    "X_val, X_test, y_val, y_test = train_test_split(X_test, y_test, test_size=0.5, random_state=43)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3    500\n",
       "2    500\n",
       "1    500\n",
       "0    500\n",
       "Name: price_range, dtype: int64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "target_data.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Baseline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LogisticRegressionCV\n",
    "from sklearn.metrics import f1_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_logreg(X, y):\n",
    "    \"\"\"\n",
    "    X: train table\n",
    "    \n",
    "    y: train target\n",
    "    \n",
    "    returns: LogisticRegreccionCV obj after fitting \n",
    "    \"\"\"\n",
    "    \n",
    "    logregCV = LogisticRegressionCV(Cs=100, cv=3, max_iter=10000, n_jobs=-1, random_state=42)\n",
    "    logregCV.fit(X, y)\n",
    "    return logregCV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9498"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "logreg = get_logreg(X_train, y_train)\n",
    "\n",
    "baseline_te_sc = round(f1_score(y_test, logreg.predict(X_test), average='macro'), 4)\n",
    "baseline_te_sc"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1) Pearson correlation criteria (PCC)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {},
   "outputs": [],
   "source": [
    "NUM_FEAT_LEAVE = 8"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.feature_selection import GenericUnivariateSelect\n",
    "import scipy\n",
    "from scipy.stats import pearsonr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {},
   "outputs": [],
   "source": [
    "def pearsonr_all(X, y):\n",
    "    val_pval = np.apply_along_axis(arr=np.array(X_train), axis=0, func1d=pearsonr, y=y_train)\n",
    "    return (val_pval[0], val_pval[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "metadata": {},
   "outputs": [],
   "source": [
    "max_val_sc = -1\n",
    "val_all_scores, test_all_scores = [], []\n",
    "\n",
    "for num_feat in np.arange(1, 15):\n",
    "    transformer = GenericUnivariateSelect(pearsonr_all, \n",
    "                                          mode='k_best', param=num_feat)\n",
    "\n",
    "    new_X_train = transformer.fit_transform(X_train, y_train)\n",
    "    new_X_val = transformer.transform(X_val)\n",
    "\n",
    "    logreg = get_logreg(new_X_train, y_train)\n",
    "    val_sc = f1_score(y_val, logreg.predict(new_X_val), average='macro')\n",
    "    \n",
    "    # save score and max-score \n",
    "    val_all_scores.append(val_sc)\n",
    "    if val_sc > max_val_sc:\n",
    "        max_val_sc = val_sc\n",
    "        best_num_feat = num_feat\n",
    "      \n",
    "    # save test values for graph\n",
    "    new_X_test = transformer.transform(X_test)\n",
    "    te_sc = f1_score(y_test, logreg.predict(new_X_test), average='macro')\n",
    "    test_all_scores.append(te_sc)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(np.arange(1, 15), val_all_scores, label='val')\n",
    "plt.plot(np.arange(1, 15), test_all_scores, label='test')\n",
    "plt.title('PCC')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "best score 0.96 with 6 features\n",
      "1.07 %\n"
     ]
    }
   ],
   "source": [
    "transformer = GenericUnivariateSelect(pearsonr_all, \n",
    "                                          mode='k_best', param=best_num_feat)\n",
    "\n",
    "new_X_train = transformer.fit_transform(X_train, y_train)\n",
    "new_X_test = transformer.transform(X_test)\n",
    "\n",
    "logreg = get_logreg(new_X_train, y_train)\n",
    "te_sc = f1_score(y_test, logreg.predict(new_X_test), average='macro')\n",
    "print('best score', round(te_sc, 4), 'with', best_num_feat, 'features')\n",
    "print(round(te_sc / baseline_te_sc * 100 - 100, 2), '%')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2) Mutual information (MI)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.feature_selection import mutual_info_classif\n",
    "\n",
    "def my_mutual_info_classif(X, y):\n",
    "    return mutual_info_classif(X, y, discrete_features=categ_ind, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [],
   "source": [
    "# для функции нужны индексы категориальных колонок, получим их\n",
    "categ_ind = []\n",
    "for name in categ_columns:\n",
    "    categ_ind += [X_train.columns.get_loc(name)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "metadata": {},
   "outputs": [],
   "source": [
    "max_val_sc = -1\n",
    "val_all_scores, test_all_scores = [], []\n",
    "\n",
    "for num_feat in np.arange(1, 15):\n",
    "    transformer = GenericUnivariateSelect(my_mutual_info_classif, \n",
    "                                          mode='k_best', param=num_feat)\n",
    "    \n",
    "    new_X_train = transformer.fit_transform(X_train, y_train)\n",
    "    new_X_val = transformer.transform(X_val)\n",
    "\n",
    "    logreg = get_logreg(new_X_train, y_train)\n",
    "    val_sc = f1_score(y_val, logreg.predict(new_X_val), average='macro')\n",
    "    \n",
    "    # save score and max-score \n",
    "    val_all_scores.append(val_sc)\n",
    "    if val_sc > max_val_sc:\n",
    "        max_val_sc = val_sc\n",
    "        best_num_feat = num_feat\n",
    "       \n",
    "    # save test values for graph\n",
    "    new_X_test = transformer.transform(X_test)\n",
    "    te_sc = f1_score(y_test, logreg.predict(new_X_test), average='macro')\n",
    "    test_all_scores.append(te_sc)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(np.arange(1, 15), val_all_scores, label='val')\n",
    "plt.plot(np.arange(1, 15), test_all_scores, label='test')\n",
    "plt.title('MI')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "best score 0.9666 with 14 features\n",
      "1.77 %\n"
     ]
    }
   ],
   "source": [
    "transformer = GenericUnivariateSelect(my_mutual_info_classif, \n",
    "                                          mode='k_best', param=best_num_feat)\n",
    "\n",
    "new_X_train = transformer.fit_transform(X_train, y_train)\n",
    "new_X_test = transformer.transform(X_test)\n",
    "\n",
    "logreg = get_logreg(new_X_train, y_train)\n",
    "te_sc = f1_score(y_test, logreg.predict(new_X_test), average='macro')\n",
    "print('best score', round(te_sc, 4), 'with', best_num_feat, 'features')\n",
    "print(round(te_sc / baseline_te_sc * 100 - 100, 2), '%')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3) Fast Correlation-based Filter (FCBF)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "from FCBF_module import FCBF, FCBFK, FCBFiP, get_i # creds https://github.com/SantiagoEG/FCBF_module"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[10]"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fcbf = FCBF()\n",
    "fcbf.fit(np.array(X_train), np.array(y_train))\n",
    "fcbf.idx_sel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "fcbfk = FCBFK(k = len(fcbf.idx_sel))\n",
    "fcbfk.fit(np.array(X_train), np.array(y_train))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7813321277878239"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_X_train = np.array(X_train)[:, fcbfk.idx_sel]\n",
    "new_X_val =  np.array(X_val)[:, fcbfk.idx_sel]\n",
    "\n",
    "\n",
    "logreg = get_logreg(new_X_train, y_train)\n",
    "val_sc = f1_score(y_val, logreg.predict(new_X_val), average='macro')\n",
    "val_sc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[4] 0.18320977759967944\n",
      "[4] 0.18320977759967944\n",
      "[0] 0.2645675563071343\n",
      "[17] 0.1721559284026749\n"
     ]
    }
   ],
   "source": [
    "npieces = get_i(np.array(X_train).shape[1])\n",
    "\n",
    "for i in npieces:\n",
    "    fcbfip = FCBFiP(npieces=i, k=len(fcbf.idx_sel))\n",
    "    fcbfip.fit(np.array(X_train), np.array(y_train))\n",
    "    \n",
    "    new_X_train = np.array(X_train)[:, fcbfip.idx_sel]\n",
    "    new_X_val =  np.array(X_val)[:, fcbfip.idx_sel]\n",
    "    logreg = get_logreg(new_X_train, y_train)\n",
    "    val_sc = f1_score(y_val, logreg.predict(new_X_val), average='macro')\n",
    "    print(fcbfip.idx_sel, val_sc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4) Minimal-Redundancy-Maximal-Relevance (MRMR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pymrmr # creds: https://github.com/fbrundu/pymrmr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "new_train = pd.concat([y_train, X_train], axis=1)\n",
    "new_val = pd.concat([y_val, X_val], axis=1)\n",
    "new_test = pd.concat([y_test, X_test], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "max_val_sc = -1\n",
    "val_all_scores, test_all_scores = [], []\n",
    "\n",
    "for num_feat in np.arange(1, 15):\n",
    "\n",
    "    feat_names = pymrmr.mRMR(new_train, 'MID', num_feat)\n",
    "    \n",
    "    logreg = get_logreg(X_train[feat_names], y_train)\n",
    "    val_sc = f1_score(y_val, logreg.predict(X_val[feat_names]), average='macro')\n",
    "    \n",
    "    # save score and max-score \n",
    "    val_all_scores.append(val_sc)\n",
    "    if val_sc > max_val_sc:\n",
    "        max_val_sc = val_sc\n",
    "        best_num_feat = num_feat\n",
    "        \n",
    "       \n",
    "    # save test values for graph\n",
    "    te_sc = f1_score(y_test, logreg.predict(X_test[feat_names]), average='macro')\n",
    "    test_all_scores.append(te_sc)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(np.arange(1, 15), val_all_scores, label='val')\n",
    "plt.plot(np.arange(1, 15), test_all_scores, label='test')\n",
    "plt.title('MRMR')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "best score 0.9699 with 11 features\n",
      "2.11 %\n"
     ]
    }
   ],
   "source": [
    "feat_names = pymrmr.mRMR(new_train, 'MID', best_num_feat)\n",
    "    \n",
    "logreg = get_logreg(X_train[feat_names], y_train)\n",
    "te_sc = f1_score(y_test, logreg.predict(X_test[feat_names]), average='macro')\n",
    "print('best score', round(te_sc, 4), 'with', best_num_feat, 'features')\n",
    "print(round(te_sc / baseline_te_sc * 100 - 100, 2), '%')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}