Result Analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "afffca84-948e-475f-a7c8-928ab3307a39",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1dd66b84",
   "metadata": {},
   "source": [
    "## Loading Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "4ec0297d-c446-41e6-8410-d12f61a6e6a8",
   "metadata": {},
   "outputs": [],
   "source": [
    "metafeatures = pd.read_csv(\"metafeatures_clean.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "4d6390f3",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "f__pymfe.general.attr_to_inst\n",
      "f__pymfe.general.cat_to_num\n",
      "f__pymfe.general.freq_class.count\n",
      "f__pymfe.general.freq_class.histogram.0\n",
      "f__pymfe.general.freq_class.histogram.1\n",
      "f__pymfe.general.freq_class.histogram.2\n",
      "f__pymfe.general.freq_class.histogram.3\n",
      "f__pymfe.general.freq_class.histogram.4\n",
      "f__pymfe.general.freq_class.histogram.5\n",
      "f__pymfe.general.freq_class.histogram.6\n",
      "f__pymfe.general.freq_class.histogram.7\n",
      "f__pymfe.general.freq_class.histogram.8\n",
      "f__pymfe.general.freq_class.histogram.9\n",
      "f__pymfe.general.freq_class.iq_range\n",
      "f__pymfe.general.freq_class.kurtosis\n",
      "f__pymfe.general.freq_class.max\n",
      "f__pymfe.general.freq_class.mean\n",
      "f__pymfe.general.freq_class.median\n",
      "f__pymfe.general.freq_class.min\n",
      "f__pymfe.general.freq_class.quantiles.0\n",
      "f__pymfe.general.freq_class.quantiles.1\n",
      "f__pymfe.general.freq_class.quantiles.2\n",
      "f__pymfe.general.freq_class.quantiles.3\n",
      "f__pymfe.general.freq_class.quantiles.4\n",
      "f__pymfe.general.freq_class.range\n",
      "f__pymfe.general.freq_class.sd\n",
      "f__pymfe.general.freq_class.skewness\n",
      "f__pymfe.general.inst_to_attr\n",
      "f__pymfe.general.nr_attr\n",
      "f__pymfe.general.nr_bin\n",
      "f__pymfe.general.nr_cat\n",
      "f__pymfe.general.nr_class\n",
      "f__pymfe.general.nr_inst\n",
      "f__pymfe.general.nr_num\n",
      "f__pymfe.general.num_to_cat\n"
     ]
    }
   ],
   "source": [
    "for column in metafeatures.columns:\n",
    "    if \"general\" in column:\n",
    "        print(column)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "50429359-e52c-4018-a96d-b1676be5be29",
   "metadata": {},
   "outputs": [],
   "source": [
    "metafeatures_datasets = metafeatures[[\"dataset_name\",\"f__pymfe.general.nr_inst\", \"f__pymfe.general.nr_cat\", \n",
    "                                        \"f__pymfe.general.nr_num\",\"f__pymfe.general.nr_class\",\n",
    "                                       \"f__pymfe.general.nr_attr\"]].groupby(\"dataset_name\").max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "769347db",
   "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>f__pymfe.general.nr_inst</th>\n",
       "      <th>f__pymfe.general.nr_cat</th>\n",
       "      <th>f__pymfe.general.nr_num</th>\n",
       "      <th>f__pymfe.general.nr_class</th>\n",
       "      <th>f__pymfe.general.nr_attr</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dataset_name</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>openml__APSFailure__168868</th>\n",
       "      <td>60800</td>\n",
       "      <td>0</td>\n",
       "      <td>170</td>\n",
       "      <td>2</td>\n",
       "      <td>170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__Amazon_employee_access__34539</th>\n",
       "      <td>26216</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__Australian__146818</th>\n",
       "      <td>552</td>\n",
       "      <td>8</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__Bioresponse__9910</th>\n",
       "      <td>3001</td>\n",
       "      <td>0</td>\n",
       "      <td>1776</td>\n",
       "      <td>2</td>\n",
       "      <td>1776</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__CIFAR_10__167124</th>\n",
       "      <td>48000</td>\n",
       "      <td>0</td>\n",
       "      <td>3072</td>\n",
       "      <td>10</td>\n",
       "      <td>3072</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__walking-activity__9945</th>\n",
       "      <td>119466</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>22</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__wall-robot-navigation__9960</th>\n",
       "      <td>4366</td>\n",
       "      <td>0</td>\n",
       "      <td>24</td>\n",
       "      <td>4</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__wdbc__9946</th>\n",
       "      <td>456</td>\n",
       "      <td>0</td>\n",
       "      <td>30</td>\n",
       "      <td>2</td>\n",
       "      <td>30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__wilt__146820</th>\n",
       "      <td>3872</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__yeast__145793</th>\n",
       "      <td>1016</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>176 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                       f__pymfe.general.nr_inst  \\\n",
       "dataset_name                                                      \n",
       "openml__APSFailure__168868                                60800   \n",
       "openml__Amazon_employee_access__34539                     26216   \n",
       "openml__Australian__146818                                  552   \n",
       "openml__Bioresponse__9910                                  3001   \n",
       "openml__CIFAR_10__167124                                  48000   \n",
       "...                                                         ...   \n",
       "openml__walking-activity__9945                           119466   \n",
       "openml__wall-robot-navigation__9960                        4366   \n",
       "openml__wdbc__9946                                          456   \n",
       "openml__wilt__146820                                       3872   \n",
       "openml__yeast__145793                                      1016   \n",
       "\n",
       "                                       f__pymfe.general.nr_cat  \\\n",
       "dataset_name                                                     \n",
       "openml__APSFailure__168868                                   0   \n",
       "openml__Amazon_employee_access__34539                        9   \n",
       "openml__Australian__146818                                   8   \n",
       "openml__Bioresponse__9910                                    0   \n",
       "openml__CIFAR_10__167124                                     0   \n",
       "...                                                        ...   \n",
       "openml__walking-activity__9945                               0   \n",
       "openml__wall-robot-navigation__9960                          0   \n",
       "openml__wdbc__9946                                           0   \n",
       "openml__wilt__146820                                         0   \n",
       "openml__yeast__145793                                        0   \n",
       "\n",
       "                                       f__pymfe.general.nr_num  \\\n",
       "dataset_name                                                     \n",
       "openml__APSFailure__168868                                 170   \n",
       "openml__Amazon_employee_access__34539                        0   \n",
       "openml__Australian__146818                                   6   \n",
       "openml__Bioresponse__9910                                 1776   \n",
       "openml__CIFAR_10__167124                                  3072   \n",
       "...                                                        ...   \n",
       "openml__walking-activity__9945                               4   \n",
       "openml__wall-robot-navigation__9960                         24   \n",
       "openml__wdbc__9946                                          30   \n",
       "openml__wilt__146820                                         5   \n",
       "openml__yeast__145793                                        8   \n",
       "\n",
       "                                       f__pymfe.general.nr_class  \\\n",
       "dataset_name                                                       \n",
       "openml__APSFailure__168868                                     2   \n",
       "openml__Amazon_employee_access__34539                          2   \n",
       "openml__Australian__146818                                     2   \n",
       "openml__Bioresponse__9910                                      2   \n",
       "openml__CIFAR_10__167124                                      10   \n",
       "...                                                          ...   \n",
       "openml__walking-activity__9945                                22   \n",
       "openml__wall-robot-navigation__9960                            4   \n",
       "openml__wdbc__9946                                             2   \n",
       "openml__wilt__146820                                           2   \n",
       "openml__yeast__145793                                          4   \n",
       "\n",
       "                                       f__pymfe.general.nr_attr  \n",
       "dataset_name                                                     \n",
       "openml__APSFailure__168868                                  170  \n",
       "openml__Amazon_employee_access__34539                         9  \n",
       "openml__Australian__146818                                   14  \n",
       "openml__Bioresponse__9910                                  1776  \n",
       "openml__CIFAR_10__167124                                   3072  \n",
       "...                                                         ...  \n",
       "openml__walking-activity__9945                                4  \n",
       "openml__wall-robot-navigation__9960                          24  \n",
       "openml__wdbc__9946                                           30  \n",
       "openml__wilt__146820                                          5  \n",
       "openml__yeast__145793                                         8  \n",
       "\n",
       "[176 rows x 5 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metafeatures_datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "09ba13df-b1ad-4381-82c6-106d7e615f04",
   "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>dataset_fold_id</th>\n",
       "      <th>dataset_name</th>\n",
       "      <th>target_type</th>\n",
       "      <th>alg_name</th>\n",
       "      <th>hparam_source</th>\n",
       "      <th>Log Loss__train</th>\n",
       "      <th>Log Loss__val</th>\n",
       "      <th>Log Loss__test</th>\n",
       "      <th>AUC__train</th>\n",
       "      <th>AUC__val</th>\n",
       "      <th>...</th>\n",
       "      <th>Accuracy__train</th>\n",
       "      <th>Accuracy__val</th>\n",
       "      <th>Accuracy__test</th>\n",
       "      <th>F1__train</th>\n",
       "      <th>F1__val</th>\n",
       "      <th>F1__test</th>\n",
       "      <th>training_time</th>\n",
       "      <th>eval-time__train</th>\n",
       "      <th>eval-time__val</th>\n",
       "      <th>eval-time__test</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>openml__APSFailure__168868__fold_0</td>\n",
       "      <td>openml__APSFailure__168868</td>\n",
       "      <td>binary</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>default</td>\n",
       "      <td>0.018501</td>\n",
       "      <td>0.026880</td>\n",
       "      <td>0.020194</td>\n",
       "      <td>0.990573</td>\n",
       "      <td>0.986760</td>\n",
       "      <td>...</td>\n",
       "      <td>0.994523</td>\n",
       "      <td>0.991053</td>\n",
       "      <td>0.993421</td>\n",
       "      <td>0.994523</td>\n",
       "      <td>0.991053</td>\n",
       "      <td>0.993421</td>\n",
       "      <td>4.039096</td>\n",
       "      <td>0.185535</td>\n",
       "      <td>0.028121</td>\n",
       "      <td>0.034166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>openml__APSFailure__168868__fold_0</td>\n",
       "      <td>openml__APSFailure__168868</td>\n",
       "      <td>binary</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>random_1_s0</td>\n",
       "      <td>0.011968</td>\n",
       "      <td>0.024195</td>\n",
       "      <td>0.017324</td>\n",
       "      <td>0.995348</td>\n",
       "      <td>0.988589</td>\n",
       "      <td>...</td>\n",
       "      <td>0.997533</td>\n",
       "      <td>0.991974</td>\n",
       "      <td>0.994605</td>\n",
       "      <td>0.997533</td>\n",
       "      <td>0.991974</td>\n",
       "      <td>0.994605</td>\n",
       "      <td>3.624039</td>\n",
       "      <td>0.129900</td>\n",
       "      <td>0.020647</td>\n",
       "      <td>0.017871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>openml__APSFailure__168868__fold_0</td>\n",
       "      <td>openml__APSFailure__168868</td>\n",
       "      <td>binary</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>random_10_s0</td>\n",
       "      <td>0.022589</td>\n",
       "      <td>0.029668</td>\n",
       "      <td>0.023737</td>\n",
       "      <td>0.990179</td>\n",
       "      <td>0.985547</td>\n",
       "      <td>...</td>\n",
       "      <td>0.992566</td>\n",
       "      <td>0.989605</td>\n",
       "      <td>0.990921</td>\n",
       "      <td>0.992566</td>\n",
       "      <td>0.989605</td>\n",
       "      <td>0.990921</td>\n",
       "      <td>2.494685</td>\n",
       "      <td>0.074099</td>\n",
       "      <td>0.021543</td>\n",
       "      <td>0.019580</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>openml__APSFailure__168868__fold_0</td>\n",
       "      <td>openml__APSFailure__168868</td>\n",
       "      <td>binary</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>random_11_s0</td>\n",
       "      <td>0.028515</td>\n",
       "      <td>0.034531</td>\n",
       "      <td>0.028539</td>\n",
       "      <td>0.987448</td>\n",
       "      <td>0.985578</td>\n",
       "      <td>...</td>\n",
       "      <td>0.990641</td>\n",
       "      <td>0.986974</td>\n",
       "      <td>0.990132</td>\n",
       "      <td>0.990641</td>\n",
       "      <td>0.986974</td>\n",
       "      <td>0.990132</td>\n",
       "      <td>2.532277</td>\n",
       "      <td>0.061874</td>\n",
       "      <td>0.014086</td>\n",
       "      <td>0.013869</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>openml__APSFailure__168868__fold_0</td>\n",
       "      <td>openml__APSFailure__168868</td>\n",
       "      <td>binary</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>random_12_s0</td>\n",
       "      <td>0.015119</td>\n",
       "      <td>0.025570</td>\n",
       "      <td>0.019329</td>\n",
       "      <td>0.993757</td>\n",
       "      <td>0.989098</td>\n",
       "      <td>...</td>\n",
       "      <td>0.996382</td>\n",
       "      <td>0.991447</td>\n",
       "      <td>0.993684</td>\n",
       "      <td>0.996382</td>\n",
       "      <td>0.991447</td>\n",
       "      <td>0.993684</td>\n",
       "      <td>4.220030</td>\n",
       "      <td>0.121606</td>\n",
       "      <td>0.034948</td>\n",
       "      <td>0.033927</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>985305</th>\n",
       "      <td>openml__yeast__145793__fold_9</td>\n",
       "      <td>openml__yeast__145793</td>\n",
       "      <td>classification</td>\n",
       "      <td>rtdl_ResNet</td>\n",
       "      <td>random_5_s0</td>\n",
       "      <td>1.075331</td>\n",
       "      <td>1.460503</td>\n",
       "      <td>1.362272</td>\n",
       "      <td>0.895086</td>\n",
       "      <td>0.847196</td>\n",
       "      <td>...</td>\n",
       "      <td>0.716535</td>\n",
       "      <td>0.606299</td>\n",
       "      <td>0.626984</td>\n",
       "      <td>0.715092</td>\n",
       "      <td>0.609722</td>\n",
       "      <td>0.626736</td>\n",
       "      <td>6.675143</td>\n",
       "      <td>0.048872</td>\n",
       "      <td>0.043895</td>\n",
       "      <td>0.042311</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>985306</th>\n",
       "      <td>openml__yeast__145793__fold_9</td>\n",
       "      <td>openml__yeast__145793</td>\n",
       "      <td>classification</td>\n",
       "      <td>rtdl_ResNet</td>\n",
       "      <td>random_6_s0</td>\n",
       "      <td>1.891564</td>\n",
       "      <td>2.155562</td>\n",
       "      <td>2.337385</td>\n",
       "      <td>0.873311</td>\n",
       "      <td>0.846246</td>\n",
       "      <td>...</td>\n",
       "      <td>0.692913</td>\n",
       "      <td>0.590551</td>\n",
       "      <td>0.603175</td>\n",
       "      <td>0.692256</td>\n",
       "      <td>0.590747</td>\n",
       "      <td>0.602349</td>\n",
       "      <td>3.138584</td>\n",
       "      <td>0.048738</td>\n",
       "      <td>0.041508</td>\n",
       "      <td>0.049839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>985307</th>\n",
       "      <td>openml__yeast__145793__fold_9</td>\n",
       "      <td>openml__yeast__145793</td>\n",
       "      <td>classification</td>\n",
       "      <td>rtdl_ResNet</td>\n",
       "      <td>random_7_s0</td>\n",
       "      <td>1.018551</td>\n",
       "      <td>1.379100</td>\n",
       "      <td>1.344363</td>\n",
       "      <td>0.896400</td>\n",
       "      <td>0.846951</td>\n",
       "      <td>...</td>\n",
       "      <td>0.713583</td>\n",
       "      <td>0.590551</td>\n",
       "      <td>0.611111</td>\n",
       "      <td>0.712877</td>\n",
       "      <td>0.592445</td>\n",
       "      <td>0.612003</td>\n",
       "      <td>9.424769</td>\n",
       "      <td>0.053142</td>\n",
       "      <td>0.048155</td>\n",
       "      <td>0.046927</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>985308</th>\n",
       "      <td>openml__yeast__145793__fold_9</td>\n",
       "      <td>openml__yeast__145793</td>\n",
       "      <td>classification</td>\n",
       "      <td>rtdl_ResNet</td>\n",
       "      <td>random_8_s0</td>\n",
       "      <td>2.329324</td>\n",
       "      <td>2.902949</td>\n",
       "      <td>2.762660</td>\n",
       "      <td>0.887312</td>\n",
       "      <td>0.854178</td>\n",
       "      <td>...</td>\n",
       "      <td>0.730315</td>\n",
       "      <td>0.637795</td>\n",
       "      <td>0.603175</td>\n",
       "      <td>0.730012</td>\n",
       "      <td>0.641874</td>\n",
       "      <td>0.604561</td>\n",
       "      <td>7.380742</td>\n",
       "      <td>0.052364</td>\n",
       "      <td>0.045521</td>\n",
       "      <td>0.043943</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>985309</th>\n",
       "      <td>openml__yeast__145793__fold_9</td>\n",
       "      <td>openml__yeast__145793</td>\n",
       "      <td>classification</td>\n",
       "      <td>rtdl_ResNet</td>\n",
       "      <td>random_9_s0</td>\n",
       "      <td>1.004896</td>\n",
       "      <td>1.330171</td>\n",
       "      <td>1.325328</td>\n",
       "      <td>0.894865</td>\n",
       "      <td>0.850567</td>\n",
       "      <td>...</td>\n",
       "      <td>0.703740</td>\n",
       "      <td>0.590551</td>\n",
       "      <td>0.619048</td>\n",
       "      <td>0.702840</td>\n",
       "      <td>0.593238</td>\n",
       "      <td>0.620122</td>\n",
       "      <td>9.830205</td>\n",
       "      <td>0.052547</td>\n",
       "      <td>0.048357</td>\n",
       "      <td>0.046934</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>985310 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                           dataset_fold_id                dataset_name  \\\n",
       "0       openml__APSFailure__168868__fold_0  openml__APSFailure__168868   \n",
       "1       openml__APSFailure__168868__fold_0  openml__APSFailure__168868   \n",
       "2       openml__APSFailure__168868__fold_0  openml__APSFailure__168868   \n",
       "3       openml__APSFailure__168868__fold_0  openml__APSFailure__168868   \n",
       "4       openml__APSFailure__168868__fold_0  openml__APSFailure__168868   \n",
       "...                                    ...                         ...   \n",
       "985305       openml__yeast__145793__fold_9       openml__yeast__145793   \n",
       "985306       openml__yeast__145793__fold_9       openml__yeast__145793   \n",
       "985307       openml__yeast__145793__fold_9       openml__yeast__145793   \n",
       "985308       openml__yeast__145793__fold_9       openml__yeast__145793   \n",
       "985309       openml__yeast__145793__fold_9       openml__yeast__145793   \n",
       "\n",
       "           target_type     alg_name hparam_source  Log Loss__train  \\\n",
       "0               binary     CatBoost       default         0.018501   \n",
       "1               binary     CatBoost   random_1_s0         0.011968   \n",
       "2               binary     CatBoost  random_10_s0         0.022589   \n",
       "3               binary     CatBoost  random_11_s0         0.028515   \n",
       "4               binary     CatBoost  random_12_s0         0.015119   \n",
       "...                ...          ...           ...              ...   \n",
       "985305  classification  rtdl_ResNet   random_5_s0         1.075331   \n",
       "985306  classification  rtdl_ResNet   random_6_s0         1.891564   \n",
       "985307  classification  rtdl_ResNet   random_7_s0         1.018551   \n",
       "985308  classification  rtdl_ResNet   random_8_s0         2.329324   \n",
       "985309  classification  rtdl_ResNet   random_9_s0         1.004896   \n",
       "\n",
       "        Log Loss__val  Log Loss__test  AUC__train  AUC__val  ...  \\\n",
       "0            0.026880        0.020194    0.990573  0.986760  ...   \n",
       "1            0.024195        0.017324    0.995348  0.988589  ...   \n",
       "2            0.029668        0.023737    0.990179  0.985547  ...   \n",
       "3            0.034531        0.028539    0.987448  0.985578  ...   \n",
       "4            0.025570        0.019329    0.993757  0.989098  ...   \n",
       "...               ...             ...         ...       ...  ...   \n",
       "985305       1.460503        1.362272    0.895086  0.847196  ...   \n",
       "985306       2.155562        2.337385    0.873311  0.846246  ...   \n",
       "985307       1.379100        1.344363    0.896400  0.846951  ...   \n",
       "985308       2.902949        2.762660    0.887312  0.854178  ...   \n",
       "985309       1.330171        1.325328    0.894865  0.850567  ...   \n",
       "\n",
       "        Accuracy__train  Accuracy__val  Accuracy__test  F1__train   F1__val  \\\n",
       "0              0.994523       0.991053        0.993421   0.994523  0.991053   \n",
       "1              0.997533       0.991974        0.994605   0.997533  0.991974   \n",
       "2              0.992566       0.989605        0.990921   0.992566  0.989605   \n",
       "3              0.990641       0.986974        0.990132   0.990641  0.986974   \n",
       "4              0.996382       0.991447        0.993684   0.996382  0.991447   \n",
       "...                 ...            ...             ...        ...       ...   \n",
       "985305         0.716535       0.606299        0.626984   0.715092  0.609722   \n",
       "985306         0.692913       0.590551        0.603175   0.692256  0.590747   \n",
       "985307         0.713583       0.590551        0.611111   0.712877  0.592445   \n",
       "985308         0.730315       0.637795        0.603175   0.730012  0.641874   \n",
       "985309         0.703740       0.590551        0.619048   0.702840  0.593238   \n",
       "\n",
       "        F1__test  training_time  eval-time__train  eval-time__val  \\\n",
       "0       0.993421       4.039096          0.185535        0.028121   \n",
       "1       0.994605       3.624039          0.129900        0.020647   \n",
       "2       0.990921       2.494685          0.074099        0.021543   \n",
       "3       0.990132       2.532277          0.061874        0.014086   \n",
       "4       0.993684       4.220030          0.121606        0.034948   \n",
       "...          ...            ...               ...             ...   \n",
       "985305  0.626736       6.675143          0.048872        0.043895   \n",
       "985306  0.602349       3.138584          0.048738        0.041508   \n",
       "985307  0.612003       9.424769          0.053142        0.048155   \n",
       "985308  0.604561       7.380742          0.052364        0.045521   \n",
       "985309  0.620122       9.830205          0.052547        0.048357   \n",
       "\n",
       "        eval-time__test  \n",
       "0              0.034166  \n",
       "1              0.017871  \n",
       "2              0.019580  \n",
       "3              0.013869  \n",
       "4              0.033927  \n",
       "...                 ...  \n",
       "985305         0.042311  \n",
       "985306         0.049839  \n",
       "985307         0.046927  \n",
       "985308         0.043943  \n",
       "985309         0.046934  \n",
       "\n",
       "[985310 rows x 21 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metadataset = pd.read_csv(\"metadataset_clean.csv\")\n",
    "metadataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "5730f883-0d87-4a06-a56b-c95094587aaa",
   "metadata": {},
   "outputs": [],
   "source": [
    "classDF = pd.DataFrame({\n",
    "    \"alg_name\" : metadataset.alg_name.unique(),\n",
    "    \"Group\" : [\"EnsembleTree\", \"Baselines\", \"NN\", \"Baselines\", \"EnsembleTree\", \"Baselines\", \"NN\", \"Baselines\", \"NN\",\"Baselines\", \"NN\", \"NN\", \"NN\",\"EnsembleTree\",\"NN\",\"NN\",\"NN\",\"NN\",\"NN\",\"NN\",\"NN\",\"NN\"]\n",
    "})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "7d34b458-b54f-48a4-981f-26430ce8540c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Accuracy__test                  0.821960\n",
       "f__pymfe.general.nr_inst    24454.613636\n",
       "f__pymfe.general.nr_cat        17.204545\n",
       "dtype: float64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## Average over all datasets\n",
    "metadataset.groupby([\"dataset_name\"]).first().join(metafeatures_datasets)[[\"Accuracy__test\", \"f__pymfe.general.nr_inst\", \"f__pymfe.general.nr_cat\"]].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "bf72ab17-47a5-47e8-90b2-6dc65fead7c3",
   "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>Accuracy__test</th>\n",
       "      <th>F1__test</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>alg_name</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>XGBoost</th>\n",
       "      <td>0.836729</td>\n",
       "      <td>0.832453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TabPFNModel</th>\n",
       "      <td>0.831847</td>\n",
       "      <td>0.827732</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CatBoost</th>\n",
       "      <td>0.827972</td>\n",
       "      <td>0.821764</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rtdl_ResNet</th>\n",
       "      <td>0.815581</td>\n",
       "      <td>0.807456</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DeepFM</th>\n",
       "      <td>0.808717</td>\n",
       "      <td>0.808717</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LightGBM</th>\n",
       "      <td>0.798470</td>\n",
       "      <td>0.795931</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DANet</th>\n",
       "      <td>0.798240</td>\n",
       "      <td>0.787423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NODE</th>\n",
       "      <td>0.787931</td>\n",
       "      <td>0.778137</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RandomForest</th>\n",
       "      <td>0.775992</td>\n",
       "      <td>0.762300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SAINT</th>\n",
       "      <td>0.772333</td>\n",
       "      <td>0.765437</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LinearModel</th>\n",
       "      <td>0.766261</td>\n",
       "      <td>0.756347</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KNN</th>\n",
       "      <td>0.763523</td>\n",
       "      <td>0.754418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rtdl_FTTransformer</th>\n",
       "      <td>0.758962</td>\n",
       "      <td>0.736049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DecisionTree</th>\n",
       "      <td>0.732140</td>\n",
       "      <td>0.717648</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SVM</th>\n",
       "      <td>0.729996</td>\n",
       "      <td>0.710238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NAM</th>\n",
       "      <td>0.727418</td>\n",
       "      <td>0.727418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>MLP</th>\n",
       "      <td>0.725751</td>\n",
       "      <td>0.697412</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rtdl_MLP</th>\n",
       "      <td>0.725056</td>\n",
       "      <td>0.696820</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TabNet</th>\n",
       "      <td>0.691911</td>\n",
       "      <td>0.681696</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>STG</th>\n",
       "      <td>0.671661</td>\n",
       "      <td>0.642759</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>VIME</th>\n",
       "      <td>0.655369</td>\n",
       "      <td>0.625133</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TabTransformer</th>\n",
       "      <td>0.626831</td>\n",
       "      <td>0.596181</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    Accuracy__test  F1__test\n",
       "alg_name                                    \n",
       "XGBoost                   0.836729  0.832453\n",
       "TabPFNModel               0.831847  0.827732\n",
       "CatBoost                  0.827972  0.821764\n",
       "rtdl_ResNet               0.815581  0.807456\n",
       "DeepFM                    0.808717  0.808717\n",
       "LightGBM                  0.798470  0.795931\n",
       "DANet                     0.798240  0.787423\n",
       "NODE                      0.787931  0.778137\n",
       "RandomForest              0.775992  0.762300\n",
       "SAINT                     0.772333  0.765437\n",
       "LinearModel               0.766261  0.756347\n",
       "KNN                       0.763523  0.754418\n",
       "rtdl_FTTransformer        0.758962  0.736049\n",
       "DecisionTree              0.732140  0.717648\n",
       "SVM                       0.729996  0.710238\n",
       "NAM                       0.727418  0.727418\n",
       "MLP                       0.725751  0.697412\n",
       "rtdl_MLP                  0.725056  0.696820\n",
       "TabNet                    0.691911  0.681696\n",
       "STG                       0.671661  0.642759\n",
       "VIME                      0.655369  0.625133\n",
       "TabTransformer            0.626831  0.596181"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Average perdormance of algorihtms\n",
    "metadataset.groupby([\"alg_name\"]).agg({\n",
    "    \"Accuracy__test\" : \"mean\",\n",
    "    \"F1__test\" : \"mean\"\n",
    "}).sort_values(\"Accuracy__test\", ascending =False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0fcacca7",
   "metadata": {},
   "source": [
    "## Splitting the Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "48ffabf3",
   "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>f__pymfe.general.nr_inst</th>\n",
       "      <th>f__pymfe.general.nr_cat</th>\n",
       "      <th>f__pymfe.general.nr_num</th>\n",
       "      <th>f__pymfe.general.nr_class</th>\n",
       "      <th>f__pymfe.general.nr_attr</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dataset_name</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>openml__lung-cancer__146024</th>\n",
       "      <td>26</td>\n",
       "      <td>56</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__labor__4</th>\n",
       "      <td>47</td>\n",
       "      <td>8</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__fl2000__3566</th>\n",
       "      <td>55</td>\n",
       "      <td>1</td>\n",
       "      <td>14</td>\n",
       "      <td>5</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__postoperative-patient-data__146210</th>\n",
       "      <td>72</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__fri_c0_100_5__3620</th>\n",
       "      <td>80</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__fri_c3_100_5__3779</th>\n",
       "      <td>80</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__analcatdata_chlamydia__3739</th>\n",
       "      <td>80</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__fertility__9984</th>\n",
       "      <td>80</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__visualizing_environmental__3602</th>\n",
       "      <td>89</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__rabe_266__3647</th>\n",
       "      <td>96</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__analcatdata_boxing1__3540</th>\n",
       "      <td>96</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__acute-inflammations__10089</th>\n",
       "      <td>96</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__visualizing_livestock__3731</th>\n",
       "      <td>104</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__transplant__3748</th>\n",
       "      <td>105</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>openml__lymph__10</th>\n",
       "      <td>120</td>\n",
       "      <td>15</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                            f__pymfe.general.nr_inst  \\\n",
       "dataset_name                                                           \n",
       "openml__lung-cancer__146024                                       26   \n",
       "openml__labor__4                                                  47   \n",
       "openml__fl2000__3566                                              55   \n",
       "openml__postoperative-patient-data__146210                        72   \n",
       "openml__fri_c0_100_5__3620                                        80   \n",
       "openml__fri_c3_100_5__3779                                        80   \n",
       "openml__analcatdata_chlamydia__3739                               80   \n",
       "openml__fertility__9984                                           80   \n",
       "openml__visualizing_environmental__3602                           89   \n",
       "openml__rabe_266__3647                                            96   \n",
       "openml__analcatdata_boxing1__3540                                 96   \n",
       "openml__acute-inflammations__10089                                96   \n",
       "openml__visualizing_livestock__3731                              104   \n",
       "openml__transplant__3748                                         105   \n",
       "openml__lymph__10                                                120   \n",
       "\n",
       "                                            f__pymfe.general.nr_cat  \\\n",
       "dataset_name                                                          \n",
       "openml__lung-cancer__146024                                      56   \n",
       "openml__labor__4                                                  8   \n",
       "openml__fl2000__3566                                              1   \n",
       "openml__postoperative-patient-data__146210                        8   \n",
       "openml__fri_c0_100_5__3620                                        0   \n",
       "openml__fri_c3_100_5__3779                                        0   \n",
       "openml__analcatdata_chlamydia__3739                               3   \n",
       "openml__fertility__9984                                           0   \n",
       "openml__visualizing_environmental__3602                           0   \n",
       "openml__rabe_266__3647                                            0   \n",
       "openml__analcatdata_boxing1__3540                                 3   \n",
       "openml__acute-inflammations__10089                                5   \n",
       "openml__visualizing_livestock__3731                               2   \n",
       "openml__transplant__3748                                          0   \n",
       "openml__lymph__10                                                15   \n",
       "\n",
       "                                            f__pymfe.general.nr_num  \\\n",
       "dataset_name                                                          \n",
       "openml__lung-cancer__146024                                       0   \n",
       "openml__labor__4                                                  8   \n",
       "openml__fl2000__3566                                             14   \n",
       "openml__postoperative-patient-data__146210                        0   \n",
       "openml__fri_c0_100_5__3620                                        5   \n",
       "openml__fri_c3_100_5__3779                                        5   \n",
       "openml__analcatdata_chlamydia__3739                               0   \n",
       "openml__fertility__9984                                           9   \n",
       "openml__visualizing_environmental__3602                           3   \n",
       "openml__rabe_266__3647                                            2   \n",
       "openml__analcatdata_boxing1__3540                                 0   \n",
       "openml__acute-inflammations__10089                                1   \n",
       "openml__visualizing_livestock__3731                               0   \n",
       "openml__transplant__3748                                          3   \n",
       "openml__lymph__10                                                 3   \n",
       "\n",
       "                                            f__pymfe.general.nr_class  \\\n",
       "dataset_name                                                            \n",
       "openml__lung-cancer__146024                                         3   \n",
       "openml__labor__4                                                    2   \n",
       "openml__fl2000__3566                                                5   \n",
       "openml__postoperative-patient-data__146210                          2   \n",
       "openml__fri_c0_100_5__3620                                          2   \n",
       "openml__fri_c3_100_5__3779                                          2   \n",
       "openml__analcatdata_chlamydia__3739                                 2   \n",
       "openml__fertility__9984                                             2   \n",
       "openml__visualizing_environmental__3602                             2   \n",
       "openml__rabe_266__3647                                              2   \n",
       "openml__analcatdata_boxing1__3540                                   2   \n",
       "openml__acute-inflammations__10089                                  2   \n",
       "openml__visualizing_livestock__3731                                 2   \n",
       "openml__transplant__3748                                            2   \n",
       "openml__lymph__10                                                   4   \n",
       "\n",
       "                                            f__pymfe.general.nr_attr  \n",
       "dataset_name                                                          \n",
       "openml__lung-cancer__146024                                       56  \n",
       "openml__labor__4                                                  16  \n",
       "openml__fl2000__3566                                              15  \n",
       "openml__postoperative-patient-data__146210                         8  \n",
       "openml__fri_c0_100_5__3620                                         5  \n",
       "openml__fri_c3_100_5__3779                                         5  \n",
       "openml__analcatdata_chlamydia__3739                                3  \n",
       "openml__fertility__9984                                            9  \n",
       "openml__visualizing_environmental__3602                            3  \n",
       "openml__rabe_266__3647                                             2  \n",
       "openml__analcatdata_boxing1__3540                                  3  \n",
       "openml__acute-inflammations__10089                                 6  \n",
       "openml__visualizing_livestock__3731                                2  \n",
       "openml__transplant__3748                                           3  \n",
       "openml__lymph__10                                                 18  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "small_datasets = metafeatures_datasets[metafeatures_datasets[\"f__pymfe.general.nr_inst\"] < 1000 ].sort_values(\"f__pymfe.general.nr_inst\")\n",
    "datasets = metafeatures_datasets[metafeatures_datasets[\"f__pymfe.general.nr_inst\"] >= 1000 ].reset_index()\n",
    "small_datasets.head(15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "ffda5446",
   "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>dataset_name</th>\n",
       "      <th>f__pymfe.general.nr_inst</th>\n",
       "      <th>f__pymfe.general.nr_cat</th>\n",
       "      <th>f__pymfe.general.nr_num</th>\n",
       "      <th>f__pymfe.general.nr_class</th>\n",
       "      <th>f__pymfe.general.nr_attr</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>openml__APSFailure__168868</td>\n",
       "      <td>60800</td>\n",
       "      <td>0</td>\n",
       "      <td>170</td>\n",
       "      <td>2</td>\n",
       "      <td>170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>openml__Amazon_employee_access__34539</td>\n",
       "      <td>26216</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>openml__Bioresponse__9910</td>\n",
       "      <td>3001</td>\n",
       "      <td>0</td>\n",
       "      <td>1776</td>\n",
       "      <td>2</td>\n",
       "      <td>1776</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>openml__CIFAR_10__167124</td>\n",
       "      <td>48000</td>\n",
       "      <td>0</td>\n",
       "      <td>3072</td>\n",
       "      <td>10</td>\n",
       "      <td>3072</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>openml__Census-Income__168340</td>\n",
       "      <td>239429</td>\n",
       "      <td>28</td>\n",
       "      <td>13</td>\n",
       "      <td>2</td>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>openml__volkert__168331</td>\n",
       "      <td>46648</td>\n",
       "      <td>0</td>\n",
       "      <td>180</td>\n",
       "      <td>10</td>\n",
       "      <td>180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>openml__walking-activity__9945</td>\n",
       "      <td>119466</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>22</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>103</th>\n",
       "      <td>openml__wall-robot-navigation__9960</td>\n",
       "      <td>4366</td>\n",
       "      <td>0</td>\n",
       "      <td>24</td>\n",
       "      <td>4</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>104</th>\n",
       "      <td>openml__wilt__146820</td>\n",
       "      <td>3872</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>openml__yeast__145793</td>\n",
       "      <td>1016</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>106 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                              dataset_name  f__pymfe.general.nr_inst  \\\n",
       "0               openml__APSFailure__168868                     60800   \n",
       "1    openml__Amazon_employee_access__34539                     26216   \n",
       "2                openml__Bioresponse__9910                      3001   \n",
       "3                 openml__CIFAR_10__167124                     48000   \n",
       "4            openml__Census-Income__168340                    239429   \n",
       "..                                     ...                       ...   \n",
       "101                openml__volkert__168331                     46648   \n",
       "102         openml__walking-activity__9945                    119466   \n",
       "103    openml__wall-robot-navigation__9960                      4366   \n",
       "104                   openml__wilt__146820                      3872   \n",
       "105                  openml__yeast__145793                      1016   \n",
       "\n",
       "     f__pymfe.general.nr_cat  f__pymfe.general.nr_num  \\\n",
       "0                          0                      170   \n",
       "1                          9                        0   \n",
       "2                          0                     1776   \n",
       "3                          0                     3072   \n",
       "4                         28                       13   \n",
       "..                       ...                      ...   \n",
       "101                        0                      180   \n",
       "102                        0                        4   \n",
       "103                        0                       24   \n",
       "104                        0                        5   \n",
       "105                        0                        8   \n",
       "\n",
       "     f__pymfe.general.nr_class  f__pymfe.general.nr_attr  \n",
       "0                            2                       170  \n",
       "1                            2                         9  \n",
       "2                            2                      1776  \n",
       "3                           10                      3072  \n",
       "4                            2                        41  \n",
       "..                         ...                       ...  \n",
       "101                         10                       180  \n",
       "102                         22                         4  \n",
       "103                          4                        24  \n",
       "104                          2                         5  \n",
       "105                          4                         8  \n",
       "\n",
       "[106 rows x 6 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "205602bb",
   "metadata": {},
   "outputs": [],
   "source": [
    "grouped_dataset = metadataset.groupby([\"dataset_name\", \"alg_name\"]).agg({\n",
    "    \"Accuracy__test\" : \"max\",\n",
    "    \"F1__test\" : \"max\"\n",
    "})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "767ced63-66e5-4aa7-a8d0-cf1d624c95ae",
   "metadata": {},
   "outputs": [],
   "source": [
    "aggregated_dataset = grouped_dataset.reset_index().merge( datasets, on = \"dataset_name\", how = \"inner\")\n",
    "aggregated_dataset = aggregated_dataset.merge(classDF, on = \"alg_name\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "9c601e66",
   "metadata": {},
   "outputs": [],
   "source": [
    "aggregated_dataset_small = grouped_dataset.reset_index().merge( small_datasets, on = \"dataset_name\", how = \"inner\")\n",
    "aggregated_dataset_small = aggregated_dataset_small.merge(classDF, on = \"alg_name\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "7e3c7aba-f9ae-428d-a38a-895f40ae4e19",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Commented to not introduce any bias\n",
    "\n",
    "#Dropping the dataset where the Linear Model has a performance grater that .95\n",
    "# Step 1: Identify the datasets where 'LinearModel' has 'Accuracy__test' > 0.95\n",
    "#high_perf_datasets = aggregated_dataset.query(\"(alg_name == 'LinearModel' or alg_name == 'DecisionTree') and Accuracy__test > 0.95\")['dataset_name'].unique()\n",
    "\n",
    "# Step 2: Filter out all rows from these datasets using a query that negates the inclusion of 'high_perf_datasets'\n",
    "#aggregated_dataset_difficult = aggregated_dataset.query(\"dataset_name not in @high_perf_datasets\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0cd5b592",
   "metadata": {},
   "source": [
    "## Finding Best Algorithms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "fce868af-4e65-4461-b3ec-311c576e73a0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Reset index so 'dataset_name' and 'alg_name' are back as columns\n",
    "grouped = aggregated_dataset.reset_index(drop=True)\n",
    "grouped_small = aggregated_dataset_small.reset_index(drop=True)\n",
    "\n",
    "# Find the max 'Accuracy_Test' within each 'dataset_name'\n",
    "max_accuracy_per_dataset = grouped.groupby('dataset_name')['Accuracy__test'].transform(\"max\")\n",
    "max_accuracy_per_dataset_small = grouped_small.groupby('dataset_name')['Accuracy__test'].transform(\"max\")\n",
    "\n",
    "# Filter rows where 'Accuracy_Test' matches the max 'Accuracy_Test' within each 'dataset_name'\n",
    "best_algs = grouped[grouped['Accuracy__test'] == max_accuracy_per_dataset]\n",
    "best_algs_small = grouped_small[grouped_small['Accuracy__test'] == max_accuracy_per_dataset_small]\n",
    "\n",
    "#Taking just the datasets where there are not ties\n",
    "tie_dataset = list(best_algs.groupby(\"dataset_name\").nunique().query(\"Group > 1\").index)\n",
    "tie_dataset_small = list(best_algs_small.groupby(\"dataset_name\").nunique().query(\"Group > 1\").index)\n",
    "\n",
    "best_algs = best_algs.query(\" dataset_name not in @tie_dataset \")\n",
    "best_algs_small = best_algs_small.query(\" dataset_name not in @tie_dataset_small \")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "1f8ad801",
   "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>dataset_name</th>\n",
       "      <th>alg_name</th>\n",
       "      <th>Accuracy__test</th>\n",
       "      <th>F1__test</th>\n",
       "      <th>f__pymfe.general.nr_inst</th>\n",
       "      <th>f__pymfe.general.nr_cat</th>\n",
       "      <th>f__pymfe.general.nr_num</th>\n",
       "      <th>f__pymfe.general.nr_class</th>\n",
       "      <th>f__pymfe.general.nr_attr</th>\n",
       "      <th>Group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [dataset_name, alg_name, Accuracy__test, F1__test, f__pymfe.general.nr_inst, f__pymfe.general.nr_cat, f__pymfe.general.nr_num, f__pymfe.general.nr_class, f__pymfe.general.nr_attr, Group]\n",
       "Index: []"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_algs.query(\" dataset_name == 'openml__jungle_chess_2pcs_raw_endgame_complete' \")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "932fca5b",
   "metadata": {},
   "outputs": [],
   "source": [
    "group_performance = aggregated_dataset.groupby([\"dataset_name\", \"Group\"]).agg({\n",
    "                                                        \"Accuracy__test\" : \"max\",\n",
    "                                                        \"F1__test\" : \"max\",\n",
    "                                                        \"f__pymfe.general.nr_inst\" : \"first\"\n",
    "                                                }).reset_index()\n",
    "max_per_dataset = group_performance.groupby(\"dataset_name\")[\"Accuracy__test\"].max().reset_index().rename(columns = {\"Accuracy__test\" : \"Accuracy_max\"})\n",
    "group_performance = group_performance.merge(max_per_dataset, on = \"dataset_name\")\n",
    "group_performance[\"gap\"] = group_performance.Accuracy__test - group_performance.Accuracy_max"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "d4a17c68",
   "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>dataset_name</th>\n",
       "      <th>Group</th>\n",
       "      <th>Accuracy__test</th>\n",
       "      <th>F1__test</th>\n",
       "      <th>f__pymfe.general.nr_inst</th>\n",
       "      <th>Accuracy_max</th>\n",
       "      <th>gap</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>178</th>\n",
       "      <td>openml__ldpa__9974</td>\n",
       "      <td>NN</td>\n",
       "      <td>0.622104</td>\n",
       "      <td>0.571127</td>\n",
       "      <td>131888</td>\n",
       "      <td>0.982713</td>\n",
       "      <td>-0.360609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65</th>\n",
       "      <td>openml__artificial-characters__14964</td>\n",
       "      <td>NN</td>\n",
       "      <td>0.723092</td>\n",
       "      <td>0.721667</td>\n",
       "      <td>8176</td>\n",
       "      <td>0.965720</td>\n",
       "      <td>-0.242628</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>104</th>\n",
       "      <td>openml__covertype__7593</td>\n",
       "      <td>NN</td>\n",
       "      <td>0.827011</td>\n",
       "      <td>0.802147</td>\n",
       "      <td>464810</td>\n",
       "      <td>0.972823</td>\n",
       "      <td>-0.145812</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>307</th>\n",
       "      <td>openml__walking-activity__9945</td>\n",
       "      <td>NN</td>\n",
       "      <td>0.560905</td>\n",
       "      <td>0.550102</td>\n",
       "      <td>119466</td>\n",
       "      <td>0.679971</td>\n",
       "      <td>-0.119065</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>openml__GesturePhaseSegmentationProcessed__14969</td>\n",
       "      <td>NN</td>\n",
       "      <td>0.618034</td>\n",
       "      <td>0.598264</td>\n",
       "      <td>7899</td>\n",
       "      <td>0.720365</td>\n",
       "      <td>-0.102330</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>184</th>\n",
       "      <td>openml__madelon__9976</td>\n",
       "      <td>NN</td>\n",
       "      <td>0.811538</td>\n",
       "      <td>0.811538</td>\n",
       "      <td>2080</td>\n",
       "      <td>0.903846</td>\n",
       "      <td>-0.092308</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>122</th>\n",
       "      <td>openml__eye_movements__3897</td>\n",
       "      <td>NN</td>\n",
       "      <td>0.695334</td>\n",
       "      <td>0.695029</td>\n",
       "      <td>8750</td>\n",
       "      <td>0.780622</td>\n",
       "      <td>-0.085288</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>116</th>\n",
       "      <td>openml__electricity__219</td>\n",
       "      <td>NN</td>\n",
       "      <td>0.867579</td>\n",
       "      <td>0.867579</td>\n",
       "      <td>36250</td>\n",
       "      <td>0.945487</td>\n",
       "      <td>-0.077908</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128</th>\n",
       "      <td>openml__first-order-theorem-proving__9985</td>\n",
       "      <td>NN</td>\n",
       "      <td>0.607843</td>\n",
       "      <td>0.600594</td>\n",
       "      <td>4896</td>\n",
       "      <td>0.669935</td>\n",
       "      <td>-0.062092</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140</th>\n",
       "      <td>openml__guillermo__168337</td>\n",
       "      <td>NN</td>\n",
       "      <td>0.797500</td>\n",
       "      <td>0.797500</td>\n",
       "      <td>16000</td>\n",
       "      <td>0.855000</td>\n",
       "      <td>-0.057500</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         dataset_name Group  Accuracy__test  \\\n",
       "178                                openml__ldpa__9974    NN        0.622104   \n",
       "65               openml__artificial-characters__14964    NN        0.723092   \n",
       "104                           openml__covertype__7593    NN        0.827011   \n",
       "307                    openml__walking-activity__9945    NN        0.560905   \n",
       "26   openml__GesturePhaseSegmentationProcessed__14969    NN        0.618034   \n",
       "184                             openml__madelon__9976    NN        0.811538   \n",
       "122                       openml__eye_movements__3897    NN        0.695334   \n",
       "116                          openml__electricity__219    NN        0.867579   \n",
       "128         openml__first-order-theorem-proving__9985    NN        0.607843   \n",
       "140                         openml__guillermo__168337    NN        0.797500   \n",
       "\n",
       "     F1__test  f__pymfe.general.nr_inst  Accuracy_max       gap  \n",
       "178  0.571127                    131888      0.982713 -0.360609  \n",
       "65   0.721667                      8176      0.965720 -0.242628  \n",
       "104  0.802147                    464810      0.972823 -0.145812  \n",
       "307  0.550102                    119466      0.679971 -0.119065  \n",
       "26   0.598264                      7899      0.720365 -0.102330  \n",
       "184  0.811538                      2080      0.903846 -0.092308  \n",
       "122  0.695029                      8750      0.780622 -0.085288  \n",
       "116  0.867579                     36250      0.945487 -0.077908  \n",
       "128  0.600594                      4896      0.669935 -0.062092  \n",
       "140  0.797500                     16000      0.855000 -0.057500  "
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "group_performance.query(\" Group == 'NN' \").sort_values('gap').head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "bda8487a",
   "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>dataset_name</th>\n",
       "      <th>Group</th>\n",
       "      <th>Accuracy__test</th>\n",
       "      <th>F1__test</th>\n",
       "      <th>f__pymfe.general.nr_inst</th>\n",
       "      <th>Accuracy_max</th>\n",
       "      <th>gap</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>204</th>\n",
       "      <td>openml__mfeat-zernike__22</td>\n",
       "      <td>EnsembleTree</td>\n",
       "      <td>0.850000</td>\n",
       "      <td>0.845493</td>\n",
       "      <td>1600</td>\n",
       "      <td>0.945000</td>\n",
       "      <td>-0.095000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>165</th>\n",
       "      <td>openml__jungle_chess_2pcs_raw_endgame_complete...</td>\n",
       "      <td>EnsembleTree</td>\n",
       "      <td>0.881080</td>\n",
       "      <td>0.880875</td>\n",
       "      <td>35856</td>\n",
       "      <td>0.961847</td>\n",
       "      <td>-0.080768</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>openml__Devnagari-Script__167121</td>\n",
       "      <td>EnsembleTree</td>\n",
       "      <td>0.910217</td>\n",
       "      <td>0.910079</td>\n",
       "      <td>73600</td>\n",
       "      <td>0.945000</td>\n",
       "      <td>-0.034783</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>openml__cmc__23</td>\n",
       "      <td>EnsembleTree</td>\n",
       "      <td>0.646259</td>\n",
       "      <td>0.645347</td>\n",
       "      <td>1179</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>-0.020408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>192</th>\n",
       "      <td>openml__mfeat-fourier__14</td>\n",
       "      <td>EnsembleTree</td>\n",
       "      <td>0.895000</td>\n",
       "      <td>0.894224</td>\n",
       "      <td>1600</td>\n",
       "      <td>0.915000</td>\n",
       "      <td>-0.020000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195</th>\n",
       "      <td>openml__mfeat-karhunen__16</td>\n",
       "      <td>EnsembleTree</td>\n",
       "      <td>0.980000</td>\n",
       "      <td>0.980110</td>\n",
       "      <td>1600</td>\n",
       "      <td>0.995000</td>\n",
       "      <td>-0.015000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>openml__bank-marketing__9899</td>\n",
       "      <td>EnsembleTree</td>\n",
       "      <td>0.922566</td>\n",
       "      <td>0.922566</td>\n",
       "      <td>3617</td>\n",
       "      <td>0.935841</td>\n",
       "      <td>-0.013274</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153</th>\n",
       "      <td>openml__isolet__3481</td>\n",
       "      <td>EnsembleTree</td>\n",
       "      <td>0.976923</td>\n",
       "      <td>0.976717</td>\n",
       "      <td>6239</td>\n",
       "      <td>0.987163</td>\n",
       "      <td>-0.010240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>189</th>\n",
       "      <td>openml__mfeat-factors__12</td>\n",
       "      <td>EnsembleTree</td>\n",
       "      <td>0.990000</td>\n",
       "      <td>0.989994</td>\n",
       "      <td>1600</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>-0.010000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>openml__CIFAR_10__167124</td>\n",
       "      <td>EnsembleTree</td>\n",
       "      <td>0.550833</td>\n",
       "      <td>0.549081</td>\n",
       "      <td>48000</td>\n",
       "      <td>0.559833</td>\n",
       "      <td>-0.009000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          dataset_name         Group  \\\n",
       "204                          openml__mfeat-zernike__22  EnsembleTree   \n",
       "165  openml__jungle_chess_2pcs_raw_endgame_complete...  EnsembleTree   \n",
       "19                    openml__Devnagari-Script__167121  EnsembleTree   \n",
       "97                                     openml__cmc__23  EnsembleTree   \n",
       "192                          openml__mfeat-fourier__14  EnsembleTree   \n",
       "195                         openml__mfeat-karhunen__16  EnsembleTree   \n",
       "70                        openml__bank-marketing__9899  EnsembleTree   \n",
       "153                               openml__isolet__3481  EnsembleTree   \n",
       "189                          openml__mfeat-factors__12  EnsembleTree   \n",
       "10                            openml__CIFAR_10__167124  EnsembleTree   \n",
       "\n",
       "     Accuracy__test  F1__test  f__pymfe.general.nr_inst  Accuracy_max  \\\n",
       "204        0.850000  0.845493                      1600      0.945000   \n",
       "165        0.881080  0.880875                     35856      0.961847   \n",
       "19         0.910217  0.910079                     73600      0.945000   \n",
       "97         0.646259  0.645347                      1179      0.666667   \n",
       "192        0.895000  0.894224                      1600      0.915000   \n",
       "195        0.980000  0.980110                      1600      0.995000   \n",
       "70         0.922566  0.922566                      3617      0.935841   \n",
       "153        0.976923  0.976717                      6239      0.987163   \n",
       "189        0.990000  0.989994                      1600      1.000000   \n",
       "10         0.550833  0.549081                     48000      0.559833   \n",
       "\n",
       "          gap  \n",
       "204 -0.095000  \n",
       "165 -0.080768  \n",
       "19  -0.034783  \n",
       "97  -0.020408  \n",
       "192 -0.020000  \n",
       "195 -0.015000  \n",
       "70  -0.013274  \n",
       "153 -0.010240  \n",
       "189 -0.010000  \n",
       "10  -0.009000  "
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "group_performance.query(\" Group == 'EnsembleTree' \").sort_values('gap').head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "c03b7f26-d71a-4ade-a201-54839c5a0e3c",
   "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>Accuracy__test</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>alg_name</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>XGBoost</th>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LightGBM</th>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CatBoost</th>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rtdl_ResNet</th>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SAINT</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rtdl_FTTransformer</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DANet</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SVM</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TabNet</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TabTransformer</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rtdl_MLP</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DecisionTree</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KNN</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NODE</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>STG</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    Accuracy__test\n",
       "alg_name                          \n",
       "XGBoost                         21\n",
       "LightGBM                        17\n",
       "CatBoost                        12\n",
       "rtdl_ResNet                      6\n",
       "SAINT                            4\n",
       "rtdl_FTTransformer               4\n",
       "DANet                            3\n",
       "SVM                              3\n",
       "TabNet                           3\n",
       "TabTransformer                   2\n",
       "rtdl_MLP                         2\n",
       "DecisionTree                     1\n",
       "KNN                              1\n",
       "NODE                             1\n",
       "STG                              1"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_algs.groupby(\"alg_name\").agg({\n",
    "    \"Accuracy__test\": \"count\"\n",
    "}).sort_values(\"Accuracy__test\", ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "3ac3fd08",
   "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>Accuracy__test</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>alg_name</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>CatBoost</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TabNet</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>XGBoost</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rtdl_FTTransformer</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rtdl_ResNet</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DeepFM</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>LightGBM</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>RandomForest</th>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DANet</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DecisionTree</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KNN</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SAINT</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TabTransformer</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rtdl_MLP</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    Accuracy__test\n",
       "alg_name                          \n",
       "CatBoost                         3\n",
       "TabNet                           3\n",
       "XGBoost                          3\n",
       "rtdl_FTTransformer               3\n",
       "rtdl_ResNet                      3\n",
       "DeepFM                           2\n",
       "LightGBM                         2\n",
       "RandomForest                     2\n",
       "DANet                            1\n",
       "DecisionTree                     1\n",
       "KNN                              1\n",
       "SAINT                            1\n",
       "TabTransformer                   1\n",
       "rtdl_MLP                         1"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_algs_small.groupby(\"alg_name\").agg({\n",
    "    \"Accuracy__test\": \"count\"\n",
    "}).sort_values(\"Accuracy__test\", ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "7ea060cb",
   "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>dataset_name</th>\n",
       "      <th>alg_name</th>\n",
       "      <th>Accuracy__test</th>\n",
       "      <th>F1__test</th>\n",
       "      <th>f__pymfe.general.nr_inst</th>\n",
       "      <th>f__pymfe.general.nr_cat</th>\n",
       "      <th>f__pymfe.general.nr_num</th>\n",
       "      <th>f__pymfe.general.nr_class</th>\n",
       "      <th>f__pymfe.general.nr_attr</th>\n",
       "      <th>Group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1243</th>\n",
       "      <td>openml__ldpa__9974</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.982713</td>\n",
       "      <td>0.982614</td>\n",
       "      <td>131888</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>11</td>\n",
       "      <td>7</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>397</th>\n",
       "      <td>openml__ldpa__9974</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.870860</td>\n",
       "      <td>0.865336</td>\n",
       "      <td>131888</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>11</td>\n",
       "      <td>7</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>openml__ldpa__9974</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.775264</td>\n",
       "      <td>0.750566</td>\n",
       "      <td>131888</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>11</td>\n",
       "      <td>7</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>702</th>\n",
       "      <td>openml__ldpa__9974</td>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.724979</td>\n",
       "      <td>0.683700</td>\n",
       "      <td>131888</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>11</td>\n",
       "      <td>7</td>\n",
       "      <td>Baselines</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>302</th>\n",
       "      <td>openml__ldpa__9974</td>\n",
       "      <td>KNN</td>\n",
       "      <td>0.712908</td>\n",
       "      <td>0.697544</td>\n",
       "      <td>131888</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>11</td>\n",
       "      <td>7</td>\n",
       "      <td>Baselines</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>155</th>\n",
       "      <td>openml__ldpa__9974</td>\n",
       "      <td>DecisionTree</td>\n",
       "      <td>0.668264</td>\n",
       "      <td>0.639008</td>\n",
       "      <td>131888</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>11</td>\n",
       "      <td>7</td>\n",
       "      <td>Baselines</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1455</th>\n",
       "      <td>openml__ldpa__9974</td>\n",
       "      <td>rtdl_ResNet</td>\n",
       "      <td>0.622104</td>\n",
       "      <td>0.571127</td>\n",
       "      <td>131888</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>11</td>\n",
       "      <td>7</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1057</th>\n",
       "      <td>openml__ldpa__9974</td>\n",
       "      <td>TabTransformer</td>\n",
       "      <td>0.518804</td>\n",
       "      <td>0.431799</td>\n",
       "      <td>131888</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>11</td>\n",
       "      <td>7</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1800</th>\n",
       "      <td>openml__ldpa__9974</td>\n",
       "      <td>rtdl_FTTransformer</td>\n",
       "      <td>0.483986</td>\n",
       "      <td>0.419209</td>\n",
       "      <td>131888</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>11</td>\n",
       "      <td>7</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>598</th>\n",
       "      <td>openml__ldpa__9974</td>\n",
       "      <td>MLP</td>\n",
       "      <td>0.471673</td>\n",
       "      <td>0.377708</td>\n",
       "      <td>131888</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>11</td>\n",
       "      <td>7</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1349</th>\n",
       "      <td>openml__ldpa__9974</td>\n",
       "      <td>rtdl_MLP</td>\n",
       "      <td>0.410894</td>\n",
       "      <td>0.277612</td>\n",
       "      <td>131888</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>11</td>\n",
       "      <td>7</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>495</th>\n",
       "      <td>openml__ldpa__9974</td>\n",
       "      <td>LinearModel</td>\n",
       "      <td>0.396215</td>\n",
       "      <td>0.326269</td>\n",
       "      <td>131888</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>11</td>\n",
       "      <td>7</td>\n",
       "      <td>Baselines</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1145</th>\n",
       "      <td>openml__ldpa__9974</td>\n",
       "      <td>VIME</td>\n",
       "      <td>0.330462</td>\n",
       "      <td>0.164161</td>\n",
       "      <td>131888</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>11</td>\n",
       "      <td>7</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            dataset_name            alg_name  Accuracy__test  F1__test  \\\n",
       "1243  openml__ldpa__9974             XGBoost        0.982713  0.982614   \n",
       "397   openml__ldpa__9974            LightGBM        0.870860  0.865336   \n",
       "53    openml__ldpa__9974            CatBoost        0.775264  0.750566   \n",
       "702   openml__ldpa__9974        RandomForest        0.724979  0.683700   \n",
       "302   openml__ldpa__9974                 KNN        0.712908  0.697544   \n",
       "155   openml__ldpa__9974        DecisionTree        0.668264  0.639008   \n",
       "1455  openml__ldpa__9974         rtdl_ResNet        0.622104  0.571127   \n",
       "1057  openml__ldpa__9974      TabTransformer        0.518804  0.431799   \n",
       "1800  openml__ldpa__9974  rtdl_FTTransformer        0.483986  0.419209   \n",
       "598   openml__ldpa__9974                 MLP        0.471673  0.377708   \n",
       "1349  openml__ldpa__9974            rtdl_MLP        0.410894  0.277612   \n",
       "495   openml__ldpa__9974         LinearModel        0.396215  0.326269   \n",
       "1145  openml__ldpa__9974                VIME        0.330462  0.164161   \n",
       "\n",
       "      f__pymfe.general.nr_inst  f__pymfe.general.nr_cat  \\\n",
       "1243                    131888                        2   \n",
       "397                     131888                        2   \n",
       "53                      131888                        2   \n",
       "702                     131888                        2   \n",
       "302                     131888                        2   \n",
       "155                     131888                        2   \n",
       "1455                    131888                        2   \n",
       "1057                    131888                        2   \n",
       "1800                    131888                        2   \n",
       "598                     131888                        2   \n",
       "1349                    131888                        2   \n",
       "495                     131888                        2   \n",
       "1145                    131888                        2   \n",
       "\n",
       "      f__pymfe.general.nr_num  f__pymfe.general.nr_class  \\\n",
       "1243                        5                         11   \n",
       "397                         5                         11   \n",
       "53                          5                         11   \n",
       "702                         5                         11   \n",
       "302                         5                         11   \n",
       "155                         5                         11   \n",
       "1455                        5                         11   \n",
       "1057                        5                         11   \n",
       "1800                        5                         11   \n",
       "598                         5                         11   \n",
       "1349                        5                         11   \n",
       "495                         5                         11   \n",
       "1145                        5                         11   \n",
       "\n",
       "      f__pymfe.general.nr_attr         Group  \n",
       "1243                         7  EnsembleTree  \n",
       "397                          7  EnsembleTree  \n",
       "53                           7  EnsembleTree  \n",
       "702                          7     Baselines  \n",
       "302                          7     Baselines  \n",
       "155                          7     Baselines  \n",
       "1455                         7            NN  \n",
       "1057                         7            NN  \n",
       "1800                         7            NN  \n",
       "598                          7            NN  \n",
       "1349                         7            NN  \n",
       "495                          7     Baselines  \n",
       "1145                         7            NN  "
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "aggregated_dataset.query(\"dataset_name == 'openml__ldpa__9974'\").sort_values(\"Accuracy__test\", ascending = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "9d876bfb-90ac-49b9-9a1d-4f7a2096d630",
   "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>dataset_name</th>\n",
       "      <th>alg_name</th>\n",
       "      <th>Accuracy__test</th>\n",
       "      <th>F1__test</th>\n",
       "      <th>f__pymfe.general.nr_inst</th>\n",
       "      <th>f__pymfe.general.nr_cat</th>\n",
       "      <th>f__pymfe.general.nr_num</th>\n",
       "      <th>f__pymfe.general.nr_class</th>\n",
       "      <th>f__pymfe.general.nr_attr</th>\n",
       "      <th>Group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>openml__dna__167140</td>\n",
       "      <td>STG</td>\n",
       "      <td>0.990596</td>\n",
       "      <td>0.990616</td>\n",
       "      <td>2550</td>\n",
       "      <td>180</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>180</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>openml__churn__167141</td>\n",
       "      <td>TabNet</td>\n",
       "      <td>0.974000</td>\n",
       "      <td>0.974000</td>\n",
       "      <td>4000</td>\n",
       "      <td>4</td>\n",
       "      <td>16</td>\n",
       "      <td>2</td>\n",
       "      <td>20</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>openml__elevators__3711</td>\n",
       "      <td>TabNet</td>\n",
       "      <td>0.915663</td>\n",
       "      <td>0.915663</td>\n",
       "      <td>13280</td>\n",
       "      <td>0</td>\n",
       "      <td>18</td>\n",
       "      <td>2</td>\n",
       "      <td>18</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>openml__sylvine__168912</td>\n",
       "      <td>TabNet</td>\n",
       "      <td>0.966797</td>\n",
       "      <td>0.966797</td>\n",
       "      <td>4100</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "      <td>2</td>\n",
       "      <td>20</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>openml__CIFAR_10__167124</td>\n",
       "      <td>TabTransformer</td>\n",
       "      <td>0.559833</td>\n",
       "      <td>0.559060</td>\n",
       "      <td>48000</td>\n",
       "      <td>0</td>\n",
       "      <td>3072</td>\n",
       "      <td>10</td>\n",
       "      <td>3072</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>openml__dilbert__168909</td>\n",
       "      <td>TabTransformer</td>\n",
       "      <td>0.995000</td>\n",
       "      <td>0.995000</td>\n",
       "      <td>8000</td>\n",
       "      <td>0</td>\n",
       "      <td>2000</td>\n",
       "      <td>5</td>\n",
       "      <td>2000</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>openml__numerai28.6__167120</td>\n",
       "      <td>rtdl_MLP</td>\n",
       "      <td>0.533430</td>\n",
       "      <td>0.533430</td>\n",
       "      <td>77056</td>\n",
       "      <td>0</td>\n",
       "      <td>21</td>\n",
       "      <td>2</td>\n",
       "      <td>21</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>openml__scene__3485</td>\n",
       "      <td>rtdl_MLP</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1927</td>\n",
       "      <td>5</td>\n",
       "      <td>294</td>\n",
       "      <td>2</td>\n",
       "      <td>299</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>openml__Devnagari-Script__167121</td>\n",
       "      <td>rtdl_ResNet</td>\n",
       "      <td>0.945000</td>\n",
       "      <td>0.944899</td>\n",
       "      <td>73600</td>\n",
       "      <td>0</td>\n",
       "      <td>1024</td>\n",
       "      <td>46</td>\n",
       "      <td>1024</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>openml__cmc__23</td>\n",
       "      <td>rtdl_ResNet</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.665618</td>\n",
       "      <td>1179</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>9</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>openml__higgs__146606</td>\n",
       "      <td>rtdl_ResNet</td>\n",
       "      <td>0.740031</td>\n",
       "      <td>0.740031</td>\n",
       "      <td>78440</td>\n",
       "      <td>0</td>\n",
       "      <td>28</td>\n",
       "      <td>2</td>\n",
       "      <td>28</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>openml__mfeat-factors__12</td>\n",
       "      <td>rtdl_ResNet</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1600</td>\n",
       "      <td>0</td>\n",
       "      <td>216</td>\n",
       "      <td>10</td>\n",
       "      <td>216</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>openml__mnist_784__3573</td>\n",
       "      <td>rtdl_ResNet</td>\n",
       "      <td>0.984000</td>\n",
       "      <td>0.983997</td>\n",
       "      <td>56000</td>\n",
       "      <td>0</td>\n",
       "      <td>784</td>\n",
       "      <td>10</td>\n",
       "      <td>784</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>openml__pollen__3735</td>\n",
       "      <td>rtdl_ResNet</td>\n",
       "      <td>0.578125</td>\n",
       "      <td>0.578125</td>\n",
       "      <td>3080</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>openml__gina_agnostic__3891</td>\n",
       "      <td>DANet</td>\n",
       "      <td>0.994236</td>\n",
       "      <td>0.994236</td>\n",
       "      <td>2776</td>\n",
       "      <td>0</td>\n",
       "      <td>970</td>\n",
       "      <td>2</td>\n",
       "      <td>970</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>openml__har__14970</td>\n",
       "      <td>DANet</td>\n",
       "      <td>0.998058</td>\n",
       "      <td>0.998058</td>\n",
       "      <td>8240</td>\n",
       "      <td>0</td>\n",
       "      <td>561</td>\n",
       "      <td>6</td>\n",
       "      <td>561</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>openml__scene__3485</td>\n",
       "      <td>DANet</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1927</td>\n",
       "      <td>5</td>\n",
       "      <td>294</td>\n",
       "      <td>2</td>\n",
       "      <td>299</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>openml__scene__3485</td>\n",
       "      <td>NODE</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1927</td>\n",
       "      <td>5</td>\n",
       "      <td>294</td>\n",
       "      <td>2</td>\n",
       "      <td>299</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>openml__helena__168329</td>\n",
       "      <td>SAINT</td>\n",
       "      <td>0.378988</td>\n",
       "      <td>0.354696</td>\n",
       "      <td>52158</td>\n",
       "      <td>0</td>\n",
       "      <td>27</td>\n",
       "      <td>100</td>\n",
       "      <td>27</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>openml__jungle_chess_2pcs_raw_endgame_complete...</td>\n",
       "      <td>SAINT</td>\n",
       "      <td>0.961847</td>\n",
       "      <td>0.961740</td>\n",
       "      <td>35856</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>openml__spambase__43</td>\n",
       "      <td>SAINT</td>\n",
       "      <td>0.980435</td>\n",
       "      <td>0.980435</td>\n",
       "      <td>3681</td>\n",
       "      <td>0</td>\n",
       "      <td>57</td>\n",
       "      <td>2</td>\n",
       "      <td>57</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>openml__splice__45</td>\n",
       "      <td>SAINT</td>\n",
       "      <td>0.987461</td>\n",
       "      <td>0.987517</td>\n",
       "      <td>2552</td>\n",
       "      <td>60</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>60</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>openml__PhishingWebsites__14952</td>\n",
       "      <td>rtdl_FTTransformer</td>\n",
       "      <td>0.983710</td>\n",
       "      <td>0.983710</td>\n",
       "      <td>8845</td>\n",
       "      <td>30</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>30</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>openml__bank-marketing__14965</td>\n",
       "      <td>rtdl_FTTransformer</td>\n",
       "      <td>0.920150</td>\n",
       "      <td>0.920150</td>\n",
       "      <td>36169</td>\n",
       "      <td>9</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>16</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>openml__bank-marketing__9899</td>\n",
       "      <td>rtdl_FTTransformer</td>\n",
       "      <td>0.935841</td>\n",
       "      <td>0.935841</td>\n",
       "      <td>3617</td>\n",
       "      <td>9</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>16</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>openml__house_16H__3686</td>\n",
       "      <td>rtdl_FTTransformer</td>\n",
       "      <td>0.906936</td>\n",
       "      <td>0.906936</td>\n",
       "      <td>18228</td>\n",
       "      <td>0</td>\n",
       "      <td>16</td>\n",
       "      <td>2</td>\n",
       "      <td>16</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         dataset_name            alg_name  \\\n",
       "0                                 openml__dna__167140                 STG   \n",
       "1                               openml__churn__167141              TabNet   \n",
       "2                             openml__elevators__3711              TabNet   \n",
       "3                             openml__sylvine__168912              TabNet   \n",
       "4                            openml__CIFAR_10__167124      TabTransformer   \n",
       "5                             openml__dilbert__168909      TabTransformer   \n",
       "6                         openml__numerai28.6__167120            rtdl_MLP   \n",
       "7                                 openml__scene__3485            rtdl_MLP   \n",
       "8                    openml__Devnagari-Script__167121         rtdl_ResNet   \n",
       "9                                     openml__cmc__23         rtdl_ResNet   \n",
       "10                              openml__higgs__146606         rtdl_ResNet   \n",
       "11                          openml__mfeat-factors__12         rtdl_ResNet   \n",
       "12                            openml__mnist_784__3573         rtdl_ResNet   \n",
       "13                               openml__pollen__3735         rtdl_ResNet   \n",
       "14                        openml__gina_agnostic__3891               DANet   \n",
       "15                                 openml__har__14970               DANet   \n",
       "16                                openml__scene__3485               DANet   \n",
       "17                                openml__scene__3485                NODE   \n",
       "18                             openml__helena__168329               SAINT   \n",
       "19  openml__jungle_chess_2pcs_raw_endgame_complete...               SAINT   \n",
       "20                               openml__spambase__43               SAINT   \n",
       "21                                 openml__splice__45               SAINT   \n",
       "22                    openml__PhishingWebsites__14952  rtdl_FTTransformer   \n",
       "23                      openml__bank-marketing__14965  rtdl_FTTransformer   \n",
       "24                       openml__bank-marketing__9899  rtdl_FTTransformer   \n",
       "25                            openml__house_16H__3686  rtdl_FTTransformer   \n",
       "\n",
       "    Accuracy__test  F1__test  f__pymfe.general.nr_inst  \\\n",
       "0         0.990596  0.990616                      2550   \n",
       "1         0.974000  0.974000                      4000   \n",
       "2         0.915663  0.915663                     13280   \n",
       "3         0.966797  0.966797                      4100   \n",
       "4         0.559833  0.559060                     48000   \n",
       "5         0.995000  0.995000                      8000   \n",
       "6         0.533430  0.533430                     77056   \n",
       "7         1.000000  1.000000                      1927   \n",
       "8         0.945000  0.944899                     73600   \n",
       "9         0.666667  0.665618                      1179   \n",
       "10        0.740031  0.740031                     78440   \n",
       "11        1.000000  1.000000                      1600   \n",
       "12        0.984000  0.983997                     56000   \n",
       "13        0.578125  0.578125                      3080   \n",
       "14        0.994236  0.994236                      2776   \n",
       "15        0.998058  0.998058                      8240   \n",
       "16        1.000000  1.000000                      1927   \n",
       "17        1.000000  1.000000                      1927   \n",
       "18        0.378988  0.354696                     52158   \n",
       "19        0.961847  0.961740                     35856   \n",
       "20        0.980435  0.980435                      3681   \n",
       "21        0.987461  0.987517                      2552   \n",
       "22        0.983710  0.983710                      8845   \n",
       "23        0.920150  0.920150                     36169   \n",
       "24        0.935841  0.935841                      3617   \n",
       "25        0.906936  0.906936                     18228   \n",
       "\n",
       "    f__pymfe.general.nr_cat  f__pymfe.general.nr_num  \\\n",
       "0                       180                        0   \n",
       "1                         4                       16   \n",
       "2                         0                       18   \n",
       "3                         0                       20   \n",
       "4                         0                     3072   \n",
       "5                         0                     2000   \n",
       "6                         0                       21   \n",
       "7                         5                      294   \n",
       "8                         0                     1024   \n",
       "9                         7                        2   \n",
       "10                        0                       28   \n",
       "11                        0                      216   \n",
       "12                        0                      784   \n",
       "13                        0                        5   \n",
       "14                        0                      970   \n",
       "15                        0                      561   \n",
       "16                        5                      294   \n",
       "17                        5                      294   \n",
       "18                        0                       27   \n",
       "19                        0                        6   \n",
       "20                        0                       57   \n",
       "21                       60                        0   \n",
       "22                       30                        0   \n",
       "23                        9                        7   \n",
       "24                        9                        7   \n",
       "25                        0                       16   \n",
       "\n",
       "    f__pymfe.general.nr_class  f__pymfe.general.nr_attr Group  \n",
       "0                           3                       180    NN  \n",
       "1                           2                        20    NN  \n",
       "2                           2                        18    NN  \n",
       "3                           2                        20    NN  \n",
       "4                          10                      3072    NN  \n",
       "5                           5                      2000    NN  \n",
       "6                           2                        21    NN  \n",
       "7                           2                       299    NN  \n",
       "8                          46                      1024    NN  \n",
       "9                           3                         9    NN  \n",
       "10                          2                        28    NN  \n",
       "11                         10                       216    NN  \n",
       "12                         10                       784    NN  \n",
       "13                          2                         5    NN  \n",
       "14                          2                       970    NN  \n",
       "15                          6                       561    NN  \n",
       "16                          2                       299    NN  \n",
       "17                          2                       299    NN  \n",
       "18                        100                        27    NN  \n",
       "19                          3                         6    NN  \n",
       "20                          2                        57    NN  \n",
       "21                          3                        60    NN  \n",
       "22                          2                        30    NN  \n",
       "23                          2                        16    NN  \n",
       "24                          2                        16    NN  \n",
       "25                          2                        16    NN  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_algs.query(\" Group == 'NN' \").reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "9f9bfd03-7bd2-44bc-ada7-a25d08048d8d",
   "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>dataset_name</th>\n",
       "      <th>alg_name</th>\n",
       "      <th>Accuracy__test</th>\n",
       "      <th>F1__test</th>\n",
       "      <th>f__pymfe.general.nr_inst</th>\n",
       "      <th>f__pymfe.general.nr_cat</th>\n",
       "      <th>f__pymfe.general.nr_num</th>\n",
       "      <th>f__pymfe.general.nr_class</th>\n",
       "      <th>f__pymfe.general.nr_attr</th>\n",
       "      <th>Group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>openml__Census-Income__168340</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.959837</td>\n",
       "      <td>0.959837</td>\n",
       "      <td>239429</td>\n",
       "      <td>28</td>\n",
       "      <td>13</td>\n",
       "      <td>2</td>\n",
       "      <td>41</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>openml__MagicTelescope__3954</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.901682</td>\n",
       "      <td>0.901682</td>\n",
       "      <td>15216</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>openml__albert__189356</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.711551</td>\n",
       "      <td>0.711551</td>\n",
       "      <td>340192</td>\n",
       "      <td>52</td>\n",
       "      <td>26</td>\n",
       "      <td>2</td>\n",
       "      <td>78</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>openml__jasmine__168911</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.859532</td>\n",
       "      <td>0.859532</td>\n",
       "      <td>2388</td>\n",
       "      <td>136</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>144</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>openml__madelon__9976</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.903846</td>\n",
       "      <td>0.903846</td>\n",
       "      <td>2080</td>\n",
       "      <td>0</td>\n",
       "      <td>500</td>\n",
       "      <td>2</td>\n",
       "      <td>500</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>openml__magic__146206</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.901682</td>\n",
       "      <td>0.901682</td>\n",
       "      <td>15216</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>openml__mfeat-morphological__18</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.805000</td>\n",
       "      <td>0.804774</td>\n",
       "      <td>1600</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>openml__one-hundred-plants-texture__9956</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.893082</td>\n",
       "      <td>0.884486</td>\n",
       "      <td>1280</td>\n",
       "      <td>0</td>\n",
       "      <td>64</td>\n",
       "      <td>100</td>\n",
       "      <td>64</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>openml__page-blocks__30</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.987203</td>\n",
       "      <td>0.987188</td>\n",
       "      <td>4379</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>10</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>openml__pc4__3902</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.958904</td>\n",
       "      <td>0.958904</td>\n",
       "      <td>1168</td>\n",
       "      <td>0</td>\n",
       "      <td>37</td>\n",
       "      <td>2</td>\n",
       "      <td>37</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>openml__steel-plates-fault__146817</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.860825</td>\n",
       "      <td>0.855989</td>\n",
       "      <td>1553</td>\n",
       "      <td>0</td>\n",
       "      <td>27</td>\n",
       "      <td>7</td>\n",
       "      <td>27</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>openml__sylva_agnostic__3889</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.998611</td>\n",
       "      <td>0.998611</td>\n",
       "      <td>11517</td>\n",
       "      <td>0</td>\n",
       "      <td>216</td>\n",
       "      <td>2</td>\n",
       "      <td>216</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>openml__Amazon_employee_access__34539</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.957888</td>\n",
       "      <td>0.957888</td>\n",
       "      <td>26216</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>9</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>openml__Click_prediction_small__190408</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.844055</td>\n",
       "      <td>0.844055</td>\n",
       "      <td>31960</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>11</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>openml__GesturePhaseSegmentationProcessed__14969</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.720365</td>\n",
       "      <td>0.713288</td>\n",
       "      <td>7899</td>\n",
       "      <td>0</td>\n",
       "      <td>32</td>\n",
       "      <td>5</td>\n",
       "      <td>32</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>openml__MiniBooNE__168335</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.951638</td>\n",
       "      <td>0.951638</td>\n",
       "      <td>104052</td>\n",
       "      <td>0</td>\n",
       "      <td>50</td>\n",
       "      <td>2</td>\n",
       "      <td>50</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>openml__ada_agnostic__3896</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.890351</td>\n",
       "      <td>0.890351</td>\n",
       "      <td>3650</td>\n",
       "      <td>0</td>\n",
       "      <td>48</td>\n",
       "      <td>2</td>\n",
       "      <td>48</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>openml__connect-4__146195</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.872114</td>\n",
       "      <td>0.861120</td>\n",
       "      <td>54047</td>\n",
       "      <td>42</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>42</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>openml__covertype__7593</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.972823</td>\n",
       "      <td>0.972781</td>\n",
       "      <td>464810</td>\n",
       "      <td>44</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>54</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>openml__electricity__219</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.945487</td>\n",
       "      <td>0.945487</td>\n",
       "      <td>36250</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>openml__eye_movements__3897</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.780622</td>\n",
       "      <td>0.781400</td>\n",
       "      <td>8750</td>\n",
       "      <td>3</td>\n",
       "      <td>24</td>\n",
       "      <td>3</td>\n",
       "      <td>27</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>openml__first-order-theorem-proving__9985</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.669935</td>\n",
       "      <td>0.658304</td>\n",
       "      <td>4896</td>\n",
       "      <td>0</td>\n",
       "      <td>51</td>\n",
       "      <td>6</td>\n",
       "      <td>51</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>openml__jasmine__168911</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.859532</td>\n",
       "      <td>0.859532</td>\n",
       "      <td>2388</td>\n",
       "      <td>136</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>144</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>openml__nomao__9977</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.979692</td>\n",
       "      <td>0.979692</td>\n",
       "      <td>27573</td>\n",
       "      <td>29</td>\n",
       "      <td>89</td>\n",
       "      <td>2</td>\n",
       "      <td>118</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>openml__ozone-level-8hr__9978</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.972332</td>\n",
       "      <td>0.972332</td>\n",
       "      <td>2028</td>\n",
       "      <td>0</td>\n",
       "      <td>72</td>\n",
       "      <td>2</td>\n",
       "      <td>72</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>openml__philippine__190410</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.808219</td>\n",
       "      <td>0.808219</td>\n",
       "      <td>4666</td>\n",
       "      <td>0</td>\n",
       "      <td>308</td>\n",
       "      <td>2</td>\n",
       "      <td>308</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>openml__phoneme__9952</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.937037</td>\n",
       "      <td>0.937037</td>\n",
       "      <td>4324</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>openml__skin-segmentation__9965</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.999878</td>\n",
       "      <td>0.999878</td>\n",
       "      <td>196047</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>openml__volkert__168331</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.721660</td>\n",
       "      <td>0.715689</td>\n",
       "      <td>46648</td>\n",
       "      <td>0</td>\n",
       "      <td>180</td>\n",
       "      <td>10</td>\n",
       "      <td>180</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>openml__APSFailure__168868</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.996447</td>\n",
       "      <td>0.996447</td>\n",
       "      <td>60800</td>\n",
       "      <td>0</td>\n",
       "      <td>170</td>\n",
       "      <td>2</td>\n",
       "      <td>170</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>openml__Bioresponse__9910</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.848000</td>\n",
       "      <td>0.848000</td>\n",
       "      <td>3001</td>\n",
       "      <td>0</td>\n",
       "      <td>1776</td>\n",
       "      <td>2</td>\n",
       "      <td>1776</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>openml__Fashion-MNIST__146825</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.911000</td>\n",
       "      <td>0.910720</td>\n",
       "      <td>56000</td>\n",
       "      <td>0</td>\n",
       "      <td>784</td>\n",
       "      <td>10</td>\n",
       "      <td>784</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>openml__SpeedDating__146607</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.894988</td>\n",
       "      <td>0.894988</td>\n",
       "      <td>6704</td>\n",
       "      <td>61</td>\n",
       "      <td>59</td>\n",
       "      <td>2</td>\n",
       "      <td>120</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>openml__adult-census__3953</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.883292</td>\n",
       "      <td>0.883292</td>\n",
       "      <td>26049</td>\n",
       "      <td>8</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>14</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>openml__adult__7592</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.883907</td>\n",
       "      <td>0.883907</td>\n",
       "      <td>39074</td>\n",
       "      <td>8</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>14</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>openml__airlines__189354</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.675133</td>\n",
       "      <td>0.675133</td>\n",
       "      <td>431507</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>7</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>openml__artificial-characters__14964</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.965720</td>\n",
       "      <td>0.965664</td>\n",
       "      <td>8176</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>openml__chess__3952</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.935139</td>\n",
       "      <td>0.935230</td>\n",
       "      <td>22446</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>18</td>\n",
       "      <td>6</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>openml__christine__168908</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.773063</td>\n",
       "      <td>0.773063</td>\n",
       "      <td>4336</td>\n",
       "      <td>37</td>\n",
       "      <td>1599</td>\n",
       "      <td>2</td>\n",
       "      <td>1636</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>openml__fabert__168910</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.743932</td>\n",
       "      <td>0.739393</td>\n",
       "      <td>6591</td>\n",
       "      <td>0</td>\n",
       "      <td>800</td>\n",
       "      <td>7</td>\n",
       "      <td>800</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>openml__guillermo__168337</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.855000</td>\n",
       "      <td>0.855000</td>\n",
       "      <td>16000</td>\n",
       "      <td>0</td>\n",
       "      <td>4296</td>\n",
       "      <td>2</td>\n",
       "      <td>4296</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>openml__jannis__168330</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.730563</td>\n",
       "      <td>0.724170</td>\n",
       "      <td>66987</td>\n",
       "      <td>0</td>\n",
       "      <td>54</td>\n",
       "      <td>4</td>\n",
       "      <td>54</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>openml__jm1__3904</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.833792</td>\n",
       "      <td>0.833792</td>\n",
       "      <td>8709</td>\n",
       "      <td>0</td>\n",
       "      <td>21</td>\n",
       "      <td>2</td>\n",
       "      <td>21</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>openml__kc1__3917</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.900474</td>\n",
       "      <td>0.900474</td>\n",
       "      <td>1688</td>\n",
       "      <td>0</td>\n",
       "      <td>21</td>\n",
       "      <td>2</td>\n",
       "      <td>21</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>openml__kropt__2076</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.935139</td>\n",
       "      <td>0.935230</td>\n",
       "      <td>22446</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>18</td>\n",
       "      <td>6</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>openml__ldpa__9974</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.982713</td>\n",
       "      <td>0.982614</td>\n",
       "      <td>131888</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>11</td>\n",
       "      <td>7</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>openml__poker-hand__9890</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.946498</td>\n",
       "      <td>0.936650</td>\n",
       "      <td>820008</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "      <td>10</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>openml__riccardo__168338</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.999000</td>\n",
       "      <td>0.999000</td>\n",
       "      <td>16000</td>\n",
       "      <td>0</td>\n",
       "      <td>4296</td>\n",
       "      <td>2</td>\n",
       "      <td>4296</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>openml__robert__168332</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.516000</td>\n",
       "      <td>0.514839</td>\n",
       "      <td>8000</td>\n",
       "      <td>0</td>\n",
       "      <td>7200</td>\n",
       "      <td>10</td>\n",
       "      <td>7200</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>openml__walking-activity__9945</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.679971</td>\n",
       "      <td>0.680082</td>\n",
       "      <td>119466</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>22</td>\n",
       "      <td>4</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                        dataset_name  alg_name  \\\n",
       "0                      openml__Census-Income__168340  CatBoost   \n",
       "1                       openml__MagicTelescope__3954  CatBoost   \n",
       "2                             openml__albert__189356  CatBoost   \n",
       "3                            openml__jasmine__168911  CatBoost   \n",
       "4                              openml__madelon__9976  CatBoost   \n",
       "5                              openml__magic__146206  CatBoost   \n",
       "6                    openml__mfeat-morphological__18  CatBoost   \n",
       "7           openml__one-hundred-plants-texture__9956  CatBoost   \n",
       "8                            openml__page-blocks__30  CatBoost   \n",
       "9                                  openml__pc4__3902  CatBoost   \n",
       "10                openml__steel-plates-fault__146817  CatBoost   \n",
       "11                      openml__sylva_agnostic__3889  CatBoost   \n",
       "12             openml__Amazon_employee_access__34539  LightGBM   \n",
       "13            openml__Click_prediction_small__190408  LightGBM   \n",
       "14  openml__GesturePhaseSegmentationProcessed__14969  LightGBM   \n",
       "15                         openml__MiniBooNE__168335  LightGBM   \n",
       "16                        openml__ada_agnostic__3896  LightGBM   \n",
       "17                         openml__connect-4__146195  LightGBM   \n",
       "18                           openml__covertype__7593  LightGBM   \n",
       "19                          openml__electricity__219  LightGBM   \n",
       "20                       openml__eye_movements__3897  LightGBM   \n",
       "21         openml__first-order-theorem-proving__9985  LightGBM   \n",
       "22                           openml__jasmine__168911  LightGBM   \n",
       "23                               openml__nomao__9977  LightGBM   \n",
       "24                     openml__ozone-level-8hr__9978  LightGBM   \n",
       "25                        openml__philippine__190410  LightGBM   \n",
       "26                             openml__phoneme__9952  LightGBM   \n",
       "27                   openml__skin-segmentation__9965  LightGBM   \n",
       "28                           openml__volkert__168331  LightGBM   \n",
       "29                        openml__APSFailure__168868   XGBoost   \n",
       "30                         openml__Bioresponse__9910   XGBoost   \n",
       "31                     openml__Fashion-MNIST__146825   XGBoost   \n",
       "32                       openml__SpeedDating__146607   XGBoost   \n",
       "33                        openml__adult-census__3953   XGBoost   \n",
       "34                               openml__adult__7592   XGBoost   \n",
       "35                          openml__airlines__189354   XGBoost   \n",
       "36              openml__artificial-characters__14964   XGBoost   \n",
       "37                               openml__chess__3952   XGBoost   \n",
       "38                         openml__christine__168908   XGBoost   \n",
       "39                            openml__fabert__168910   XGBoost   \n",
       "40                         openml__guillermo__168337   XGBoost   \n",
       "41                            openml__jannis__168330   XGBoost   \n",
       "42                                 openml__jm1__3904   XGBoost   \n",
       "43                                 openml__kc1__3917   XGBoost   \n",
       "44                               openml__kropt__2076   XGBoost   \n",
       "45                                openml__ldpa__9974   XGBoost   \n",
       "46                          openml__poker-hand__9890   XGBoost   \n",
       "47                          openml__riccardo__168338   XGBoost   \n",
       "48                            openml__robert__168332   XGBoost   \n",
       "49                    openml__walking-activity__9945   XGBoost   \n",
       "\n",
       "    Accuracy__test  F1__test  f__pymfe.general.nr_inst  \\\n",
       "0         0.959837  0.959837                    239429   \n",
       "1         0.901682  0.901682                     15216   \n",
       "2         0.711551  0.711551                    340192   \n",
       "3         0.859532  0.859532                      2388   \n",
       "4         0.903846  0.903846                      2080   \n",
       "5         0.901682  0.901682                     15216   \n",
       "6         0.805000  0.804774                      1600   \n",
       "7         0.893082  0.884486                      1280   \n",
       "8         0.987203  0.987188                      4379   \n",
       "9         0.958904  0.958904                      1168   \n",
       "10        0.860825  0.855989                      1553   \n",
       "11        0.998611  0.998611                     11517   \n",
       "12        0.957888  0.957888                     26216   \n",
       "13        0.844055  0.844055                     31960   \n",
       "14        0.720365  0.713288                      7899   \n",
       "15        0.951638  0.951638                    104052   \n",
       "16        0.890351  0.890351                      3650   \n",
       "17        0.872114  0.861120                     54047   \n",
       "18        0.972823  0.972781                    464810   \n",
       "19        0.945487  0.945487                     36250   \n",
       "20        0.780622  0.781400                      8750   \n",
       "21        0.669935  0.658304                      4896   \n",
       "22        0.859532  0.859532                      2388   \n",
       "23        0.979692  0.979692                     27573   \n",
       "24        0.972332  0.972332                      2028   \n",
       "25        0.808219  0.808219                      4666   \n",
       "26        0.937037  0.937037                      4324   \n",
       "27        0.999878  0.999878                    196047   \n",
       "28        0.721660  0.715689                     46648   \n",
       "29        0.996447  0.996447                     60800   \n",
       "30        0.848000  0.848000                      3001   \n",
       "31        0.911000  0.910720                     56000   \n",
       "32        0.894988  0.894988                      6704   \n",
       "33        0.883292  0.883292                     26049   \n",
       "34        0.883907  0.883907                     39074   \n",
       "35        0.675133  0.675133                    431507   \n",
       "36        0.965720  0.965664                      8176   \n",
       "37        0.935139  0.935230                     22446   \n",
       "38        0.773063  0.773063                      4336   \n",
       "39        0.743932  0.739393                      6591   \n",
       "40        0.855000  0.855000                     16000   \n",
       "41        0.730563  0.724170                     66987   \n",
       "42        0.833792  0.833792                      8709   \n",
       "43        0.900474  0.900474                      1688   \n",
       "44        0.935139  0.935230                     22446   \n",
       "45        0.982713  0.982614                    131888   \n",
       "46        0.946498  0.936650                    820008   \n",
       "47        0.999000  0.999000                     16000   \n",
       "48        0.516000  0.514839                      8000   \n",
       "49        0.679971  0.680082                    119466   \n",
       "\n",
       "    f__pymfe.general.nr_cat  f__pymfe.general.nr_num  \\\n",
       "0                        28                       13   \n",
       "1                         0                       10   \n",
       "2                        52                       26   \n",
       "3                       136                        8   \n",
       "4                         0                      500   \n",
       "5                         0                       10   \n",
       "6                         0                        6   \n",
       "7                         0                       64   \n",
       "8                         0                       10   \n",
       "9                         0                       37   \n",
       "10                        0                       27   \n",
       "11                        0                      216   \n",
       "12                        9                        0   \n",
       "13                        6                        5   \n",
       "14                        0                       32   \n",
       "15                        0                       50   \n",
       "16                        0                       48   \n",
       "17                       42                        0   \n",
       "18                       44                       10   \n",
       "19                        1                        7   \n",
       "20                        3                       24   \n",
       "21                        0                       51   \n",
       "22                      136                        8   \n",
       "23                       29                       89   \n",
       "24                        0                       72   \n",
       "25                        0                      308   \n",
       "26                        0                        5   \n",
       "27                        0                        3   \n",
       "28                        0                      180   \n",
       "29                        0                      170   \n",
       "30                        0                     1776   \n",
       "31                        0                      784   \n",
       "32                       61                       59   \n",
       "33                        8                        6   \n",
       "34                        8                        6   \n",
       "35                        4                        3   \n",
       "36                        0                        7   \n",
       "37                        6                        0   \n",
       "38                       37                     1599   \n",
       "39                        0                      800   \n",
       "40                        0                     4296   \n",
       "41                        0                       54   \n",
       "42                        0                       21   \n",
       "43                        0                       21   \n",
       "44                        6                        0   \n",
       "45                        2                        5   \n",
       "46                        0                       10   \n",
       "47                        0                     4296   \n",
       "48                        0                     7200   \n",
       "49                        0                        4   \n",
       "\n",
       "    f__pymfe.general.nr_class  f__pymfe.general.nr_attr         Group  \n",
       "0                           2                        41  EnsembleTree  \n",
       "1                           2                        10  EnsembleTree  \n",
       "2                           2                        78  EnsembleTree  \n",
       "3                           2                       144  EnsembleTree  \n",
       "4                           2                       500  EnsembleTree  \n",
       "5                           2                        10  EnsembleTree  \n",
       "6                          10                         6  EnsembleTree  \n",
       "7                         100                        64  EnsembleTree  \n",
       "8                           5                        10  EnsembleTree  \n",
       "9                           2                        37  EnsembleTree  \n",
       "10                          7                        27  EnsembleTree  \n",
       "11                          2                       216  EnsembleTree  \n",
       "12                          2                         9  EnsembleTree  \n",
       "13                          2                        11  EnsembleTree  \n",
       "14                          5                        32  EnsembleTree  \n",
       "15                          2                        50  EnsembleTree  \n",
       "16                          2                        48  EnsembleTree  \n",
       "17                          3                        42  EnsembleTree  \n",
       "18                          7                        54  EnsembleTree  \n",
       "19                          2                         8  EnsembleTree  \n",
       "20                          3                        27  EnsembleTree  \n",
       "21                          6                        51  EnsembleTree  \n",
       "22                          2                       144  EnsembleTree  \n",
       "23                          2                       118  EnsembleTree  \n",
       "24                          2                        72  EnsembleTree  \n",
       "25                          2                       308  EnsembleTree  \n",
       "26                          2                         5  EnsembleTree  \n",
       "27                          2                         3  EnsembleTree  \n",
       "28                         10                       180  EnsembleTree  \n",
       "29                          2                       170  EnsembleTree  \n",
       "30                          2                      1776  EnsembleTree  \n",
       "31                         10                       784  EnsembleTree  \n",
       "32                          2                       120  EnsembleTree  \n",
       "33                          2                        14  EnsembleTree  \n",
       "34                          2                        14  EnsembleTree  \n",
       "35                          2                         7  EnsembleTree  \n",
       "36                         10                         7  EnsembleTree  \n",
       "37                         18                         6  EnsembleTree  \n",
       "38                          2                      1636  EnsembleTree  \n",
       "39                          7                       800  EnsembleTree  \n",
       "40                          2                      4296  EnsembleTree  \n",
       "41                          4                        54  EnsembleTree  \n",
       "42                          2                        21  EnsembleTree  \n",
       "43                          2                        21  EnsembleTree  \n",
       "44                         18                         6  EnsembleTree  \n",
       "45                         11                         7  EnsembleTree  \n",
       "46                         10                        10  EnsembleTree  \n",
       "47                          2                      4296  EnsembleTree  \n",
       "48                         10                      7200  EnsembleTree  \n",
       "49                         22                         4  EnsembleTree  "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_algs.query(\" Group == 'EnsembleTree' \").reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "f4d4200f-013c-455e-a935-06f5c541d6ef",
   "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>dataset_name</th>\n",
       "      <th>alg_name</th>\n",
       "      <th>Accuracy__test</th>\n",
       "      <th>F1__test</th>\n",
       "      <th>f__pymfe.general.nr_inst</th>\n",
       "      <th>f__pymfe.general.nr_cat</th>\n",
       "      <th>f__pymfe.general.nr_num</th>\n",
       "      <th>f__pymfe.general.nr_class</th>\n",
       "      <th>f__pymfe.general.nr_attr</th>\n",
       "      <th>Group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>openml__sick__3021</td>\n",
       "      <td>DecisionTree</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3018</td>\n",
       "      <td>22</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>28</td>\n",
       "      <td>Baselines</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>openml__eeg-eye-state__14951</td>\n",
       "      <td>KNN</td>\n",
       "      <td>0.970628</td>\n",
       "      <td>0.970628</td>\n",
       "      <td>11984</td>\n",
       "      <td>0</td>\n",
       "      <td>14</td>\n",
       "      <td>2</td>\n",
       "      <td>14</td>\n",
       "      <td>Baselines</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>openml__isolet__3481</td>\n",
       "      <td>SVM</td>\n",
       "      <td>0.987163</td>\n",
       "      <td>0.987089</td>\n",
       "      <td>6239</td>\n",
       "      <td>0</td>\n",
       "      <td>617</td>\n",
       "      <td>26</td>\n",
       "      <td>617</td>\n",
       "      <td>Baselines</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>openml__letter__6</td>\n",
       "      <td>SVM</td>\n",
       "      <td>0.979500</td>\n",
       "      <td>0.979543</td>\n",
       "      <td>16000</td>\n",
       "      <td>0</td>\n",
       "      <td>16</td>\n",
       "      <td>26</td>\n",
       "      <td>16</td>\n",
       "      <td>Baselines</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>openml__mfeat-zernike__22</td>\n",
       "      <td>SVM</td>\n",
       "      <td>0.945000</td>\n",
       "      <td>0.943850</td>\n",
       "      <td>1600</td>\n",
       "      <td>0</td>\n",
       "      <td>47</td>\n",
       "      <td>10</td>\n",
       "      <td>47</td>\n",
       "      <td>Baselines</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   dataset_name      alg_name  Accuracy__test  F1__test  \\\n",
       "0            openml__sick__3021  DecisionTree        1.000000  1.000000   \n",
       "1  openml__eeg-eye-state__14951           KNN        0.970628  0.970628   \n",
       "2          openml__isolet__3481           SVM        0.987163  0.987089   \n",
       "3             openml__letter__6           SVM        0.979500  0.979543   \n",
       "4     openml__mfeat-zernike__22           SVM        0.945000  0.943850   \n",
       "\n",
       "   f__pymfe.general.nr_inst  f__pymfe.general.nr_cat  f__pymfe.general.nr_num  \\\n",
       "0                      3018                       22                        6   \n",
       "1                     11984                        0                       14   \n",
       "2                      6239                        0                      617   \n",
       "3                     16000                        0                       16   \n",
       "4                      1600                        0                       47   \n",
       "\n",
       "   f__pymfe.general.nr_class  f__pymfe.general.nr_attr      Group  \n",
       "0                          2                        28  Baselines  \n",
       "1                          2                        14  Baselines  \n",
       "2                         26                       617  Baselines  \n",
       "3                         26                        16  Baselines  \n",
       "4                         10                        47  Baselines  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_algs.query(\" Group == 'Baselines' \").reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "c2a13362-d534-406d-b495-5b6d013c593d",
   "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>dataset_name</th>\n",
       "      <th>alg_name</th>\n",
       "      <th>Accuracy__test</th>\n",
       "      <th>F1__test</th>\n",
       "      <th>f__pymfe.general.nr_inst</th>\n",
       "      <th>f__pymfe.general.nr_cat</th>\n",
       "      <th>f__pymfe.general.nr_num</th>\n",
       "      <th>f__pymfe.general.nr_class</th>\n",
       "      <th>f__pymfe.general.nr_attr</th>\n",
       "      <th>Group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>849</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>SVM</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>Baselines</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1508</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>DANet</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1410</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>rtdl_ResNet</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1304</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>rtdl_MLP</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1200</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>XGBoost</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1025</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>TabTransformer</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>DecisionTree</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>Baselines</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>929</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>TabNet</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>658</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>RandomForest</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>Baselines</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>553</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>MLP</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>357</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>LightGBM</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>208</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>DeepFM</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1765</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>rtdl_FTTransformer</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>0.998039</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>759</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>STG</td>\n",
       "      <td>0.996078</td>\n",
       "      <td>0.996078</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1104</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>VIME</td>\n",
       "      <td>0.996078</td>\n",
       "      <td>0.996078</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>453</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>LinearModel</td>\n",
       "      <td>0.996078</td>\n",
       "      <td>0.996078</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>Baselines</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>260</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>KNN</td>\n",
       "      <td>0.996078</td>\n",
       "      <td>0.996078</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>Baselines</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1622</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>NODE</td>\n",
       "      <td>0.996078</td>\n",
       "      <td>0.996078</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1692</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>SAINT</td>\n",
       "      <td>0.992157</td>\n",
       "      <td>0.992157</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1581</th>\n",
       "      <td>openml__Satellite__167211</td>\n",
       "      <td>NAM</td>\n",
       "      <td>0.986275</td>\n",
       "      <td>0.986275</td>\n",
       "      <td>4080</td>\n",
       "      <td>0</td>\n",
       "      <td>36</td>\n",
       "      <td>2</td>\n",
       "      <td>36</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   dataset_name            alg_name  Accuracy__test  F1__test  \\\n",
       "12    openml__Satellite__167211            CatBoost        0.998039  0.998039   \n",
       "849   openml__Satellite__167211                 SVM        0.998039  0.998039   \n",
       "1508  openml__Satellite__167211               DANet        0.998039  0.998039   \n",
       "1410  openml__Satellite__167211         rtdl_ResNet        0.998039  0.998039   \n",
       "1304  openml__Satellite__167211            rtdl_MLP        0.998039  0.998039   \n",
       "1200  openml__Satellite__167211             XGBoost        0.998039  0.998039   \n",
       "1025  openml__Satellite__167211      TabTransformer        0.998039  0.998039   \n",
       "110   openml__Satellite__167211        DecisionTree        0.998039  0.998039   \n",
       "929   openml__Satellite__167211              TabNet        0.998039  0.998039   \n",
       "658   openml__Satellite__167211        RandomForest        0.998039  0.998039   \n",
       "553   openml__Satellite__167211                 MLP        0.998039  0.998039   \n",
       "357   openml__Satellite__167211            LightGBM        0.998039  0.998039   \n",
       "208   openml__Satellite__167211              DeepFM        0.998039  0.998039   \n",
       "1765  openml__Satellite__167211  rtdl_FTTransformer        0.998039  0.998039   \n",
       "759   openml__Satellite__167211                 STG        0.996078  0.996078   \n",
       "1104  openml__Satellite__167211                VIME        0.996078  0.996078   \n",
       "453   openml__Satellite__167211         LinearModel        0.996078  0.996078   \n",
       "260   openml__Satellite__167211                 KNN        0.996078  0.996078   \n",
       "1622  openml__Satellite__167211                NODE        0.996078  0.996078   \n",
       "1692  openml__Satellite__167211               SAINT        0.992157  0.992157   \n",
       "1581  openml__Satellite__167211                 NAM        0.986275  0.986275   \n",
       "\n",
       "      f__pymfe.general.nr_inst  f__pymfe.general.nr_cat  \\\n",
       "12                        4080                        0   \n",
       "849                       4080                        0   \n",
       "1508                      4080                        0   \n",
       "1410                      4080                        0   \n",
       "1304                      4080                        0   \n",
       "1200                      4080                        0   \n",
       "1025                      4080                        0   \n",
       "110                       4080                        0   \n",
       "929                       4080                        0   \n",
       "658                       4080                        0   \n",
       "553                       4080                        0   \n",
       "357                       4080                        0   \n",
       "208                       4080                        0   \n",
       "1765                      4080                        0   \n",
       "759                       4080                        0   \n",
       "1104                      4080                        0   \n",
       "453                       4080                        0   \n",
       "260                       4080                        0   \n",
       "1622                      4080                        0   \n",
       "1692                      4080                        0   \n",
       "1581                      4080                        0   \n",
       "\n",
       "      f__pymfe.general.nr_num  f__pymfe.general.nr_class  \\\n",
       "12                         36                          2   \n",
       "849                        36                          2   \n",
       "1508                       36                          2   \n",
       "1410                       36                          2   \n",
       "1304                       36                          2   \n",
       "1200                       36                          2   \n",
       "1025                       36                          2   \n",
       "110                        36                          2   \n",
       "929                        36                          2   \n",
       "658                        36                          2   \n",
       "553                        36                          2   \n",
       "357                        36                          2   \n",
       "208                        36                          2   \n",
       "1765                       36                          2   \n",
       "759                        36                          2   \n",
       "1104                       36                          2   \n",
       "453                        36                          2   \n",
       "260                        36                          2   \n",
       "1622                       36                          2   \n",
       "1692                       36                          2   \n",
       "1581                       36                          2   \n",
       "\n",
       "      f__pymfe.general.nr_attr         Group  \n",
       "12                          36  EnsembleTree  \n",
       "849                         36     Baselines  \n",
       "1508                        36            NN  \n",
       "1410                        36            NN  \n",
       "1304                        36            NN  \n",
       "1200                        36  EnsembleTree  \n",
       "1025                        36            NN  \n",
       "110                         36     Baselines  \n",
       "929                         36            NN  \n",
       "658                         36     Baselines  \n",
       "553                         36            NN  \n",
       "357                         36  EnsembleTree  \n",
       "208                         36            NN  \n",
       "1765                        36            NN  \n",
       "759                         36            NN  \n",
       "1104                        36            NN  \n",
       "453                         36     Baselines  \n",
       "260                         36     Baselines  \n",
       "1622                        36            NN  \n",
       "1692                        36            NN  \n",
       "1581                        36            NN  "
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "aggregated_dataset.query(\"dataset_name == 'openml__Satellite__167211'\").sort_values(\"Accuracy__test\", ascending = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "7b9b8ba9-fa88-44d9-ada2-e823fa17a60f",
   "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>dataset_name</th>\n",
       "      <th>alg_name</th>\n",
       "      <th>Accuracy__test</th>\n",
       "      <th>F1__test</th>\n",
       "      <th>f__pymfe.general.nr_inst</th>\n",
       "      <th>f__pymfe.general.nr_cat</th>\n",
       "      <th>f__pymfe.general.nr_num</th>\n",
       "      <th>f__pymfe.general.nr_class</th>\n",
       "      <th>f__pymfe.general.nr_attr</th>\n",
       "      <th>Group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>openml__Census-Income__168340</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.959837</td>\n",
       "      <td>0.959837</td>\n",
       "      <td>239429</td>\n",
       "      <td>28</td>\n",
       "      <td>13</td>\n",
       "      <td>2</td>\n",
       "      <td>41</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>openml__MagicTelescope__3954</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.901682</td>\n",
       "      <td>0.901682</td>\n",
       "      <td>15216</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>openml__albert__189356</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.711551</td>\n",
       "      <td>0.711551</td>\n",
       "      <td>340192</td>\n",
       "      <td>52</td>\n",
       "      <td>26</td>\n",
       "      <td>2</td>\n",
       "      <td>78</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>openml__jasmine__168911</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.859532</td>\n",
       "      <td>0.859532</td>\n",
       "      <td>2388</td>\n",
       "      <td>136</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>144</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>openml__madelon__9976</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.903846</td>\n",
       "      <td>0.903846</td>\n",
       "      <td>2080</td>\n",
       "      <td>0</td>\n",
       "      <td>500</td>\n",
       "      <td>2</td>\n",
       "      <td>500</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>openml__magic__146206</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.901682</td>\n",
       "      <td>0.901682</td>\n",
       "      <td>15216</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>openml__mfeat-morphological__18</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.805000</td>\n",
       "      <td>0.804774</td>\n",
       "      <td>1600</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>openml__one-hundred-plants-texture__9956</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.893082</td>\n",
       "      <td>0.884486</td>\n",
       "      <td>1280</td>\n",
       "      <td>0</td>\n",
       "      <td>64</td>\n",
       "      <td>100</td>\n",
       "      <td>64</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>openml__page-blocks__30</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.987203</td>\n",
       "      <td>0.987188</td>\n",
       "      <td>4379</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "      <td>10</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>openml__pc4__3902</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.958904</td>\n",
       "      <td>0.958904</td>\n",
       "      <td>1168</td>\n",
       "      <td>0</td>\n",
       "      <td>37</td>\n",
       "      <td>2</td>\n",
       "      <td>37</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>openml__steel-plates-fault__146817</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.860825</td>\n",
       "      <td>0.855989</td>\n",
       "      <td>1553</td>\n",
       "      <td>0</td>\n",
       "      <td>27</td>\n",
       "      <td>7</td>\n",
       "      <td>27</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>openml__sylva_agnostic__3889</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.998611</td>\n",
       "      <td>0.998611</td>\n",
       "      <td>11517</td>\n",
       "      <td>0</td>\n",
       "      <td>216</td>\n",
       "      <td>2</td>\n",
       "      <td>216</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                dataset_name  alg_name  Accuracy__test  \\\n",
       "0              openml__Census-Income__168340  CatBoost        0.959837   \n",
       "1               openml__MagicTelescope__3954  CatBoost        0.901682   \n",
       "2                     openml__albert__189356  CatBoost        0.711551   \n",
       "3                    openml__jasmine__168911  CatBoost        0.859532   \n",
       "4                      openml__madelon__9976  CatBoost        0.903846   \n",
       "5                      openml__magic__146206  CatBoost        0.901682   \n",
       "6            openml__mfeat-morphological__18  CatBoost        0.805000   \n",
       "7   openml__one-hundred-plants-texture__9956  CatBoost        0.893082   \n",
       "8                    openml__page-blocks__30  CatBoost        0.987203   \n",
       "9                          openml__pc4__3902  CatBoost        0.958904   \n",
       "10        openml__steel-plates-fault__146817  CatBoost        0.860825   \n",
       "11              openml__sylva_agnostic__3889  CatBoost        0.998611   \n",
       "\n",
       "    F1__test  f__pymfe.general.nr_inst  f__pymfe.general.nr_cat  \\\n",
       "0   0.959837                    239429                       28   \n",
       "1   0.901682                     15216                        0   \n",
       "2   0.711551                    340192                       52   \n",
       "3   0.859532                      2388                      136   \n",
       "4   0.903846                      2080                        0   \n",
       "5   0.901682                     15216                        0   \n",
       "6   0.804774                      1600                        0   \n",
       "7   0.884486                      1280                        0   \n",
       "8   0.987188                      4379                        0   \n",
       "9   0.958904                      1168                        0   \n",
       "10  0.855989                      1553                        0   \n",
       "11  0.998611                     11517                        0   \n",
       "\n",
       "    f__pymfe.general.nr_num  f__pymfe.general.nr_class  \\\n",
       "0                        13                          2   \n",
       "1                        10                          2   \n",
       "2                        26                          2   \n",
       "3                         8                          2   \n",
       "4                       500                          2   \n",
       "5                        10                          2   \n",
       "6                         6                         10   \n",
       "7                        64                        100   \n",
       "8                        10                          5   \n",
       "9                        37                          2   \n",
       "10                       27                          7   \n",
       "11                      216                          2   \n",
       "\n",
       "    f__pymfe.general.nr_attr         Group  \n",
       "0                         41  EnsembleTree  \n",
       "1                         10  EnsembleTree  \n",
       "2                         78  EnsembleTree  \n",
       "3                        144  EnsembleTree  \n",
       "4                        500  EnsembleTree  \n",
       "5                         10  EnsembleTree  \n",
       "6                          6  EnsembleTree  \n",
       "7                         64  EnsembleTree  \n",
       "8                         10  EnsembleTree  \n",
       "9                         37  EnsembleTree  \n",
       "10                        27  EnsembleTree  \n",
       "11                       216  EnsembleTree  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_algs.query(\" alg_name == 'CatBoost' \").reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "6634f82c-9e45-46c5-8a61-459cd8c93db3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Series([], Name: Accuracy__test, dtype: float64)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "aggregated_dataset.query(\"dataset_name == 'openml__sonar__39'\").groupby(\"alg_name\")[\"Accuracy__test\"].max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "accd1d61-056e-4d9b-a8bd-04c06e4ae92a",
   "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>dataset_name</th>\n",
       "      <th>alg_name</th>\n",
       "      <th>Accuracy__test</th>\n",
       "      <th>F1__test</th>\n",
       "      <th>f__pymfe.general.nr_inst</th>\n",
       "      <th>f__pymfe.general.nr_cat</th>\n",
       "      <th>f__pymfe.general.nr_num</th>\n",
       "      <th>f__pymfe.general.nr_class</th>\n",
       "      <th>f__pymfe.general.nr_attr</th>\n",
       "      <th>Group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>openml__sick__3021</td>\n",
       "      <td>DecisionTree</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3018</td>\n",
       "      <td>22</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>28</td>\n",
       "      <td>Baselines</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         dataset_name      alg_name  Accuracy__test  F1__test  \\\n",
       "0  openml__sick__3021  DecisionTree             1.0       1.0   \n",
       "\n",
       "   f__pymfe.general.nr_inst  f__pymfe.general.nr_cat  f__pymfe.general.nr_num  \\\n",
       "0                      3018                       22                        6   \n",
       "\n",
       "   f__pymfe.general.nr_class  f__pymfe.general.nr_attr      Group  \n",
       "0                          2                        28  Baselines  "
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_algs.query(\" alg_name == 'DecisionTree' \").reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "14a6f607",
   "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>dataset_name</th>\n",
       "      <th>alg_name</th>\n",
       "      <th>Accuracy__test</th>\n",
       "      <th>F1__test</th>\n",
       "      <th>f__pymfe.general.nr_inst</th>\n",
       "      <th>f__pymfe.general.nr_cat</th>\n",
       "      <th>f__pymfe.general.nr_num</th>\n",
       "      <th>f__pymfe.general.nr_class</th>\n",
       "      <th>f__pymfe.general.nr_attr</th>\n",
       "      <th>Group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>openml__Census-Income__168340</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.959837</td>\n",
       "      <td>0.959837</td>\n",
       "      <td>239429</td>\n",
       "      <td>28</td>\n",
       "      <td>13</td>\n",
       "      <td>2</td>\n",
       "      <td>41</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>openml__MagicTelescope__3954</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.901682</td>\n",
       "      <td>0.901682</td>\n",
       "      <td>15216</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>openml__albert__189356</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.711551</td>\n",
       "      <td>0.711551</td>\n",
       "      <td>340192</td>\n",
       "      <td>52</td>\n",
       "      <td>26</td>\n",
       "      <td>2</td>\n",
       "      <td>78</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>openml__jasmine__168911</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.859532</td>\n",
       "      <td>0.859532</td>\n",
       "      <td>2388</td>\n",
       "      <td>136</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>144</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>openml__madelon__9976</td>\n",
       "      <td>CatBoost</td>\n",
       "      <td>0.903846</td>\n",
       "      <td>0.903846</td>\n",
       "      <td>2080</td>\n",
       "      <td>0</td>\n",
       "      <td>500</td>\n",
       "      <td>2</td>\n",
       "      <td>500</td>\n",
       "      <td>EnsembleTree</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1752</th>\n",
       "      <td>openml__splice__45</td>\n",
       "      <td>SAINT</td>\n",
       "      <td>0.987461</td>\n",
       "      <td>0.987517</td>\n",
       "      <td>2552</td>\n",
       "      <td>60</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>60</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1764</th>\n",
       "      <td>openml__PhishingWebsites__14952</td>\n",
       "      <td>rtdl_FTTransformer</td>\n",
       "      <td>0.983710</td>\n",
       "      <td>0.983710</td>\n",
       "      <td>8845</td>\n",
       "      <td>30</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>30</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1772</th>\n",
       "      <td>openml__bank-marketing__14965</td>\n",
       "      <td>rtdl_FTTransformer</td>\n",
       "      <td>0.920150</td>\n",
       "      <td>0.920150</td>\n",
       "      <td>36169</td>\n",
       "      <td>9</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>16</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1773</th>\n",
       "      <td>openml__bank-marketing__9899</td>\n",
       "      <td>rtdl_FTTransformer</td>\n",
       "      <td>0.935841</td>\n",
       "      <td>0.935841</td>\n",
       "      <td>3617</td>\n",
       "      <td>9</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>16</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1793</th>\n",
       "      <td>openml__house_16H__3686</td>\n",
       "      <td>rtdl_FTTransformer</td>\n",
       "      <td>0.906936</td>\n",
       "      <td>0.906936</td>\n",
       "      <td>18228</td>\n",
       "      <td>0</td>\n",
       "      <td>16</td>\n",
       "      <td>2</td>\n",
       "      <td>16</td>\n",
       "      <td>NN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>81 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                         dataset_name            alg_name  Accuracy__test  \\\n",
       "3       openml__Census-Income__168340            CatBoost        0.959837   \n",
       "9        openml__MagicTelescope__3954            CatBoost        0.901682   \n",
       "18             openml__albert__189356            CatBoost        0.711551   \n",
       "47            openml__jasmine__168911            CatBoost        0.859532   \n",
       "55              openml__madelon__9976            CatBoost        0.903846   \n",
       "...                               ...                 ...             ...   \n",
       "1752               openml__splice__45               SAINT        0.987461   \n",
       "1764  openml__PhishingWebsites__14952  rtdl_FTTransformer        0.983710   \n",
       "1772    openml__bank-marketing__14965  rtdl_FTTransformer        0.920150   \n",
       "1773     openml__bank-marketing__9899  rtdl_FTTransformer        0.935841   \n",
       "1793          openml__house_16H__3686  rtdl_FTTransformer        0.906936   \n",
       "\n",
       "      F1__test  f__pymfe.general.nr_inst  f__pymfe.general.nr_cat  \\\n",
       "3     0.959837                    239429                       28   \n",
       "9     0.901682                     15216                        0   \n",
       "18    0.711551                    340192                       52   \n",
       "47    0.859532                      2388                      136   \n",
       "55    0.903846                      2080                        0   \n",
       "...        ...                       ...                      ...   \n",
       "1752  0.987517                      2552                       60   \n",
       "1764  0.983710                      8845                       30   \n",
       "1772  0.920150                     36169                        9   \n",
       "1773  0.935841                      3617                        9   \n",
       "1793  0.906936                     18228                        0   \n",
       "\n",
       "      f__pymfe.general.nr_num  f__pymfe.general.nr_class  \\\n",
       "3                          13                          2   \n",
       "9                          10                          2   \n",
       "18                         26                          2   \n",
       "47                          8                          2   \n",
       "55                        500                          2   \n",
       "...                       ...                        ...   \n",
       "1752                        0                          3   \n",
       "1764                        0                          2   \n",
       "1772                        7                          2   \n",
       "1773                        7                          2   \n",
       "1793                       16                          2   \n",
       "\n",
       "      f__pymfe.general.nr_attr         Group  \n",
       "3                           41  EnsembleTree  \n",
       "9                           10  EnsembleTree  \n",
       "18                          78  EnsembleTree  \n",
       "47                         144  EnsembleTree  \n",
       "55                         500  EnsembleTree  \n",
       "...                        ...           ...  \n",
       "1752                        60            NN  \n",
       "1764                        30            NN  \n",
       "1772                        16            NN  \n",
       "1773                        16            NN  \n",
       "1793                        16            NN  \n",
       "\n",
       "[81 rows x 10 columns]"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "best_algs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "ea8a3576",
   "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>dataset_name</th>\n",
       "      <th>Group</th>\n",
       "      <th>Accuracy__test</th>\n",
       "      <th>F1__test</th>\n",
       "      <th>f__pymfe.general.nr_inst</th>\n",
       "      <th>Accuracy_max</th>\n",
       "      <th>gap</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>openml__APSFailure__168868</td>\n",
       "      <td>Baselines</td>\n",
       "      <td>0.994342</td>\n",
       "      <td>0.994342</td>\n",
       "      <td>60800</td>\n",
       "      <td>0.996447</td>\n",
       "      <td>-0.002105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>openml__APSFailure__168868</td>\n",
       "      <td>EnsembleTree</td>\n",
       "      <td>0.996447</td>\n",
       "      <td>0.996447</td>\n",
       "      <td>60800</td>\n",
       "      <td>0.996447</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>openml__APSFailure__168868</td>\n",
       "      <td>NN</td>\n",
       "      <td>0.994474</td>\n",
       "      <td>0.994474</td>\n",
       "      <td>60800</td>\n",
       "      <td>0.996447</td>\n",
       "      <td>-0.001974</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>openml__Amazon_employee_access__34539</td>\n",
       "      <td>Baselines</td>\n",
       "      <td>0.946903</td>\n",
       "      <td>0.946903</td>\n",
       "      <td>26216</td>\n",
       "      <td>0.957888</td>\n",
       "      <td>-0.010986</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>openml__Amazon_employee_access__34539</td>\n",
       "      <td>EnsembleTree</td>\n",
       "      <td>0.957888</td>\n",
       "      <td>0.957888</td>\n",
       "      <td>26216</td>\n",
       "      <td>0.957888</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>312</th>\n",
       "      <td>openml__wilt__146820</td>\n",
       "      <td>EnsembleTree</td>\n",
       "      <td>0.995868</td>\n",
       "      <td>0.995868</td>\n",
       "      <td>3872</td>\n",
       "      <td>0.995868</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>313</th>\n",
       "      <td>openml__wilt__146820</td>\n",
       "      <td>NN</td>\n",
       "      <td>0.995868</td>\n",
       "      <td>0.995868</td>\n",
       "      <td>3872</td>\n",
       "      <td>0.995868</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>314</th>\n",
       "      <td>openml__yeast__145793</td>\n",
       "      <td>Baselines</td>\n",
       "      <td>0.724409</td>\n",
       "      <td>0.727976</td>\n",
       "      <td>1016</td>\n",
       "      <td>0.755906</td>\n",
       "      <td>-0.031496</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>315</th>\n",
       "      <td>openml__yeast__145793</td>\n",
       "      <td>EnsembleTree</td>\n",
       "      <td>0.755906</td>\n",
       "      <td>0.757876</td>\n",
       "      <td>1016</td>\n",
       "      <td>0.755906</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>316</th>\n",
       "      <td>openml__yeast__145793</td>\n",
       "      <td>NN</td>\n",
       "      <td>0.755906</td>\n",
       "      <td>0.757341</td>\n",
       "      <td>1016</td>\n",
       "      <td>0.755906</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>317 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                              dataset_name         Group  Accuracy__test  \\\n",
       "0               openml__APSFailure__168868     Baselines        0.994342   \n",
       "1               openml__APSFailure__168868  EnsembleTree        0.996447   \n",
       "2               openml__APSFailure__168868            NN        0.994474   \n",
       "3    openml__Amazon_employee_access__34539     Baselines        0.946903   \n",
       "4    openml__Amazon_employee_access__34539  EnsembleTree        0.957888   \n",
       "..                                     ...           ...             ...   \n",
       "312                   openml__wilt__146820  EnsembleTree        0.995868   \n",
       "313                   openml__wilt__146820            NN        0.995868   \n",
       "314                  openml__yeast__145793     Baselines        0.724409   \n",
       "315                  openml__yeast__145793  EnsembleTree        0.755906   \n",
       "316                  openml__yeast__145793            NN        0.755906   \n",
       "\n",
       "     F1__test  f__pymfe.general.nr_inst  Accuracy_max       gap  \n",
       "0    0.994342                     60800      0.996447 -0.002105  \n",
       "1    0.996447                     60800      0.996447  0.000000  \n",
       "2    0.994474                     60800      0.996447 -0.001974  \n",
       "3    0.946903                     26216      0.957888 -0.010986  \n",
       "4    0.957888                     26216      0.957888  0.000000  \n",
       "..        ...                       ...           ...       ...  \n",
       "312  0.995868                      3872      0.995868  0.000000  \n",
       "313  0.995868                      3872      0.995868  0.000000  \n",
       "314  0.727976                      1016      0.755906 -0.031496  \n",
       "315  0.757876                      1016      0.755906  0.000000  \n",
       "316  0.757341                      1016      0.755906  0.000000  \n",
       "\n",
       "[317 rows x 7 columns]"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "group_performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "id": "bb1e10fe",
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset_big_gap = np.unique(group_performance.sort_values(\"gap\").head(50)[\"dataset_name\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "id": "29297a7a-f161-40d7-af4a-e50f674ec5e1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set(style=\"whitegrid\")\n",
    "\n",
    "to_plot = group_performance.query(\" dataset_name in @dataset_big_gap\")\n",
    "plt.figure(figsize=(20, 10))\n",
    "\n",
    "# Plot vertical lines\n",
    "for x in to_plot['f__pymfe.general.nr_inst'].unique():\n",
    "    plt.axvline(x=x, color='black', linestyle='-')  # Non-dotted black vertical line\n",
    "\n",
    "scatter = sns.scatterplot(x='f__pymfe.general.nr_inst',\n",
    "                          y='Accuracy__test', \n",
    "                          hue='Group', \n",
    "                          data=to_plot, \n",
    "                          palette='coolwarm', \n",
    "                          s=150,\n",
    "                          alpha=1  # adjusted alpha for transparency\n",
    "                         )\n",
    "\n",
    "plt.title('Scatter Plot of nr_inst vs accuracy Colored by Group')\n",
    "plt.xlabel('Number of Instances (nr_inst)')\n",
    "plt.ylabel('Test Accuracy (Accuracy__test)')\n",
    "\n",
    "plt.xscale(value = \"log\")\n",
    "\n",
    "# Setting the limits for x and y axes\n",
    "#plt.xlim(-1000, 10000)\n",
    "#plt.ylim(-1, 20)\n",
    "\n",
    "plt.grid(visible=False)\n",
    "plt.legend(title='Group', bbox_to_anchor=(1.05, 1), loc=2)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "7d82aacd",
   "metadata": {},
   "outputs": [],
   "source": [
    "wrost_NN = group_performance.query(\" Group == 'NN' \").sort_values('gap')[\"dataset_name\"].head(10)\n",
    "wrost_GDBT = group_performance.query(\" Group == 'EnsembleTree' \").sort_values('gap')[\"dataset_name\"].head(10)\n",
    "wrost_baseline = group_performance.query(\" Group == 'Baselines' \").sort_values('gap')[\"dataset_name\"].head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "0a5f428f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10, 6))\n",
    "\n",
    "to_plot = group_performance.query(\" dataset_name in @wrost_NN \")\n",
    "\n",
    "sns.barplot(data=to_plot, x='dataset_name', y='Accuracy__test', hue='Group')\n",
    "plt.title('Accuracy Test Results on 10 wrost datataset for NN')\n",
    "plt.xlabel('Dataset Name')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend(title='Group')\n",
    "plt.xticks(rotation=90)  # Rotate dataset names to vertical\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "id": "bb5cfe9d",
   "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>dataset_name</th>\n",
       "      <th>Group</th>\n",
       "      <th>Accuracy__test</th>\n",
       "      <th>F1__test</th>\n",
       "      <th>f__pymfe.general.nr_inst</th>\n",
       "      <th>Accuracy_max</th>\n",
       "      <th>gap</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [dataset_name, Group, Accuracy__test, F1__test, f__pymfe.general.nr_inst, Accuracy_max, gap]\n",
       "Index: []"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "to_plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "id": "124bb1da",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10, 6))\n",
    "\n",
    "to_plot = group_performance.query(\" dataset_name in @wrost_GDBT \")\n",
    "\n",
    "sns.barplot(data=to_plot, x='dataset_name', y='Accuracy__test', hue='Group')\n",
    "plt.title('Accuracy Test Results on 10 wrost datataset for GDBT')\n",
    "plt.xlabel('Dataset Name')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend(title='Group')\n",
    "plt.xticks(rotation=90)  # Rotate dataset names to vertical\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "c91d33f1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10, 6))\n",
    "\n",
    "to_plot = group_performance.query(\" dataset_name in @wrost_baseline \")\n",
    "\n",
    "sns.barplot(data=to_plot, x='dataset_name', y='Accuracy__test', hue='Group')\n",
    "plt.title('Accuracy Test Results on 10 wrost datataset for Baselines')\n",
    "plt.xlabel('Dataset Name')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend(title='Group')\n",
    "plt.xticks(rotation=90)  # Rotate dataset names to vertical\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7646d747-8965-4730-843a-85d06e1d44c5",
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots(figsize=(10, 6))\n",
    "hb = ax.hexbin(merged_df['f__pymfe.general.nr_inst'], merged_df['f__pymfe.general.nr_cat'], gridsize=50, cmap='Greys', bins='log')\n",
    "ax.set_xlim([0, 17000])\n",
    "ax.set_ylim([0, 20])\n",
    "ax.set_xlabel('Number of Instances (nr_inst)')\n",
    "ax.set_ylabel('Number of Categories (nr_cat)')\n",
    "ax.set_title('Hexbin Plot of nr_inst vs nr_cat')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f61a0646",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}