result.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.cm as cm\n",
    "import tikzplotlib\n",
    "# plt.style.use(['science','bright'])\n",
    "# plt.style.use(['fivethirtyeight'])\n",
    "# plt.style.use(['seaborn-paper'])\n",
    "# plt.style.use(['ggplot'])\n",
    "plt.style.use(['default'])\n",
    "plt.rcParams['text.usetex'] = True\n",
    "# plt.rcParams['font.size'] = 15\n",
    "plt.rcParams[\"font.family\"] = \"Times New Roman\"\n",
    "\n",
    "from os import path\n",
    "import pickle\n",
    "import itertools\n",
    "from functools import partial\n",
    "import pprint\n",
    "pp = pprint.PrettyPrinter(indent=4)\n",
    "\n",
    "from util import get_loader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'NOTES:\\nexp = loader.find_by_id(2)\\nlen(qs)\\nqs[0].to_dict()\\nexp.metrics.keys()\\nloader.find_latest().to_dict()\\n'"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loader = get_loader()\n",
    "\n",
    "'''NOTES:\n",
    "exp = loader.find_by_id(2)\n",
    "len(qs)\n",
    "qs[0].to_dict()\n",
    "exp.metrics.keys()\n",
    "loader.find_latest().to_dict()\n",
    "'''"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.set_option('display.max_columns', 500)  # or 1000\n",
    "pd.set_option('display.max_rows', 500)  # or 1000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset_names = {'phishing-websites':'phishing',\n",
    "                 'speed-dating':'speed',\n",
    "                 'bank-marketing':'bank',\n",
    "                }"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": false
   },
   "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>dataset</th>\n",
       "      <th>n</th>\n",
       "      <th>nvar</th>\n",
       "      <th>ncls</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>exp_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6269</th>\n",
       "      <td>iris</td>\n",
       "      <td>150</td>\n",
       "      <td>20</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6301</th>\n",
       "      <td>ilpd</td>\n",
       "      <td>583</td>\n",
       "      <td>46</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6291</th>\n",
       "      <td>breast-w</td>\n",
       "      <td>699</td>\n",
       "      <td>45</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6296</th>\n",
       "      <td>tic-tac-toe</td>\n",
       "      <td>958</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6302</th>\n",
       "      <td>phoneme</td>\n",
       "      <td>5404</td>\n",
       "      <td>25</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            dataset     n  nvar  ncls\n",
       "exp_id                               \n",
       "6269           iris   150    20     3\n",
       "6301           ilpd   583    46     2\n",
       "6291       breast-w   699    45     2\n",
       "6296    tic-tac-toe   958    27     2\n",
       "6302        phoneme  5404    25     2"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# loader.find_latest().to_dict()\n",
    "version = 15\n",
    "\n",
    "query = {\"$and\": [\n",
    "    {\"status\": 'COMPLETED'},\n",
    "    {\"config.ver\": version},\n",
    "    {\"config.iter\": 1},\n",
    "    {\"config.prefix\": ''},\n",
    "    {\"config.criterion\": 'entropy'},\n",
    "    {\"config.costtype\": 'unit'},\n",
    "#     {\"config.tradeoff\": 1},\n",
    "#     {\"_id\": 9},\n",
    "#     {\"config.dup_times\": 1},\n",
    "#     {\"config.dataset\": \"dota\"}\n",
    "        ]}\n",
    "qs = loader.find(query)\n",
    "df = qs.project(on=[\n",
    "    'config.dataset',\n",
    "    \n",
    "    'info.n',\n",
    "    'info.nvar',\n",
    "    'info.ncls',\n",
    "    ])\n",
    "\n",
    "\n",
    "df = df.sort_values(by=['n','nvar','ncls'])\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([['iris', 150, 20, 3],\n",
       "       ['ilpd', 583, 46, 2],\n",
       "       ['breast-w', 699, 45, 2],\n",
       "       ['tic-tac-toe', 958, 27, 2],\n",
       "       ['phoneme', 5404, 25, 2]], dtype=object)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Comparison"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Best trade"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# loader.find_latest().to_dict()\n",
    "# version = 11\n",
    "# version = 12\n",
    "# version = 13\n",
    "# version = 15\n",
    "version = 16\n",
    "costtype = 'unit'\n",
    "# costtype = 'random'\n",
    "\n",
    "query = {\"$and\": [\n",
    "    {\"status\": 'COMPLETED'},\n",
    "    {\"config.ver\": version},\n",
    "#     {\"config.costtype\": 'unit'},\n",
    "    {\"config.costtype\": costtype},\n",
    "#     {'config.prefix': 'spc'},\n",
    "    {'config.prefix': {\"$in\": ['ep']}},\n",
    "#     {\"_id\": 9},\n",
    "#     {\"config.dup_times\": 1},\n",
    "#     {\"config.dataset\": \"iris\"}\n",
    "        ]}\n",
    "qs = loader.find(query)\n",
    "df = qs.project(on=[\n",
    "    'config.ver',\n",
    "    'config.iter',\n",
    "    'config.dataset',\n",
    "    'config.costtype',\n",
    "    'config.rn',\n",
    "    'config.prefix',\n",
    "    'config.criterion',\n",
    "    'config.tradeoff',\n",
    "    \n",
    "    'info.n',\n",
    "    'info.nvar',\n",
    "    'info.ncls',\n",
    "    'info.auc',\n",
    "    'info.cost',\n",
    "    'info.auc2',\n",
    "    'info.cost2',\n",
    "    'info.best_ccp_alpha',\n",
    "    'config.ccp_alphas',\n",
    "    'info.ccp_aucs',\n",
    "    'info.ccp_cost',\n",
    "    'info.runtime',\n",
    "    'info.best_trade',\n",
    "    'info.trade_aucs',\n",
    "    'info.trade_cost',\n",
    "    ])\n",
    "\n",
    "df['name'] = df['prefix']+df['criterion']\n",
    "# df = df.drop(columns=['prefix', 'criterion'])\n",
    "# df = df.sort_values(by=['rn','dataset','costtype','name'])\n",
    "df['dataset'].replace(dataset_names, inplace=True)\n",
    "df = df.sort_values(by=['rn','dataset','name'])\n",
    "df.groupby(['rn','dataset','criterion', 'prefix'])['ver'].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "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>ver</th>\n",
       "      <th>iter</th>\n",
       "      <th>dataset</th>\n",
       "      <th>costtype</th>\n",
       "      <th>rn</th>\n",
       "      <th>prefix</th>\n",
       "      <th>criterion</th>\n",
       "      <th>tradeoff</th>\n",
       "      <th>n</th>\n",
       "      <th>nvar</th>\n",
       "      <th>ncls</th>\n",
       "      <th>auc</th>\n",
       "      <th>cost</th>\n",
       "      <th>auc2</th>\n",
       "      <th>cost2</th>\n",
       "      <th>best_ccp_alpha</th>\n",
       "      <th>ccp_alphas</th>\n",
       "      <th>ccp_aucs</th>\n",
       "      <th>ccp_cost</th>\n",
       "      <th>runtime</th>\n",
       "      <th>best_trade</th>\n",
       "      <th>trade_aucs</th>\n",
       "      <th>trade_cost</th>\n",
       "      <th>name</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>exp_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>11031.0</th>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>bank</td>\n",
       "      <td>unit</td>\n",
       "      <td>42</td>\n",
       "      <td>ep</td>\n",
       "      <td>gini</td>\n",
       "      <td>(0.01, 0.1, 1, 10, 100)</td>\n",
       "      <td>45211</td>\n",
       "      <td>76</td>\n",
       "      <td>2</td>\n",
       "      <td>0.858937</td>\n",
       "      <td>8.149493</td>\n",
       "      <td>0.850157</td>\n",
       "      <td>5.686827</td>\n",
       "      <td>0.00001</td>\n",
       "      <td>(1e-05, 0.0001, 0.001, 0.01, 0.1, 1)</td>\n",
       "      <td>(0.8652446670278175, 0.8077374029722961, 0.5, ...</td>\n",
       "      <td>(5.686826555439601, 2.1884854804558214, 0.0, 0...</td>\n",
       "      <td>221.740165</td>\n",
       "      <td>100.0</td>\n",
       "      <td>(0.8692957749146687, 0.8472254054422645)</td>\n",
       "      <td>(8.149492842921918, 8.15808765443606)</td>\n",
       "      <td>epgini</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11024.0</th>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>bioresponse</td>\n",
       "      <td>unit</td>\n",
       "      <td>42</td>\n",
       "      <td>ep</td>\n",
       "      <td>gini</td>\n",
       "      <td>(0.01, 0.1, 1, 10, 100)</td>\n",
       "      <td>3751</td>\n",
       "      <td>5333</td>\n",
       "      <td>2</td>\n",
       "      <td>0.768556</td>\n",
       "      <td>7.790476</td>\n",
       "      <td>0.767967</td>\n",
       "      <td>7.522667</td>\n",
       "      <td>0.00010</td>\n",
       "      <td>(1e-05, 0.0001, 0.001, 0.01, 0.1, 1)</td>\n",
       "      <td>(0.7517900103104594, 0.7563437965402682, 0.726...</td>\n",
       "      <td>(7.790476190476041, 7.5226666666665825, 0.9999...</td>\n",
       "      <td>3819.320715</td>\n",
       "      <td>100.0</td>\n",
       "      <td>(0.6828244930690802, 0.6583944323519302)</td>\n",
       "      <td>(4.931558935361212, 5.3307984790874485)</td>\n",
       "      <td>epgini</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11036.0</th>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>breast-w</td>\n",
       "      <td>unit</td>\n",
       "      <td>42</td>\n",
       "      <td>ep</td>\n",
       "      <td>gini</td>\n",
       "      <td>(0.01, 0.1, 1, 10, 100)</td>\n",
       "      <td>699</td>\n",
       "      <td>45</td>\n",
       "      <td>2</td>\n",
       "      <td>0.950408</td>\n",
       "      <td>4.419223</td>\n",
       "      <td>0.950408</td>\n",
       "      <td>4.419223</td>\n",
       "      <td>0.00001</td>\n",
       "      <td>(1e-05, 0.0001, 0.001, 0.01, 0.1, 1)</td>\n",
       "      <td>(0.9433876811594203, 0.9433876811594203, 0.920...</td>\n",
       "      <td>(4.419222903885485, 4.419222903885485, 2.99182...</td>\n",
       "      <td>1.040579</td>\n",
       "      <td>1.0</td>\n",
       "      <td>(0.9139492753623188, 0.9375, 0.943387681159420...</td>\n",
       "      <td>(4.312883435582818, 4.543967280163605, 4.41922...</td>\n",
       "      <td>epgini</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11042.0</th>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>covertype</td>\n",
       "      <td>unit</td>\n",
       "      <td>42</td>\n",
       "      <td>ep</td>\n",
       "      <td>gini</td>\n",
       "      <td>(0.01, 0.1, 1, 10, 100)</td>\n",
       "      <td>581012</td>\n",
       "      <td>94</td>\n",
       "      <td>7</td>\n",
       "      <td>0.914436</td>\n",
       "      <td>8.533844</td>\n",
       "      <td>0.913999</td>\n",
       "      <td>8.162279</td>\n",
       "      <td>0.00001</td>\n",
       "      <td>(1e-05, 0.0001, 0.001, 0.01, 0.1, 1)</td>\n",
       "      <td>(0.9154288353415502, 0.9021275257544071, 0.5, ...</td>\n",
       "      <td>(8.162278987074275, 5.225403054277907, 0.0, 0....</td>\n",
       "      <td>1403.338135</td>\n",
       "      <td>10.0</td>\n",
       "      <td>(0.9156433404421559, 0.9100385789374632, 0.836...</td>\n",
       "      <td>(9.23235229033109, 8.523493398244076, 8.217206...</td>\n",
       "      <td>epgini</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11038.0</th>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>default</td>\n",
       "      <td>unit</td>\n",
       "      <td>42</td>\n",
       "      <td>ep</td>\n",
       "      <td>gini</td>\n",
       "      <td>(0.01, 0.1, 1, 10, 100)</td>\n",
       "      <td>30000</td>\n",
       "      <td>112</td>\n",
       "      <td>2</td>\n",
       "      <td>0.704621</td>\n",
       "      <td>9.529762</td>\n",
       "      <td>0.704370</td>\n",
       "      <td>7.104714</td>\n",
       "      <td>0.00001</td>\n",
       "      <td>(1e-05, 0.0001, 0.001, 0.01, 0.1, 1)</td>\n",
       "      <td>(0.69840642535897, 0.6425230029707872, 0.5, 0....</td>\n",
       "      <td>(7.104714285713722, 1.0000000000002534, 0.0, 0...</td>\n",
       "      <td>159.714092</td>\n",
       "      <td>100.0</td>\n",
       "      <td>(0.6960690579097211, 0.6329085606123124)</td>\n",
       "      <td>(9.529761904761441, 8.565619047620784)</td>\n",
       "      <td>epgini</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11028.0</th>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>diabetic</td>\n",
       "      <td>unit</td>\n",
       "      <td>42</td>\n",
       "      <td>ep</td>\n",
       "      <td>gini</td>\n",
       "      <td>(0.01, 0.1, 1, 10, 100)</td>\n",
       "      <td>101766</td>\n",
       "      <td>264</td>\n",
       "      <td>3</td>\n",
       "      <td>0.609315</td>\n",
       "      <td>8.627220</td>\n",
       "      <td>0.611071</td>\n",
       "      <td>7.375770</td>\n",
       "      <td>0.00001</td>\n",
       "      <td>(1e-05, 0.0001, 0.001, 0.01, 0.1, 1)</td>\n",
       "      <td>(0.6001264914297221, 0.5, 0.5, 0.5, 0.5, 0.5)</td>\n",
       "      <td>(7.3757703376144965, 0.0, 0.0, 0.0, 0.0, 0.0)</td>\n",
       "      <td>2580.077058</td>\n",
       "      <td>100.0</td>\n",
       "      <td>(0.5764183176321384, 0.5456512058221596)</td>\n",
       "      <td>(8.625491297023817, 8.171532846716072)</td>\n",
       "      <td>epgini</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11034.0</th>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>iris</td>\n",
       "      <td>unit</td>\n",
       "      <td>42</td>\n",
       "      <td>ep</td>\n",
       "      <td>gini</td>\n",
       "      <td>(0.01, 0.1, 1, 10, 100)</td>\n",
       "      <td>150</td>\n",
       "      <td>20</td>\n",
       "      <td>3</td>\n",
       "      <td>0.945000</td>\n",
       "      <td>3.228571</td>\n",
       "      <td>0.945000</td>\n",
       "      <td>3.228571</td>\n",
       "      <td>0.00001</td>\n",
       "      <td>(1e-05, 0.0001, 0.001, 0.01, 0.1, 1)</td>\n",
       "      <td>(1.0, 1.0, 1.0, 1.0, 0.5, 0.5)</td>\n",
       "      <td>(3.228571428571426, 3.228571428571426, 3.22857...</td>\n",
       "      <td>0.100925</td>\n",
       "      <td>1.0</td>\n",
       "      <td>(0.9500000000000001, 1.0, 1.0, 0.85)</td>\n",
       "      <td>(2.7999999999999985, 3.0095238095238077, 3.228...</td>\n",
       "      <td>epgini</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11039.0</th>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>letter</td>\n",
       "      <td>unit</td>\n",
       "      <td>42</td>\n",
       "      <td>ep</td>\n",
       "      <td>gini</td>\n",
       "      <td>(0.01, 0.1, 1, 10, 100)</td>\n",
       "      <td>20000</td>\n",
       "      <td>80</td>\n",
       "      <td>26</td>\n",
       "      <td>0.949501</td>\n",
       "      <td>8.434071</td>\n",
       "      <td>0.949059</td>\n",
       "      <td>7.917500</td>\n",
       "      <td>0.00010</td>\n",
       "      <td>(1e-05, 0.0001, 0.001, 0.01, 0.1, 1)</td>\n",
       "      <td>(0.9458325463225262, 0.9462609843311832, 0.913...</td>\n",
       "      <td>(8.33428571428629, 7.917500000000579, 5.134071...</td>\n",
       "      <td>89.184625</td>\n",
       "      <td>100.0</td>\n",
       "      <td>(0.9459661183676435, 0.9109750457553224)</td>\n",
       "      <td>(8.43407142857196, 8.204071428572469)</td>\n",
       "      <td>epgini</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11027.0</th>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>musk</td>\n",
       "      <td>unit</td>\n",
       "      <td>42</td>\n",
       "      <td>ep</td>\n",
       "      <td>gini</td>\n",
       "      <td>(0.01, 0.1, 1, 10, 100)</td>\n",
       "      <td>6598</td>\n",
       "      <td>830</td>\n",
       "      <td>2</td>\n",
       "      <td>0.927270</td>\n",
       "      <td>7.037246</td>\n",
       "      <td>0.926788</td>\n",
       "      <td>6.899091</td>\n",
       "      <td>0.00001</td>\n",
       "      <td>(1e-05, 0.0001, 0.001, 0.01, 0.1, 1)</td>\n",
       "      <td>(0.9509716072809054, 0.9440754796542273, 0.746...</td>\n",
       "      <td>(6.899090515374974, 5.7516240796883, 1.3098744...</td>\n",
       "      <td>416.740418</td>\n",
       "      <td>100.0</td>\n",
       "      <td>(0.95221027479092, 0.9021716213367067)</td>\n",
       "      <td>(7.037245560849177, 6.99740147249882)</td>\n",
       "      <td>epgini</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11026.0</th>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>phishing</td>\n",
       "      <td>unit</td>\n",
       "      <td>42</td>\n",
       "      <td>ep</td>\n",
       "      <td>gini</td>\n",
       "      <td>(0.01, 0.1, 1, 10, 100)</td>\n",
       "      <td>11055</td>\n",
       "      <td>46</td>\n",
       "      <td>2</td>\n",
       "      <td>0.984303</td>\n",
       "      <td>6.487464</td>\n",
       "      <td>0.983574</td>\n",
       "      <td>5.955932</td>\n",
       "      <td>0.00001</td>\n",
       "      <td>(1e-05, 0.0001, 0.001, 0.01, 0.1, 1)</td>\n",
       "      <td>(0.9789723032069972, 0.9651553803339518, 0.903...</td>\n",
       "      <td>(5.955931765313925, 3.7109072111656776, 1.4307...</td>\n",
       "      <td>19.153621</td>\n",
       "      <td>100.0</td>\n",
       "      <td>(0.9790551285449245, 0.965367413199046)</td>\n",
       "      <td>(6.487464461100822, 6.112044455931645)</td>\n",
       "      <td>epgini</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11044.0</th>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>phoneme</td>\n",
       "      <td>unit</td>\n",
       "      <td>42</td>\n",
       "      <td>ep</td>\n",
       "      <td>gini</td>\n",
       "      <td>(0.01, 0.1, 1, 10, 100)</td>\n",
       "      <td>5404</td>\n",
       "      <td>25</td>\n",
       "      <td>2</td>\n",
       "      <td>0.800180</td>\n",
       "      <td>6.322052</td>\n",
       "      <td>0.800149</td>\n",
       "      <td>6.315970</td>\n",
       "      <td>0.00001</td>\n",
       "      <td>(1e-05, 0.0001, 0.001, 0.01, 0.1, 1)</td>\n",
       "      <td>(0.7945437781948698, 0.7887648589021701, 0.749...</td>\n",
       "      <td>(6.315970386039141, 6.030671602326823, 4.62136...</td>\n",
       "      <td>1.740524</td>\n",
       "      <td>10.0</td>\n",
       "      <td>(0.7791497909051994, 0.7944367611709309, 0.782...</td>\n",
       "      <td>(7.044420941300896, 6.322051824431526, 6.48175...</td>\n",
       "      <td>epgini</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11035.0</th>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>shoppers</td>\n",
       "      <td>unit</td>\n",
       "      <td>42</td>\n",
       "      <td>ep</td>\n",
       "      <td>gini</td>\n",
       "      <td>(0.01, 0.1, 1, 10, 100)</td>\n",
       "      <td>12330</td>\n",
       "      <td>454</td>\n",
       "      <td>2</td>\n",
       "      <td>0.848818</td>\n",
       "      <td>8.527285</td>\n",
       "      <td>0.841444</td>\n",
       "      <td>6.821342</td>\n",
       "      <td>0.00001</td>\n",
       "      <td>(1e-05, 0.0001, 0.001, 0.01, 0.1, 1)</td>\n",
       "      <td>(0.8685396589321782, 0.773610455125564, 0.5, 0...</td>\n",
       "      <td>(6.821341675356597, 1.2438883095818825, 0.0, 0...</td>\n",
       "      <td>544.943222</td>\n",
       "      <td>100.0</td>\n",
       "      <td>(0.8710820914270785, 0.8559480861412305)</td>\n",
       "      <td>(8.527285366701417, 8.50341791217696)</td>\n",
       "      <td>epgini</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11029.0</th>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>spambase</td>\n",
       "      <td>unit</td>\n",
       "      <td>42</td>\n",
       "      <td>ep</td>\n",
       "      <td>gini</td>\n",
       "      <td>(0.01, 0.1, 1, 10, 100)</td>\n",
       "      <td>4601</td>\n",
       "      <td>285</td>\n",
       "      <td>2</td>\n",
       "      <td>0.918190</td>\n",
       "      <td>9.509627</td>\n",
       "      <td>0.918168</td>\n",
       "      <td>9.452795</td>\n",
       "      <td>0.00001</td>\n",
       "      <td>(1e-05, 0.0001, 0.001, 0.01, 0.1, 1)</td>\n",
       "      <td>(0.9209489296817759, 0.9205726846076161, 0.787...</td>\n",
       "      <td>(9.452795031055933, 8.874534161490748, 1.42142...</td>\n",
       "      <td>104.682156</td>\n",
       "      <td>100.0</td>\n",
       "      <td>(0.9208301154478307, 0.8817798372244996)</td>\n",
       "      <td>(9.50962732919258, 8.790062111801076)</td>\n",
       "      <td>epgini</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11033.0</th>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>speed</td>\n",
       "      <td>unit</td>\n",
       "      <td>42</td>\n",
       "      <td>ep</td>\n",
       "      <td>gini</td>\n",
       "      <td>(0.01, 0.1, 1, 10, 100)</td>\n",
       "      <td>8378</td>\n",
       "      <td>733</td>\n",
       "      <td>2</td>\n",
       "      <td>0.735745</td>\n",
       "      <td>7.940314</td>\n",
       "      <td>0.753476</td>\n",
       "      <td>4.828274</td>\n",
       "      <td>0.00010</td>\n",
       "      <td>(1e-05, 0.0001, 0.001, 0.01, 0.1, 1)</td>\n",
       "      <td>(0.7339699792960663, 0.7528105590062112, 0.5, ...</td>\n",
       "      <td>(7.677182810368278, 4.828274215552317, 0.0, 0....</td>\n",
       "      <td>606.846513</td>\n",
       "      <td>100.0</td>\n",
       "      <td>(0.735, 0.706552795031056)</td>\n",
       "      <td>(7.9403137789904035, 7.8115620736692035)</td>\n",
       "      <td>epgini</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11032.0</th>\n",
       "      <td>16</td>\n",
       "      <td>1</td>\n",
       "      <td>tic-tac-toe</td>\n",
       "      <td>unit</td>\n",
       "      <td>42</td>\n",
       "      <td>ep</td>\n",
       "      <td>gini</td>\n",
       "      <td>(0.01, 0.1, 1, 10, 100)</td>\n",
       "      <td>958</td>\n",
       "      <td>27</td>\n",
       "      <td>2</td>\n",
       "      <td>0.961731</td>\n",
       "      <td>4.747761</td>\n",
       "      <td>0.961731</td>\n",
       "      <td>4.747761</td>\n",
       "      <td>0.00001</td>\n",
       "      <td>(1e-05, 0.0001, 0.001, 0.01, 0.1, 1)</td>\n",
       "      <td>(0.9603174603174603, 0.9603174603174603, 0.959...</td>\n",
       "      <td>(4.7477611940298665, 4.7477611940298665, 4.519...</td>\n",
       "      <td>1.362587</td>\n",
       "      <td>100.0</td>\n",
       "      <td>(0.9603174603174603, 0.9297739297739297)</td>\n",
       "      <td>(4.7477611940298665, 5.317910447761205)</td>\n",
       "      <td>epgini</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         ver  iter      dataset costtype  rn prefix criterion  \\\n",
       "exp_id                                                          \n",
       "11031.0   16     1         bank     unit  42     ep      gini   \n",
       "11024.0   16     1  bioresponse     unit  42     ep      gini   \n",
       "11036.0   16     1     breast-w     unit  42     ep      gini   \n",
       "11042.0   16     1    covertype     unit  42     ep      gini   \n",
       "11038.0   16     1      default     unit  42     ep      gini   \n",
       "11028.0   16     1     diabetic     unit  42     ep      gini   \n",
       "11034.0   16     1         iris     unit  42     ep      gini   \n",
       "11039.0   16     1       letter     unit  42     ep      gini   \n",
       "11027.0   16     1         musk     unit  42     ep      gini   \n",
       "11026.0   16     1     phishing     unit  42     ep      gini   \n",
       "11044.0   16     1      phoneme     unit  42     ep      gini   \n",
       "11035.0   16     1     shoppers     unit  42     ep      gini   \n",
       "11029.0   16     1     spambase     unit  42     ep      gini   \n",
       "11033.0   16     1        speed     unit  42     ep      gini   \n",
       "11032.0   16     1  tic-tac-toe     unit  42     ep      gini   \n",
       "\n",
       "                        tradeoff       n  nvar  ncls       auc      cost  \\\n",
       "exp_id                                                                     \n",
       "11031.0  (0.01, 0.1, 1, 10, 100)   45211    76     2  0.858937  8.149493   \n",
       "11024.0  (0.01, 0.1, 1, 10, 100)    3751  5333     2  0.768556  7.790476   \n",
       "11036.0  (0.01, 0.1, 1, 10, 100)     699    45     2  0.950408  4.419223   \n",
       "11042.0  (0.01, 0.1, 1, 10, 100)  581012    94     7  0.914436  8.533844   \n",
       "11038.0  (0.01, 0.1, 1, 10, 100)   30000   112     2  0.704621  9.529762   \n",
       "11028.0  (0.01, 0.1, 1, 10, 100)  101766   264     3  0.609315  8.627220   \n",
       "11034.0  (0.01, 0.1, 1, 10, 100)     150    20     3  0.945000  3.228571   \n",
       "11039.0  (0.01, 0.1, 1, 10, 100)   20000    80    26  0.949501  8.434071   \n",
       "11027.0  (0.01, 0.1, 1, 10, 100)    6598   830     2  0.927270  7.037246   \n",
       "11026.0  (0.01, 0.1, 1, 10, 100)   11055    46     2  0.984303  6.487464   \n",
       "11044.0  (0.01, 0.1, 1, 10, 100)    5404    25     2  0.800180  6.322052   \n",
       "11035.0  (0.01, 0.1, 1, 10, 100)   12330   454     2  0.848818  8.527285   \n",
       "11029.0  (0.01, 0.1, 1, 10, 100)    4601   285     2  0.918190  9.509627   \n",
       "11033.0  (0.01, 0.1, 1, 10, 100)    8378   733     2  0.735745  7.940314   \n",
       "11032.0  (0.01, 0.1, 1, 10, 100)     958    27     2  0.961731  4.747761   \n",
       "\n",
       "             auc2     cost2  best_ccp_alpha  \\\n",
       "exp_id                                        \n",
       "11031.0  0.850157  5.686827         0.00001   \n",
       "11024.0  0.767967  7.522667         0.00010   \n",
       "11036.0  0.950408  4.419223         0.00001   \n",
       "11042.0  0.913999  8.162279         0.00001   \n",
       "11038.0  0.704370  7.104714         0.00001   \n",
       "11028.0  0.611071  7.375770         0.00001   \n",
       "11034.0  0.945000  3.228571         0.00001   \n",
       "11039.0  0.949059  7.917500         0.00010   \n",
       "11027.0  0.926788  6.899091         0.00001   \n",
       "11026.0  0.983574  5.955932         0.00001   \n",
       "11044.0  0.800149  6.315970         0.00001   \n",
       "11035.0  0.841444  6.821342         0.00001   \n",
       "11029.0  0.918168  9.452795         0.00001   \n",
       "11033.0  0.753476  4.828274         0.00010   \n",
       "11032.0  0.961731  4.747761         0.00001   \n",
       "\n",
       "                                   ccp_alphas  \\\n",
       "exp_id                                          \n",
       "11031.0  (1e-05, 0.0001, 0.001, 0.01, 0.1, 1)   \n",
       "11024.0  (1e-05, 0.0001, 0.001, 0.01, 0.1, 1)   \n",
       "11036.0  (1e-05, 0.0001, 0.001, 0.01, 0.1, 1)   \n",
       "11042.0  (1e-05, 0.0001, 0.001, 0.01, 0.1, 1)   \n",
       "11038.0  (1e-05, 0.0001, 0.001, 0.01, 0.1, 1)   \n",
       "11028.0  (1e-05, 0.0001, 0.001, 0.01, 0.1, 1)   \n",
       "11034.0  (1e-05, 0.0001, 0.001, 0.01, 0.1, 1)   \n",
       "11039.0  (1e-05, 0.0001, 0.001, 0.01, 0.1, 1)   \n",
       "11027.0  (1e-05, 0.0001, 0.001, 0.01, 0.1, 1)   \n",
       "11026.0  (1e-05, 0.0001, 0.001, 0.01, 0.1, 1)   \n",
       "11044.0  (1e-05, 0.0001, 0.001, 0.01, 0.1, 1)   \n",
       "11035.0  (1e-05, 0.0001, 0.001, 0.01, 0.1, 1)   \n",
       "11029.0  (1e-05, 0.0001, 0.001, 0.01, 0.1, 1)   \n",
       "11033.0  (1e-05, 0.0001, 0.001, 0.01, 0.1, 1)   \n",
       "11032.0  (1e-05, 0.0001, 0.001, 0.01, 0.1, 1)   \n",
       "\n",
       "                                                  ccp_aucs  \\\n",
       "exp_id                                                       \n",
       "11031.0  (0.8652446670278175, 0.8077374029722961, 0.5, ...   \n",
       "11024.0  (0.7517900103104594, 0.7563437965402682, 0.726...   \n",
       "11036.0  (0.9433876811594203, 0.9433876811594203, 0.920...   \n",
       "11042.0  (0.9154288353415502, 0.9021275257544071, 0.5, ...   \n",
       "11038.0  (0.69840642535897, 0.6425230029707872, 0.5, 0....   \n",
       "11028.0      (0.6001264914297221, 0.5, 0.5, 0.5, 0.5, 0.5)   \n",
       "11034.0                     (1.0, 1.0, 1.0, 1.0, 0.5, 0.5)   \n",
       "11039.0  (0.9458325463225262, 0.9462609843311832, 0.913...   \n",
       "11027.0  (0.9509716072809054, 0.9440754796542273, 0.746...   \n",
       "11026.0  (0.9789723032069972, 0.9651553803339518, 0.903...   \n",
       "11044.0  (0.7945437781948698, 0.7887648589021701, 0.749...   \n",
       "11035.0  (0.8685396589321782, 0.773610455125564, 0.5, 0...   \n",
       "11029.0  (0.9209489296817759, 0.9205726846076161, 0.787...   \n",
       "11033.0  (0.7339699792960663, 0.7528105590062112, 0.5, ...   \n",
       "11032.0  (0.9603174603174603, 0.9603174603174603, 0.959...   \n",
       "\n",
       "                                                  ccp_cost      runtime  \\\n",
       "exp_id                                                                    \n",
       "11031.0  (5.686826555439601, 2.1884854804558214, 0.0, 0...   221.740165   \n",
       "11024.0  (7.790476190476041, 7.5226666666665825, 0.9999...  3819.320715   \n",
       "11036.0  (4.419222903885485, 4.419222903885485, 2.99182...     1.040579   \n",
       "11042.0  (8.162278987074275, 5.225403054277907, 0.0, 0....  1403.338135   \n",
       "11038.0  (7.104714285713722, 1.0000000000002534, 0.0, 0...   159.714092   \n",
       "11028.0      (7.3757703376144965, 0.0, 0.0, 0.0, 0.0, 0.0)  2580.077058   \n",
       "11034.0  (3.228571428571426, 3.228571428571426, 3.22857...     0.100925   \n",
       "11039.0  (8.33428571428629, 7.917500000000579, 5.134071...    89.184625   \n",
       "11027.0  (6.899090515374974, 5.7516240796883, 1.3098744...   416.740418   \n",
       "11026.0  (5.955931765313925, 3.7109072111656776, 1.4307...    19.153621   \n",
       "11044.0  (6.315970386039141, 6.030671602326823, 4.62136...     1.740524   \n",
       "11035.0  (6.821341675356597, 1.2438883095818825, 0.0, 0...   544.943222   \n",
       "11029.0  (9.452795031055933, 8.874534161490748, 1.42142...   104.682156   \n",
       "11033.0  (7.677182810368278, 4.828274215552317, 0.0, 0....   606.846513   \n",
       "11032.0  (4.7477611940298665, 4.7477611940298665, 4.519...     1.362587   \n",
       "\n",
       "         best_trade                                         trade_aucs  \\\n",
       "exp_id                                                                   \n",
       "11031.0       100.0           (0.8692957749146687, 0.8472254054422645)   \n",
       "11024.0       100.0           (0.6828244930690802, 0.6583944323519302)   \n",
       "11036.0         1.0  (0.9139492753623188, 0.9375, 0.943387681159420...   \n",
       "11042.0        10.0  (0.9156433404421559, 0.9100385789374632, 0.836...   \n",
       "11038.0       100.0           (0.6960690579097211, 0.6329085606123124)   \n",
       "11028.0       100.0           (0.5764183176321384, 0.5456512058221596)   \n",
       "11034.0         1.0               (0.9500000000000001, 1.0, 1.0, 0.85)   \n",
       "11039.0       100.0           (0.9459661183676435, 0.9109750457553224)   \n",
       "11027.0       100.0             (0.95221027479092, 0.9021716213367067)   \n",
       "11026.0       100.0            (0.9790551285449245, 0.965367413199046)   \n",
       "11044.0        10.0  (0.7791497909051994, 0.7944367611709309, 0.782...   \n",
       "11035.0       100.0           (0.8710820914270785, 0.8559480861412305)   \n",
       "11029.0       100.0           (0.9208301154478307, 0.8817798372244996)   \n",
       "11033.0       100.0                         (0.735, 0.706552795031056)   \n",
       "11032.0       100.0           (0.9603174603174603, 0.9297739297739297)   \n",
       "\n",
       "                                                trade_cost    name  \n",
       "exp_id                                                              \n",
       "11031.0              (8.149492842921918, 8.15808765443606)  epgini  \n",
       "11024.0            (4.931558935361212, 5.3307984790874485)  epgini  \n",
       "11036.0  (4.312883435582818, 4.543967280163605, 4.41922...  epgini  \n",
       "11042.0  (9.23235229033109, 8.523493398244076, 8.217206...  epgini  \n",
       "11038.0             (9.529761904761441, 8.565619047620784)  epgini  \n",
       "11028.0             (8.625491297023817, 8.171532846716072)  epgini  \n",
       "11034.0  (2.7999999999999985, 3.0095238095238077, 3.228...  epgini  \n",
       "11039.0              (8.43407142857196, 8.204071428572469)  epgini  \n",
       "11027.0              (7.037245560849177, 6.99740147249882)  epgini  \n",
       "11026.0             (6.487464461100822, 6.112044455931645)  epgini  \n",
       "11044.0  (7.044420941300896, 6.322051824431526, 6.48175...  epgini  \n",
       "11035.0              (8.527285366701417, 8.50341791217696)  epgini  \n",
       "11029.0              (9.50962732919258, 8.790062111801076)  epgini  \n",
       "11033.0           (7.9403137789904035, 7.8115620736692035)  epgini  \n",
       "11032.0            (4.7477611940298665, 5.317910447761205)  epgini  "
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df['criterion'] == 'gini']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 345,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "rn    dataset      criterion  prefix\n",
       "42    bank         entropy              1\n",
       "                              b         1\n",
       "                              ep        1\n",
       "                              ep0       1\n",
       "                   gini                 1\n",
       "                                       ..\n",
       "2021  tic-tac-toe  entropy    ep0       1\n",
       "                   gini                 1\n",
       "                              ep        1\n",
       "                   subm       sc        1\n",
       "                              spc       1\n",
       "Name: ver, Length: 774, dtype: int64"
      ]
     },
     "execution_count": 345,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# loader.find_latest().to_dict()\n",
    "# version = 11\n",
    "# version = 12\n",
    "# version = 13\n",
    "# version = 15\n",
    "version = 16\n",
    "costtype = 'unit'\n",
    "# costtype = 'random'\n",
    "\n",
    "query = {\"$and\": [\n",
    "    {\"status\": 'COMPLETED'},\n",
    "    {\"config.ver\": version},\n",
    "    {\"config.costtype\": costtype},\n",
    "#     {'config.prefix': 'spc'},\n",
    "#     {'config.prefix': {\"$in\": ['ep','ep0','spc']}},\n",
    "#     {\"_id\": 9},\n",
    "#     {\"config.dup_times\": 1},\n",
    "#     {\"config.dataset\": \"iris\"}\n",
    "        ]}\n",
    "qs = loader.find(query)\n",
    "df = qs.project(on=[\n",
    "    'config.ver',\n",
    "    'config.iter',\n",
    "    'config.dataset',\n",
    "    'config.costtype',\n",
    "    'config.rn',\n",
    "    'config.prefix',\n",
    "    'config.criterion',\n",
    "    'config.tradeoff',\n",
    "    \n",
    "    'info.n',\n",
    "    'info.nvar',\n",
    "    'info.ncls',\n",
    "    'info.auc',\n",
    "    'info.cost',\n",
    "    'info.auc2',\n",
    "    'info.cost2',\n",
    "    'info.best_ccp_alpha',\n",
    "    'config.ccp_alphas',\n",
    "    'info.ccp_aucs',\n",
    "    'info.ccp_cost',\n",
    "    'info.runtime',\n",
    "    ])\n",
    "\n",
    "df['name'] = df['prefix']+df['criterion']\n",
    "# df = df.drop(columns=['prefix', 'criterion'])\n",
    "# df = df.sort_values(by=['rn','dataset','costtype','name'])\n",
    "df['dataset'].replace(dataset_names, inplace=True)\n",
    "df = df.sort_values(by=['rn','dataset','name'])\n",
    "df.groupby(['rn','dataset','criterion', 'prefix'])['ver'].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 478,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df[df['dataset']=='dota2'].groupby(['rn','dataset','criterion'])['ver'].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 477,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# df[df['dataset']=='iris'].sort_values(by=['rn','prefix','criterion'])\n",
    "# df[df['dataset'].isin(['dota2'])].sort_values(by=['iter','prefix','criterion'])[['dataset','iter','prefix','criterion']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 346,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"7\" halign=\"left\">ROC AUC</th>\n",
       "      <th colspan=\"7\" halign=\"left\">Cost</th>\n",
       "      <th colspan=\"7\" halign=\"left\">ROC AUC (ccp)</th>\n",
       "      <th colspan=\"7\" halign=\"left\">Cost (ccp)</th>\n",
       "      <th colspan=\"7\" halign=\"left\">Runtime</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>C4.5</th>\n",
       "      <th>EC4.5</th>\n",
       "      <th>CART</th>\n",
       "      <th>ECART</th>\n",
       "      <th>ASR</th>\n",
       "      <th>IP</th>\n",
       "      <th>Bal</th>\n",
       "      <th>C4.5</th>\n",
       "      <th>EC4.5</th>\n",
       "      <th>CART</th>\n",
       "      <th>ECART</th>\n",
       "      <th>ASR</th>\n",
       "      <th>IP</th>\n",
       "      <th>Bal</th>\n",
       "      <th>C4.5</th>\n",
       "      <th>EC4.5</th>\n",
       "      <th>CART</th>\n",
       "      <th>ECART</th>\n",
       "      <th>ASR</th>\n",
       "      <th>IP</th>\n",
       "      <th>Bal</th>\n",
       "      <th>C4.5</th>\n",
       "      <th>EC4.5</th>\n",
       "      <th>CART</th>\n",
       "      <th>ECART</th>\n",
       "      <th>ASR</th>\n",
       "      <th>IP</th>\n",
       "      <th>Bal</th>\n",
       "      <th>C4.5</th>\n",
       "      <th>EC4.5</th>\n",
       "      <th>CART</th>\n",
       "      <th>ECART</th>\n",
       "      <th>ASR</th>\n",
       "      <th>IP</th>\n",
       "      <th>Bal</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dataset</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>bank</th>\n",
       "      <td>0.897834</td>\n",
       "      <td>0.874647</td>\n",
       "      <td>0.899850</td>\n",
       "      <td>0.860747</td>\n",
       "      <td>0.834065</td>\n",
       "      <td>0.855202</td>\n",
       "      <td>0.824304</td>\n",
       "      <td>13.150232</td>\n",
       "      <td>8.245344</td>\n",
       "      <td>18.692773</td>\n",
       "      <td>8.072986</td>\n",
       "      <td>8.110905</td>\n",
       "      <td>8.229936</td>\n",
       "      <td>29.713363</td>\n",
       "      <td>0.901538</td>\n",
       "      <td>0.874584</td>\n",
       "      <td>0.906445</td>\n",
       "      <td>0.851543</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>11.879325</td>\n",
       "      <td>7.949645</td>\n",
       "      <td>15.155528</td>\n",
       "      <td>5.607400</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>53.934706</td>\n",
       "      <td>232.459637</td>\n",
       "      <td>72.040812</td>\n",
       "      <td>231.279182</td>\n",
       "      <td>188.073861</td>\n",
       "      <td>34.954801</td>\n",
       "      <td>50.070138</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bioresponse</th>\n",
       "      <td>0.747969</td>\n",
       "      <td>0.767715</td>\n",
       "      <td>0.742571</td>\n",
       "      <td>0.766073</td>\n",
       "      <td>0.754778</td>\n",
       "      <td>0.757020</td>\n",
       "      <td>NaN</td>\n",
       "      <td>12.030019</td>\n",
       "      <td>7.824457</td>\n",
       "      <td>16.111543</td>\n",
       "      <td>7.811886</td>\n",
       "      <td>7.875886</td>\n",
       "      <td>7.625219</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.747969</td>\n",
       "      <td>0.768195</td>\n",
       "      <td>0.758761</td>\n",
       "      <td>0.760012</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>12.030019</td>\n",
       "      <td>7.204952</td>\n",
       "      <td>5.042362</td>\n",
       "      <td>6.375848</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4593.632122</td>\n",
       "      <td>4215.667313</td>\n",
       "      <td>6192.435255</td>\n",
       "      <td>4629.768001</td>\n",
       "      <td>4127.723269</td>\n",
       "      <td>2700.924505</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>breast-w</th>\n",
       "      <td>0.942754</td>\n",
       "      <td>0.947758</td>\n",
       "      <td>0.936141</td>\n",
       "      <td>0.947509</td>\n",
       "      <td>0.954688</td>\n",
       "      <td>0.948981</td>\n",
       "      <td>0.970811</td>\n",
       "      <td>3.662577</td>\n",
       "      <td>3.983231</td>\n",
       "      <td>4.513701</td>\n",
       "      <td>4.215133</td>\n",
       "      <td>4.240082</td>\n",
       "      <td>3.466258</td>\n",
       "      <td>23.131288</td>\n",
       "      <td>0.961821</td>\n",
       "      <td>0.953012</td>\n",
       "      <td>0.927446</td>\n",
       "      <td>0.943614</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.817178</td>\n",
       "      <td>3.233129</td>\n",
       "      <td>3.851943</td>\n",
       "      <td>4.011452</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.191986</td>\n",
       "      <td>1.001592</td>\n",
       "      <td>0.192608</td>\n",
       "      <td>1.010131</td>\n",
       "      <td>0.807835</td>\n",
       "      <td>0.151000</td>\n",
       "      <td>0.430479</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>covertype</th>\n",
       "      <td>0.945313</td>\n",
       "      <td>0.916533</td>\n",
       "      <td>0.944080</td>\n",
       "      <td>0.917318</td>\n",
       "      <td>0.848372</td>\n",
       "      <td>0.835172</td>\n",
       "      <td>0.843646</td>\n",
       "      <td>12.811080</td>\n",
       "      <td>9.247226</td>\n",
       "      <td>14.517318</td>\n",
       "      <td>8.952649</td>\n",
       "      <td>8.265283</td>\n",
       "      <td>8.348718</td>\n",
       "      <td>48.114243</td>\n",
       "      <td>0.945406</td>\n",
       "      <td>0.916550</td>\n",
       "      <td>0.943142</td>\n",
       "      <td>0.917200</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>12.747302</td>\n",
       "      <td>9.162628</td>\n",
       "      <td>13.861205</td>\n",
       "      <td>8.489265</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>341.836037</td>\n",
       "      <td>1597.349569</td>\n",
       "      <td>402.592217</td>\n",
       "      <td>1539.147105</td>\n",
       "      <td>1136.326300</td>\n",
       "      <td>268.721726</td>\n",
       "      <td>524.012348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>default</th>\n",
       "      <td>0.699991</td>\n",
       "      <td>0.717587</td>\n",
       "      <td>0.698888</td>\n",
       "      <td>0.705811</td>\n",
       "      <td>0.638023</td>\n",
       "      <td>0.681492</td>\n",
       "      <td>0.667648</td>\n",
       "      <td>18.040419</td>\n",
       "      <td>9.988438</td>\n",
       "      <td>19.931981</td>\n",
       "      <td>9.457657</td>\n",
       "      <td>8.540619</td>\n",
       "      <td>8.622629</td>\n",
       "      <td>60.468781</td>\n",
       "      <td>0.700115</td>\n",
       "      <td>0.716950</td>\n",
       "      <td>0.694239</td>\n",
       "      <td>0.701388</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>17.942152</td>\n",
       "      <td>8.985381</td>\n",
       "      <td>18.328467</td>\n",
       "      <td>6.910095</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>44.764109</td>\n",
       "      <td>174.722208</td>\n",
       "      <td>45.350617</td>\n",
       "      <td>165.505596</td>\n",
       "      <td>127.302811</td>\n",
       "      <td>24.118354</td>\n",
       "      <td>66.731410</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>diabetic</th>\n",
       "      <td>0.632339</td>\n",
       "      <td>0.611578</td>\n",
       "      <td>0.630964</td>\n",
       "      <td>0.610370</td>\n",
       "      <td>0.564986</td>\n",
       "      <td>0.569121</td>\n",
       "      <td>0.565388</td>\n",
       "      <td>17.551583</td>\n",
       "      <td>8.636224</td>\n",
       "      <td>17.830409</td>\n",
       "      <td>8.641362</td>\n",
       "      <td>8.176789</td>\n",
       "      <td>8.205961</td>\n",
       "      <td>152.077543</td>\n",
       "      <td>0.639558</td>\n",
       "      <td>0.611693</td>\n",
       "      <td>0.630972</td>\n",
       "      <td>0.611505</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.499597</td>\n",
       "      <td>8.340799</td>\n",
       "      <td>17.824112</td>\n",
       "      <td>7.478029</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1035.128554</td>\n",
       "      <td>2737.601433</td>\n",
       "      <td>1138.214176</td>\n",
       "      <td>2869.144890</td>\n",
       "      <td>2062.491981</td>\n",
       "      <td>601.082457</td>\n",
       "      <td>4489.117611</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dota2</th>\n",
       "      <td>0.552002</td>\n",
       "      <td>0.534855</td>\n",
       "      <td>0.556627</td>\n",
       "      <td>0.534855</td>\n",
       "      <td>0.534695</td>\n",
       "      <td>0.534997</td>\n",
       "      <td>0.535006</td>\n",
       "      <td>29.603068</td>\n",
       "      <td>9.145354</td>\n",
       "      <td>28.085247</td>\n",
       "      <td>9.145354</td>\n",
       "      <td>9.145354</td>\n",
       "      <td>9.146455</td>\n",
       "      <td>36.325198</td>\n",
       "      <td>0.552002</td>\n",
       "      <td>0.535639</td>\n",
       "      <td>0.556627</td>\n",
       "      <td>0.536333</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>29.603068</td>\n",
       "      <td>7.226604</td>\n",
       "      <td>28.085247</td>\n",
       "      <td>5.790070</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1694.156266</td>\n",
       "      <td>3569.622728</td>\n",
       "      <td>1573.224023</td>\n",
       "      <td>3681.001757</td>\n",
       "      <td>2941.977681</td>\n",
       "      <td>471.520335</td>\n",
       "      <td>995.454448</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>electricity</th>\n",
       "      <td>0.817044</td>\n",
       "      <td>0.806567</td>\n",
       "      <td>0.815470</td>\n",
       "      <td>0.806567</td>\n",
       "      <td>0.806298</td>\n",
       "      <td>0.804139</td>\n",
       "      <td>0.801348</td>\n",
       "      <td>10.337327</td>\n",
       "      <td>8.405057</td>\n",
       "      <td>10.950342</td>\n",
       "      <td>8.405057</td>\n",
       "      <td>8.405057</td>\n",
       "      <td>8.515616</td>\n",
       "      <td>27.563527</td>\n",
       "      <td>0.817173</td>\n",
       "      <td>0.806549</td>\n",
       "      <td>0.813637</td>\n",
       "      <td>0.805309</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10.318466</td>\n",
       "      <td>8.227903</td>\n",
       "      <td>7.950840</td>\n",
       "      <td>7.867194</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.368101</td>\n",
       "      <td>28.723873</td>\n",
       "      <td>6.364826</td>\n",
       "      <td>28.117674</td>\n",
       "      <td>24.080955</td>\n",
       "      <td>4.706966</td>\n",
       "      <td>5.879775</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>firewall</th>\n",
       "      <td>0.946090</td>\n",
       "      <td>0.950280</td>\n",
       "      <td>0.946090</td>\n",
       "      <td>0.950280</td>\n",
       "      <td>0.950280</td>\n",
       "      <td>0.945758</td>\n",
       "      <td>0.947367</td>\n",
       "      <td>3.360162</td>\n",
       "      <td>3.311609</td>\n",
       "      <td>3.360162</td>\n",
       "      <td>3.311609</td>\n",
       "      <td>3.311609</td>\n",
       "      <td>2.887105</td>\n",
       "      <td>26.162312</td>\n",
       "      <td>0.951344</td>\n",
       "      <td>0.953798</td>\n",
       "      <td>0.956854</td>\n",
       "      <td>0.957780</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.653049</td>\n",
       "      <td>2.844307</td>\n",
       "      <td>1.687075</td>\n",
       "      <td>2.703543</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.207097</td>\n",
       "      <td>2.818632</td>\n",
       "      <td>2.148224</td>\n",
       "      <td>2.786070</td>\n",
       "      <td>2.696570</td>\n",
       "      <td>2.683861</td>\n",
       "      <td>2.398550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ilpd</th>\n",
       "      <td>0.529022</td>\n",
       "      <td>0.528207</td>\n",
       "      <td>0.525124</td>\n",
       "      <td>0.510737</td>\n",
       "      <td>0.517931</td>\n",
       "      <td>0.514989</td>\n",
       "      <td>0.509709</td>\n",
       "      <td>7.310565</td>\n",
       "      <td>6.406388</td>\n",
       "      <td>7.604914</td>\n",
       "      <td>6.358722</td>\n",
       "      <td>6.551351</td>\n",
       "      <td>6.291892</td>\n",
       "      <td>34.377887</td>\n",
       "      <td>0.514989</td>\n",
       "      <td>0.508186</td>\n",
       "      <td>0.504040</td>\n",
       "      <td>0.494933</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.445209</td>\n",
       "      <td>3.866830</td>\n",
       "      <td>4.556757</td>\n",
       "      <td>4.816708</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.423081</td>\n",
       "      <td>1.963373</td>\n",
       "      <td>0.388731</td>\n",
       "      <td>1.799762</td>\n",
       "      <td>1.438511</td>\n",
       "      <td>0.327805</td>\n",
       "      <td>0.716311</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>iris</th>\n",
       "      <td>0.945000</td>\n",
       "      <td>0.957000</td>\n",
       "      <td>0.945000</td>\n",
       "      <td>0.957000</td>\n",
       "      <td>0.933833</td>\n",
       "      <td>0.938333</td>\n",
       "      <td>0.925167</td>\n",
       "      <td>2.563810</td>\n",
       "      <td>3.196190</td>\n",
       "      <td>2.554286</td>\n",
       "      <td>3.177143</td>\n",
       "      <td>4.093333</td>\n",
       "      <td>3.668571</td>\n",
       "      <td>11.727619</td>\n",
       "      <td>0.945000</td>\n",
       "      <td>0.957000</td>\n",
       "      <td>0.945000</td>\n",
       "      <td>0.957000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.563810</td>\n",
       "      <td>3.196190</td>\n",
       "      <td>2.554286</td>\n",
       "      <td>3.177143</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.019084</td>\n",
       "      <td>0.099292</td>\n",
       "      <td>0.017226</td>\n",
       "      <td>0.094784</td>\n",
       "      <td>0.099996</td>\n",
       "      <td>0.025907</td>\n",
       "      <td>0.052904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>letter</th>\n",
       "      <td>0.959188</td>\n",
       "      <td>0.956326</td>\n",
       "      <td>0.959222</td>\n",
       "      <td>0.945474</td>\n",
       "      <td>0.882100</td>\n",
       "      <td>0.928720</td>\n",
       "      <td>0.901861</td>\n",
       "      <td>9.835014</td>\n",
       "      <td>8.722757</td>\n",
       "      <td>11.928129</td>\n",
       "      <td>8.437014</td>\n",
       "      <td>8.229657</td>\n",
       "      <td>8.225571</td>\n",
       "      <td>40.663186</td>\n",
       "      <td>0.960861</td>\n",
       "      <td>0.957054</td>\n",
       "      <td>0.959993</td>\n",
       "      <td>0.945753</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>9.035957</td>\n",
       "      <td>8.421243</td>\n",
       "      <td>11.375229</td>\n",
       "      <td>8.026686</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>18.811961</td>\n",
       "      <td>96.639955</td>\n",
       "      <td>21.278312</td>\n",
       "      <td>93.525154</td>\n",
       "      <td>72.038900</td>\n",
       "      <td>13.985473</td>\n",
       "      <td>25.190675</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>musk</th>\n",
       "      <td>0.957262</td>\n",
       "      <td>0.937173</td>\n",
       "      <td>0.961903</td>\n",
       "      <td>0.925721</td>\n",
       "      <td>0.907016</td>\n",
       "      <td>0.946439</td>\n",
       "      <td>0.838284</td>\n",
       "      <td>7.015721</td>\n",
       "      <td>6.650671</td>\n",
       "      <td>7.994846</td>\n",
       "      <td>7.010871</td>\n",
       "      <td>7.326895</td>\n",
       "      <td>5.767735</td>\n",
       "      <td>305.567908</td>\n",
       "      <td>0.957815</td>\n",
       "      <td>0.935763</td>\n",
       "      <td>0.966269</td>\n",
       "      <td>0.925510</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.854353</td>\n",
       "      <td>5.843742</td>\n",
       "      <td>7.837549</td>\n",
       "      <td>6.799870</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>143.756494</td>\n",
       "      <td>441.057515</td>\n",
       "      <td>134.439432</td>\n",
       "      <td>462.342145</td>\n",
       "      <td>399.991160</td>\n",
       "      <td>207.245449</td>\n",
       "      <td>3173.947440</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>obesity</th>\n",
       "      <td>0.933344</td>\n",
       "      <td>0.940772</td>\n",
       "      <td>0.930496</td>\n",
       "      <td>0.932185</td>\n",
       "      <td>0.904429</td>\n",
       "      <td>0.925042</td>\n",
       "      <td>0.898480</td>\n",
       "      <td>8.131754</td>\n",
       "      <td>7.501016</td>\n",
       "      <td>8.923494</td>\n",
       "      <td>7.517536</td>\n",
       "      <td>7.602979</td>\n",
       "      <td>7.129316</td>\n",
       "      <td>34.919431</td>\n",
       "      <td>0.933344</td>\n",
       "      <td>0.941144</td>\n",
       "      <td>0.941539</td>\n",
       "      <td>0.934537</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.127827</td>\n",
       "      <td>7.139336</td>\n",
       "      <td>7.446581</td>\n",
       "      <td>6.662559</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.957805</td>\n",
       "      <td>9.619941</td>\n",
       "      <td>1.890502</td>\n",
       "      <td>9.423308</td>\n",
       "      <td>7.404757</td>\n",
       "      <td>1.216496</td>\n",
       "      <td>3.186684</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>phishing</th>\n",
       "      <td>0.984896</td>\n",
       "      <td>0.973436</td>\n",
       "      <td>0.983284</td>\n",
       "      <td>0.978625</td>\n",
       "      <td>0.951221</td>\n",
       "      <td>0.975835</td>\n",
       "      <td>0.918885</td>\n",
       "      <td>6.498889</td>\n",
       "      <td>6.186637</td>\n",
       "      <td>6.638537</td>\n",
       "      <td>6.653037</td>\n",
       "      <td>7.137116</td>\n",
       "      <td>5.787775</td>\n",
       "      <td>16.172163</td>\n",
       "      <td>0.984472</td>\n",
       "      <td>0.973177</td>\n",
       "      <td>0.982615</td>\n",
       "      <td>0.978064</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.027526</td>\n",
       "      <td>6.090954</td>\n",
       "      <td>5.820109</td>\n",
       "      <td>5.997338</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.200531</td>\n",
       "      <td>18.740488</td>\n",
       "      <td>3.148936</td>\n",
       "      <td>20.107271</td>\n",
       "      <td>17.751595</td>\n",
       "      <td>2.658962</td>\n",
       "      <td>3.548902</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>phoneme</th>\n",
       "      <td>0.803093</td>\n",
       "      <td>0.815406</td>\n",
       "      <td>0.798711</td>\n",
       "      <td>0.810175</td>\n",
       "      <td>0.805957</td>\n",
       "      <td>0.800090</td>\n",
       "      <td>0.820802</td>\n",
       "      <td>7.372448</td>\n",
       "      <td>6.352195</td>\n",
       "      <td>8.176467</td>\n",
       "      <td>6.704918</td>\n",
       "      <td>6.458964</td>\n",
       "      <td>6.243628</td>\n",
       "      <td>17.451772</td>\n",
       "      <td>0.806299</td>\n",
       "      <td>0.813420</td>\n",
       "      <td>0.812702</td>\n",
       "      <td>0.808923</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.312956</td>\n",
       "      <td>6.192015</td>\n",
       "      <td>6.868852</td>\n",
       "      <td>6.671920</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.485781</td>\n",
       "      <td>1.894281</td>\n",
       "      <td>0.463133</td>\n",
       "      <td>1.859440</td>\n",
       "      <td>1.467588</td>\n",
       "      <td>0.396477</td>\n",
       "      <td>0.458207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>shoppers</th>\n",
       "      <td>0.799800</td>\n",
       "      <td>0.849363</td>\n",
       "      <td>0.809192</td>\n",
       "      <td>0.851248</td>\n",
       "      <td>0.718704</td>\n",
       "      <td>0.826267</td>\n",
       "      <td>0.673752</td>\n",
       "      <td>15.950967</td>\n",
       "      <td>8.496142</td>\n",
       "      <td>24.799004</td>\n",
       "      <td>8.511250</td>\n",
       "      <td>8.281520</td>\n",
       "      <td>8.187696</td>\n",
       "      <td>367.692712</td>\n",
       "      <td>0.799800</td>\n",
       "      <td>0.854003</td>\n",
       "      <td>0.831621</td>\n",
       "      <td>0.850394</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>15.950967</td>\n",
       "      <td>7.236937</td>\n",
       "      <td>7.224632</td>\n",
       "      <td>7.145105</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>227.413054</td>\n",
       "      <td>561.655217</td>\n",
       "      <td>355.652823</td>\n",
       "      <td>569.681416</td>\n",
       "      <td>460.231885</td>\n",
       "      <td>159.721222</td>\n",
       "      <td>2685.249490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>spambase</th>\n",
       "      <td>0.920957</td>\n",
       "      <td>0.927775</td>\n",
       "      <td>0.918836</td>\n",
       "      <td>0.915848</td>\n",
       "      <td>0.882062</td>\n",
       "      <td>0.882457</td>\n",
       "      <td>0.874379</td>\n",
       "      <td>13.670000</td>\n",
       "      <td>10.188012</td>\n",
       "      <td>18.320994</td>\n",
       "      <td>9.794410</td>\n",
       "      <td>9.024037</td>\n",
       "      <td>8.905217</td>\n",
       "      <td>200.398944</td>\n",
       "      <td>0.924949</td>\n",
       "      <td>0.927605</td>\n",
       "      <td>0.918836</td>\n",
       "      <td>0.915737</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>11.826832</td>\n",
       "      <td>9.545901</td>\n",
       "      <td>18.320994</td>\n",
       "      <td>9.578509</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>29.554844</td>\n",
       "      <td>116.429532</td>\n",
       "      <td>37.637575</td>\n",
       "      <td>111.199511</td>\n",
       "      <td>86.649042</td>\n",
       "      <td>23.616803</td>\n",
       "      <td>167.770056</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>speed</th>\n",
       "      <td>0.744924</td>\n",
       "      <td>0.751240</td>\n",
       "      <td>0.729087</td>\n",
       "      <td>0.744509</td>\n",
       "      <td>0.640672</td>\n",
       "      <td>0.705158</td>\n",
       "      <td>0.590986</td>\n",
       "      <td>8.743622</td>\n",
       "      <td>7.567121</td>\n",
       "      <td>15.682196</td>\n",
       "      <td>7.834754</td>\n",
       "      <td>7.942428</td>\n",
       "      <td>7.387005</td>\n",
       "      <td>312.957333</td>\n",
       "      <td>0.778993</td>\n",
       "      <td>0.751092</td>\n",
       "      <td>0.729087</td>\n",
       "      <td>0.751650</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.850750</td>\n",
       "      <td>7.551364</td>\n",
       "      <td>15.682196</td>\n",
       "      <td>5.997851</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>183.501822</td>\n",
       "      <td>623.170238</td>\n",
       "      <td>347.466398</td>\n",
       "      <td>644.402542</td>\n",
       "      <td>551.329362</td>\n",
       "      <td>250.298566</td>\n",
       "      <td>4333.318936</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tic-tac-toe</th>\n",
       "      <td>0.935367</td>\n",
       "      <td>0.953493</td>\n",
       "      <td>0.939379</td>\n",
       "      <td>0.957827</td>\n",
       "      <td>0.882519</td>\n",
       "      <td>0.947964</td>\n",
       "      <td>0.818137</td>\n",
       "      <td>4.844478</td>\n",
       "      <td>4.862090</td>\n",
       "      <td>4.815821</td>\n",
       "      <td>4.895522</td>\n",
       "      <td>5.931940</td>\n",
       "      <td>5.057313</td>\n",
       "      <td>13.610746</td>\n",
       "      <td>0.938890</td>\n",
       "      <td>0.953421</td>\n",
       "      <td>0.941875</td>\n",
       "      <td>0.958949</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.389552</td>\n",
       "      <td>4.854328</td>\n",
       "      <td>4.800299</td>\n",
       "      <td>4.861791</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.292063</td>\n",
       "      <td>1.441155</td>\n",
       "      <td>0.267546</td>\n",
       "      <td>1.440984</td>\n",
       "      <td>1.526473</td>\n",
       "      <td>0.218031</td>\n",
       "      <td>0.470075</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              ROC AUC                                                    \\\n",
       "                 C4.5     EC4.5      CART     ECART       ASR        IP   \n",
       "dataset                                                                   \n",
       "bank         0.897834  0.874647  0.899850  0.860747  0.834065  0.855202   \n",
       "bioresponse  0.747969  0.767715  0.742571  0.766073  0.754778  0.757020   \n",
       "breast-w     0.942754  0.947758  0.936141  0.947509  0.954688  0.948981   \n",
       "covertype    0.945313  0.916533  0.944080  0.917318  0.848372  0.835172   \n",
       "default      0.699991  0.717587  0.698888  0.705811  0.638023  0.681492   \n",
       "diabetic     0.632339  0.611578  0.630964  0.610370  0.564986  0.569121   \n",
       "dota2        0.552002  0.534855  0.556627  0.534855  0.534695  0.534997   \n",
       "electricity  0.817044  0.806567  0.815470  0.806567  0.806298  0.804139   \n",
       "firewall     0.946090  0.950280  0.946090  0.950280  0.950280  0.945758   \n",
       "ilpd         0.529022  0.528207  0.525124  0.510737  0.517931  0.514989   \n",
       "iris         0.945000  0.957000  0.945000  0.957000  0.933833  0.938333   \n",
       "letter       0.959188  0.956326  0.959222  0.945474  0.882100  0.928720   \n",
       "musk         0.957262  0.937173  0.961903  0.925721  0.907016  0.946439   \n",
       "obesity      0.933344  0.940772  0.930496  0.932185  0.904429  0.925042   \n",
       "phishing     0.984896  0.973436  0.983284  0.978625  0.951221  0.975835   \n",
       "phoneme      0.803093  0.815406  0.798711  0.810175  0.805957  0.800090   \n",
       "shoppers     0.799800  0.849363  0.809192  0.851248  0.718704  0.826267   \n",
       "spambase     0.920957  0.927775  0.918836  0.915848  0.882062  0.882457   \n",
       "speed        0.744924  0.751240  0.729087  0.744509  0.640672  0.705158   \n",
       "tic-tac-toe  0.935367  0.953493  0.939379  0.957827  0.882519  0.947964   \n",
       "\n",
       "                            Cost                                            \\\n",
       "                  Bal       C4.5      EC4.5       CART     ECART       ASR   \n",
       "dataset                                                                      \n",
       "bank         0.824304  13.150232   8.245344  18.692773  8.072986  8.110905   \n",
       "bioresponse       NaN  12.030019   7.824457  16.111543  7.811886  7.875886   \n",
       "breast-w     0.970811   3.662577   3.983231   4.513701  4.215133  4.240082   \n",
       "covertype    0.843646  12.811080   9.247226  14.517318  8.952649  8.265283   \n",
       "default      0.667648  18.040419   9.988438  19.931981  9.457657  8.540619   \n",
       "diabetic     0.565388  17.551583   8.636224  17.830409  8.641362  8.176789   \n",
       "dota2        0.535006  29.603068   9.145354  28.085247  9.145354  9.145354   \n",
       "electricity  0.801348  10.337327   8.405057  10.950342  8.405057  8.405057   \n",
       "firewall     0.947367   3.360162   3.311609   3.360162  3.311609  3.311609   \n",
       "ilpd         0.509709   7.310565   6.406388   7.604914  6.358722  6.551351   \n",
       "iris         0.925167   2.563810   3.196190   2.554286  3.177143  4.093333   \n",
       "letter       0.901861   9.835014   8.722757  11.928129  8.437014  8.229657   \n",
       "musk         0.838284   7.015721   6.650671   7.994846  7.010871  7.326895   \n",
       "obesity      0.898480   8.131754   7.501016   8.923494  7.517536  7.602979   \n",
       "phishing     0.918885   6.498889   6.186637   6.638537  6.653037  7.137116   \n",
       "phoneme      0.820802   7.372448   6.352195   8.176467  6.704918  6.458964   \n",
       "shoppers     0.673752  15.950967   8.496142  24.799004  8.511250  8.281520   \n",
       "spambase     0.874379  13.670000  10.188012  18.320994  9.794410  9.024037   \n",
       "speed        0.590986   8.743622   7.567121  15.682196  7.834754  7.942428   \n",
       "tic-tac-toe  0.818137   4.844478   4.862090   4.815821  4.895522  5.931940   \n",
       "\n",
       "                                  ROC AUC (ccp)                                \\\n",
       "                   IP         Bal          C4.5     EC4.5      CART     ECART   \n",
       "dataset                                                                         \n",
       "bank         8.229936   29.713363      0.901538  0.874584  0.906445  0.851543   \n",
       "bioresponse  7.625219         NaN      0.747969  0.768195  0.758761  0.760012   \n",
       "breast-w     3.466258   23.131288      0.961821  0.953012  0.927446  0.943614   \n",
       "covertype    8.348718   48.114243      0.945406  0.916550  0.943142  0.917200   \n",
       "default      8.622629   60.468781      0.700115  0.716950  0.694239  0.701388   \n",
       "diabetic     8.205961  152.077543      0.639558  0.611693  0.630972  0.611505   \n",
       "dota2        9.146455   36.325198      0.552002  0.535639  0.556627  0.536333   \n",
       "electricity  8.515616   27.563527      0.817173  0.806549  0.813637  0.805309   \n",
       "firewall     2.887105   26.162312      0.951344  0.953798  0.956854  0.957780   \n",
       "ilpd         6.291892   34.377887      0.514989  0.508186  0.504040  0.494933   \n",
       "iris         3.668571   11.727619      0.945000  0.957000  0.945000  0.957000   \n",
       "letter       8.225571   40.663186      0.960861  0.957054  0.959993  0.945753   \n",
       "musk         5.767735  305.567908      0.957815  0.935763  0.966269  0.925510   \n",
       "obesity      7.129316   34.919431      0.933344  0.941144  0.941539  0.934537   \n",
       "phishing     5.787775   16.172163      0.984472  0.973177  0.982615  0.978064   \n",
       "phoneme      6.243628   17.451772      0.806299  0.813420  0.812702  0.808923   \n",
       "shoppers     8.187696  367.692712      0.799800  0.854003  0.831621  0.850394   \n",
       "spambase     8.905217  200.398944      0.924949  0.927605  0.918836  0.915737   \n",
       "speed        7.387005  312.957333      0.778993  0.751092  0.729087  0.751650   \n",
       "tic-tac-toe  5.057313   13.610746      0.938890  0.953421  0.941875  0.958949   \n",
       "\n",
       "                        Cost (ccp)                                             \\\n",
       "            ASR  IP Bal       C4.5     EC4.5       CART     ECART ASR  IP Bal   \n",
       "dataset                                                                         \n",
       "bank        NaN NaN NaN  11.879325  7.949645  15.155528  5.607400 NaN NaN NaN   \n",
       "bioresponse NaN NaN NaN  12.030019  7.204952   5.042362  6.375848 NaN NaN NaN   \n",
       "breast-w    NaN NaN NaN   2.817178  3.233129   3.851943  4.011452 NaN NaN NaN   \n",
       "covertype   NaN NaN NaN  12.747302  9.162628  13.861205  8.489265 NaN NaN NaN   \n",
       "default     NaN NaN NaN  17.942152  8.985381  18.328467  6.910095 NaN NaN NaN   \n",
       "diabetic    NaN NaN NaN   8.499597  8.340799  17.824112  7.478029 NaN NaN NaN   \n",
       "dota2       NaN NaN NaN  29.603068  7.226604  28.085247  5.790070 NaN NaN NaN   \n",
       "electricity NaN NaN NaN  10.318466  8.227903   7.950840  7.867194 NaN NaN NaN   \n",
       "firewall    NaN NaN NaN   2.653049  2.844307   1.687075  2.703543 NaN NaN NaN   \n",
       "ilpd        NaN NaN NaN   4.445209  3.866830   4.556757  4.816708 NaN NaN NaN   \n",
       "iris        NaN NaN NaN   2.563810  3.196190   2.554286  3.177143 NaN NaN NaN   \n",
       "letter      NaN NaN NaN   9.035957  8.421243  11.375229  8.026686 NaN NaN NaN   \n",
       "musk        NaN NaN NaN   6.854353  5.843742   7.837549  6.799870 NaN NaN NaN   \n",
       "obesity     NaN NaN NaN   8.127827  7.139336   7.446581  6.662559 NaN NaN NaN   \n",
       "phishing    NaN NaN NaN   6.027526  6.090954   5.820109  5.997338 NaN NaN NaN   \n",
       "phoneme     NaN NaN NaN   7.312956  6.192015   6.868852  6.671920 NaN NaN NaN   \n",
       "shoppers    NaN NaN NaN  15.950967  7.236937   7.224632  7.145105 NaN NaN NaN   \n",
       "spambase    NaN NaN NaN  11.826832  9.545901  18.320994  9.578509 NaN NaN NaN   \n",
       "speed       NaN NaN NaN   6.850750  7.551364  15.682196  5.997851 NaN NaN NaN   \n",
       "tic-tac-toe NaN NaN NaN   4.389552  4.854328   4.800299  4.861791 NaN NaN NaN   \n",
       "\n",
       "                 Runtime                                                      \\\n",
       "                    C4.5        EC4.5         CART        ECART          ASR   \n",
       "dataset                                                                        \n",
       "bank           53.934706   232.459637    72.040812   231.279182   188.073861   \n",
       "bioresponse  4593.632122  4215.667313  6192.435255  4629.768001  4127.723269   \n",
       "breast-w        0.191986     1.001592     0.192608     1.010131     0.807835   \n",
       "covertype     341.836037  1597.349569   402.592217  1539.147105  1136.326300   \n",
       "default        44.764109   174.722208    45.350617   165.505596   127.302811   \n",
       "diabetic     1035.128554  2737.601433  1138.214176  2869.144890  2062.491981   \n",
       "dota2        1694.156266  3569.622728  1573.224023  3681.001757  2941.977681   \n",
       "electricity     6.368101    28.723873     6.364826    28.117674    24.080955   \n",
       "firewall        2.207097     2.818632     2.148224     2.786070     2.696570   \n",
       "ilpd            0.423081     1.963373     0.388731     1.799762     1.438511   \n",
       "iris            0.019084     0.099292     0.017226     0.094784     0.099996   \n",
       "letter         18.811961    96.639955    21.278312    93.525154    72.038900   \n",
       "musk          143.756494   441.057515   134.439432   462.342145   399.991160   \n",
       "obesity         1.957805     9.619941     1.890502     9.423308     7.404757   \n",
       "phishing        3.200531    18.740488     3.148936    20.107271    17.751595   \n",
       "phoneme         0.485781     1.894281     0.463133     1.859440     1.467588   \n",
       "shoppers      227.413054   561.655217   355.652823   569.681416   460.231885   \n",
       "spambase       29.554844   116.429532    37.637575   111.199511    86.649042   \n",
       "speed         183.501822   623.170238   347.466398   644.402542   551.329362   \n",
       "tic-tac-toe     0.292063     1.441155     0.267546     1.440984     1.526473   \n",
       "\n",
       "                                       \n",
       "                      IP          Bal  \n",
       "dataset                                \n",
       "bank           34.954801    50.070138  \n",
       "bioresponse  2700.924505          NaN  \n",
       "breast-w        0.151000     0.430479  \n",
       "covertype     268.721726   524.012348  \n",
       "default        24.118354    66.731410  \n",
       "diabetic      601.082457  4489.117611  \n",
       "dota2         471.520335   995.454448  \n",
       "electricity     4.706966     5.879775  \n",
       "firewall        2.683861     2.398550  \n",
       "ilpd            0.327805     0.716311  \n",
       "iris            0.025907     0.052904  \n",
       "letter         13.985473    25.190675  \n",
       "musk          207.245449  3173.947440  \n",
       "obesity         1.216496     3.186684  \n",
       "phishing        2.658962     3.548902  \n",
       "phoneme         0.396477     0.458207  \n",
       "shoppers      159.721222  2685.249490  \n",
       "spambase       23.616803   167.770056  \n",
       "speed         250.298566  4333.318936  \n",
       "tic-tac-toe     0.218031     0.470075  "
      ]
     },
     "execution_count": 346,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2 = df.pivot_table(index=['dataset','rn'], columns='name', values=['auc','cost','auc2','cost2','runtime'])\n",
    "cols = pd.MultiIndex.from_product([['auc','cost','auc2','cost2','runtime'],\n",
    "                                   ['entropy' if costtype=='unit' else 'centropy',\n",
    "                                    'epentropy', \n",
    "                                    'gini' if costtype=='unit' else 'cgini',\n",
    "                                    'epgini', \n",
    "                                    'ep0entropy',\n",
    "                                    'spcsubm', 'bentropy'\n",
    "                                   ]])\n",
    "df2 = df2[cols]\n",
    "cols = pd.MultiIndex.from_product([['ROC AUC','Cost','ROC AUC (ccp)','Cost (ccp)','Runtime'],\n",
    "                                   ['C4.5' if costtype=='unit' else 'CC4.5',\n",
    "                                    'EC4.5',\n",
    "                                    'CART' if costtype=='unit' else 'CCART',\n",
    "                                    'ECART',\n",
    "                                    'ASR',\n",
    "                                    'IP','Bal',\n",
    "                                   ]])\n",
    "df2.columns = cols\n",
    "# df2.fillna(0, inplace=True)\n",
    "cols = pd.MultiIndex.from_product([['ROC AUC (ccp)','Cost (ccp)'],\n",
    "                                   ['ASR','IP','Bal',]])\n",
    "df2.loc[:,cols] = np.nan # pd.NA\n",
    "df2.groupby('dataset').mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 350,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-350-eba8ad588f2d>:102: UserWarning: This figure includes Axes that are not compatible with tight_layout, so results might be incorrect.\n",
      "  fig.tight_layout()\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 3000x4000 with 55 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# method = 'C4.5'\n",
    "method = 'CART'\n",
    "\n",
    "datasets = df2.groupby('dataset').mean().index.to_list()\n",
    "algs = ['{}'.format(method if costtype=='unit' else 'C'+method),\n",
    "       'E{}'.format(method),\n",
    "       'ASR', \n",
    "       'IP',\n",
    "       'Bal']\n",
    "def make_x_y(df2, metric, algs):\n",
    "    df3 = df2.groupby('dataset').mean()\n",
    "#     df3.fillna(0, inplace=True)\n",
    "    df3_err = df2.groupby('dataset').std()\n",
    "    df3_err.fillna(0, inplace=True)\n",
    "    df3 = df3[metric][algs]\n",
    "    df3_err = df3_err[metric][algs]\n",
    "    return df3, df3_err\n",
    "\n",
    "\n",
    "nrow = 4\n",
    "ncol = len(datasets)//nrow\n",
    "gridspec = dict(wspace=0.1, hspace=0.1, height_ratios=[1, 1, 0.15]*(nrow-1)+[1, 1])\n",
    "fig, ax_lst = plt.subplots(nrow*3-1, ncol, figsize=(30, 10*nrow), sharex='col', gridspec_kw=gridspec)\n",
    "for ax in ax_lst[[2,5,8]].flatten(): ax.set_visible(False) \n",
    "sz_font = 40\n",
    "cycle = plt.rcParams['axes.prop_cycle'].by_key()['color']\n",
    "width = 0.3  # the width of the bars\n",
    "\n",
    "\n",
    "metric = 'ROC AUC'\n",
    "df3, df3_err = make_x_y(df2, metric, algs)\n",
    "df3_, df3_err_ = make_x_y(df2, metric+' (ccp)', algs)\n",
    "df3.fillna(0, inplace=True)\n",
    "df3_.fillna(0, inplace=True)\n",
    "for i,(dataset,ax) in enumerate(zip(datasets, ax_lst[[0,3,6,9]].flatten())):\n",
    "    x = np.arange(len(df3.columns))  # the label locations\n",
    "\n",
    "    y = df3.loc[dataset].to_numpy()\n",
    "    yerr = df3_err.loc[dataset].to_numpy()\n",
    "    rects = ax.bar(x-width/1.5, y, width, yerr=yerr, color=cycle[0:len(x)], label='Un-pruned')\n",
    "    \n",
    "    y = df3_.loc[dataset].to_numpy()\n",
    "    yerr = df3_err_.loc[dataset].to_numpy()\n",
    "    rects = ax.bar(x+width/1.5, y, width, yerr=yerr, color=cycle[0:len(x)], hatch='+', alpha=0.5, label='Pruned')\n",
    "    \n",
    "#     ax.set_title(dataset, fontdict={'fontsize': sz_font}, rotation=25)\n",
    "    ax.set_title(dataset, fontdict={'fontsize': sz_font})\n",
    "#     ax.tick_params(axis='x', labelsize=sz_font-10, labelrotation=270)\n",
    "    ax.tick_params(axis='y', labelsize=sz_font-10, labelrotation=45)\n",
    "    ax.set_ylim(0.5,1)\n",
    "    \n",
    "    if i%ncol == 0:\n",
    "        ax.set_ylabel('{}'.format(metric), fontsize=sz_font)  \n",
    "    else:\n",
    "        ax.set_yticklabels([])\n",
    "        ax.set_xticklabels([])\n",
    "        ax.set_xticks([])\n",
    "\n",
    "\n",
    "# https://matplotlib.org/3.1.0/gallery/subplots_axes_and_figures/broken_axis.html\n",
    "\n",
    "metric = 'Cost'\n",
    "df3, df3_err = make_x_y(df2, metric, algs)\n",
    "df3_, df3_err_ = make_x_y(df2, metric+' (ccp)', algs)\n",
    "df3.fillna(1, inplace=True)\n",
    "df3_.fillna(1, inplace=True)\n",
    "d = .015\n",
    "\n",
    "ylims = np.logspace(0, 12, num=13, base=2).astype(int)\n",
    "idx_ylim = np.where(max(df3[algs[0]])*2 < ylims)[0][0]\n",
    "ylim_sep = ylims[idx_ylim]\n",
    "yticks = np.logspace(0, idx_ylim, num=idx_ylim+1, base=2).astype(int)\n",
    "yticks = [t for i,t in enumerate(yticks) if i%2==0]\n",
    "\n",
    "for i,(dataset,ax) in enumerate(zip(datasets, ax_lst[[1,4,7,10]].flatten())):\n",
    "    x = np.arange(len(df3.columns))  # the label locations\n",
    "    ax.set_yscale('log', base=2)\n",
    "\n",
    "    y = df3.loc[dataset].to_numpy()\n",
    "    yerr = df3_err.loc[dataset].to_numpy()\n",
    "    rects = ax.bar(x-width/1.5, y, width, yerr=yerr, color=cycle[0:len(x)], label='Un-pruned')\n",
    "    \n",
    "    y = df3_.loc[dataset].to_numpy()\n",
    "    yerr = df3_err_.loc[dataset].to_numpy()\n",
    "    rects = ax.bar(x+width/1.5, y, width, yerr=yerr, color=cycle[0:len(x)], hatch='+', alpha=0.5, label='Pruned')\n",
    "    \n",
    "    ax.set_ylim(1,ylim_sep)\n",
    "    ax.set_xticks(x)\n",
    "    ax.set_xticklabels(df3.columns)\n",
    "    ax.tick_params(axis='x', labelsize=sz_font-13, labelrotation=90)\n",
    "    ax.tick_params(axis='y', labelsize=sz_font-10, labelrotation=45)\n",
    "    \n",
    "    if i%ncol == 0:\n",
    "        ax.set_yticks(yticks)\n",
    "        ax.set_ylabel('{}'.format('Expected \\ntree height'), fontsize=sz_font)\n",
    "#         ax.set_xlabel('Algorithms', fontsize=sz_font)\n",
    "    else:\n",
    "        ax.set_yticklabels([])\n",
    "        \n",
    "\n",
    "plt.legend(prop={'size': sz_font-10}, loc='best', handletextpad=0.2, borderaxespad=0.1)\n",
    "fig.tight_layout()\n",
    "# fig.savefig('comparison-{}-{}.pdf'.format(method.replace('.',''),costtype))\n",
    "# tikzplotlib.save('comparison-{}-{}.tex'.format(method.replace('.',''),costtype))\n",
    "code = tikzplotlib.get_tikz_code()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 351,
   "metadata": {},
   "outputs": [],
   "source": [
    "code_ = code"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 352,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "def find_nth(haystack, needle, n):\n",
    "    start = haystack.find(needle)\n",
    "    while start >= 0 and n > 1:\n",
    "        start = haystack.find(needle, start+len(needle))\n",
    "        n -= 1\n",
    "    return start\n",
    "\n",
    "def add_opt(code, key, opt, beg=False):\n",
    "    idx = code.find(key)\n",
    "    if beg:\n",
    "        return code[:idx] + opt + code[idx:]\n",
    "    return code[:idx+len(key)] + opt + code[idx+len(key):]\n",
    "\n",
    "def add_opt_which(code, which, key, opt, beg=False):\n",
    "    idx = find_nth(code, key, which)\n",
    "    if beg:\n",
    "        return code[:idx] + opt + code[idx:]\n",
    "    return code[:idx+len(key)] + opt + code[idx+len(key):]\n",
    "\n",
    "\n",
    "code = code_\n",
    "\n",
    "key_gl = '\\\\begin{groupplot}['\n",
    "key_per = 'nextgroupplot['\n",
    "\n",
    "# code = code.replace('\\nyticklabel style={rotate=45.0}', '')\n",
    "code = code.replace('Expected', 'Expected\\\\\\\\')\n",
    "code = code.replace('ROC', 'ROC\\\\\\\\')\n",
    "code = code.replace('\\ntick pos=both,', '')\n",
    "# code = code.replace('ymin=1,', 'ymin=0,')\n",
    "code = code.replace('\\nyticklabel style={rotate=45.0},', '')\n",
    "# code = code.replace('xticklabel style={rotate=315.0}', 'xticklabel style={rotate=270.0}')\n",
    "code = code.replace('ultra thick', 'thick')\n",
    "\n",
    "# code = add_opt(code, 'minor xticklabels', '%', beg=True)\n",
    "# code = add_opt(code, 'scaled minor x ticks', '%', beg=True)\n",
    "\n",
    "opt = '''\n",
    "every axis title/.style={at={(0.5,1)},above,yshift=1pt,text depth=0pt},\n",
    "width = 0.27\\\\textwidth, height = 0.23\\\\textwidth, \n",
    "group/horizontal sep=2pt, group/vertical sep=3pt,\n",
    "yticklabels={,,}, \n",
    "xtick style={draw=none},\n",
    "tickwidth=1pt,\n",
    "tick label style={font=\\\\tiny},\n",
    "label style={font=\\\\small},\n",
    "'''\n",
    "'''\n",
    "ytick style={draw=none},\n",
    "xtick style={draw=none},\n",
    "'''\n",
    "code = add_opt(code, key_gl, opt)\n",
    "\n",
    "\n",
    "idx_auc = [3*ncol*i+1 for i in range(nrow)]\n",
    "idx_cost = [3*ncol*i + ncol + 1 for i in range(nrow)]\n",
    "idx_pad = [3*ncol*i + ncol*2 + 1 for i in range(nrow-1)]\n",
    "idx_auc2 = [j for i in idx_auc for j in range(i,i+ncol)]\n",
    "idx_cost2 = [j for i in idx_cost for j in range(i,i+ncol)]\n",
    "idx_pad2 = [j for i in idx_pad for j in range(i,i+ncol)]\n",
    "\n",
    "yticks_cost = ','.join([str(i) for i in yticks])\n",
    "\n",
    "legend = '''legend pos=north west,\n",
    "legend style={\n",
    "  fill opacity=0.6,\n",
    "  draw opacity=1,\n",
    "  text opacity=1,\n",
    "  at={(0,2)},\n",
    "  legend columns=-1,\n",
    "  draw=white!80!black\n",
    "}, '''\n",
    "code = add_opt_which(code, 1, key_per, legend)\n",
    "\n",
    "for i in idx_auc:\n",
    "    code = add_opt_which(code, i, key_per, 'ytick={0.4,0.6,0.8,1}, ')\n",
    "for i in idx_cost2:\n",
    "    idx = find_nth(code, key_per, i)\n",
    "    idx2 = find_nth(code, key_per, i+1)\n",
    "    code = code[:idx] + code[idx:idx2].replace(',0) rectangle', ',1) rectangle') + code[idx2:]\n",
    "    code = add_opt_which(code, i, key_per, 'ytick={'+yticks_cost+'}, ')\n",
    "\n",
    "# bioresponse missing result for BAL\n",
    "i = idx_cost2[1]\n",
    "idx = find_nth(code, key_per, i)\n",
    "idx2 = find_nth(code, key_per, i+1)\n",
    "code = code[:idx] + code[idx:idx2].replace('\\\\draw[draw=none,fill=color5', '%\\\\draw[draw=none,fill=color5') \\\n",
    "    + code[idx2:]\n",
    "\n",
    "for i in idx_pad2: \n",
    "    code = add_opt_which(code, i, key_per, 'group/empty plot, height = 0.16\\\\textwidth,')\n",
    "\n",
    "for i in idx_auc:\n",
    "    code = add_opt_which(code, i, key_per, \n",
    "                         'ylabel style={align=center}, yticklabels={0.4,0.6,0.8,1}, ')\n",
    "for i in idx_cost:\n",
    "    code = add_opt_which(code, i, key_per, \n",
    "                         'ylabel style={align=center}, yticklabels={'+yticks_cost+'}, ')\n",
    "    \n",
    "# print(code)\n",
    "with open('figs/comparison-{}-{}.tex'.format(method.replace('.',''),costtype), 'w') as fout:\n",
    "    fout.write(code)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Running time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CC4.5    1.000000\n",
       "EC4.5    3.680023\n",
       "CCART    1.081030\n",
       "ECART    3.542950\n",
       "ASR      2.479761\n",
       "IP       0.603995\n",
       "Bal      3.357446\n",
       "dtype: float64"
      ]
     },
     "execution_count": 151,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "running = df2.groupby('dataset').mean()['Runtime']\n",
    "cols = running.columns\n",
    "for c in cols[1:]:\n",
    "    running[c] = running[c] / running[cols[0]]\n",
    "running[cols[0]] = 1\n",
    "\n",
    "running.mean(axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.bar(cols, running.mean(axis=0))\n",
    "plt.tick_params(axis='x', labelsize=sz_font-12, labelrotation=90)\n",
    "plt.tick_params(axis='y', labelsize=sz_font-12)\n",
    "plt.xticks(cols)\n",
    "plt.ylabel('Normalized running time', fontsize=sz_font-12)\n",
    "\n",
    "code = tikzplotlib.get_tikz_code()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [],
   "source": [
    "opt = 'width = 0.8\\\\textwidth, height = 0.4\\\\textwidth, xticklabels={' + ','.join(cols) + '}, xtick={' + ','.join([str(i) for i in range(len(cols))]) + '},' \n",
    "key = '\\\\begin{axis}['\n",
    "idx = code.find(key)\n",
    "code = code[:idx+len(key)] + opt + code[idx+len(key):]\n",
    "\n",
    "with open('figs/running-time-{}-{}.tex'.format(method.replace('.',''),costtype), 'w') as fout:\n",
    "    fout.write(code)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Tex"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\begin{tabular}{lrrrrrrr}\n",
      "\\toprule\n",
      "{} &   C4.5 &  EC4.5 &   CART &  ECART &    ASR &     IP &    Bal \\\\\n",
      "dataset     &        &        &        &        &        &        &        \\\\\n",
      "\\midrule\n",
      "bank        &  0.898 &  0.875 &  0.900 &  0.861 &  0.834 &  0.855 &  0.824 \\\\\n",
      "bioresponse &  0.748 &  0.768 &  0.743 &  0.766 &  0.755 &  0.757 &    NaN \\\\\n",
      "breast-w    &  0.943 &  0.948 &  0.936 &  0.948 &  0.955 &  0.949 &  0.971 \\\\\n",
      "covertype   &  0.945 &  0.917 &  0.944 &  0.917 &  0.848 &  0.835 &  0.844 \\\\\n",
      "default     &  0.700 &  0.718 &  0.699 &  0.706 &  0.638 &  0.681 &  0.668 \\\\\n",
      "diabetic    &  0.632 &  0.612 &  0.631 &  0.610 &  0.565 &  0.569 &  0.565 \\\\\n",
      "dota2       &  0.552 &  0.535 &  0.557 &  0.535 &  0.535 &  0.535 &  0.535 \\\\\n",
      "electricity &  0.817 &  0.807 &  0.815 &  0.807 &  0.806 &  0.804 &  0.801 \\\\\n",
      "firewall    &  0.946 &  0.950 &  0.946 &  0.950 &  0.950 &  0.946 &  0.947 \\\\\n",
      "ilpd        &  0.529 &  0.528 &  0.525 &  0.511 &  0.518 &  0.515 &  0.510 \\\\\n",
      "iris        &  0.945 &  0.957 &  0.945 &  0.957 &  0.934 &  0.938 &  0.925 \\\\\n",
      "letter      &  0.959 &  0.956 &  0.959 &  0.945 &  0.882 &  0.929 &  0.902 \\\\\n",
      "musk        &  0.957 &  0.937 &  0.962 &  0.926 &  0.907 &  0.946 &  0.838 \\\\\n",
      "obesity     &  0.933 &  0.941 &  0.930 &  0.932 &  0.904 &  0.925 &  0.898 \\\\\n",
      "phishing    &  0.985 &  0.973 &  0.983 &  0.979 &  0.951 &  0.976 &  0.919 \\\\\n",
      "phoneme     &  0.803 &  0.815 &  0.799 &  0.810 &  0.806 &  0.800 &  0.821 \\\\\n",
      "shoppers    &  0.800 &  0.849 &  0.809 &  0.851 &  0.719 &  0.826 &  0.674 \\\\\n",
      "spambase    &  0.921 &  0.928 &  0.919 &  0.916 &  0.882 &  0.882 &  0.874 \\\\\n",
      "speed       &  0.745 &  0.751 &  0.729 &  0.745 &  0.641 &  0.705 &  0.591 \\\\\n",
      "tic-tac-toe &  0.935 &  0.953 &  0.939 &  0.958 &  0.883 &  0.948 &  0.818 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n",
      "\\begin{tabular}{lrrrrrrr}\n",
      "\\toprule\n",
      "{} &    C4.5 &   EC4.5 &    CART &  ECART &    ASR &     IP &      Bal \\\\\n",
      "dataset     &         &         &         &        &        &        &          \\\\\n",
      "\\midrule\n",
      "bank        &  13.150 &   8.245 &  18.693 &  8.073 &  8.111 &  8.230 &   29.713 \\\\\n",
      "bioresponse &  12.030 &   7.824 &  16.112 &  7.812 &  7.876 &  7.625 &      NaN \\\\\n",
      "breast-w    &   3.663 &   3.983 &   4.514 &  4.215 &  4.240 &  3.466 &   23.131 \\\\\n",
      "covertype   &  12.811 &   9.247 &  14.517 &  8.953 &  8.265 &  8.349 &   48.114 \\\\\n",
      "default     &  18.040 &   9.988 &  19.932 &  9.458 &  8.541 &  8.623 &   60.469 \\\\\n",
      "diabetic    &  17.552 &   8.636 &  17.830 &  8.641 &  8.177 &  8.206 &  152.078 \\\\\n",
      "dota2       &  29.603 &   9.145 &  28.085 &  9.145 &  9.145 &  9.146 &   36.325 \\\\\n",
      "electricity &  10.337 &   8.405 &  10.950 &  8.405 &  8.405 &  8.516 &   27.564 \\\\\n",
      "firewall    &   3.360 &   3.312 &   3.360 &  3.312 &  3.312 &  2.887 &   26.162 \\\\\n",
      "ilpd        &   7.311 &   6.406 &   7.605 &  6.359 &  6.551 &  6.292 &   34.378 \\\\\n",
      "iris        &   2.564 &   3.196 &   2.554 &  3.177 &  4.093 &  3.669 &   11.728 \\\\\n",
      "letter      &   9.835 &   8.723 &  11.928 &  8.437 &  8.230 &  8.226 &   40.663 \\\\\n",
      "musk        &   7.016 &   6.651 &   7.995 &  7.011 &  7.327 &  5.768 &  305.568 \\\\\n",
      "obesity     &   8.132 &   7.501 &   8.923 &  7.518 &  7.603 &  7.129 &   34.919 \\\\\n",
      "phishing    &   6.499 &   6.187 &   6.639 &  6.653 &  7.137 &  5.788 &   16.172 \\\\\n",
      "phoneme     &   7.372 &   6.352 &   8.176 &  6.705 &  6.459 &  6.244 &   17.452 \\\\\n",
      "shoppers    &  15.951 &   8.496 &  24.799 &  8.511 &  8.282 &  8.188 &  367.693 \\\\\n",
      "spambase    &  13.670 &  10.188 &  18.321 &  9.794 &  9.024 &  8.905 &  200.399 \\\\\n",
      "speed       &   8.744 &   7.567 &  15.682 &  7.835 &  7.942 &  7.387 &  312.957 \\\\\n",
      "tic-tac-toe &   4.844 &   4.862 &   4.816 &  4.896 &  5.932 &  5.057 &   13.611 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n",
      "\\begin{tabular}{lrrrrrrr}\n",
      "\\toprule\n",
      "{} &    C4.5 &   EC4.5 &    CART &   ECART &     ASR &      IP &     Bal \\\\\n",
      "dataset     &         &         &         &         &         &         &         \\\\\n",
      "\\midrule\n",
      "bank        &    53.9 &   232.5 &    72.0 &   231.3 &   188.1 &    35.0 &    50.1 \\\\\n",
      "bioresponse &  4593.6 &  4215.7 &  6192.4 &  4629.8 &  4127.7 &  2700.9 &     NaN \\\\\n",
      "breast-w    &     0.2 &     1.0 &     0.2 &     1.0 &     0.8 &     0.2 &     0.4 \\\\\n",
      "covertype   &   341.8 &  1597.3 &   402.6 &  1539.1 &  1136.3 &   268.7 &   524.0 \\\\\n",
      "default     &    44.8 &   174.7 &    45.4 &   165.5 &   127.3 &    24.1 &    66.7 \\\\\n",
      "diabetic    &  1035.1 &  2737.6 &  1138.2 &  2869.1 &  2062.5 &   601.1 &  4489.1 \\\\\n",
      "dota2       &  1694.2 &  3569.6 &  1573.2 &  3681.0 &  2942.0 &   471.5 &   995.5 \\\\\n",
      "electricity &     6.4 &    28.7 &     6.4 &    28.1 &    24.1 &     4.7 &     5.9 \\\\\n",
      "firewall    &     2.2 &     2.8 &     2.1 &     2.8 &     2.7 &     2.7 &     2.4 \\\\\n",
      "ilpd        &     0.4 &     2.0 &     0.4 &     1.8 &     1.4 &     0.3 &     0.7 \\\\\n",
      "iris        &     0.0 &     0.1 &     0.0 &     0.1 &     0.1 &     0.0 &     0.1 \\\\\n",
      "letter      &    18.8 &    96.6 &    21.3 &    93.5 &    72.0 &    14.0 &    25.2 \\\\\n",
      "musk        &   143.8 &   441.1 &   134.4 &   462.3 &   400.0 &   207.2 &  3173.9 \\\\\n",
      "obesity     &     2.0 &     9.6 &     1.9 &     9.4 &     7.4 &     1.2 &     3.2 \\\\\n",
      "phishing    &     3.2 &    18.7 &     3.1 &    20.1 &    17.8 &     2.7 &     3.5 \\\\\n",
      "phoneme     &     0.5 &     1.9 &     0.5 &     1.9 &     1.5 &     0.4 &     0.5 \\\\\n",
      "shoppers    &   227.4 &   561.7 &   355.7 &   569.7 &   460.2 &   159.7 &  2685.2 \\\\\n",
      "spambase    &    29.6 &   116.4 &    37.6 &   111.2 &    86.6 &    23.6 &   167.8 \\\\\n",
      "speed       &   183.5 &   623.2 &   347.5 &   644.4 &   551.3 &   250.3 &  4333.3 \\\\\n",
      "tic-tac-toe &     0.3 &     1.4 &     0.3 &     1.4 &     1.5 &     0.2 &     0.5 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(df2.groupby('dataset').mean()['ROC AUC'].round(3).to_latex())\n",
    "print(df2.groupby('dataset').mean()['Cost'].round(3).to_latex())\n",
    "print(df2.groupby('dataset').mean()['Runtime'].round(1).to_latex())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 351,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col2 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col6 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col7 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col8 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col10 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col11 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col12 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col0 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col1 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col6 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col10 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col11 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col12 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col0 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col1 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col2 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col9 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col10 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col11 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col0 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col1 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col5 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col9 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col10 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col11 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col0 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col1 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col2 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col3 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col4 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col5 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col6 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col7 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col8 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col10 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col11 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col13 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col3 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col4 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col5 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col7 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col8 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col13 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col0 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col1 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col6 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col10 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col11 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col12 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col3 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col4 {\n",
       "            color:  red;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col5 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col7 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col8 {\n",
       "            color:  green;\n",
       "        }    #T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col12 {\n",
       "            color:  red;\n",
       "        }</style><table id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958\" ><thead>    <tr>        <th class=\"blank level0\" ></th>        <th class=\"col_heading level0 col0\" colspan=7>ROC AUC</th>        <th class=\"col_heading level0 col7\" colspan=7>Cost</th>    </tr>    <tr>        <th class=\"blank level1\" ></th>        <th class=\"col_heading level1 col0\" >C4.5</th>        <th class=\"col_heading level1 col1\" >CC4.5</th>        <th class=\"col_heading level1 col2\" >EC4.5</th>        <th class=\"col_heading level1 col3\" >CART</th>        <th class=\"col_heading level1 col4\" >CCART</th>        <th class=\"col_heading level1 col5\" >ECART</th>        <th class=\"col_heading level1 col6\" >ASR</th>        <th class=\"col_heading level1 col7\" >C4.5</th>        <th class=\"col_heading level1 col8\" >CC4.5</th>        <th class=\"col_heading level1 col9\" >EC4.5</th>        <th class=\"col_heading level1 col10\" >CART</th>        <th class=\"col_heading level1 col11\" >CCART</th>        <th class=\"col_heading level1 col12\" >ECART</th>        <th class=\"col_heading level1 col13\" >ASR</th>    </tr>    <tr>        <th class=\"index_name level0\" >dataset</th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>        <th class=\"blank\" ></th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                        <th id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958level0_row0\" class=\"row_heading level0 row0\" >iris</th>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col0\" class=\"data row0 col0\" >0.933333</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col1\" class=\"data row0 col1\" >0.933333</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col2\" class=\"data row0 col2\" >0.938889</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col3\" class=\"data row0 col3\" >0.933333</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col4\" class=\"data row0 col4\" >0.933333</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col5\" class=\"data row0 col5\" >0.938395</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col6\" class=\"data row0 col6\" >0.932346</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col7\" class=\"data row0 col7\" >1.70159</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col8\" class=\"data row0 col8\" >1.70159</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col9\" class=\"data row0 col9\" >1.73651</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col10\" class=\"data row0 col10\" >1.70159</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col11\" class=\"data row0 col11\" >1.70159</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col12\" class=\"data row0 col12\" >1.86032</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row0_col13\" class=\"data row0 col13\" >1.85397</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958level0_row1\" class=\"row_heading level0 row1\" >obesity</th>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col0\" class=\"data row1 col0\" >0.882707</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col1\" class=\"data row1 col1\" >0.882707</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col2\" class=\"data row1 col2\" >0.861682</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col3\" class=\"data row1 col3\" >0.881246</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col4\" class=\"data row1 col4\" >0.881246</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col5\" class=\"data row1 col5\" >0.864197</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col6\" class=\"data row1 col6\" >0.82823</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col7\" class=\"data row1 col7\" >4.89438</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col8\" class=\"data row1 col8\" >4.89438</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col9\" class=\"data row1 col9\" >4.88332</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col10\" class=\"data row1 col10\" >5.08328</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col11\" class=\"data row1 col11\" >5.08328</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col12\" class=\"data row1 col12\" >4.84315</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row1_col13\" class=\"data row1 col13\" >5.07718</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958level0_row2\" class=\"row_heading level0 row2\" >shoppers</th>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col0\" class=\"data row2 col0\" >0.58779</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col1\" class=\"data row2 col1\" >0.58779</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col2\" class=\"data row2 col2\" >0.560693</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col3\" class=\"data row2 col3\" >0.58718</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col4\" class=\"data row2 col4\" >0.58718</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col5\" class=\"data row2 col5\" >0.562041</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col6\" class=\"data row2 col6\" >0.561894</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col7\" class=\"data row2 col7\" >6.01842</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col8\" class=\"data row2 col8\" >6.01842</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col9\" class=\"data row2 col9\" >4.6484</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col10\" class=\"data row2 col10\" >6.74078</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col11\" class=\"data row2 col11\" >6.74078</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col12\" class=\"data row2 col12\" >4.65153</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row2_col13\" class=\"data row2 col13\" >4.65354</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958level0_row3\" class=\"row_heading level0 row3\" >default</th>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col0\" class=\"data row3 col0\" >0.680898</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col1\" class=\"data row3 col1\" >0.680898</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col2\" class=\"data row3 col2\" >0.675584</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col3\" class=\"data row3 col3\" >0.677157</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col4\" class=\"data row3 col4\" >0.677157</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col5\" class=\"data row3 col5\" >0.647149</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col6\" class=\"data row3 col6\" >0.647749</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col7\" class=\"data row3 col7\" >10.5537</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col8\" class=\"data row3 col8\" >10.5537</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col9\" class=\"data row3 col9\" >7.61235</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col10\" class=\"data row3 col10\" >11.295</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col11\" class=\"data row3 col11\" >11.295</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col12\" class=\"data row3 col12\" >7.61548</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row3_col13\" class=\"data row3 col13\" >7.61398</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958level0_row4\" class=\"row_heading level0 row4\" >firewall</th>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col0\" class=\"data row4 col0\" >0.917959</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col1\" class=\"data row4 col1\" >0.917959</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col2\" class=\"data row4 col2\" >0.917942</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col3\" class=\"data row4 col3\" >0.917959</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col4\" class=\"data row4 col4\" >0.917959</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col5\" class=\"data row4 col5\" >0.917942</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col6\" class=\"data row4 col6\" >0.917959</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col7\" class=\"data row4 col7\" >2.38052</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col8\" class=\"data row4 col8\" >2.38052</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col9\" class=\"data row4 col9\" >2.4323</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col10\" class=\"data row4 col10\" >2.38052</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col11\" class=\"data row4 col11\" >2.38052</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col12\" class=\"data row4 col12\" >2.4323</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row4_col13\" class=\"data row4 col13\" >2.97372</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958level0_row5\" class=\"row_heading level0 row5\" >dota</th>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col0\" class=\"data row5 col0\" >0.539365</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col1\" class=\"data row5 col1\" >0.539365</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col2\" class=\"data row5 col2\" >0.526418</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col3\" class=\"data row5 col3\" >0.540485</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col4\" class=\"data row5 col4\" >0.540485</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col5\" class=\"data row5 col5\" >0.5263</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col6\" class=\"data row5 col6\" >0.526345</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col7\" class=\"data row5 col7\" >18.0754</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col8\" class=\"data row5 col8\" >18.0754</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col9\" class=\"data row5 col9\" >8.80579</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col10\" class=\"data row5 col10\" >17.2761</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col11\" class=\"data row5 col11\" >17.2761</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col12\" class=\"data row5 col12\" >8.8058</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row5_col13\" class=\"data row5 col13\" >8.80578</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958level0_row6\" class=\"row_heading level0 row6\" >diab</th>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col0\" class=\"data row6 col0\" >0.598823</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col1\" class=\"data row6 col1\" >0.598823</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col2\" class=\"data row6 col2\" >0.570652</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col3\" class=\"data row6 col3\" >0.597501</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col4\" class=\"data row6 col4\" >0.597501</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col5\" class=\"data row6 col5\" >0.569781</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col6\" class=\"data row6 col6\" >0.569502</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col7\" class=\"data row6 col7\" >7.53082</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col8\" class=\"data row6 col8\" >7.53082</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col9\" class=\"data row6 col9\" >5.52523</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col10\" class=\"data row6 col10\" >7.54899</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col11\" class=\"data row6 col11\" >7.54899</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col12\" class=\"data row6 col12\" >5.50491</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row6_col13\" class=\"data row6 col13\" >5.50537</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                        <th id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958level0_row7\" class=\"row_heading level0 row7\" >poke</th>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col0\" class=\"data row7 col0\" >0.692298</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col1\" class=\"data row7 col1\" >0.692298</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col2\" class=\"data row7 col2\" >0.694119</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col3\" class=\"data row7 col3\" >0.691758</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col4\" class=\"data row7 col4\" >0.691758</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col5\" class=\"data row7 col5\" >0.696221</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col6\" class=\"data row7 col6\" >0.693099</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col7\" class=\"data row7 col7\" >3.19289</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col8\" class=\"data row7 col8\" >3.19289</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col9\" class=\"data row7 col9\" >3.19039</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col10\" class=\"data row7 col10\" >3.18766</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col11\" class=\"data row7 col11\" >3.18766</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col12\" class=\"data row7 col12\" >3.18628</td>\n",
       "                        <td id=\"T_88c077b4_2811_11eb_bec1_3c15c2bc9958row7_col13\" class=\"data row7 col13\" >3.18903</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x7f9d4ff67588>"
      ]
     },
     "execution_count": 351,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def highlight_max_min(s, start, level=None):\n",
    "    '''\n",
    "    highlight the maximum in a Series yellow.\n",
    "    # df.columns = list(map(lambda x: md[x], df.columns))\n",
    "    # print(df.round(2).to_latex())\n",
    "    '''\n",
    "    if level is None:\n",
    "        is_max = s == s[start:].max()\n",
    "        is_min = s == s[start:].min()\n",
    "        return ['color: green' if vmax else ('color: red' if vmin else '') \\\n",
    "                for vmax,vmin in zip(is_max,is_min)]\n",
    "    else:\n",
    "        ret = []\n",
    "        for l in level:\n",
    "            is_max = s[l] == s[l][start:].max()\n",
    "            is_min = s[l] == s[l][start:].min()\n",
    "            ret = ret + ['color: green' if vmax else ('color: red' if vmin else '') \\\n",
    "                    for vmax,vmin in zip(is_max,is_min)]\n",
    "        return ret\n",
    "\n",
    "\n",
    "h = partial(highlight_max_min, start=0, level=['ROC AUC', 'Cost'])\n",
    "df2.style.apply(h, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 352,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\begin{tabular}{lrrrrrrr}\n",
      "\\toprule\n",
      "{} &      C4.5 &     CC4.5 &     EC4.5 &      CART &     CCART &     ECART &       ASR \\\\\n",
      "dataset  &           &           &           &           &           &           &           \\\\\n",
      "\\midrule\n",
      "iris     &  0.933333 &  0.933333 &  0.938889 &  0.933333 &  0.933333 &  0.938395 &  0.932346 \\\\\n",
      "obesity  &  0.882707 &  0.882707 &  0.861682 &  0.881246 &  0.881246 &  0.864197 &  0.828230 \\\\\n",
      "shoppers &  0.587790 &  0.587790 &  0.560693 &  0.587180 &  0.587180 &  0.562041 &  0.561894 \\\\\n",
      "default  &  0.680898 &  0.680898 &  0.675584 &  0.677157 &  0.677157 &  0.647149 &  0.647749 \\\\\n",
      "firewall &  0.917959 &  0.917959 &  0.917942 &  0.917959 &  0.917959 &  0.917942 &  0.917959 \\\\\n",
      "dota     &  0.539365 &  0.539365 &  0.526418 &  0.540485 &  0.540485 &  0.526300 &  0.526345 \\\\\n",
      "diab     &  0.598823 &  0.598823 &  0.570652 &  0.597501 &  0.597501 &  0.569781 &  0.569502 \\\\\n",
      "poke     &  0.692298 &  0.692298 &  0.694119 &  0.691758 &  0.691758 &  0.696221 &  0.693099 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n",
      "\n",
      "iris\t& 0.933\t& 0.933\t& \\bf{0.939}\t& 0.933\t& 0.933\t& 0.938\t& 0.932\\\\\n",
      "obesity\t& \\bf{0.883}\t& \\bf{0.883}\t& 0.862\t& 0.881\t& 0.881\t& 0.864\t& 0.828\\\\\n",
      "shoppers\t& \\bf{0.588}\t& \\bf{0.588}\t& 0.561\t& 0.587\t& 0.587\t& 0.562\t& 0.562\\\\\n",
      "default\t& \\bf{0.681}\t& \\bf{0.681}\t& 0.676\t& 0.677\t& 0.677\t& 0.647\t& 0.648\\\\\n",
      "firewall\t& \\bf{0.918}\t& \\bf{0.918}\t& 0.918\t& \\bf{0.918}\t& \\bf{0.918}\t& 0.918\t& \\bf{0.918}\\\\\n",
      "dota\t& 0.539\t& 0.539\t& 0.526\t& \\bf{0.54}\t& \\bf{0.54}\t& 0.526\t& 0.526\\\\\n",
      "diab\t& \\bf{0.599}\t& \\bf{0.599}\t& 0.571\t& 0.598\t& 0.598\t& 0.57\t& 0.57\\\\\n",
      "poke\t& 0.692\t& 0.692\t& 0.694\t& 0.692\t& 0.692\t& \\bf{0.696}\t& 0.693\n",
      "\n",
      "\\begin{tabular}{lrrrrrrr}\n",
      "\\toprule\n",
      "{} &       C4.5 &      CC4.5 &     EC4.5 &       CART &      CCART &     ECART &       ASR \\\\\n",
      "dataset  &            &            &           &            &            &           &           \\\\\n",
      "\\midrule\n",
      "iris     &   1.701587 &   1.701587 &  1.736508 &   1.701587 &   1.701587 &  1.860317 &  1.853968 \\\\\n",
      "obesity  &   4.894381 &   4.894381 &  4.883322 &   5.083277 &   5.083277 &  4.843151 &  5.077183 \\\\\n",
      "shoppers &   6.018422 &   6.018422 &  4.648399 &   6.740779 &   6.740779 &  4.651527 &  4.653536 \\\\\n",
      "default  &  10.553730 &  10.553730 &  7.612349 &  11.295032 &  11.295032 &  7.615476 &  7.613984 \\\\\n",
      "firewall &   2.380515 &   2.380515 &  2.432297 &   2.380515 &   2.380515 &  2.432297 &  2.973717 \\\\\n",
      "dota     &  18.075441 &  18.075441 &  8.805789 &  17.276087 &  17.276087 &  8.805805 &  8.805784 \\\\\n",
      "diab     &   7.530818 &   7.530818 &  5.525226 &   7.548987 &   7.548987 &  5.504913 &  5.505367 \\\\\n",
      "poke     &   3.192893 &   3.192893 &  3.190385 &   3.187657 &   3.187657 &  3.186284 &  3.189034 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n",
      "\n",
      "iris\t& \\bf{1.702}\t& \\bf{1.702}\t& 1.737\t& \\bf{1.702}\t& \\bf{1.702}\t& 1.86\t& 1.854\\\\\n",
      "obesity\t& 4.894\t& 4.894\t& 4.883\t& 5.083\t& 5.083\t& \\bf{4.843}\t& 5.077\\\\\n",
      "shoppers\t& 6.018\t& 6.018\t& \\bf{4.648}\t& 6.741\t& 6.741\t& 4.652\t& 4.654\\\\\n",
      "default\t& 10.554\t& 10.554\t& \\bf{7.612}\t& 11.295\t& 11.295\t& 7.615\t& 7.614\\\\\n",
      "firewall\t& \\bf{2.381}\t& \\bf{2.381}\t& 2.432\t& \\bf{2.381}\t& \\bf{2.381}\t& 2.432\t& 2.974\\\\\n",
      "dota\t& 18.075\t& 18.075\t& 8.806\t& 17.276\t& 17.276\t& 8.806\t& \\bf{8.806}\\\\\n",
      "diab\t& 7.531\t& 7.531\t& 5.525\t& 7.549\t& 7.549\t& \\bf{5.505}\t& 5.505\\\\\n",
      "poke\t& 3.193\t& 3.193\t& 3.19\t& 3.188\t& 3.188\t& \\bf{3.186}\t& 3.189\n"
     ]
    }
   ],
   "source": [
    "def latex_strip(tex):\n",
    "    tex = tex[tex.find('midrule') + 8:]\n",
    "    return tex[:tex.find('\\\\bottomrule')]\n",
    "\n",
    "def isnum(s):\n",
    "    try: \n",
    "        float(s)\n",
    "        return True\n",
    "    except ValueError:\n",
    "        return False\n",
    "\n",
    "def xval(line, start=1, use_max=True, prec=3):\n",
    "    line = line.replace(' ','').strip('\\\\')\n",
    "    f = max if use_max else min\n",
    "    items = line.split('&')\n",
    "    nums = [float(i) for i in items[start:] if isnum(i)] # drop non-numerical\n",
    "    x = f(nums)\n",
    "    newitems = ['\\\\bf{'+str(round(n,prec))+'}' if n==x else str(round(n,prec)) for n in nums]\n",
    "    return '\\t& '.join(items[:start] + newitems)\n",
    "\n",
    "tex = df2['ROC AUC'].to_latex(escape=False)\n",
    "print(tex)\n",
    "print()\n",
    "tex = latex_strip(tex)\n",
    "tex = '\\\\\\\\\\n'.join([xval(line,start=1,use_max=True,prec=3) for line in tex.split('\\n') if line.strip() != ''])\n",
    "print(tex)\n",
    "print()\n",
    "\n",
    "tex = df2['Cost'].to_latex(escape=False)\n",
    "print(tex)\n",
    "print()\n",
    "tex = latex_strip(tex)\n",
    "tex = '\\\\\\\\\\n'.join([xval(line,start=1,use_max=False,prec=3) for line in tex.split('\\n') if line.strip() != ''])\n",
    "print(tex)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Post-doc tests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "import scipy.stats as ss\n",
    "import scikit_posthocs as sp\n",
    "from Orange.evaluation import compute_CD, graph_ranks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 307,
   "metadata": {},
   "outputs": [],
   "source": [
    "# alpha='0.1'\n",
    "alpha='0.05'\n",
    "comm_cols = ['ASR', 'IP', 'Bal']\n",
    "\n",
    "# full_cols = ['C4.5', 'EC4.5', 'CART', 'ECART', 'ASR', 'IP', 'Bal'] + \\\n",
    "#     ['pC4.5', 'pEC4.5', 'pCART', 'pECART', 'pASR', 'pIP', 'pBal']\n",
    "\n",
    "# typ = 'unit-c45'\n",
    "# cols = ['C4.5', 'EC4.5']\n",
    "# pcols = ['p'+c for c in cols]\n",
    "\n",
    "# typ = 'unit-cart'\n",
    "# cols = ['CART', 'ECART']\n",
    "# pcols = ['p'+c for c in cols]\n",
    "\n",
    "\n",
    "full_cols = ['CC4.5', 'EC4.5', 'CCART', 'ECART', 'ASR', 'IP', 'Bal'] + \\\n",
    "    ['pCC4.5', 'pEC4.5', 'pCCART', 'pECART', 'pASR', 'pIP', 'pBal']\n",
    "\n",
    "# typ = 'random-c45'\n",
    "# cols = ['CC4.5', 'EC4.5']\n",
    "# pcols = ['p'+c for c in cols]\n",
    "\n",
    "typ = 'random-cart'\n",
    "cols = ['CCART', 'ECART']\n",
    "pcols = ['p'+c for c in cols]\n",
    "\n",
    "\n",
    "cols = cols + comm_cols\n",
    "baseline = len(cols) + 2 - 1 # pEC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 308,
   "metadata": {},
   "outputs": [],
   "source": [
    "def tikz_cd(code, height=0.35):\n",
    "    code = code.replace('scale=0.5', 'scale=0.7')\n",
    "#     code = code.replace('0.133333333333333', '0.2')\n",
    "#     code = code.replace('0.166666666666667', '0.2')\n",
    "#     code = code.replace('0.833333333333333', '0.8')\n",
    "#     code = code.replace('0.866666666666667', '0.8')\n",
    "    \n",
    "    opt = '''\n",
    "width = 0.6\\\\textwidth, height = {}\\\\textwidth,'''.format(height)\n",
    "    key = '\\\\begin{axis}['\n",
    "    idx = code.find(key)\n",
    "    code = code[:idx+len(key)] + opt + code[idx+len(key):]\n",
    "    for i in range(1,12+1,2):\n",
    "        code = code.replace('{'+str(i)+'};', '{};')\n",
    "\n",
    "    return code"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## AUC"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 309,
   "metadata": {},
   "outputs": [],
   "source": [
    "# data: dataset-by-alg\n",
    "data = df2.groupby('dataset').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 310,
   "metadata": {},
   "outputs": [],
   "source": [
    "auc = data[['ROC AUC', 'ROC AUC (ccp)']]\n",
    "auc.columns = full_cols\n",
    "auc = auc[cols + pcols]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 311,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "FriedmanchisquareResult(statistic=23.55834829443449, pvalue=0.0006295011219514167)"
      ]
     },
     "execution_count": 311,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ss.friedmanchisquare(*auc.values.T)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 312,
   "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>CCART</th>\n",
       "      <th>ECART</th>\n",
       "      <th>ASR</th>\n",
       "      <th>IP</th>\n",
       "      <th>Bal</th>\n",
       "      <th>pCCART</th>\n",
       "      <th>pECART</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>CCART</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.738735</td>\n",
       "      <td>0.001505</td>\n",
       "      <td>0.317933</td>\n",
       "      <td>0.223038</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.717216</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ECART</th>\n",
       "      <td>0.738735</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.163405</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.652657</td>\n",
       "      <td>0.900000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ASR</th>\n",
       "      <td>0.001505</td>\n",
       "      <td>0.163405</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.545064</td>\n",
       "      <td>0.653790</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.176967</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>IP</th>\n",
       "      <td>0.317933</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.545064</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.240902</td>\n",
       "      <td>0.900000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Bal</th>\n",
       "      <td>0.223038</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.653790</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.162688</td>\n",
       "      <td>0.900000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pCCART</th>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.652657</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.240902</td>\n",
       "      <td>0.162688</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.631139</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pECART</th>\n",
       "      <td>0.717216</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.176967</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.631139</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           CCART     ECART       ASR        IP       Bal    pCCART    pECART\n",
       "CCART   1.000000  0.738735  0.001505  0.317933  0.223038  0.900000  0.717216\n",
       "ECART   0.738735  1.000000  0.163405  0.900000  0.900000  0.652657  0.900000\n",
       "ASR     0.001505  0.163405  1.000000  0.545064  0.653790  0.001000  0.176967\n",
       "IP      0.317933  0.900000  0.545064  1.000000  0.900000  0.240902  0.900000\n",
       "Bal     0.223038  0.900000  0.653790  0.900000  1.000000  0.162688  0.900000\n",
       "pCCART  0.900000  0.652657  0.001000  0.240902  0.162688  1.000000  0.631139\n",
       "pECART  0.717216  0.900000  0.176967  0.900000  0.900000  0.631139  1.000000"
      ]
     },
     "execution_count": 312,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sp.posthoc_nemenyi_friedman(auc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 313,
   "metadata": {},
   "outputs": [],
   "source": [
    "# https://edisciplinas.usp.br/pluginfile.php/4129451/mod_resource/content/1/model_selection_evaluation.pdf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 314,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculating the ranks of the algorithms for each dataset. The value of p is multipled by -1\n",
    "# because the rankdata method ranks from the smallest to the greatest performance values.\n",
    "# Since we are considering AUC as our performance measure, we want larger values to be best ranked.\n",
    "ranks = np.array([ss.rankdata(-p) for p in auc.values])\n",
    "# Calculating the average ranks.\n",
    "average_ranks = np.mean(ranks, axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 315,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CCART average rank: 2.875\n",
      "ECART average rank: 3.875\n",
      "ASR average rank: 5.575\n",
      "IP average rank: 4.35\n",
      "Bal average rank: 4.65\n",
      "pCCART average rank: 2.775\n",
      "pECART average rank: 3.9\n"
     ]
    }
   ],
   "source": [
    "print('\\n'.join('{} average rank: {}'.format(a, r) for a, r in zip(auc.columns, average_ranks)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 316,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 600x195 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# This method computes the critical difference for Nemenyi test with alpha=0.1.\n",
    "# For some reason, this method only accepts alpha='0.05' or alpha='0.1'.\n",
    "cd = compute_CD(average_ranks, \n",
    "                n=auc.shape[0],\n",
    "                alpha=alpha,\n",
    "                test='nemenyi')\n",
    "# This method generates the plot.\n",
    "graph_ranks(average_ranks,\n",
    "            names=auc.columns,\n",
    "            cd=cd,\n",
    "            width=6,\n",
    "            textspace=2,\n",
    "            reverse=True,\n",
    "            filename='figs/cd-nemenyi-auc-{}-{}.pdf'.format(typ,alpha.replace('.','')))\n",
    "# plt.show()\n",
    "code = tikzplotlib.get_tikz_code()\n",
    "code = tikz_cd(code)\n",
    "with open('figs/cd-nemenyi-auc-{}-{}.tex'.format(typ,alpha.replace('.','')), 'w') as fout:\n",
    "    fout.write(code)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 317,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 600x160 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# This method computes the critical difference for Bonferroni-Dunn test with alpha=0.05.\n",
    "# For some reason, this method only accepts alpha='0.05' or alpha='0.1'.\n",
    "cd = compute_CD(average_ranks,\n",
    "                n=auc.shape[0],\n",
    "                alpha=alpha,\n",
    "                test='bonferroni-dunn')\n",
    "# This method generates the plot.\n",
    "graph_ranks(average_ranks,\n",
    "            names=auc.columns,\n",
    "            cd=cd,\n",
    "            cdmethod=baseline,\n",
    "            width=6,\n",
    "            textspace=2,            \n",
    "            reverse=True,\n",
    "            filename='figs/cd-bonferroni-auc-{}-{}.pdf'.format(typ,alpha.replace('.','')))\n",
    "# plt.show()\n",
    "code = tikzplotlib.get_tikz_code()\n",
    "code = tikz_cd(code)\n",
    "with open('figs/cd-bonferroni-auc-{}-{}.tex'.format(typ,alpha.replace('.','')), 'w') as fout:\n",
    "    fout.write(code)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Cost"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 318,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = df2.groupby('dataset').mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 319,
   "metadata": {},
   "outputs": [],
   "source": [
    "cost = data[['Cost', 'Cost (ccp)']]\n",
    "cost.columns = full_cols\n",
    "cost = cost[cols + pcols]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 320,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "FriedmanchisquareResult(statistic=105.73321396598035, pvalue=1.5925137965963477e-20)"
      ]
     },
     "execution_count": 320,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ss.friedmanchisquare(*cost.values.T)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 321,
   "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>CCART</th>\n",
       "      <th>ECART</th>\n",
       "      <th>ASR</th>\n",
       "      <th>IP</th>\n",
       "      <th>Bal</th>\n",
       "      <th>pCCART</th>\n",
       "      <th>pECART</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>CCART</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.011844</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.738735</td>\n",
       "      <td>0.001000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ECART</th>\n",
       "      <td>0.011844</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.176967</td>\n",
       "      <td>0.115802</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.455921</td>\n",
       "      <td>0.150125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ASR</th>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.176967</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.900000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>IP</th>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.115802</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.900000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Bal</th>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.216029</td>\n",
       "      <td>0.001000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pCCART</th>\n",
       "      <td>0.738735</td>\n",
       "      <td>0.455921</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.216029</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.001000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pECART</th>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.150125</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           CCART     ECART       ASR        IP       Bal    pCCART    pECART\n",
       "CCART   1.000000  0.011844  0.001000  0.001000  0.900000  0.738735  0.001000\n",
       "ECART   0.011844  1.000000  0.176967  0.115802  0.001000  0.455921  0.150125\n",
       "ASR     0.001000  0.176967  1.000000  0.900000  0.001000  0.001000  0.900000\n",
       "IP      0.001000  0.115802  0.900000  1.000000  0.001000  0.001000  0.900000\n",
       "Bal     0.900000  0.001000  0.001000  0.001000  1.000000  0.216029  0.001000\n",
       "pCCART  0.738735  0.455921  0.001000  0.001000  0.216029  1.000000  0.001000\n",
       "pECART  0.001000  0.150125  0.900000  0.900000  0.001000  0.001000  1.000000"
      ]
     },
     "execution_count": 321,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sp.posthoc_nemenyi_friedman(cost)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 322,
   "metadata": {},
   "outputs": [],
   "source": [
    "# https://edisciplinas.usp.br/pluginfile.php/4129451/mod_resource/content/1/model_selection_evaluation.pdf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 323,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculating the ranks of the algorithms for each dataset. The value of p is multipled by -1\n",
    "# because the rankdata method ranks from the smallest to the greatest performance values.\n",
    "# Since we are considering AUC as our performance measure, we want larger values to be best ranked.\n",
    "ranks = np.array([ss.rankdata(p) for p in cost.values])\n",
    "# Calculating the average ranks.\n",
    "average_ranks = np.mean(ranks, axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 324,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CCART average rank: 6.125\n",
      "ECART average rank: 3.8\n",
      "ASR average rank: 2.125\n",
      "IP average rank: 2.0\n",
      "Bal average rank: 6.75\n",
      "pCCART average rank: 5.125\n",
      "pECART average rank: 2.075\n"
     ]
    }
   ],
   "source": [
    "print('\\n'.join('{} average rank: {}'.format(a, r) for a, r in zip(cost.columns, average_ranks)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 325,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 600x205 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# This method computes the critical difference for Nemenyi test with alpha=0.1.\n",
    "# For some reason, this method only accepts alpha='0.05' or alpha='0.1'.\n",
    "cd = compute_CD(average_ranks, \n",
    "                n=cost.shape[0],\n",
    "                alpha=alpha,\n",
    "                test='nemenyi')\n",
    "# This method generates the plot.\n",
    "graph_ranks(average_ranks,\n",
    "            names=cost.columns,\n",
    "            cd=cd,\n",
    "            width=6,\n",
    "            textspace=2,\n",
    "            reverse=True,\n",
    "            filename='figs/cd-nemenyi-cost-{}-{}.pdf'.format(typ,alpha.replace('.','')))\n",
    "# plt.show()\n",
    "code = tikzplotlib.get_tikz_code()\n",
    "code = tikz_cd(code)\n",
    "with open('figs/cd-nemenyi-cost-{}-{}.tex'.format(typ,alpha.replace('.','')), 'w') as fout:\n",
    "    fout.write(code)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 326,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 600x160 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# This method computes the critical difference for Bonferroni-Dunn test with alpha=0.05.\n",
    "# For some reason, this method only accepts alpha='0.05' or alpha='0.1'.\n",
    "cd = compute_CD(average_ranks,\n",
    "                n=cost.shape[0],\n",
    "                alpha=alpha,\n",
    "                test='bonferroni-dunn')\n",
    "# This method generates the plot.\n",
    "graph_ranks(average_ranks,\n",
    "            names=cost.columns,\n",
    "            cd=cd,\n",
    "            cdmethod=baseline,\n",
    "            width=6,\n",
    "            textspace=2,\n",
    "            reverse=True,\n",
    "            filename='figs/cd-bonferroni-cost-{}-{}.pdf'.format(typ,alpha.replace('.','')))\n",
    "# plt.show()\n",
    "code = tikzplotlib.get_tikz_code()\n",
    "code = tikz_cd(code)\n",
    "with open('figs/cd-bonferroni-cost-{}-{}.tex'.format(typ,alpha.replace('.','')), 'w') as fout:\n",
    "    fout.write(code)"
   ]
  }
 ],
 "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": "3.8.2"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "165px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}