mushroom-report-v1.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "np.random.seed(1337)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "DataSet Size : (8124, 23)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>class</th>\n",
       "      <th>cap-shape</th>\n",
       "      <th>cap-surface</th>\n",
       "      <th>cap-color</th>\n",
       "      <th>bruises</th>\n",
       "      <th>odor</th>\n",
       "      <th>gill-attachment</th>\n",
       "      <th>gill-spacing</th>\n",
       "      <th>gill-size</th>\n",
       "      <th>gill-color</th>\n",
       "      <th>...</th>\n",
       "      <th>stalk-surface-below-ring</th>\n",
       "      <th>stalk-color-above-ring</th>\n",
       "      <th>stalk-color-below-ring</th>\n",
       "      <th>veil-type</th>\n",
       "      <th>veil-color</th>\n",
       "      <th>ring-number</th>\n",
       "      <th>ring-type</th>\n",
       "      <th>spore-print-color</th>\n",
       "      <th>population</th>\n",
       "      <th>habitat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>p</td>\n",
       "      <td>x</td>\n",
       "      <td>s</td>\n",
       "      <td>n</td>\n",
       "      <td>t</td>\n",
       "      <td>p</td>\n",
       "      <td>f</td>\n",
       "      <td>c</td>\n",
       "      <td>n</td>\n",
       "      <td>k</td>\n",
       "      <td>...</td>\n",
       "      <td>s</td>\n",
       "      <td>w</td>\n",
       "      <td>w</td>\n",
       "      <td>p</td>\n",
       "      <td>w</td>\n",
       "      <td>o</td>\n",
       "      <td>p</td>\n",
       "      <td>k</td>\n",
       "      <td>s</td>\n",
       "      <td>u</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>e</td>\n",
       "      <td>x</td>\n",
       "      <td>s</td>\n",
       "      <td>y</td>\n",
       "      <td>t</td>\n",
       "      <td>a</td>\n",
       "      <td>f</td>\n",
       "      <td>c</td>\n",
       "      <td>b</td>\n",
       "      <td>k</td>\n",
       "      <td>...</td>\n",
       "      <td>s</td>\n",
       "      <td>w</td>\n",
       "      <td>w</td>\n",
       "      <td>p</td>\n",
       "      <td>w</td>\n",
       "      <td>o</td>\n",
       "      <td>p</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>g</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>e</td>\n",
       "      <td>b</td>\n",
       "      <td>s</td>\n",
       "      <td>w</td>\n",
       "      <td>t</td>\n",
       "      <td>l</td>\n",
       "      <td>f</td>\n",
       "      <td>c</td>\n",
       "      <td>b</td>\n",
       "      <td>n</td>\n",
       "      <td>...</td>\n",
       "      <td>s</td>\n",
       "      <td>w</td>\n",
       "      <td>w</td>\n",
       "      <td>p</td>\n",
       "      <td>w</td>\n",
       "      <td>o</td>\n",
       "      <td>p</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>m</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>p</td>\n",
       "      <td>x</td>\n",
       "      <td>y</td>\n",
       "      <td>w</td>\n",
       "      <td>t</td>\n",
       "      <td>p</td>\n",
       "      <td>f</td>\n",
       "      <td>c</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>...</td>\n",
       "      <td>s</td>\n",
       "      <td>w</td>\n",
       "      <td>w</td>\n",
       "      <td>p</td>\n",
       "      <td>w</td>\n",
       "      <td>o</td>\n",
       "      <td>p</td>\n",
       "      <td>k</td>\n",
       "      <td>s</td>\n",
       "      <td>u</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>e</td>\n",
       "      <td>x</td>\n",
       "      <td>s</td>\n",
       "      <td>g</td>\n",
       "      <td>f</td>\n",
       "      <td>n</td>\n",
       "      <td>f</td>\n",
       "      <td>w</td>\n",
       "      <td>b</td>\n",
       "      <td>k</td>\n",
       "      <td>...</td>\n",
       "      <td>s</td>\n",
       "      <td>w</td>\n",
       "      <td>w</td>\n",
       "      <td>p</td>\n",
       "      <td>w</td>\n",
       "      <td>o</td>\n",
       "      <td>e</td>\n",
       "      <td>n</td>\n",
       "      <td>a</td>\n",
       "      <td>g</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "  class cap-shape cap-surface cap-color bruises odor gill-attachment  \\\n",
       "0     p         x           s         n       t    p               f   \n",
       "1     e         x           s         y       t    a               f   \n",
       "2     e         b           s         w       t    l               f   \n",
       "3     p         x           y         w       t    p               f   \n",
       "4     e         x           s         g       f    n               f   \n",
       "\n",
       "  gill-spacing gill-size gill-color   ...   stalk-surface-below-ring  \\\n",
       "0            c         n          k   ...                          s   \n",
       "1            c         b          k   ...                          s   \n",
       "2            c         b          n   ...                          s   \n",
       "3            c         n          n   ...                          s   \n",
       "4            w         b          k   ...                          s   \n",
       "\n",
       "  stalk-color-above-ring stalk-color-below-ring veil-type veil-color  \\\n",
       "0                      w                      w         p          w   \n",
       "1                      w                      w         p          w   \n",
       "2                      w                      w         p          w   \n",
       "3                      w                      w         p          w   \n",
       "4                      w                      w         p          w   \n",
       "\n",
       "  ring-number ring-type spore-print-color population habitat  \n",
       "0           o         p                 k          s       u  \n",
       "1           o         p                 n          n       g  \n",
       "2           o         p                 n          n       m  \n",
       "3           o         p                 k          s       u  \n",
       "4           o         e                 n          a       g  \n",
       "\n",
       "[5 rows x 23 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset = \"./data/mushrooms.csv\"\n",
    "\n",
    "data = pd.read_csv(dataset)  # read DataSet\n",
    "\n",
    "print(\"DataSet Size : {}\".format(data.shape))\n",
    "data.head(5)  # top 5 contents"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "poison : 3916 edible : 4208\n"
     ]
    }
   ],
   "source": [
    "p = data.loc[data['class'] == 'p']\n",
    "e = data.loc[data['class'] == 'e']\n",
    "\n",
    "print(\"poison : {} edible : {}\".format(len(p), len(e)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def draw_bar(p, e, name):\n",
    "    classes = ['poison', 'edible']\n",
    "    counts = [len(p), len(e)]\n",
    "\n",
    "    fig = plt.figure(figsize=(5, 5))\n",
    "    ax = fig.add_subplot(111)\n",
    "\n",
    "    rects = plt.bar(np.arange(2), counts, align='center', width=.75)\n",
    "    plt.xticks(np.arange(2), classes)\n",
    "\n",
    "    for i, rect in enumerate(rects):\n",
    "        ax.text(rect.get_x() + rect.get_width() / 2.0, 0.95 * rect.get_height(), str(counts[i]) + \" ({:.1f}%)\".format(counts[i] / 8124 * 100), ha='center')\n",
    "\n",
    "    plt.xlabel(name)\n",
    "    plt.ylabel('Counts')\n",
    "\n",
    "draw_bar(p, e, \"Classes\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "stalk-root    True\n",
       "dtype: bool"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res = data.isin(['?']).any()\n",
    "res.loc[res == True]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2480\n"
     ]
    }
   ],
   "source": [
    "print(len(data.loc[data['stalk-root'] == '?']))\n",
    "data.drop('stalk-root', axis=1, inplace=True)  # drop stalk-shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "n = 5\n",
    "top_p = p.sample(n)\n",
    "top_e = e.sample(n)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>class</th>\n",
       "      <th>cap-shape</th>\n",
       "      <th>cap-surface</th>\n",
       "      <th>cap-color</th>\n",
       "      <th>bruises</th>\n",
       "      <th>odor</th>\n",
       "      <th>gill-attachment</th>\n",
       "      <th>gill-spacing</th>\n",
       "      <th>gill-size</th>\n",
       "      <th>gill-color</th>\n",
       "      <th>...</th>\n",
       "      <th>stalk-surface-below-ring</th>\n",
       "      <th>stalk-color-above-ring</th>\n",
       "      <th>stalk-color-below-ring</th>\n",
       "      <th>veil-type</th>\n",
       "      <th>veil-color</th>\n",
       "      <th>ring-number</th>\n",
       "      <th>ring-type</th>\n",
       "      <th>spore-print-color</th>\n",
       "      <th>population</th>\n",
       "      <th>habitat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4487</th>\n",
       "      <td>p</td>\n",
       "      <td>x</td>\n",
       "      <td>y</td>\n",
       "      <td>g</td>\n",
       "      <td>f</td>\n",
       "      <td>f</td>\n",
       "      <td>f</td>\n",
       "      <td>c</td>\n",
       "      <td>b</td>\n",
       "      <td>p</td>\n",
       "      <td>...</td>\n",
       "      <td>k</td>\n",
       "      <td>p</td>\n",
       "      <td>p</td>\n",
       "      <td>p</td>\n",
       "      <td>w</td>\n",
       "      <td>o</td>\n",
       "      <td>l</td>\n",
       "      <td>h</td>\n",
       "      <td>v</td>\n",
       "      <td>d</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5575</th>\n",
       "      <td>p</td>\n",
       "      <td>x</td>\n",
       "      <td>s</td>\n",
       "      <td>b</td>\n",
       "      <td>t</td>\n",
       "      <td>f</td>\n",
       "      <td>f</td>\n",
       "      <td>c</td>\n",
       "      <td>b</td>\n",
       "      <td>h</td>\n",
       "      <td>...</td>\n",
       "      <td>s</td>\n",
       "      <td>w</td>\n",
       "      <td>w</td>\n",
       "      <td>p</td>\n",
       "      <td>w</td>\n",
       "      <td>o</td>\n",
       "      <td>p</td>\n",
       "      <td>h</td>\n",
       "      <td>v</td>\n",
       "      <td>g</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6462</th>\n",
       "      <td>p</td>\n",
       "      <td>f</td>\n",
       "      <td>y</td>\n",
       "      <td>e</td>\n",
       "      <td>f</td>\n",
       "      <td>s</td>\n",
       "      <td>f</td>\n",
       "      <td>c</td>\n",
       "      <td>n</td>\n",
       "      <td>b</td>\n",
       "      <td>...</td>\n",
       "      <td>k</td>\n",
       "      <td>w</td>\n",
       "      <td>p</td>\n",
       "      <td>p</td>\n",
       "      <td>w</td>\n",
       "      <td>o</td>\n",
       "      <td>e</td>\n",
       "      <td>w</td>\n",
       "      <td>v</td>\n",
       "      <td>p</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7065</th>\n",
       "      <td>p</td>\n",
       "      <td>x</td>\n",
       "      <td>y</td>\n",
       "      <td>n</td>\n",
       "      <td>f</td>\n",
       "      <td>m</td>\n",
       "      <td>a</td>\n",
       "      <td>c</td>\n",
       "      <td>b</td>\n",
       "      <td>w</td>\n",
       "      <td>...</td>\n",
       "      <td>y</td>\n",
       "      <td>c</td>\n",
       "      <td>c</td>\n",
       "      <td>p</td>\n",
       "      <td>w</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>w</td>\n",
       "      <td>c</td>\n",
       "      <td>d</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5981</th>\n",
       "      <td>p</td>\n",
       "      <td>f</td>\n",
       "      <td>s</td>\n",
       "      <td>w</td>\n",
       "      <td>t</td>\n",
       "      <td>f</td>\n",
       "      <td>f</td>\n",
       "      <td>c</td>\n",
       "      <td>b</td>\n",
       "      <td>w</td>\n",
       "      <td>...</td>\n",
       "      <td>s</td>\n",
       "      <td>w</td>\n",
       "      <td>w</td>\n",
       "      <td>p</td>\n",
       "      <td>w</td>\n",
       "      <td>o</td>\n",
       "      <td>p</td>\n",
       "      <td>h</td>\n",
       "      <td>s</td>\n",
       "      <td>u</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     class cap-shape cap-surface cap-color bruises odor gill-attachment  \\\n",
       "4487     p         x           y         g       f    f               f   \n",
       "5575     p         x           s         b       t    f               f   \n",
       "6462     p         f           y         e       f    s               f   \n",
       "7065     p         x           y         n       f    m               a   \n",
       "5981     p         f           s         w       t    f               f   \n",
       "\n",
       "     gill-spacing gill-size gill-color   ...   stalk-surface-below-ring  \\\n",
       "4487            c         b          p   ...                          k   \n",
       "5575            c         b          h   ...                          s   \n",
       "6462            c         n          b   ...                          k   \n",
       "7065            c         b          w   ...                          y   \n",
       "5981            c         b          w   ...                          s   \n",
       "\n",
       "     stalk-color-above-ring stalk-color-below-ring veil-type veil-color  \\\n",
       "4487                      p                      p         p          w   \n",
       "5575                      w                      w         p          w   \n",
       "6462                      w                      p         p          w   \n",
       "7065                      c                      c         p          w   \n",
       "5981                      w                      w         p          w   \n",
       "\n",
       "     ring-number ring-type spore-print-color population habitat  \n",
       "4487           o         l                 h          v       d  \n",
       "5575           o         p                 h          v       g  \n",
       "6462           o         e                 w          v       p  \n",
       "7065           n         n                 w          c       d  \n",
       "5981           o         p                 h          s       u  \n",
       "\n",
       "[5 rows x 23 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_p"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>class</th>\n",
       "      <th>cap-shape</th>\n",
       "      <th>cap-surface</th>\n",
       "      <th>cap-color</th>\n",
       "      <th>bruises</th>\n",
       "      <th>odor</th>\n",
       "      <th>gill-attachment</th>\n",
       "      <th>gill-spacing</th>\n",
       "      <th>gill-size</th>\n",
       "      <th>gill-color</th>\n",
       "      <th>...</th>\n",
       "      <th>stalk-surface-below-ring</th>\n",
       "      <th>stalk-color-above-ring</th>\n",
       "      <th>stalk-color-below-ring</th>\n",
       "      <th>veil-type</th>\n",
       "      <th>veil-color</th>\n",
       "      <th>ring-number</th>\n",
       "      <th>ring-type</th>\n",
       "      <th>spore-print-color</th>\n",
       "      <th>population</th>\n",
       "      <th>habitat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2523</th>\n",
       "      <td>e</td>\n",
       "      <td>x</td>\n",
       "      <td>y</td>\n",
       "      <td>n</td>\n",
       "      <td>t</td>\n",
       "      <td>n</td>\n",
       "      <td>f</td>\n",
       "      <td>c</td>\n",
       "      <td>b</td>\n",
       "      <td>n</td>\n",
       "      <td>...</td>\n",
       "      <td>s</td>\n",
       "      <td>g</td>\n",
       "      <td>w</td>\n",
       "      <td>p</td>\n",
       "      <td>w</td>\n",
       "      <td>o</td>\n",
       "      <td>p</td>\n",
       "      <td>k</td>\n",
       "      <td>y</td>\n",
       "      <td>d</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7744</th>\n",
       "      <td>e</td>\n",
       "      <td>k</td>\n",
       "      <td>f</td>\n",
       "      <td>w</td>\n",
       "      <td>f</td>\n",
       "      <td>n</td>\n",
       "      <td>f</td>\n",
       "      <td>w</td>\n",
       "      <td>b</td>\n",
       "      <td>w</td>\n",
       "      <td>...</td>\n",
       "      <td>s</td>\n",
       "      <td>w</td>\n",
       "      <td>w</td>\n",
       "      <td>p</td>\n",
       "      <td>w</td>\n",
       "      <td>t</td>\n",
       "      <td>p</td>\n",
       "      <td>w</td>\n",
       "      <td>n</td>\n",
       "      <td>g</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1849</th>\n",
       "      <td>e</td>\n",
       "      <td>x</td>\n",
       "      <td>f</td>\n",
       "      <td>n</td>\n",
       "      <td>t</td>\n",
       "      <td>n</td>\n",
       "      <td>f</td>\n",
       "      <td>c</td>\n",
       "      <td>b</td>\n",
       "      <td>w</td>\n",
       "      <td>...</td>\n",
       "      <td>s</td>\n",
       "      <td>g</td>\n",
       "      <td>w</td>\n",
       "      <td>p</td>\n",
       "      <td>w</td>\n",
       "      <td>o</td>\n",
       "      <td>p</td>\n",
       "      <td>k</td>\n",
       "      <td>v</td>\n",
       "      <td>d</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3517</th>\n",
       "      <td>e</td>\n",
       "      <td>f</td>\n",
       "      <td>y</td>\n",
       "      <td>n</td>\n",
       "      <td>t</td>\n",
       "      <td>n</td>\n",
       "      <td>f</td>\n",
       "      <td>c</td>\n",
       "      <td>b</td>\n",
       "      <td>u</td>\n",
       "      <td>...</td>\n",
       "      <td>s</td>\n",
       "      <td>g</td>\n",
       "      <td>w</td>\n",
       "      <td>p</td>\n",
       "      <td>w</td>\n",
       "      <td>o</td>\n",
       "      <td>p</td>\n",
       "      <td>k</td>\n",
       "      <td>y</td>\n",
       "      <td>d</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3260</th>\n",
       "      <td>e</td>\n",
       "      <td>f</td>\n",
       "      <td>y</td>\n",
       "      <td>g</td>\n",
       "      <td>t</td>\n",
       "      <td>n</td>\n",
       "      <td>f</td>\n",
       "      <td>c</td>\n",
       "      <td>b</td>\n",
       "      <td>p</td>\n",
       "      <td>...</td>\n",
       "      <td>s</td>\n",
       "      <td>g</td>\n",
       "      <td>p</td>\n",
       "      <td>p</td>\n",
       "      <td>w</td>\n",
       "      <td>o</td>\n",
       "      <td>p</td>\n",
       "      <td>k</td>\n",
       "      <td>y</td>\n",
       "      <td>d</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     class cap-shape cap-surface cap-color bruises odor gill-attachment  \\\n",
       "2523     e         x           y         n       t    n               f   \n",
       "7744     e         k           f         w       f    n               f   \n",
       "1849     e         x           f         n       t    n               f   \n",
       "3517     e         f           y         n       t    n               f   \n",
       "3260     e         f           y         g       t    n               f   \n",
       "\n",
       "     gill-spacing gill-size gill-color   ...   stalk-surface-below-ring  \\\n",
       "2523            c         b          n   ...                          s   \n",
       "7744            w         b          w   ...                          s   \n",
       "1849            c         b          w   ...                          s   \n",
       "3517            c         b          u   ...                          s   \n",
       "3260            c         b          p   ...                          s   \n",
       "\n",
       "     stalk-color-above-ring stalk-color-below-ring veil-type veil-color  \\\n",
       "2523                      g                      w         p          w   \n",
       "7744                      w                      w         p          w   \n",
       "1849                      g                      w         p          w   \n",
       "3517                      g                      w         p          w   \n",
       "3260                      g                      p         p          w   \n",
       "\n",
       "     ring-number ring-type spore-print-color population habitat  \n",
       "2523           o         p                 k          y       d  \n",
       "7744           t         p                 w          n       g  \n",
       "1849           o         p                 k          v       d  \n",
       "3517           o         p                 k          y       d  \n",
       "3260           o         p                 k          y       d  \n",
       "\n",
       "[5 rows x 23 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_e"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def pd_to_docx(df, name):\n",
    "    import docx\n",
    "    \n",
    "    doc = docx.Document()\n",
    "\n",
    "    t = doc.add_table(df.shape[0] + 1, df.shape[1])\n",
    "\n",
    "    for j in range(df.shape[-1]):\n",
    "        t.cell(0, j).text = df.columns[j]\n",
    "\n",
    "    for i in range(df.shape[0]):\n",
    "        for j in range(df.shape[-1]):\n",
    "            t.cell(i + 1, j).text = str(df.values[i,j])\n",
    "\n",
    "    doc.save(name)\n",
    "\n",
    "# pd_to_docx(top_e, './top-%d-edible-table.docx' % n)\n",
    "# pd_to_docx(top_p, './top-%d-poison-table.docx' % n)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Column : cap-shape\n",
      "[*] key x : 1948/1708 (0.533/0.467) (total 3656)\n",
      "[*] key b : 404/48 (0.894/0.106) (total 452)\n",
      "[*] key s : 32/0 (1.000/0.000) (total 32)\n",
      "[*] key f : 1596/1556 (0.506/0.494) (total 3152)\n",
      "[*] key k : 228/600 (0.275/0.725) (total 828)\n",
      "[*] key c : 0/4 (0.000/1.000) (total 4)\n",
      "Column : cap-surface\n",
      "[*] key s : 1144/1412 (0.448/0.552) (total 2556)\n",
      "[*] key y : 1504/1740 (0.464/0.536) (total 3244)\n",
      "[*] key f : 1560/760 (0.672/0.328) (total 2320)\n",
      "[*] key g : 0/4 (0.000/1.000) (total 4)\n",
      "Column : cap-color\n",
      "[*] key n : 1264/1020 (0.553/0.447) (total 2284)\n",
      "[*] key y : 400/672 (0.373/0.627) (total 1072)\n",
      "[*] key w : 720/320 (0.692/0.308) (total 1040)\n",
      "[*] key g : 1032/808 (0.561/0.439) (total 1840)\n",
      "[*] key e : 624/876 (0.416/0.584) (total 1500)\n",
      "[*] key p : 56/88 (0.389/0.611) (total 144)\n",
      "[*] key b : 48/120 (0.286/0.714) (total 168)\n",
      "[*] key u : 16/0 (1.000/0.000) (total 16)\n",
      "[*] key c : 32/12 (0.727/0.273) (total 44)\n",
      "[*] key r : 16/0 (1.000/0.000) (total 16)\n",
      "Column : bruises\n",
      "[*] key t : 2752/624 (0.815/0.185) (total 3376)\n",
      "[*] key f : 1456/3292 (0.307/0.693) (total 4748)\n",
      "Column : odor\n",
      "[*] key p : 0/256 (0.000/1.000) (total 256)\n",
      "[*] key a : 400/0 (1.000/0.000) (total 400)\n",
      "[*] key l : 400/0 (1.000/0.000) (total 400)\n",
      "[*] key n : 3408/120 (0.966/0.034) (total 3528)\n",
      "[*] key f : 0/2160 (0.000/1.000) (total 2160)\n",
      "[*] key c : 0/192 (0.000/1.000) (total 192)\n",
      "[*] key y : 0/576 (0.000/1.000) (total 576)\n",
      "[*] key s : 0/576 (0.000/1.000) (total 576)\n",
      "[*] key m : 0/36 (0.000/1.000) (total 36)\n",
      "Column : gill-attachment\n",
      "[*] key f : 4016/3898 (0.507/0.493) (total 7914)\n",
      "[*] key a : 192/18 (0.914/0.086) (total 210)\n",
      "Column : gill-spacing\n",
      "[*] key c : 3008/3804 (0.442/0.558) (total 6812)\n",
      "[*] key w : 1200/112 (0.915/0.085) (total 1312)\n",
      "Column : gill-size\n",
      "[*] key n : 288/2224 (0.115/0.885) (total 2512)\n",
      "[*] key b : 3920/1692 (0.699/0.301) (total 5612)\n",
      "Column : gill-color\n",
      "[*] key k : 344/64 (0.843/0.157) (total 408)\n",
      "[*] key n : 936/112 (0.893/0.107) (total 1048)\n",
      "[*] key g : 248/504 (0.330/0.670) (total 752)\n",
      "[*] key p : 852/640 (0.571/0.429) (total 1492)\n",
      "[*] key w : 956/246 (0.795/0.205) (total 1202)\n",
      "[*] key h : 204/528 (0.279/0.721) (total 732)\n",
      "[*] key u : 444/48 (0.902/0.098) (total 492)\n",
      "[*] key e : 96/0 (1.000/0.000) (total 96)\n",
      "[*] key b : 0/1728 (0.000/1.000) (total 1728)\n",
      "[*] key r : 0/24 (0.000/1.000) (total 24)\n",
      "[*] key y : 64/22 (0.744/0.256) (total 86)\n",
      "[*] key o : 64/0 (1.000/0.000) (total 64)\n",
      "Column : stalk-shape\n",
      "[*] key e : 1616/1900 (0.460/0.540) (total 3516)\n",
      "[*] key t : 2592/2016 (0.562/0.438) (total 4608)\n",
      "Column : stalk-surface-above-ring\n",
      "[*] key s : 3640/1536 (0.703/0.297) (total 5176)\n",
      "[*] key f : 408/144 (0.739/0.261) (total 552)\n",
      "[*] key k : 144/2228 (0.061/0.939) (total 2372)\n",
      "[*] key y : 16/8 (0.667/0.333) (total 24)\n",
      "Column : stalk-surface-below-ring\n",
      "[*] key s : 3400/1536 (0.689/0.311) (total 4936)\n",
      "[*] key f : 456/144 (0.760/0.240) (total 600)\n",
      "[*] key y : 208/76 (0.732/0.268) (total 284)\n",
      "[*] key k : 144/2160 (0.062/0.938) (total 2304)\n",
      "Column : stalk-color-above-ring\n",
      "[*] key w : 2752/1712 (0.616/0.384) (total 4464)\n",
      "[*] key g : 576/0 (1.000/0.000) (total 576)\n",
      "[*] key p : 576/1296 (0.308/0.692) (total 1872)\n",
      "[*] key n : 16/432 (0.036/0.964) (total 448)\n",
      "[*] key b : 0/432 (0.000/1.000) (total 432)\n",
      "[*] key e : 96/0 (1.000/0.000) (total 96)\n",
      "[*] key o : 192/0 (1.000/0.000) (total 192)\n",
      "[*] key c : 0/36 (0.000/1.000) (total 36)\n",
      "[*] key y : 0/8 (0.000/1.000) (total 8)\n",
      "Column : stalk-color-below-ring\n",
      "[*] key w : 2704/1680 (0.617/0.383) (total 4384)\n",
      "[*] key p : 576/1296 (0.308/0.692) (total 1872)\n",
      "[*] key g : 576/0 (1.000/0.000) (total 576)\n",
      "[*] key b : 0/432 (0.000/1.000) (total 432)\n",
      "[*] key n : 64/448 (0.125/0.875) (total 512)\n",
      "[*] key e : 96/0 (1.000/0.000) (total 96)\n",
      "[*] key y : 0/24 (0.000/1.000) (total 24)\n",
      "[*] key o : 192/0 (1.000/0.000) (total 192)\n",
      "[*] key c : 0/36 (0.000/1.000) (total 36)\n",
      "Column : veil-type\n",
      "[*] key p : 4208/3916 (0.518/0.482) (total 8124)\n",
      "Column : veil-color\n",
      "[*] key w : 4016/3908 (0.507/0.493) (total 7924)\n",
      "[*] key n : 96/0 (1.000/0.000) (total 96)\n",
      "[*] key o : 96/0 (1.000/0.000) (total 96)\n",
      "[*] key y : 0/8 (0.000/1.000) (total 8)\n",
      "Column : ring-number\n",
      "[*] key o : 3680/3808 (0.491/0.509) (total 7488)\n",
      "[*] key t : 528/72 (0.880/0.120) (total 600)\n",
      "[*] key n : 0/36 (0.000/1.000) (total 36)\n",
      "Column : ring-type\n",
      "[*] key p : 3152/816 (0.794/0.206) (total 3968)\n",
      "[*] key e : 1008/1768 (0.363/0.637) (total 2776)\n",
      "[*] key l : 0/1296 (0.000/1.000) (total 1296)\n",
      "[*] key f : 48/0 (1.000/0.000) (total 48)\n",
      "[*] key n : 0/36 (0.000/1.000) (total 36)\n",
      "Column : spore-print-color\n",
      "[*] key k : 1648/224 (0.880/0.120) (total 1872)\n",
      "[*] key n : 1744/224 (0.886/0.114) (total 1968)\n",
      "[*] key u : 48/0 (1.000/0.000) (total 48)\n",
      "[*] key h : 48/1584 (0.029/0.971) (total 1632)\n",
      "[*] key w : 576/1812 (0.241/0.759) (total 2388)\n",
      "[*] key r : 0/72 (0.000/1.000) (total 72)\n",
      "[*] key o : 48/0 (1.000/0.000) (total 48)\n",
      "[*] key y : 48/0 (1.000/0.000) (total 48)\n",
      "[*] key b : 48/0 (1.000/0.000) (total 48)\n",
      "Column : population\n",
      "[*] key s : 880/368 (0.705/0.295) (total 1248)\n",
      "[*] key n : 400/0 (1.000/0.000) (total 400)\n",
      "[*] key a : 384/0 (1.000/0.000) (total 384)\n",
      "[*] key v : 1192/2848 (0.295/0.705) (total 4040)\n",
      "[*] key y : 1064/648 (0.621/0.379) (total 1712)\n",
      "[*] key c : 288/52 (0.847/0.153) (total 340)\n",
      "Column : habitat\n",
      "[*] key u : 96/272 (0.261/0.739) (total 368)\n",
      "[*] key g : 1408/740 (0.655/0.345) (total 2148)\n",
      "[*] key m : 256/36 (0.877/0.123) (total 292)\n",
      "[*] key d : 1880/1268 (0.597/0.403) (total 3148)\n",
      "[*] key p : 136/1008 (0.119/0.881) (total 1144)\n",
      "[*] key w : 192/0 (1.000/0.000) (total 192)\n",
      "[*] key l : 240/592 (0.288/0.712) (total 832)\n"
     ]
    }
   ],
   "source": [
    "for c in data.columns[1:]:\n",
    "    print(\"Column : %s\" % c)\n",
    "    keys = data[c].unique()\n",
    "    for k in keys:\n",
    "        d = pd.concat([data['class'], data[c]], axis=1)\n",
    "        e = d.loc[(d['class'] == 'e') & (data[c] == k)]\n",
    "        p = d.loc[(d['class'] == 'p') & (data[c] == k)]\n",
    "        e_len, p_len = len(e), len(p)\n",
    "        total = e_len + p_len\n",
    "        print(\"[*] key %s : %d/%d (%.3f/%.3f) (total %d)\" % \n",
    "              (k, e_len, p_len, (e_len/total), (p_len/total), total))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Data Labeling\n",
    "from sklearn.preprocessing import LabelEncoder, OneHotEncoder\n",
    "\n",
    "le = LabelEncoder()  # robust way\n",
    "for i in data.columns:\n",
    "    data[i] = le.fit_transform(data[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEUCAYAAADEGSquAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAGVJJREFUeJzt3X+cXXV95/HXOxOSEEmCkqCYRIIWrIiu2pG1stsi1i5oha1rLVhbbV1Yt1Lrj9pi2xVr3cf2oe36UJf+SK1SXZUi7bbRojy2/qptxTIoPww/s5QfCSBBMBDJr0k++8e9c3ozTGbuJHPmZpLX8/GYx9zzPd9z7udObuY93/M995xUFZIkAcwbdAGSpIOHoSBJahgKkqSGoSBJahgKkqSGoSBJahgKak2S65IcuY91dyY5ZZr7+2qSO7r7vS7JL85Mpc3+z07yge7j05OMdB+vSfLgfuxvTZILxrW9NcmxB1jnpUkuPJB9HODzvyfJgkE9v9plKKg1VfW8qto2w7t9S3e/z6uqj8/kjqtqXVW9cwZ3uQa4YFzbW4EDCoWDwMWAoXCIMhTUmiSV5Kju43+f5MYk/5zkw0C67cd2Rw3D3eXXJ/mHJPNbqmlBkrVJbus+z/9KckV33RvGHk9zn59KMtJ9ff8nyRO7qy4BTu6Oaq5I8lvAU4Erum0nJ3lpkm8k+XZ3+3N79rsyyV8muaH79a6epz0lyZeT3J7kE0nGfp6XJvmT7rq7knwwyRlJvt79Of9qz/6fmeQLSa5Jcn3vyKv7b/eb3XV3JPlP3fZLul3+qfsajp7uz0sHuaryy69WvoACjgIWApuA07vtr+muO6W7fDpwG/Ai4G5g9T7291XgZuBG4H8DK/fRbxi4ch/rfgX4IjAfWARcDVzRXfeGnsenAyPdx2uAByd5nct7Hr8P+L3x++hZf+fY6+4uPxEY6j5+MrAReGJ3+SvAO8c/D3Ap8A/d+hcA64GXjVu3EFgMPAB8jM4fgCuBrd1/k/nAtcAPd7dbAtzas1zAhd3HpwGbxv+7Dvr95Vc7X44UNBueCTxWVV8FqKrLgS1jK7vtn6bzy+zNVXXPPvbz81X1LOB5wC3AX0zUqapGqurl+9jHS4BPVtVoVW0HPjP9l/M4v5Dk2iQ3Aq/t1tevFXRGDt8BrgKeBDyzO8J6MfDBsY5V1Tuv8ddVtb2qdgLfAp4xbt2OqnqMzi/6K6tqT1VtAh4GVgEnAc8CLktyHfB1OkHyrJ79XNb9fjXw1CSLpvG6NEe1MkSXxkkffZ4PbKbzC2tCY2FRVbuTfAh4T5J5VbVnmrXs9wW/uodYxg7BfIDOyOa/Ai+uqs1JXsvj5xEm80fAOuBVVVVJbqMzApjK9p7Hu9n7//L4dRP1LTqjn8kCbDs0P2/w98VhwZGCZsMtwJFJfgwgyauBZWMrk7yNzmGQFwAXJXncL6ok85M8uafpPODGaQYCdA7JvK67v0XAz05n46r6eP3rRPengKPpjHq+l2Qh8Es93R+h53Xuo+1o4M5uILwM+KHu82wF/gl421jHJMunU+sUbgUeS/LzPfv/4SRL+9j2UR7/unSIMBTUuqraQeeX+CVJ/pnOMf+7AZKcCrwF+IWqug/4z3QOaSwZt5uFwN92J1xvBF4FnMsEkgwnuXIf5fwxcB+d4/Cfp3Ncfcs++vbjC8D/oxN8X6BzKGfMDcCtSb7TM4H9YeDjYxPNwEXA7yf5BvDq7jZjXgec1t3+euCNB1DnXqpqFHglcG73Z7oe+EP6O6voD4AvO9F8aEqVl87W4SXJkqp6tPuX/Trgs1X10UHXJR0MPEaow9HfdQNhEfB3dM7YkYQjBUlSD+cUJEkNQ0GS1JhzcwrLly+vNWvWDLoMSZpTrr322gerasVU/eZcKKxZs4aRkZFBlyFJc0qSu/rp5+EjSVLDUJAkNQwFSVLDUJAkNQwFSVLDUJAkNQwFSVLDUJAkNVr78FqSjwE/BTxQVadMsD7Ah4CXA48Bb6iqb43vp8e78W//YNAlPM7RJ57B6pOeP+gyNGC3XfNFdjywftBlPM5zXvGOQZcwZ7Q5UrgUOHOS9WcBJ3a/LqBzW0JN4cYv/smgS5jQ92//8qBL0EHgYAwEgH+54SuDLmHOaG2kUFV/n2TNJF3OAT5RnWt3X53k6CTHde++1ZotW7Zw2223tfkUUzr22GOZP38/f/S7t85sMTNo06ZN+7Xd6OgoDzzwwAxXM30nnXQSy5YN7i6TB8N7E/b//blzx7YWqpkZW+/5FpuOOWm/tj0Y3p+z+d4c5JzCSuCenuWN3bbHSXJBkpEkI5s3b56V4iRNz/wjFg66BM2AQV4QLxO0TXjHn6paC6wFGB4ePqC7Ai1btowXvvCFB7KLgTrmmDdz+5cuGXQZE1q5csJM78vxxx8/g5XMTXP9vQnw/Rum7jMIzz7rbcybt/9/Ax9O789BjhQ2Aqt7llcB9w6oljlj0aJFB9+k2VFPO/hq0kB03gdHDbqMvTznFe84oEA43AxypLAOuDDJZcC/Bba0PZ9wKPGXsA5Wz3nFfxl0CToAbZ6S+hngdGB5ko3AxcARAFX1x8CVdE5H3UDnlNRfbKsWSVJ/2jz76Lwp1hfw5raeX5I0fR5okyQ1DAVJUsNQkCQ1DAVJUsNQkCQ1DAVJUsNQkCQ1DAVJUsNQkCQ1DAVJUsNQkCQ1DAVJUsNQkCQ1DAVJUsNQkCQ1DAVJUsNQkCQ1DAVJUsNQkCQ1DAVJUsNQkCQ1DAVJUsNQkCQ1DAVJUsNQkCQ1DAVJUsNQkCQ1DAVJUsNQkCQ1Wg2FJGcmuTXJhiQXTbD+aUm+kuTbSW5I8vI265EkTa61UEgyBFwCnAWcDJyX5ORx3X4buLyqng+cC/xhW/VIkqbW5kjhVGBDVd1RVTuBy4BzxvUpYGn38TLg3hbrkSRNoc1QWAnc07O8sdvW6z3A65JsBK4EfmWiHSW5IMlIkpHNmze3UaskiXZDIRO01bjl84BLq2oV8HLgk0keV1NVra2q4aoaXrFiRQulSpKg3VDYCKzuWV7F4w8PvRG4HKCqvgEsApa3WJMkaRJthsI1wIlJTkiygM5E8rpxfe4GXgqQ5Fl0QsHjQ5I0IK2FQlWNAhcCVwE30znLaH2S9yY5u9vtHcD5Sa4HPgO8oarGH2KSJM2S+W3uvKqupDOB3Nv27p7HNwGntVmDJKl/fqJZktQwFCRJDUNBktQwFCRJDUNBktQwFCRJDUNBktQwFCRJDUNBktQwFCRJDUNBktQwFCRJDUNBktQwFCRJDUNBktQwFCRJjb5CIcnjboQzUZskaW7rd6TwkT7bJElz2KS340zyo8CLgRVJ3t6zaikw1GZhkqTZN9U9mhcAR3X7LelpfwR4dVtFSZIGY9JQqKqvAV9LcmlV3TVLNUmSBmSqkcKYhUnWAmt6t6mqM9ooSpI0GP2GwmeBPwY+CuxurxxJ0iD1GwqjVfVHrVYiSRq4fk9J/VySX05yXJInjX21Wpkkadb1O1J4fff7O3vaCnj6zJYjSRqkvkKhqk5ouxBJ0uD1e5mLxUl+u3sGEklOTPJT7ZYmSZpt/c4pfBzYSefTzQAbgfe1UpEkaWD6DYVnVNX7gV0AVbUNSGtVSZIGot9Q2JnkSDqTyyR5BrBjqo2SnJnk1iQbkly0jz6vSXJTkvVJPt135ZKkGdfv2UcXA18EVif5FHAa8IbJNkgyBFwCvIzO4aZrkqyrqpt6+pwIvAs4raoeTnLs9F+CJGmm9Hv20f9N8i3gRXQOG/1qVT04xWanAhuq6g6AJJcB5wA39fQ5H7ikqh7uPs8D06xfkjSDpnPntZV0Lpe9APixJK/qo/89Pcsbu229TgJOSvKPSa5OcuZEO0pyQZKRJCObN2+eRsmSpOnoa6SQ5GPAc4H1wJ5ucwF/NdlmE7TVBM9/InA6sAr4epJTqur7e21UtRZYCzA8PDx+H5KkGdLvnMKLqurkae57I7C6Z3kVcO8Efa6uql3AvyS5lU5IXDPN55IkzYB+Dx99I8l0Q+Ea4MQkJyRZAJwLrBvX56+BlwAkWU7ncNId03weSdIM6Xek8Od0guF+OqeiBqiqeu6+Nqiq0SQXAlfRmYv4WFWtT/JeYKSq1nXX/WSSm+hckvudVfW9A3g9kqQDkKqpD9En2QC8HbiRf51TYBB3YxseHq6RkZHZflpJmtOSXFtVw1P163ekcHf3L3tJ0iGs31C4pftp48/R80nmqprs7CNJ0hzTbygcSScMfrKnbapTUiVJc0y/n2j+xbYLkSQNXr8fXltB55IUa3q3qapfaqcsSdIg9Hv46G+ArwN/R+fUUUnSIajfUFhcVb/RaiWSpIHr9xPNn0/y8lYrkSQN3KQjhSSP0jnLKMBvJtlB5+5rY59oXtp+iZKk2TJpKFTVktkqRJI0eH0dPkry00mW9SwfneQ/tleWJGkQ+p1TuLiqtowtdO93cHE7JUmSBqXfUJioX79nLkmS5oh+Q2Ekyf9M8owkT0/yQeDaNguTJM2+fkPhV4CdwF8AnwW2A29uqyhJ0mD0e+2jHwAXtVyLJGnApnPto18Hng0sGmuvqjNaqkuSNAD9Hj76FHALcALwO8CddO7BLEk6hPQbCsdU1Z8Bu6rqa92ro76oxbokSQPQ72mlu7rf70vyCuBeYFU7JUmSBqXfUHhf9xPN7wA+AiwF3tpaVZKkgeg3FB7ufqJ5C/ASgCSntVaVJGkg+p1T+EifbZKkOWyqS2f/KPBiYEWSt/esWgoMtVmYJGn2TXX4aAFwVLdf72W0HwFe3VZRkqTBmOp+Cl8DvpZkW1W9v3ddkp8Bbm+zOEnS7Op3TuHcCdreNZOFSJIGb6o5hbOAlwMrk3y4Z9VSYLTNwiRJs2+qOYV7gRHgbPa+VPajwNvaKkqSNBhTzSlcD1yf5NNVtWuyvhNJcibwITpnKn20qn5vH/1eTeeS3C+sqpHpPo8kaWb0++G1NUn+B3Aye18l9en72iDJEHAJ8DJgI3BNknVVddO4fkuAtwDfnGbtkqQZ1u9E88eBP6Izj/AS4BPAJ6fY5lRgQ1XdUVU7gcuAcybo97vA++ncuEeSNED9hsKRVfUlIFV1V1W9B5jqXgorgXt6ljd22xpJng+srqrPT7ajJBckGUkysnnz5j5LliRNV7+hsD3JPOD2JBcm+Wng2Cm2yQRt1azs7O+DdC6yN6mqWltVw1U1vGLFij5LliRNV7+h8FZgMZ1j/z8CvA54/RTbbARW9yyvonM205glwCnAV5PcSef+DOuSDPdZkyRphvV7j+axu6xtTfKuqrq/j82uAU5McgKwic4H4F7bs88twPKx5SRfBX7Ns48kaXD6HSn0urKfTlU1ClwIXAXcDFxeVeuTvDfJ2fvxvJKklvV7SmqvieYKJlRVVzIuRKrq3fvoe/p+1CJJmkH7M1L40xmvQpJ0UJjq2kdPmqD5srH2qnqolaokSQMx1eGja+mcRtp7yGhsuYB9fqJZkjT3THXtoxNmqxBJ0uBNdfjoBZOtr6pvzWw5kqRBmurw0R9Msq6Y+lIXkqQ5ZKrDRy+ZrUIkSYM31eGjM6rqy0leNdH6qvqrdsqSJA3CVIePfhz4MvBK9j7raOy7oSBJh5CpDh9d3H34HfY+NbWALUmeV1XXtVifJGkW9fuJ5h8B3gQcBzwVuAA4HfjTJL/eTmmSpNnW77WPjgFeUFVbAZJcDFwB/BidD7i9v53yJEmzqd+RwtOAnT3Lu4Djq2obsGPGq5IkDUS/I4VPA1cn+Zvu8iuBzyR5AnBTK5VJkmZdvzfZ+d0kVwL/js5k85t6bobzc20VJ0maXX3fT6GqrqUzfyBJOkTtz/0UJEmHKENBktQwFCRJDUNBktQwFCRJDUNBktQwFCRJDUNBktQwFCRJDUNBktQwFCRJDUNBktQwFCRJDUNBktRoNRSSnJnk1iQbklw0wfq3J7kpyQ1JvpTk+DbrkSRNrrVQSDIEXAKcBZwMnJfk5HHdvg0MV9Vz6dzz2Xs9S9IAtTlSOBXYUFV3VNVO4DLgnN4OVfWVqnqsu3g1sKrFeiRJU2gzFFYC9/Qsb+y27csbgS9MtCLJBUlGkoxs3rx5BkuUJPVqMxQyQVtN2DF5HTAMfGCi9VW1tqqGq2p4xYoVM1iiJKlX3/do3g8bgdU9y6uAe8d3SvITwG8BP15VO1qsR5I0hTZHCtcAJyY5IckC4FxgXW+HJM8H/gQ4u6oeaLEWSVIfWguFqhoFLgSuAm4GLq+q9Unem+TsbrcPAEcBn01yXZJ1+9idJGkWtHn4iKq6ErhyXNu7ex7/RJvPL0maHj/RLElqGAqSpIahIElqGAqSpIahIElqGAqSpIahIElqGAqSpIahIElqGAqSpIahIElqGAqSpIahIElqGAqSpIahIElqGAqSpIahIElqGAqSpIahIElqGAqSpIahIElqGAqSpIahIElqGAqSpIahIElqGAqSpIahIElqGAqSpMb8QRegwXnlK1/ZPP7c5z43wEqkvZ1//vncf//9HHfccaxdu3bQ5RxWWg2FJGcCHwKGgI9W1e+NW78Q+ATwI8D3gJ+tqjvbrOlQsWvHNh669xZ2bf8BixYdxYInPImqXRTzoXazZ88oCxYvZeejDzG08Eiqwu7tjzB0xBM44sgnMH/hkaw5binLj17Id7+3g3tv/yYLFi5h965tLF72ZLZv28Loth8AMH/xUnY/toXFT1zJ7tHH2LVjB1sf3sSio45mydLjWPLk4wf809DBavfoTjJviGQee0Z3APPYs2eUIxYu3qvfrp3b2DO6k3nz5jN/4WIee+Rh5iVs3fIgu0d3MTT/iM7+dm1n3vyFJAFgz+5RRnduo2oPCxYtYc/uXcybv6BZr+lLVbWz42QIuA14GbARuAY4r6pu6unzy8Bzq+pNSc4Ffrqqfnay/Q4PD9fIyEgrNc8VN39pLaPbHx10GXt52gvOYdlxPzToMnSQ2LN7lI03XMWWe28lQ0PMmzef3bu2N+uPOHIJz3jxaxk6YhEb/vFT7Hj0wWZdVZFkr+/Lnvosdj32MNu2fJcFi5ex6t+cxSP3386D/3Lt4577iEVHsfK5/4ElK9bMxkudM5JcW1XDU/Vrc07hVGBDVd1RVTuBy4BzxvU5B/jz7uMrgJfGiJ/U9+/bcNAFAsDd31o36BJ0EHno7hvYcu8tQFG7R/cKBIBd2x5l4/VXcf8tX98rEIDmr/ze74/cdwvbtnwXgJ2PbeHub39+wkAA2LV9K/dcdyV79uye4Vd1eGgzFFYC9/Qsb+y2TdinqkaBLcAx43eU5IIkI0lGNm/e3FK5c8P3N60fdAn70M6IU3PT2C/wyWx/dDOPPXzvfu1/dPvWSdfv3rntoPzjaS5oMxQm+ot//G+OfvpQVWurariqhlesWDEjxc1Vx/7QiwZdwsQyNOgKdBA5avnU80xHLT+eZU/Zv0OOC5csn3T9gsVHc8SRy/Zr34e7NieaNwKre5ZXAeP/LBjrszHJfGAZ8FCLNc15i49+Mk9Y8XR+sPmOA9rPRHNJB3Lk7qQff/2BlKNDzBNXncyu7Y/y0N03kqH5DA3NZ8fWh9izexSAJxyzmpWn/AQZms9jj2zmkftuY+zvwV2jexia969zCqO79/C0U85g+9bNbH3wLo5ceixPPeWl/OChe9j0na9Qu3d2njRDDM0/gkVLj+Wpz36Jk837qc2J5vl0JppfCmyiM9H82qpa39PnzcBzeiaaX1VVr5lsv040zxxPSdXByvfmzBv4RHN3juBC4CrgZuDyqlqf5L1Jzu52+zPgmCQbgLcDF7VVj6S54ylPeQoAxx133IArOfy0NlJoiyMFSZq+gY8UJElzj6EgSWoYCpKkhqEgSWoYCpKkhqEgSWoYCpKkxpz7nEKSzcBdg67jELIceHDKXtLs8705s46vqikvHjfnQkEzK8lIPx9okWab783B8PCRJKlhKEiSGoaCvCu6Dla+NwfAOQVJUsORgiSpYShIkhqGwmEqyZlJbk2yIYk3N9JBI8nHkjyQ5DuDruVwZCgchpIMAZcAZwEnA+clOXmwVUmNS4EzB13E4cpQODydCmyoqjuqaidwGXDOgGuSAKiqvwceGnQdhytD4fC0ErinZ3ljt03SYc5QODxlgjbPTZZkKBymNgKre5ZXAfcOqBZJBxFD4fB0DXBikhOSLADOBdYNuCZJBwFD4TBUVaPAhcBVwM3A5VW1frBVSR1JPgN8A3hmko1J3jjomg4nXuZCktRwpCBJahgKkqSGoSBJahgKkqSGoSBJahgK0jQkeU+SXxt0HVJbDAVJUsNQkCaR5BeS3JDk+iSfHLfu/CTXdNf9ZZLF3fafSfKdbvvfd9ueneSfk1zX3d+Jg3g90lT88Jq0D0meDfwVcFpVPZjkScBbgK1V9ftJjqmq73X7vg/4blV9JMmNwJlVtSnJ0VX1/SQfAa6uqk91Ly0yVFXbBvXapH1xpCDt2xnAFVX1IEBVjb/G/ylJvt4NgZ8Dnt1t/0fg0iTnA0Pdtm8Av5nkN4DjDQQdrAwFad/C5JcUvxS4sKqeA/wOsAigqt4E/DadK9Fe1x1RfBo4G9gGXJXkjDYLl/aXoSDt25eA1yQ5BqB7+KjXEuC+JEfQGSnQ7feMqvpmVb0beBBYneTpwB1V9WE6V6R97qy8Amma5g+6AOlgVVXrk/x34GtJdgPfBu7s6fLfgG8CdwE30gkJgA90J5JDJ1iuBy4CXpdkF3A/8N5ZeRHSNDnRLElqePhIktQwFCRJDUNBktQwFCRJDUNBktQwFCRJDUNBktT4/8n3rMqKMS3tAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEUCAYAAADEGSquAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzt3XmcHWWd7/HPt7vT6aydkAQSQichyr5jwygIIs4ooMALWQQdEJ2RQWHEcbkyjorDjPfeuer4YmAYDKMCDgMKqAMIoojsa8CEkIQlBCQhidnIvvXyu39U9eGk00v1Ul29fN+v13l11VNPPfU7p7vPr556alFEYGZmBlBRdABmZtZ/OCmYmVmJk4KZmZU4KZiZWYmTgpmZlTgpmJlZiZPCECJpjqQR7Sx7XdLBXWzvy5JektQs6SOtlu0h6TeSXpY0V9Kf9ST2Vm2fIOmDGeteKOn2svVm91YcGbY9QdLj6ef+lb7abhtxzJC0uqjttybpNEnfKToOa1tV0QFY34mIw3u5yYeAXwL/2cay/wM8HBEflPRe4GZJ+0TvXBhzAjAa+E0vtJULSZXAnwNvRcQxRcfTX0iqiog7gTuLjsXa5p7CECIpJI1Op4+TNE/S05L+DVBavnvaa6hP5z8p6VFJu+xARMQzEbGonc2dA1yX1nsU2AbUdzHe/SQ9kfY0Xkh7JocAFwMXpHvgl0uqknSfpNmS5kv6saTqTtoeJ+kBSV9oY1mFpGslvZhu+7G0fKeeRvl8Oj1H0tWSngROAb4DHJuWHyfp45KekvSH9PWBsrYOSHtWz6e/l0+m5VMk3Z7+nuZJ+loH7+m7kp5JY/6dpOltLG9p57iy8gvSsucl/ULS7mn5K5IOK6v3t5J+XPa7ubdse5/qIK6Q9BVJDwJXtNF7myPpB+n250o6oGzdb0talH5u/9KXPb0hKyL8GiIvIEj2sIcDbwInpOXnpMsOTudPAF4G3g28AdR10u6DwEfK5icAm1vVuQf4aBvr7gnMaafdq4BvlM2PT39+C/huWbmACWXTNwEXp/MXAreXva/ZwHTgWeCsdrZ7RPr+K1pt9wRgdlm90nw63QS8p2x5adtln4vS6f2Apel0Vbq9s8vrpj9/CxyfTlcDjwB/0U7cE8um/xq4NZ2ekf5+L0jn3wcsTf8ODgaWAVPSZf8E/DSd/gbwr2VtPgscn8b7LLB/Wj4GeKllvp2/u6+29bmkn1sDcEQ6/w/Azen0qcBcYBTJDuzPyz9/v/J5+fDR0LQfsCUiHgSIiJ9JmtWyMCIelPTfwKPAGRGxJK9AImIZ0N5hrYeB76Z7/b9PX22pAL4s6WSgEhgPbGmn7pS0nQsi6cG0ZXHazg8lPQDc3ekbSbwSEU90sPwdwC2SppJ8EU6WNJkkWVRFxG0tFSNijaRRJF+akyS1LBoDHECSLFo7WdIlJIm/9f/2DuC/0rYfkrSV5O/gfcA9EbE8rfcDki9igBuBpyT9L2B/oJYkKR2Qvm4ti2t4WvZiO+/9xnbKAV6KiD+k00+SJAOA9wM/i4jNAJJuJElUliMfPhqa1HkVjgBWAXt1tfGIWAMgaWJZ8TSgS8klIu4AjgVeBS4HftJO1Y8D7wWOi4hDgGuBmnbqvkWyV3tKS4GkD6WHMOZI+kpErAcOAn4KHArMT7+8G9n5f6b1NjZ18pZuAa6NiIOAI9P2amj/91FBspd9VEQcnr7eERH/1jrm9FDR94HzIuJg4NMdfAak24yyn+WS3fuIN4AFwMnAp4AbI9mFF7C6LKbDI2JGRPxC0qfK4vpExs9mW9l0E28ntLZis5w5KQxNLwIjJB0PIOkskr1A0vm/IzlUcSRwuaTuDFDfRnLsHyUDzSNIDjlkJumdwIqIuAH4R+DodNGG8niBcSRfUhsl1ZIkifZsA04HDpB0lSRFxH1lX27fkTQJGBERvyZJRuuBmcBrwExJ45XsIp/XlfeTxvlaOv1XJHvXkPw+GiWdXfbeJ0TERpI988vLyuskTW4dMzCWpDewQlIF6Wdfprrlc0nHE2pIkuPvgFPSpAfwGeD+svVuIDkUdR5v7+2/BGyRdH5ZXPtLGhsRPy6L6+Yufj6t/R44W9LI9D2d39kK1nNOCkNQRGwn+Sf/d0lPkwwAvwEg6Wjg8ySHV5aTHpuWNKZ1O+ke6lLgPcANkpZKGpsuvhw4QdIrJHvu50dEcxtt7ClpTjuhngPMk/QH4GrgsrT8F0B9ujd6OckYwhhJ80mS0SOdvP8dwNnAHsCs9AunXB1wv6S5wPPAvcCTEfEm8D2S5HY/sJyu+QLwS0mPkhznX5PG00iSqC5OB3zn8nZP5hPAgWn5PJLey7g23tO89L3PBx7g7eTTYg2wj6SnSH4f50XEjoiYD/w98FtJzwOH8fbnDHAHySGsBWnPoSXeU4Fz08Hh+WmbHQ7ud1UkZyndR3I46wFgEUmCthy1DHqZmfU7ksakPcAKklOfl0XE14uOazBzUjCzfkvSL0h6VS2HHz8bERsKDWqQc1IwM7MSjymYmVmJk4KZmZUMuIvXJk6cGDNmzCg6DDOzAeXZZ59dHRGTOqs34JLCjBkzmD3btz8xM+sKSX/MUs+Hj8zMrMRJwczMSpwUzMysxEnBzMxKnBTMzKzEScHMzEqcFMzMrGTAXadgveeSSy7hjTfeYMaMGVx99dVFh2Nm/UBuSUFSDcnjFIen27k9Iq5oVWc4yb3w30Vyv/ePRcTrecU0GEU08+YLv2PD8ldQZRUVFZU0NWxjRO1kJu59JGuXzIMIJux9JDs2v8XyhQ/T3LgdVVTxkaMmsm7/MTz2/NKd2ly//BWWL3yIxm0bARg2spa6w09mWM1oVr7yJDu2rGPk+L3YsXU9jds3M37qgYybekBb4Zl1asfWDbzx3N1sXb8CIqBiGHWHn0zN6N1Y9epTNO7Yym51B1M7Zb9O24oIli98iHVvLqSisoopBxyfaT17W253SU2fTDUqIjZJGkbyvN/LIuLJsjqfAw6NiIslnUvyPOCPddRufX19+Irmt735wv2s/ePczitKyT9cGxqbmrn9sWV8+1++z5a3lvHq47e0sXoF1aPGs33TmjbbmHbkqdRO2bdLsZtFBC/+7gc0bt+8y7LKYSNoathamp9efwZj95jZYXvLX3yY1a8+s1PZO997PiNqd++dgAcwSc9GRH1n9XLrKaTPcm15Luuw9NX6W+l04Fvp9O3ANenjEX0/74zWL385W8UOPtKqygpqh20HYN2yl9pZvbndhNCynpPC4LB+/XoaGxv7ZFsNW95qMyEAOyUEgJWvP09DVW2bdVusXTJ/l7JlrzzNuL3f0+0Yq6qqqK3teLuDSa4DzZIq00ctrgR+GxFPtaoylfRh7ukj/tYDE9po5yJJsyXNXrVqVZ4hDzhVw0f1SjtvbUyenV49sqM//vaeLw/VI8e2u8ysPZXDRmSvm+FvvbJ61zpVI3Z5eql1INeB5ohoAg6XNA74haSDI+KFsiptfcvssksbEbOAWZAcPsol2AFqr0M/yOLHf0ryUe+sZsxEtm1cDcCoCXVs27iGph1bdqm34LXVLHgtqbdb3SGsfeP5XXoFY3afyagJdax48RGIZiqqqmlu3AHA8NETmDiz016pDRB9vVfc9I6jWfXq0zuVVQ4fQ834OjavWABAzdhJ1B14LFXVIztsa+QRp/DqYzcTzUlPp3rUeOr2fzcVlT6nJqs+e/KapCuAzRHx3bKy+4BvRcQTkqqAFcCkjg4feUxhV81NjWxY8QrDRo6lqnoEm9a8Se3kd1JVPYJtm9ZABDVjJhIRbFqzlA1/epX/uu0uXl68lAhYtW7LLmcgbV77Jtu3rKe5cTujxk8tHZNt2LaZhq0bGFG7Bw3bN9G0Yys1Y3cnGUIy654dWzewYcUitm9ex9jdZ7KjcgwAY0ZU0dSwrUt/Y9HcxPoVr1BVPZLRE6flGfaAknVMIc+B5klAQ0SskzQC+A3wLxFxd1mdS4BDygaaPxoR53TUrpNC7zn11FNL03fddVeBkZjtbM2apKc6YcIuR5Otm7ImhTzHFKYAv5f0PPAMyZjC3ZKulHRaWueHwARJi4AvApfnGI+1Mm1ashflhxaZWYs+O3zUW9xTMBv83FPoff2hp2BmZgOMk4KZmZU4KZiZWYmTgpmZlTgpmJlZiZOCmZmVOCmYmVmJk4KZmZU4KZiZWYmTgpmZlTgpmJlZiZOCmZmVOCmYmVmJk4KZmZU4KZiZWYmTgpmZlTgpmJlZiZOCmZmVOCmYmVmJk4KZmZU4KZiZWYmTgpmZlTgpmJlZSW5JQVKdpN9LWihpvqTL2qhzgqT1kuakr2/mFY+ZmXWuKse2G4EvRcRzksYAz0r6bUQsaFXvkYj4SI5xmJlZRrn1FCJieUQ8l05vBBYCU/PanpmZ9VyfjClImgEcATzVxuL3SJor6V5JB7Wz/kWSZkuavWrVqhwjNTMb2nJPCpJGA3cAX4iIDa0WPwdMj4jDgKuBX7bVRkTMioj6iKifNGlSvgGbmQ1huSYFScNIEsLNEfHz1ssjYkNEbEqn7wGGSZqYZ0xmZta+PM8+EvBDYGFE/Gs7dSan9ZB0dBrPmrxiMjOzjuV59tGxwPnAPElz0rKvAdMAIuI64Czgs5Iaga3AuREROcZkZmYdyC0pRMSjgDqpcw1wTV4xmJlZ12RKCpJ2a6N4Y0Q09HI8ZmZWoKxjCs8Bq4CXgVfS6dckPSfpXXkFZ2ZmfStrUvg1cEpETIyICcDJwM+AzwHX5hWcmZn1raxJoT4i7muZiYjfAMdHxJPA8FwiMzOzPpd1oHmtpK8Ct6bzHwPeklQJNOcSmZmZ9bmsPYWPA3uRXHH8PySnlX4cqATOySc0MzPra5l6ChGxGvjbdhYv6r1wzMysSFlPSd0X+DIwo3ydiDgxn7DMzKwIWccUbgOuA/4TaMovHDMzK1LWpNAYEf+RayRmZla4rAPNd0n6nKQpknZreeUamZmZ9bmsPYVPpj+/UlYWwMzeDcfMzIqU9eyjvfMOxMzMitdhUpB0YkQ8IOmjbS1v68E5ZmY2cHXWU3gf8ABwahvLAnBSMDMbRDpMChFxhaQK4N6I+FkfxWRmZgXp9OyjiGgGLu2DWMzMrGBZT0n9raQvS6rzKalmZoNX1lNSP53+vKSszKekmpkNMj4l1czMSrIePiqRNCuPQMzMrHhdTgpAfa9HYWZm/UJ3ksLKXo/CzMz6hawDzQBIGhURJ2WsWwfcBEwmeWTnrIi4qlUdAVcBpwBbgAsj4rmuxDQULXz61zSuml90GDvZ+31/w+jRo4sOw/qBeb/6Xq+1tawX2hi91yHsfdgHe6GloSFTT0HSMZIWAAvT+cMkXdvJao3AlyLiAODdwCWSDmxV52Rgn/R1EeDbc3eioaGh3yUEgNce+kHRIVg/0JsJobdsWjqPtUv63/9Mf5W1p/B94EPAnQARMVfS8R2tEBHLgeXp9EZJC4GpwIKyaqcDN0VEAE9KGidpSrpuLtavX8/LL7+cV/OZ7L777lRVdamTVrJ2zq29HE3veeHJXzG+7vAur9fY2MjKlcUfldx3332pra0tbPvXX389ixcvLmz7veG8YyeQHADoX5bMuYfvXPtfRYfRbTNnzuQzn/lMn2wr8zdTRCxp9cvO/AQ2STOAI4CnWi2aCiwpm1+alu2UFCRdRNKTYNq0aVk3a32somp40SEMaIsXL2buq6+yberUokPptvOKDqAdzcBT27YVHUa31Lz5Zp9uL2tSWCLpGCAkVQOfJz2U1BlJo4E7gC9ExIbWi9tYJXYpiJgFzAKor6/fZXlX1NbWctRRR/WkiUJNnfqlftlFBziw/s+7ve706dN7MZKBa9vUqbx22WVFh9FtjfPvoIq2/7GLEsDX647ltcv2LDqUbtn7qqs6r9SLsp59dDHJ1cxTSfbmD2fnq5vbJGkYSUK4uZ3bbC8F6srm96J3xpYGtUmHtnXT2mId8uEvFR2C9QNnHnQma0j2zPvL63/vcQQLxg3MhFCErD0FRcQnutJwembRD4GFEfGv7VS7E7hU0q3AnwHr8xxPGCwm1+3L5Dp/CVv/9OmDziw6BOuBrEnhcUmvAT8F7oiIdRnWORY4H5gnaU5a9jVgGkBEXAfcQ3I66iKSU1I/1YXYzcysl2W999E+ko4GzgX+IT099daIaHc4PyIepZNDi+lZR50ehjIzs76R+YrmiHg6Ir4IHA2sBW7MLSozMytE1ovXxkr6pKR7gcdJThk9OtfIzMysz2UdU5gL/BK4MiKeyDEeMzMrUNakMDMiQtIYSaMjYlOuUZmZWSGyjikcJOkPwAvAAknPSjo4x7jMzKwAWZPCLOCLETE9IqYBX0rLzMxsEMmaFEZFxO9bZiLiQWBULhGZmVlhso4pLJb0DeAn6fxfAq/lE5KZmRUla0/h08Ak4OfAL9JpX31sZjbIZL2i+S3g85JqgeaI2JhvWGZmVoSsF68dJWkeyfUK8yTNlfSufEMzM7O+lnVM4YfA5yLiEQBJ7wV+DByaV2BmZtb3so4pbGxJCFC62Z0PIZmZDTId9hQkHZlOPi3pB8AtJA8y+hjwYL6hmZlZX+vs8FHr5z5eUTbdo8dimplZ/9NhUoiI9/dVIGZmVrysA81I+jBwEFDTUhYRV+YRlJmZFSPrKanXkYwj/C3J09TOBqbnGJeZmRUg69lHx0TEBcBbEfGPwHuAuvzCMjOzImRNClvTn1sk7Qk0AHvnE5KZmRUl65jC3ZLGAd8BniM58+j63KIyM7NCZL330T+lk3dIuhuoiYj1+YVlZmZFyHr4qNzVTghmZoNTd5JCfZZKkn4kaaWkF9pZfoKk9ZLmpK9vdiMWMzPrRZmvUyizMmO9G4BrgJs6qPNIRHykGzGYmVkOutRTkDQqIk7KUjciHgbWdisqMzMrRNaL146RtABYmM4fJunaXtj+e9JnM9wr6aAOtn+RpNmSZq9ataoXNmtmZm3J2lP4PvAhYA1ARMwFju/htp8DpkfEYcDVwC/bqxgRsyKiPiLqJ02a1MPNmplZezIfPoqIJa2Kmnqy4YjYEBGb0ul7gGGSJvakTTMz65msSWGJpGOAkFQt6cukh5K6S9JkSUqnj05jWdOTNs3MrGeynn10MXAVMBVYCvwGuKSjFSTdApwATJS0lORZDMMAIuI64Czgs5IaSW6jcW5E+BkNZmYFynpF82rgE11pOCLO62T5NSSnrJqZWT+R9eyjG9N7H7XMj5f0o/zCMjOzImQdUzg0Ita1zETEW8AR+YRkZmZFyZoUKiSNb5mRtBvduxrazMz6saxf7N8DHpd0ezp/NvDtfEIyM7OiZB1ovknSs8D7SR7H+dGIWJBrZGZm1ucyHwKKiPmSVgE1AJKmRcQbuUVmZmZ9LuvZR6dJegV4DXgIeB24N8e4zMysAFkHmv8JeDfwckTsDXwAeCy3qMzMrBBZk0JDRKwhOQupIiJ+DxyeY1xmZlaArGMK6ySNBh4Gbpa0EmjMLywzMytChz0FScPTydNJ7k/0d8CvgVeBU/MNzczM+lpnPYUngCOB6yLi/LTsxnxDMjOzonSWFKolfRI4RtJHWy+MiJ/nE5aZmRWhs6RwMcndUcex6+GiAJwUzMwGkQ6TQkQ8KulxYGlE+LYWZmaDXKenpEZEM/CRPojFzMwKlvU6hd9IOrPl8ZlmZjY4Zb1O4YvAKKBR0jaSm+JFRIzNLTIzM+tzWe+SOibvQMzMrHiZkoKk49sqj4iHezccMzMrUtbDR18pm64BjgaeBU7s9YjMzKwwWQ8f7XSNgqQ64P/lEpGZmRWmu89ZXgoc3JuBWN879dS3c/1dd91VYCRm1l9kHVO4muQKZkhOYz0cmNvJOj8iub5hZUTskkDS01uvAk4BtgAXRsRz2UMf2rZtWsPaP85j68ZVNG7fQtWwGpqaG2huagAqadi6Hpp2UDFsBEg0NzZBNKa/xSY0fAyfPeMIaqqrWLdxG68/9yuiuZEta5dCBKqspqmxAZq2JRusrIamHaXpkbVTaGrcRkXVMOoOP5nhI3wimiUuePkRzmhYWXQYQPLnfuZ+pxNV3d3/HXqyflKzy6YbgVsiorOH7NwAXAPc1M7yk4F90tefAf+R/rROLF/wEKtfm71T2Y526jY3bG2zPLZvpG6PWgAmjR/FxuUv7lyhcfvO8007dpresvaPpdmXH7ieKQd9gIkz/IiNoe62+XdQTXLOen/xy5f+h0unHceSMbsXHcqAkHVMoXRnVEnjgboM6zwsaUYHVU4HboqIAJ6UNE7SlIhYniWmoaqpYdsuCaE/WLHgQSeFHjruuOP44OjRNP/pT0WH0m39LSFAEs/VbzzC6onvLTqUbqk44wy2btrUZ9vLevjoQeC0tP4cYJWkhyLiiz3Y9lRgSdn80rRsl6Qg6SLgIoBp06b1YJMDX1PD9s4rFSCiqegQBrza2lqqq6uLDmPQGluR9QYO/cyIEVRXVvbZ5rIePqqNiA2S/hr4cURcIen5Hm67rR2KaKOMiJgFzAKor69vs85QUT2ylurRu7Fj09qiQ9nJiPF7Fh3CgHfssccWHUKPzftV/7x0acS4qRx21FFFhzEgZE2dVZKmAOcAd/fStpey82GovYBlvdT2oPbOY85jzOR96ElHPSJ2ejU3N3e7rRHj9uSdx5zX7fVt8DjgQ5cVHUKb9j323KJDGDCy9hSuBO4DHo2IZyTNBF7p4bbvBC6VdCvJAPN6jydkUzmshhnvOq3H7fiUVOttVVVVHPLhL/W4nTVr1gAwYcKEHrdlXZN1oPk24Lay+cXAmR2tI+kW4ARgoqSlwBXAsHT964B7SE5HXURySuqnuh6+mZn1pi6fvCvpuYg4srN6EdHh8YT0rKNLurp96z3uHZhZa90Zju9vZ5yZmVkv6U5S+FWvR2FmZv1CpqQg6eSW6Yj4elp2cV5BmZlZMbL2FL4hqXSbbElfJbki2czMBpGsA82nAXdL+gpwErB/WmZmZoNI1lNSV0s6Dbif5OE6Z6VnD5mZ2SDSYVKQtJHk1hNKf1YDM4GzJEVE+H7JZmaDSIdJISLG9FUgZmZWvM56Ch1epOaH4piZDS6djSl8r4NlAZzYwXIzMxtgOjt89P6+CsTMzIqX+d5Hkg4GDgRqWsoior1HbZqZ2QCU9clrV5Dc8fRAkrubngw8SvvPXzYzswEo6xXNZwEfAFZExKeAw4DhuUVlZmaFyJoUtkZEM9AoaSywkuR6BTMzG0SyjinMljQOuJ7kiuZNwNO5RWVmZoXIepuLz6WT10n6NTA2Ip7PLywzMytC1ltn/65lOiJej4jny8vMzGxw6OyK5hpgJMlzlsfz9lPXxgJ75hybmZn1sc4OH/0N8AWSBPAsb98YbyNwTb6hmZlZX+vw8FFEXBURewPfBg5Pp38MLAae6IP4zMysD2W+TiEiNkh6L/AXwA3Af+QWlZmZFSJrUmhKf34YuC4i/ofk2QpmZjaIZE0Kb0r6AXAOcI+k4VnWlXSSpJckLZJ0eRvLL5S0StKc9PXXXQvfzMx6U9aL184heTbzdyNinaQpwFc6WkFSJfDvJIeblgLPSLozIha0qvrTiLi0i3GbmVkOsl68tgX4edn8cmB5J6sdDSyKiMUAkm4FTgdaJwUzM+snsh4+6o6pwJKy+aVpWWtnSnpe0u2S6tpqSNJFkmZLmr1q1ao8YjUzM/JNCmqjLFrN3wXMiIhDgfuBG9tqKCJmRUR9RNRPmjSpl8M0M7MWeSaFpUD5nv9ewLLyChGxJiK2p7PXA+/KMR4zM+tEnknhGWAfSXtLqgbOBe4sr5AOWLc4DViYYzxmZtaJzI/j7KqIaJR0KXAfUAn8KCLmS7oSmB0RdwKfl3Qa0AisBS7MKx4zM+tcbkkBICLuIXl8Z3nZN8um/x74+zxjMDOz7PI8fGRmZgOMk4KZmZU4KZiZWYmTgpmZlTgpmJlZiZOCmZmVOCmYmVmJk4KZmZU4KZiZWYmTgpmZlTgpmJlZiZOCmZmVOCmYmVmJk4KZmZU4KZiZWYmTgpmZlTgpmJlZiZOCmZmVOCmYmVmJk4KZmZU4KZiZWYmTgpmZlVQVHYAV59RTTy1N33XXXQVGYmb9Ra5JQdJJwFVAJfCfEfF/Wy0fDtwEvAtYA3wsIl7PM6bBorFhBxtXvspbK14DmoEKEDTsaKRm5Fiamxpoam6iurqGhh1bqayopGrEGLZvWs2w4bU0NWznpKOnMbyykobmZl58/Haqq0fQXFnB9o1rqayqoULB8JHjaW7aTjMimpqJpma2bVrDiNoJ1IyuZfiocUzY62AqKtzptLdt37KexobtNG7bRMO2LQwbVQtNjVQNH0lTUwMNm9ezdcsGaqprqB49gYphw2jYsglVVRHRwFvLl6HKYVQ17UXTjs0Mq6kFmqiqGUPj9s007tiKQgwfXQsVlURTM00NW2mKZpq3baKqZiwVFWLUblORVPTHMaAoIvJpWKoEXgb+AlgKPAOcFxELyup8Djg0Ii6WdC5wRkR8rKN26+vrY/bs2bnEPFAseuxmtq5bUXQYOxm75/5MP+LDRYdhBWtuamLh/dfS3Lij6FBKJh/wPibNrC86jMJJejYiOv0g8uwpHA0siojFaUC3AqcDC8rqnA58K52+HbhGkiKvTDUIbN+8rt8lBIANy16k+ZAPUlE1rOhQrIfWr19PY2Njt9Zd/eJv+1VCAFix8CE0dka3ewxVVVXU1tb2clT9V559/qnAkrL5pWlZm3UiohFYD0xo3ZCkiyTNljR71apVOYU7MKxb/nLRIbRr26Y1RYdgBWvc3D//BqKpoegQBow8ewptpeXWPYAsdYiIWcAsSA4f9Ty0gWvC9MNY+dIjRYfRphG1exQdgvWCnuwVb528H28tmduL0fQCVTBpjylFRzFg5NlTWArUlc3vBSxrr46kKqAWWJtjTANe1bDh7Dbt8B63ExG7vHpi8gEnekDP2OvQP2f46F06+8VRBTOOOrPoKAaUPHsKzwD7SNobeBM4F/h4qzp3Ap8EngDOAh7weELnph7yAaYe8oEet+NTUi0P+77vwqJDsB7ILSlERKOkS4H7SE5J/VFEzJfqHIdFAAADDElEQVR0JTA7Iu4Efgj8RNIikh7CuXnFY2ZmncvtlNS8+JRUM7Ouy3pKqq84MjOzEicFMzMrcVIwM7MSJwUzMytxUjAzsxInBTMzK3FSMDOzkgF3nYKkVcAfi45jEJkIrC46CLM2+G+zd02PiEmdVRpwScF6l6TZWS5oMetr/tsshg8fmZlZiZOCmZmVOCnYrKIDMGuH/zYL4DEFMzMrcU/BzMxKnBTMzKzESWGIknSSpJckLZJ0edHxmLWQ9CNJKyW9UHQsQ5GTwhAkqRL4d+Bk4EDgPEkHFhuVWckNwElFBzFUOSkMTUcDiyJicUTsAG4FTi84JjMAIuJhksfzWgGcFIamqcCSsvmlaZmZDXFOCkOT2ijzuclm5qQwRC0F6srm9wKWFRSLmfUjTgpD0zPAPpL2llQNnAvcWXBMZtYPOCkMQRHRCFwK3AcsBH4WEfOLjcosIekW4AlgP0lLJf1V0TENJb7NhZmZlbinYGZmJU4KZmZW4qRgZmYlTgpmZlbipGBmZiVOCmZdJOlbkr5cdBxmeXBSMDOzEicFs05IukDS85LmSvpJq2WfkfRMuuwOSSPT8rMlvZCWP5yWHSTpaUlz0vb2KeL9mHXEF6+ZdUDSQcDPgWMjYrWk3YDPA5si4ruSJkTEmrTuPwN/ioirJc0DToqINyWNi4h1kq4GnoyIm9Pbi1RGxNai3ptZW9xTMOvYicDtEbEaICJa3+f/YEmPpEngE8BBafljwA2SPgNUpmVPAF+T9FVguhOC9UdOCmYdEx3fVvwG4NKIOAT4R6AGICIuBr5OcjfaOWmP4r+B04CtwH2STswzcLPucFIw69jvgHMkTQBIDx+VGwMslzSMpKdAWu8dEfFURHwTWA3USZoJLI6IfyO5K+2hffIOzLqgqugAzPqziJgv6dvAQ5KagD8Ar5dV+QbwFPBHYB5JkgD4TjqQLJLEMhe4HPhLSQ3ACuDKPnkTZl3ggWYzMyvx4SMzMytxUjAzsxInBTMzK3FSMDOzEicFMzMrcVIwM7MSJwUzMyv5//GBQNRNMB5bAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "import os\n",
    "\n",
    "\n",
    "if not os.path.exists('./box_plot/'):\n",
    "    os.mkdir('./box_plot/')\n",
    "\n",
    "def boxplot(name=\"\", title=\"\"):\n",
    "    global data\n",
    "\n",
    "    ax = sns.boxplot(x=\"class\", y=name, data=data, color='cyan')\n",
    "    plt.setp(ax.lines, color=\".1\")\n",
    "    \n",
    "    ax = sns.stripplot(x=\"class\", y=name, data=data, jitter=True,\n",
    "                       color='tan', edgecolor='red')\n",
    "    plt.setp(ax.lines, color=\".9\")\n",
    "    \n",
    "    # plt.savefig('./box_plot/%s' % name)\n",
    "    plt.title(title, fontsize=11)\n",
    "    plt.show()\n",
    "\n",
    "\n",
    "for idx, col in enumerate(data.columns[1:]):\n",
    "    boxplot(name=col, title=\"idx %d : %s\" % (idx, col))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(8124,) (8124, 21)\n"
     ]
    }
   ],
   "source": [
    "## Data State\n",
    "train_y = data['class']\n",
    "data = data.drop(['class'], axis=1)\n",
    "train_x = data\n",
    "\"\"\"\n",
    "pd.concat(\n",
    "    [data['odor'], data['gill-attachment'],\n",
    "     data['spore-print-color'], data['stalk-surface-above-ring'],\n",
    "     data['population']])\n",
    "\"\"\"\n",
    "print(train_y.shape, train_x.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>cap-shape</th>\n",
       "      <th>cap-surface</th>\n",
       "      <th>cap-color</th>\n",
       "      <th>bruises</th>\n",
       "      <th>odor</th>\n",
       "      <th>gill-attachment</th>\n",
       "      <th>gill-spacing</th>\n",
       "      <th>gill-size</th>\n",
       "      <th>gill-color</th>\n",
       "      <th>stalk-shape</th>\n",
       "      <th>...</th>\n",
       "      <th>stalk-surface-below-ring</th>\n",
       "      <th>stalk-color-above-ring</th>\n",
       "      <th>stalk-color-below-ring</th>\n",
       "      <th>veil-type</th>\n",
       "      <th>veil-color</th>\n",
       "      <th>ring-number</th>\n",
       "      <th>ring-type</th>\n",
       "      <th>spore-print-color</th>\n",
       "      <th>population</th>\n",
       "      <th>habitat</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.0</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.000000</td>\n",
       "      <td>8124.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>3.348104</td>\n",
       "      <td>1.827671</td>\n",
       "      <td>4.504677</td>\n",
       "      <td>0.415559</td>\n",
       "      <td>4.144756</td>\n",
       "      <td>0.974151</td>\n",
       "      <td>0.161497</td>\n",
       "      <td>0.309207</td>\n",
       "      <td>4.810684</td>\n",
       "      <td>0.567208</td>\n",
       "      <td>...</td>\n",
       "      <td>1.603644</td>\n",
       "      <td>5.816347</td>\n",
       "      <td>5.794682</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.965534</td>\n",
       "      <td>1.069424</td>\n",
       "      <td>2.291974</td>\n",
       "      <td>3.596750</td>\n",
       "      <td>3.644018</td>\n",
       "      <td>1.508616</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.604329</td>\n",
       "      <td>1.229873</td>\n",
       "      <td>2.545821</td>\n",
       "      <td>0.492848</td>\n",
       "      <td>2.103729</td>\n",
       "      <td>0.158695</td>\n",
       "      <td>0.368011</td>\n",
       "      <td>0.462195</td>\n",
       "      <td>3.540359</td>\n",
       "      <td>0.495493</td>\n",
       "      <td>...</td>\n",
       "      <td>0.675974</td>\n",
       "      <td>1.901747</td>\n",
       "      <td>1.907291</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.242669</td>\n",
       "      <td>0.271064</td>\n",
       "      <td>1.801672</td>\n",
       "      <td>2.382663</td>\n",
       "      <td>1.252082</td>\n",
       "      <td>1.719975</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</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>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>5.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>2.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>5.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>6.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         cap-shape  cap-surface    cap-color      bruises         odor  \\\n",
       "count  8124.000000  8124.000000  8124.000000  8124.000000  8124.000000   \n",
       "mean      3.348104     1.827671     4.504677     0.415559     4.144756   \n",
       "std       1.604329     1.229873     2.545821     0.492848     2.103729   \n",
       "min       0.000000     0.000000     0.000000     0.000000     0.000000   \n",
       "25%       2.000000     0.000000     3.000000     0.000000     2.000000   \n",
       "50%       3.000000     2.000000     4.000000     0.000000     5.000000   \n",
       "75%       5.000000     3.000000     8.000000     1.000000     5.000000   \n",
       "max       5.000000     3.000000     9.000000     1.000000     8.000000   \n",
       "\n",
       "       gill-attachment  gill-spacing    gill-size   gill-color  stalk-shape  \\\n",
       "count      8124.000000   8124.000000  8124.000000  8124.000000  8124.000000   \n",
       "mean          0.974151      0.161497     0.309207     4.810684     0.567208   \n",
       "std           0.158695      0.368011     0.462195     3.540359     0.495493   \n",
       "min           0.000000      0.000000     0.000000     0.000000     0.000000   \n",
       "25%           1.000000      0.000000     0.000000     2.000000     0.000000   \n",
       "50%           1.000000      0.000000     0.000000     5.000000     1.000000   \n",
       "75%           1.000000      0.000000     1.000000     7.000000     1.000000   \n",
       "max           1.000000      1.000000     1.000000    11.000000     1.000000   \n",
       "\n",
       "          ...       stalk-surface-below-ring  stalk-color-above-ring  \\\n",
       "count     ...                    8124.000000             8124.000000   \n",
       "mean      ...                       1.603644                5.816347   \n",
       "std       ...                       0.675974                1.901747   \n",
       "min       ...                       0.000000                0.000000   \n",
       "25%       ...                       1.000000                6.000000   \n",
       "50%       ...                       2.000000                7.000000   \n",
       "75%       ...                       2.000000                7.000000   \n",
       "max       ...                       3.000000                8.000000   \n",
       "\n",
       "       stalk-color-below-ring  veil-type   veil-color  ring-number  \\\n",
       "count             8124.000000     8124.0  8124.000000  8124.000000   \n",
       "mean                 5.794682        0.0     1.965534     1.069424   \n",
       "std                  1.907291        0.0     0.242669     0.271064   \n",
       "min                  0.000000        0.0     0.000000     0.000000   \n",
       "25%                  6.000000        0.0     2.000000     1.000000   \n",
       "50%                  7.000000        0.0     2.000000     1.000000   \n",
       "75%                  7.000000        0.0     2.000000     1.000000   \n",
       "max                  8.000000        0.0     3.000000     2.000000   \n",
       "\n",
       "         ring-type  spore-print-color   population      habitat  \n",
       "count  8124.000000        8124.000000  8124.000000  8124.000000  \n",
       "mean      2.291974           3.596750     3.644018     1.508616  \n",
       "std       1.801672           2.382663     1.252082     1.719975  \n",
       "min       0.000000           0.000000     0.000000     0.000000  \n",
       "25%       0.000000           2.000000     3.000000     0.000000  \n",
       "50%       2.000000           3.000000     4.000000     1.000000  \n",
       "75%       4.000000           7.000000     4.000000     2.000000  \n",
       "max       4.000000           8.000000     5.000000     6.000000  \n",
       "\n",
       "[8 rows x 21 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_x.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Data Normalize\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler, RobustScaler, MaxAbsScaler, MinMaxScaler \n",
    "\n",
    "scalers = {\"Standard\": StandardScaler(), \"Robust\": RobustScaler(),\n",
    "           \"MaxAbs\": MaxAbsScaler(), \"MinMax\": MinMaxScaler()}\n",
    "\n",
    "x_scaled = []\n",
    "for _, v in scalers.items():\n",
    "    x_scaled.append(v.fit_transform(train_x))\n",
    "\n",
    "x_scaled.append(train_x)  # not nomalized data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Feature Engineering\n",
    "\n",
    "from sklearn.decomposition import PCA, TruncatedSVD\n",
    "from sklearn.manifold import TSNE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# scree plot for selecting good PCA dim\n",
    "A = x_scaled[0]  # StandardScaler()\n",
    "U, S, V = np.linalg.svd(A)\n",
    "eigvals = S ** 2 / np.cumsum(S)[-1]\n",
    "eigvals2 = S ** 2 / np.sum(S)\n",
    "assert (eigvals == eigvals2).all()\n",
    "\n",
    "fig = plt.figure(figsize=(10, 6))\n",
    "sing_vals = np.arange(len(eigvals)) + 1\n",
    "plt.plot(sing_vals, eigvals, 'ro-', linewidth=2)\n",
    "plt.title('Scree Plot')\n",
    "plt.xlabel('Principal Component')\n",
    "plt.ylabel('Eigenvalue')\n",
    "#I don't like the default legend so I typically make mine like below, e.g.\n",
    "#with smaller fonts and a bit transparent so I do not cover up data, and make\n",
    "#it moveable by the viewer in case upper-right is a bad place for it\n",
    "leg = plt.legend(['Eigenvalues from SVD'], loc='best', borderpad=0.3,\n",
    "                 shadow=False, markerscale=0.4)\n",
    "leg.get_frame().set_alpha(0.4)\n",
    "leg.draggable(state=True)\n",
    "plt.savefig('./scree_plot.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 1. PCA\n",
    "for idx, x in enumerate(x_scaled[:1]):\n",
    "    pca = PCA()\n",
    "    pca.fit_transform(x)\n",
    "    \n",
    "    plt.figure(figsize=(6, 4))\n",
    "    plt.bar(range(21), pca.explained_variance_, alpha=0.5,\n",
    "            align='center', label='individual explained variance')\n",
    "    \n",
    "    # plt.title('%s Scaler' % (list(scalers.keys())[idx] if not idx == 4 else \"None\"))\n",
    "    plt.ylabel('Explained variance ratio')\n",
    "    plt.xlabel('Principal Components')\n",
    "    plt.legend(loc='best')\n",
    "    plt.tight_layout()\n",
    "    plt.savefig('./variance_pcs.png')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 1-1. PCA visualize with k-mean cluserting (only 2 features)\n",
    "pca = PCA(n_components=7)\n",
    "pca_x = pca.fit_transform(data.values)\n",
    "# plt.figure(figsize=(6, 5))\n",
    "# plt.scatter(x[:,0], x[:,1])  # by class\n",
    "# plt.tight_layout()\n",
    "# plt.show()\n",
    "\n",
    "from sklearn.cluster import KMeans\n",
    "\n",
    "kmeans = KMeans(n_clusters=2, random_state=42)\n",
    "X_clustered = kmeans.fit_predict(data.values)\n",
    "\n",
    "LABEL_COLOR_MAP = {0 : 'r', 1 : 'y'}\n",
    "label_color = [LABEL_COLOR_MAP[l] for l in X_clustered]\n",
    "\n",
    "plt.figure(figsize=(5, 5))\n",
    "plt.scatter(x[:,0], x[:,1], c=label_color)  # by class\n",
    "plt.tight_layout()\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 1-2. PCA visualize with k-mean cluserting (only 1 features)\n",
    "kmeans = KMeans(n_clusters=2, random_state=42)\n",
    "X_clustered = kmeans.fit_predict(data.values)\n",
    "\n",
    "LABEL_COLOR_MAP = {0 : 'r', 1 : 'y'}\n",
    "label_color = [LABEL_COLOR_MAP[l] for l in X_clustered]\n",
    "\n",
    "pca = PCA(n_components=2)\n",
    "x = pca.fit_transform(data.values)\n",
    "plt.figure(figsize=(5, 5))\n",
    "\n",
    "xf = x[:,0]\n",
    "yf = len(xf)*[0]\n",
    "plt.scatter(xf, yf, c=label_color)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[t-SNE] Computing 121 nearest neighbors...\n",
      "[t-SNE] Indexed 8124 samples in 0.005s...\n",
      "[t-SNE] Computed neighbors for 8124 samples in 0.439s...\n",
      "[t-SNE] Computed conditional probabilities for sample 1000 / 8124\n",
      "[t-SNE] Computed conditional probabilities for sample 2000 / 8124\n",
      "[t-SNE] Computed conditional probabilities for sample 3000 / 8124\n",
      "[t-SNE] Computed conditional probabilities for sample 4000 / 8124\n",
      "[t-SNE] Computed conditional probabilities for sample 5000 / 8124\n",
      "[t-SNE] Computed conditional probabilities for sample 6000 / 8124\n",
      "[t-SNE] Computed conditional probabilities for sample 7000 / 8124\n",
      "[t-SNE] Computed conditional probabilities for sample 8000 / 8124\n",
      "[t-SNE] Computed conditional probabilities for sample 8124 / 8124\n",
      "[t-SNE] Mean sigma: 0.643779\n",
      "[t-SNE] KL divergence after 250 iterations with early exaggeration: 12.692382\n",
      "[t-SNE] Error after 1000 iterations: 0.413679\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 2-1. t_SNE visualize\n",
    "from sklearn.manifold import TSNE\n",
    "\n",
    "model = TSNE(n_components=2,\n",
    "             learning_rate=1000,\n",
    "             perplexity=40,\n",
    "             early_exaggeration=4,\n",
    "             n_iter=1000,\n",
    "             metric='l2',\n",
    "             random_state=42,\n",
    "             verbose=1)\n",
    "transformed = model.fit_transform(pca_x)  # PCA\n",
    "\n",
    "xs = transformed[:, 0]\n",
    "ys = transformed[:, 1]\n",
    "plt.title('t-SNE')\n",
    "plt.scatter(xs, ys, c=label_color)\n",
    "plt.tight_layout()\n",
    "plt.savefig('./t-SNE.png')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "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>class_e</th>\n",
       "      <th>class_p</th>\n",
       "      <th>cap-shape_b</th>\n",
       "      <th>cap-shape_c</th>\n",
       "      <th>cap-shape_f</th>\n",
       "      <th>cap-shape_k</th>\n",
       "      <th>cap-shape_s</th>\n",
       "      <th>cap-shape_x</th>\n",
       "      <th>cap-surface_f</th>\n",
       "      <th>cap-surface_g</th>\n",
       "      <th>...</th>\n",
       "      <th>population_s</th>\n",
       "      <th>population_v</th>\n",
       "      <th>population_y</th>\n",
       "      <th>habitat_d</th>\n",
       "      <th>habitat_g</th>\n",
       "      <th>habitat_l</th>\n",
       "      <th>habitat_m</th>\n",
       "      <th>habitat_p</th>\n",
       "      <th>habitat_u</th>\n",
       "      <th>habitat_w</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>7 rows × 114 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   class_e  class_p  cap-shape_b  cap-shape_c  cap-shape_f  cap-shape_k  \\\n",
       "0        0        1            0            0            0            0   \n",
       "1        1        0            0            0            0            0   \n",
       "2        1        0            1            0            0            0   \n",
       "3        0        1            0            0            0            0   \n",
       "4        1        0            0            0            0            0   \n",
       "5        1        0            0            0            0            0   \n",
       "6        1        0            1            0            0            0   \n",
       "\n",
       "   cap-shape_s  cap-shape_x  cap-surface_f  cap-surface_g    ...      \\\n",
       "0            0            1              0              0    ...       \n",
       "1            0            1              0              0    ...       \n",
       "2            0            0              0              0    ...       \n",
       "3            0            1              0              0    ...       \n",
       "4            0            1              0              0    ...       \n",
       "5            0            1              0              0    ...       \n",
       "6            0            0              0              0    ...       \n",
       "\n",
       "   population_s  population_v  population_y  habitat_d  habitat_g  habitat_l  \\\n",
       "0             1             0             0          0          0          0   \n",
       "1             0             0             0          0          1          0   \n",
       "2             0             0             0          0          0          0   \n",
       "3             1             0             0          0          0          0   \n",
       "4             0             0             0          0          1          0   \n",
       "5             0             0             0          0          1          0   \n",
       "6             0             0             0          0          0          0   \n",
       "\n",
       "   habitat_m  habitat_p  habitat_u  habitat_w  \n",
       "0          0          0          1          0  \n",
       "1          0          0          0          0  \n",
       "2          1          0          0          0  \n",
       "3          0          0          1          0  \n",
       "4          0          0          0          0  \n",
       "5          0          0          0          0  \n",
       "6          1          0          0          0  \n",
       "\n",
       "[7 rows x 114 columns]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# again... with one-hot-encoded data\n",
    "dataset = \"./data/mushrooms.csv\"\n",
    "\n",
    "df = pd.read_csv(dataset)  # read DataSet\n",
    "df.drop('stalk-root', axis=1, inplace=True)\n",
    "\n",
    "for col in df.columns:\n",
    "    df = pd.get_dummies(df, columns=[col])\n",
    "df.head(7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8124, 114)"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape  # 114 dims..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_y = df['class_e']\n",
    "df.drop(['class_p', 'class_e'], axis=1, inplace=True)\n",
    "train_x = df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_x = StandardScaler().fit_transform(train_x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "A = train_x\n",
    "U, S, V = np.linalg.svd(A)\n",
    "eigvals = S ** 2 / np.cumsum(S)[-1]\n",
    "eigvals2 = S ** 2 / np.sum(S)\n",
    "assert (eigvals == eigvals2).all()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x792 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(30, 11))\n",
    "sing_vals = np.arange(len(eigvals)) + 1\n",
    "\n",
    "plt.plot(sing_vals, eigvals, 'ro-', linewidth=3)\n",
    "plt.title('Scree Plot')\n",
    "plt.xlabel('Principal Component')\n",
    "plt.ylabel('Eigenvalue')\n",
    "\n",
    "leg = plt.legend(['Eigenvalues from SVD'], loc='best', borderpad=0.5,\n",
    "                 shadow=False, markerscale=1)\n",
    "leg.get_frame().set_alpha(0.3)\n",
    "leg.draggable(state=True)\n",
    "plt.savefig('./scree_plot.png')\n",
    "plt.grid(True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "pca = PCA(n_components=13)\n",
    "pca_x = pca.fit_transform(train_x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[t-SNE] Computing 121 nearest neighbors...\n",
      "[t-SNE] Indexed 8124 samples in 0.008s...\n",
      "[t-SNE] Computed neighbors for 8124 samples in 0.803s...\n",
      "[t-SNE] Computed conditional probabilities for sample 1000 / 8124\n",
      "[t-SNE] Computed conditional probabilities for sample 2000 / 8124\n",
      "[t-SNE] Computed conditional probabilities for sample 3000 / 8124\n",
      "[t-SNE] Computed conditional probabilities for sample 4000 / 8124\n",
      "[t-SNE] Computed conditional probabilities for sample 5000 / 8124\n",
      "[t-SNE] Computed conditional probabilities for sample 6000 / 8124\n",
      "[t-SNE] Computed conditional probabilities for sample 7000 / 8124\n",
      "[t-SNE] Computed conditional probabilities for sample 8000 / 8124\n",
      "[t-SNE] Computed conditional probabilities for sample 8124 / 8124\n",
      "[t-SNE] Mean sigma: 0.552771\n",
      "[t-SNE] KL divergence after 250 iterations with early exaggeration: 15.301887\n",
      "[t-SNE] Error after 2000 iterations: 0.673558\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.manifold import TSNE\n",
    "\n",
    "model = TSNE(n_components=2,\n",
    "             learning_rate=1000,\n",
    "             perplexity=40,\n",
    "             early_exaggeration=4,\n",
    "             n_iter=2000,\n",
    "             metric='l2',\n",
    "             random_state=42,\n",
    "             verbose=1)\n",
    "transformed = model.fit_transform(pca_x)  # PCA\n",
    "\n",
    "xs = transformed[:, 0]\n",
    "ys = transformed[:, 1]\n",
    "plt.title('t-SNE')\n",
    "plt.scatter(xs, ys, c=label_color)\n",
    "plt.tight_layout()\n",
    "plt.savefig('./t-SNE-with-one-hot.png')\n",
    "plt.show()"
   ]
  },
  {
   "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": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}