stage_comparisons.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1. Packages and Functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import neurokit2 as nk2\n",
    "import biosppy as bs\n",
    "# import biobss as bo\n",
    "from sklearn.metrics import mean_squared_error, mean_absolute_error\n",
    "import warnings\n",
    "from copy import deepcopy\n",
    "import os\n",
    "import pandas as pd\n",
    "from tqdm import tqdm\n",
    "from collections import Counter\n",
    "plt.rcParams['figure.figsize'] = [20, 5]  # Bigger images\n",
    "plt.rcParams['font.size']= 14\n",
    "pd.set_option('display.max_columns', None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_ecg(signal, peaks=None, title=\"ECG\"):\n",
    "    colors = [\"#9F0162\", \"#009F81\", \"#FF5AAF\", \"#00FCCF\", \"#8400CD\", \"#008DF9\", \"#00C2F9\", \"#FFB2FD\", \"#A40122\", \"#E20134\", \"#FF6E3A\", \"#FFC33B\"]\n",
    "    plt.plot(signal)\n",
    "    plt.title(title)\n",
    "    legend = [\"Signal\"]\n",
    "    if peaks is not None:\n",
    "        for clr, (peak_name, peak_points) in enumerate(peaks.items()):\n",
    "            print(peak_name, peak_points)\n",
    "            plt.scatter(peak_points, signal[peak_points], s=200, c=colors[clr % len(colors)], marker='.')\n",
    "            legend.append(peak_name)\n",
    "    plt.legend(legend, loc=\"upper right\")\n",
    "    plt.show()\n",
    "\n",
    "def plot_multiple_ecg(signals, titles):\n",
    "    fig, axs = plt.subplots(len(signals))\n",
    "    for i, signal in enumerate(signals):\n",
    "        axs[i].plot(signal)\n",
    "        axs[i].set_xlabel(\"Time\")\n",
    "        axs[i].set_ylabel(titles[i])\n",
    "    for ax in axs.flat:\n",
    "        ax.label_outer()\n",
    "    fig.savefig(\"figs/denoised\", dpi=500, transparent=True)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. Get Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_path = \"./data/\"\n",
    "comparison_folder = \"./comparisons/\"\n",
    "signals = np.load(data_path + \"signals.npy\", allow_pickle=True)\n",
    "labels = np.load(data_path + \"labels.npy\", allow_pickle=True)\n",
    "r_peaks = np.load(data_path + \"peaks.npy\", allow_pickle=True)\n",
    "likelihoods = np.load(data_path + \"likelihoods.npy\", allow_pickle=True)\n",
    "sr = 100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "r_peaks [2, 100, 203, 301, 407, 516, 624, 730, 826]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 2000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# np.random.seed(42)\n",
    "idx = np.random.randint(0, signals.shape[0])\n",
    "signal = signals[idx, :, 0]\n",
    "signal_peaks = {\"r_peaks\": r_peaks[idx]}\n",
    "plot_ecg(signal, signal_peaks, \"Raw ECG and Validated R Peaks (Label: \" + str(labels[idx]) +\")\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. Denoising"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "methods = [\"Raw\", \"Neurokit\", \"Biosppy\", \"pantompkins1985\", \"hamilton2002\", \"elgendi2010\", \"koka2022\"]\n",
    "cleaned_signals = [signal]\n",
    "for method in methods[1:]:\n",
    "    cleaned_signals.append(nk2.ecg_clean(signal, sampling_rate=sr, method=method))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# plot_ecg(signal, title=\"Raw\")\n",
    "# plot_ecg(nk2.ecg_clean(signal, sampling_rate=sr, method=\"neurokit\"), title=\"Cleaned (Neurokit)\")\n",
    "# plot_ecg(nk2.ecg_clean(signal, sampling_rate=sr, method=\"biosppy\"), title=\"Cleaned (Biosppy-nk)\")\n",
    "# plot_ecg(bs.signals.ecg.ecg(signal, sampling_rate=sr, show=False)[1], title=\"Cleaned (Biosppy-org)\")\n",
    "# plot_ecg(nk2.ecg_clean(signal, sampling_rate=sr, method=\"pantompkins1985\"), title=\"Cleaned (pantompkins1985)\")\n",
    "# plot_ecg(nk2.ecg_clean(signal, sampling_rate=sr, method=\"hamilton2002\"), title=\"Cleaned (hamilton2002)\")\n",
    "# plot_ecg(nk2.ecg_clean(signal, sampling_rate=sr, method=\"elgendi2010\"), title=\"Cleaned (elgendi2010)\")\n",
    "# plot_ecg(nk2.ecg_clean(signal, sampling_rate=sr, method=\"koka2022\"), title=\"Cleaned (koka2022)\")\n",
    "plot_multiple_ecg(cleaned_signals, titles=methods)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 4. Feature Engineering Method Comparison"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4.1 R Peak Detection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "methods_denoising = [\"Raw\", \"Neurokit\", \"Biosppy\", \"pantompkins1985\", \"hamilton2002\", \"elgendi2010\", \"koka2022\"]\n",
    "methods_R_detection = [\"Neurokit\", \"pantompkins1985\", \"hamilton2002\", \"zong2003\", \"martinez2004\", \"christov2004\", \"elgendi2010\",\"kalidas2017\", \"nabian2018\", \"rodrigues2021\", \"koka2022\", \"promac\"]\n",
    "promac_methods=[\n",
    "        \"neurokit\",\n",
    "        \"ssf\",\n",
    "        \"zong\",\n",
    "        \"elgendi\",\n",
    "        \"kalidas\",\n",
    "        \"martinez\",\n",
    "        \"rodrigues\",\n",
    "    ]\n",
    "kwargs = {'promac_methods': promac_methods}\n",
    "R_info_template = {\n",
    "    \"MSE\": [],\n",
    "    \"MAE\": [],\n",
    "    \"Correct R #\": [],\n",
    "    \"Detected R #\": [],\n",
    "    \"TP\": [],\n",
    "    \"FP\": [],\n",
    "    \"Performance\": [],\n",
    "    \"Acc\": [],\n",
    "    \"Err\": [],\n",
    "    \"Match\": [],\n",
    "    \"Deviation\": [],\n",
    "    \"Mean deviation\": [],\n",
    "    \"Std deviation\": [],\n",
    "    \"mean_ref_ibi\": [],\n",
    "    \"std_ref_ibi\": [],\n",
    "    \"mean_test_ibi\": [],\n",
    "    \"std_test_ibi\": [],\n",
    "    \"Label\": [],\n",
    "    \"Likelihoods\": []\n",
    "}\n",
    "R_info = {}\n",
    "for cleaner in methods_denoising:\n",
    "    for detector in methods_R_detection:\n",
    "        for ca in [True, False]:\n",
    "            name = cleaner + \"+\" + detector\n",
    "            if ca:\n",
    "                name += \"+Kubios\"\n",
    "            info = deepcopy(R_info_template)\n",
    "            print(\"Starting \" + name)\n",
    "            for signal, r_correct, label, likelihood in tqdm(zip(signals, r_peaks, labels, likelihoods)):\n",
    "                signal = signal[:,0]\n",
    "                with warnings.catch_warnings():\n",
    "                    warnings.simplefilter(\"ignore\", category=RuntimeWarning)\n",
    "                    cleaned = nk2.ecg_clean(signal, sampling_rate=sr, method=cleaner) if cleaner != \"Raw\" else signal.copy()\n",
    "                    if detector == \"promac\":\n",
    "                        r_detected = nk2.ecg_peaks(cleaned, sr, detector, ca, **kwargs)[1][\"ECG_R_Peaks\"]\n",
    "                    else:\n",
    "                        r_detected = nk2.ecg_peaks(cleaned, sr, detector, ca)[1][\"ECG_R_Peaks\"]\n",
    "\n",
    "                if(len(r_detected) == len(r_correct)):\n",
    "                    info['MSE'].append(mean_squared_error(r_correct, r_detected))\n",
    "                    info['MAE'].append(mean_absolute_error(r_correct, r_detected))\n",
    "                else:\n",
    "                    info['MSE'].append(np.Inf)\n",
    "                    info['MAE'].append(np.Inf)\n",
    "\n",
    "                info['Correct R #'].append(len(r_correct))\n",
    "                info['Detected R #'].append(len(r_detected))\n",
    "                info['Label'].append(str(label))\n",
    "                info['Likelihoods'].append(str(likelihood))\n",
    "                try:\n",
    "                    comparison = dict(bs.signals.ecg.compare_segmentation(r_correct, r_detected, sr))\n",
    "                    info['TP'].append(comparison['TP'])\n",
    "                    info['FP'].append(comparison['FP'])\n",
    "                    info['Performance'].append(comparison['performance'])\n",
    "                    info['Acc'].append(comparison['acc'])\n",
    "                    info['Err'].append(comparison['err'])\n",
    "                    info['Match'].append(str(comparison['match']))\n",
    "                    info['Deviation'].append(str(comparison['deviation']))\n",
    "                    info['Mean deviation'].append(comparison['mean_deviation'])\n",
    "                    info['Std deviation'].append(comparison['std_deviation'])\n",
    "                    info[\"mean_ref_ibi\"].append(comparison[\"mean_ref_ibi\"])\n",
    "                    info[\"std_ref_ibi\"].append(comparison[\"std_ref_ibi\"])\n",
    "                    info[\"mean_test_ibi\"].append(comparison[\"mean_test_ibi\"])\n",
    "                    info[\"std_test_ibi\"].append(comparison[\"std_test_ibi\"])\n",
    "                except ZeroDivisionError:\n",
    "                    info['TP'].append(0)\n",
    "                    info['FP'].append(0)\n",
    "                    info['Performance'].append(0)\n",
    "                    info['Acc'].append(0)\n",
    "                    info['Err'].append(0)\n",
    "                    info['Match'].append(str(0))\n",
    "                    info['Deviation'].append(str(0))\n",
    "                    info['Mean deviation'].append(0)\n",
    "                    info['Std deviation'].append(0)\n",
    "                    info[\"mean_ref_ibi\"].append(0)\n",
    "                    info[\"std_ref_ibi\"].append(0)\n",
    "                    info[\"mean_test_ibi\"].append(0)\n",
    "                    info[\"std_test_ibi\"].append(0)\n",
    "            df = pd.DataFrame.from_dict(info)\n",
    "            display(df.head())\n",
    "            df.to_csv(name + '.csv', index=False)\n",
    "            R_info[name] = info"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 4.1.1 Compare R Peak Detection Algorithms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "metric_files = os.listdir(comparison_folder)\n",
    "\n",
    "# str_labels = list(map(str, labels))\n",
    "# str_likelihoods = list(map(str, likelihoods))\n",
    "\n",
    "\n",
    "# for metric_file in tqdm(metric_files):\n",
    "#     df = pd.read_csv(os.path.join(comparison_folder, metric_file))\n",
    "#     df['Label'] = str_labels\n",
    "#     df['Likelihoods'] = str_likelihoods\n",
    "#     df.to_csv(os.path.join(comparison_folder, metric_file), index=False)\n",
    "\n",
    "nk_promac_ca = pd.read_csv(os.path.join(comparison_folder + \"koka2022+rodrigues2021.csv\"))\n",
    "display(nk_promac_ca[:10])\n",
    "print(\"Neurokit+promac+Kubios.csv\"[:-4])\n",
    "Exactly_matched_rows = nk_promac_ca[nk_promac_ca[\"Correct R #\"] == nk_promac_ca[\"Detected R #\"]]\n",
    "print(len(Exactly_matched_rows))\n",
    "print(Exactly_matched_rows['MSE'].mean())\n",
    "print(Exactly_matched_rows['MSE'].max())\n",
    "print(Exactly_matched_rows['MSE'].min())\n",
    "print(Exactly_matched_rows['MAE'].mean())\n",
    "print(Exactly_matched_rows['MAE'].max())\n",
    "print(Exactly_matched_rows['MAE'].min())\n",
    "print(nk_promac_ca['Correct R #'].mean())\n",
    "print(nk_promac_ca['Detected R #'].mean())\n",
    "print(nk_promac_ca['TP'].mean())\n",
    "print(nk_promac_ca['FP'].mean())\n",
    "print(nk_promac_ca['Performance'].mean())\n",
    "print(nk_promac_ca['Acc'].mean())\n",
    "print(nk_promac_ca['Err'].mean())\n",
    "print(nk_promac_ca['Mean deviation'].mean())\n",
    "print(nk_promac_ca['Std deviation'].mean())\n",
    "print(nk_promac_ca['mean_ref_ibi'].mean())\n",
    "print(nk_promac_ca['mean_ref_ibi'].max())\n",
    "print(nk_promac_ca['mean_ref_ibi'].min())\n",
    "print(nk_promac_ca['std_ref_ibi'].mean())\n",
    "print(nk_promac_ca['std_ref_ibi'].max())\n",
    "print(nk_promac_ca['std_ref_ibi'].min())\n",
    "print(nk_promac_ca['mean_test_ibi'].mean())\n",
    "print(nk_promac_ca['mean_test_ibi'].max())\n",
    "print(nk_promac_ca['mean_test_ibi'].min())\n",
    "print(nk_promac_ca['std_test_ibi'].mean())\n",
    "print(nk_promac_ca['std_test_ibi'].max())\n",
    "print(nk_promac_ca['std_test_ibi'].min())"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Name of the method\n",
    "* Exact Matches: # of signals where detected R # == correct R #\n",
    "* Mean MSE: mean of exact matched signals\n",
    "* Max MSE\n",
    "* Min MSE\n",
    "* Mean MAE: mean of exact matched signals\n",
    "* Max MAE\n",
    "* Min MAE\n",
    "* Mean Correct R #: all signals\n",
    "* Mean Detected R #: all signals\n",
    "\n",
    "---- Below can be set to 0 because of the zerodivision error\n",
    "* Mean TP: mean of correctly detected peaks with tolerence of +- 0.05 seconds\n",
    "* Mean FP: mean of incrorrectly detected peaks at the beginning (intermediate incorrect detections can be removed by min RR check)\n",
    "* Mean Performance: mean of (TP/ Correct R #)\n",
    "* Mean Acc: mean of (TP/ (TP+FP))\n",
    "* Mean Err: mean of (FP / (TP+FP))\n",
    "* Mean mean_deviation:  Mean error (seconds)\n",
    "* Mean std_deviation: Standard deviation of error (seconds)\n",
    "* Mean mean_ref_ibi: Mean of the reference interbeat intervals (seconds)\n",
    "* Min mean_ref_ibi: Mean of the reference interbeat intervals (seconds)\n",
    "* Mean std_ref_ibi: Standard deviation of the reference interbeat intervals (seconds)\n",
    "* Min std_ref_ibi: Standard deviation of the reference interbeat intervals (seconds)\n",
    "* Mean mean_test_ibi: Mean of the test interbeat intervals (seconds)\n",
    "* Min mean_test_ibi: Mean of the test interbeat intervals (seconds)\n",
    "* Mean std_test_ibi: Standard deviation of the test interbeat intervals (seconds)\n",
    "* Min std_test_ibi: Standard deviation of the test interbeat intervals (seconds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 182/182 [01:31<00:00,  1.98it/s]\n",
      "100%|██████████| 182/182 [01:07<00:00,  2.70it/s]\n",
      "100%|██████████| 182/182 [01:06<00:00,  2.73it/s]\n",
      "100%|██████████| 182/182 [01:00<00:00,  3.01it/s]\n"
     ]
    }
   ],
   "source": [
    "stats_template = {\n",
    "    \"Name\": [],\n",
    "    \"Exact Matches #\": [],\n",
    "    \"Mean MSE\": [],\n",
    "    \"Max MSE\": [],\n",
    "    \"Min MSE\": [],\n",
    "    \"Mean MAE\": [],\n",
    "    \"Max MAE\": [],\n",
    "    \"Min MAE\": [],\n",
    "    \"Mean Correct R #\": [],\n",
    "    \"Mean Detected R #\": [],\n",
    "    \"Mean TP\": [],\n",
    "    \"Mean FP\": [],\n",
    "    \"Mean Performance\": [],\n",
    "    \"Mean Acc\": [],\n",
    "    \"Mean Err\": [],\n",
    "    \"Mean mean_deviation\": [],\n",
    "    \"Mean std_deviation\": [],\n",
    "    \"Mean mean_ref_ibi\": [],\n",
    "    \"Max mean_ref_ibi\": [],\n",
    "    \"Min mean_ref_ibi\": [],\n",
    "    \"Mean std_ref_ibi\": [],\n",
    "    \"Max std_ref_ibi\": [],\n",
    "    \"Min std_ref_ibi\": [],\n",
    "    \"Mean mean_test_ibi\": [],\n",
    "    \"Max mean_test_ibi\": [],\n",
    "    \"Min mean_test_ibi\": [],\n",
    "    \"Mean std_test_ibi\": [],\n",
    "    \"Max std_test_ibi\": [],\n",
    "    \"Min std_test_ibi\": []\n",
    "}\n",
    "def create_stats(metric_files, name=\"stats\", label_filter = None):\n",
    "    stats = deepcopy(stats_template)\n",
    "    for metric_file in tqdm(metric_files):\n",
    "        stats['Name'].append(metric_file[:-4])\n",
    "        df = pd.read_csv(os.path.join(comparison_folder, metric_file))\n",
    "\n",
    "        if label_filter is not None:\n",
    "            df = df[label_filter == df['Label']]\n",
    "        \n",
    "        exactly_matched_rows = df[df[\"Correct R #\"] == df[\"Detected R #\"]]\n",
    "        stats['Exact Matches #'].append(len(exactly_matched_rows))\n",
    "        \n",
    "        stats['Mean MSE'].append(exactly_matched_rows['MSE'].mean())\n",
    "        stats['Max MSE'].append(exactly_matched_rows['MSE'].max())\n",
    "        stats['Min MSE'].append(exactly_matched_rows['MSE'].min())\n",
    "\n",
    "        stats['Mean MAE'].append(exactly_matched_rows['MAE'].mean())\n",
    "        stats['Max MAE'].append(exactly_matched_rows['MAE'].max())\n",
    "        stats['Min MAE'].append(exactly_matched_rows['MAE'].min())\n",
    "\n",
    "        stats['Mean Correct R #'].append(df['Correct R #'].mean())\n",
    "        stats['Mean Detected R #'].append(df['Detected R #'].mean())\n",
    "        stats['Mean TP'].append(df['TP'].mean())\n",
    "        stats['Mean FP'].append(df['FP'].mean())\n",
    "        \n",
    "        stats['Mean Performance'].append(df['Performance'].mean())\n",
    "        stats['Mean Acc'].append(df['Acc'].mean())\n",
    "        stats['Mean Err'].append(df['Err'].mean())\n",
    "\n",
    "        stats['Mean mean_deviation'].append(df['Mean deviation'].mean())\n",
    "        stats['Mean std_deviation'].append(df['Std deviation'].mean())\n",
    "\n",
    "        stats['Mean mean_ref_ibi'].append(df['mean_ref_ibi'].mean())\n",
    "        stats['Max mean_ref_ibi'].append(df['mean_ref_ibi'].max())\n",
    "        stats['Min mean_ref_ibi'].append(df['mean_ref_ibi'].min())\n",
    "\n",
    "        stats['Mean std_ref_ibi'].append(df['std_ref_ibi'].mean())\n",
    "        stats['Max std_ref_ibi'].append(df['std_ref_ibi'].max())\n",
    "        stats['Min std_ref_ibi'].append(df['std_ref_ibi'].min())\n",
    "\n",
    "        stats['Mean mean_test_ibi'].append(df['mean_test_ibi'].mean())\n",
    "        stats['Max mean_test_ibi'].append(df['mean_test_ibi'].max())\n",
    "        stats['Min mean_test_ibi'].append(df['mean_test_ibi'].min())\n",
    "\n",
    "        stats['Mean std_test_ibi'].append(df['std_test_ibi'].mean())\n",
    "        stats['Max std_test_ibi'].append(df['std_test_ibi'].max())\n",
    "        stats['Min std_test_ibi'].append(df['std_test_ibi'].min())\n",
    "\n",
    "    df_stats = pd.DataFrame.from_dict(stats)\n",
    "    df_stats.to_csv(os.path.join(\"./comparison_stats/\", name + '.csv'), index=False)\n",
    "\n",
    "create_stats(metric_files)\n",
    "create_stats(metric_files, \"Healthy_stats\", \"[0]\")\n",
    "create_stats(metric_files, \"MI_stats\", \"[1]\")\n",
    "create_stats(metric_files, \"STEMI_stats\", \"[1, 2]\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>Name</th>\n",
       "      <th>Exact Matches #</th>\n",
       "      <th>Mean MSE</th>\n",
       "      <th>Max MSE</th>\n",
       "      <th>Min MSE</th>\n",
       "      <th>Mean MAE</th>\n",
       "      <th>Max MAE</th>\n",
       "      <th>Min MAE</th>\n",
       "      <th>Mean Correct R #</th>\n",
       "      <th>Mean Detected R #</th>\n",
       "      <th>Mean TP</th>\n",
       "      <th>Mean FP</th>\n",
       "      <th>Mean Performance</th>\n",
       "      <th>Mean Acc</th>\n",
       "      <th>Mean Err</th>\n",
       "      <th>Mean mean_deviation</th>\n",
       "      <th>Mean std_deviation</th>\n",
       "      <th>Mean mean_ref_ibi</th>\n",
       "      <th>Max mean_ref_ibi</th>\n",
       "      <th>Min mean_ref_ibi</th>\n",
       "      <th>Mean std_ref_ibi</th>\n",
       "      <th>Max std_ref_ibi</th>\n",
       "      <th>Min std_ref_ibi</th>\n",
       "      <th>Mean mean_test_ibi</th>\n",
       "      <th>Max mean_test_ibi</th>\n",
       "      <th>Min mean_test_ibi</th>\n",
       "      <th>Mean std_test_ibi</th>\n",
       "      <th>Max std_test_ibi</th>\n",
       "      <th>Min std_test_ibi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>179</th>\n",
       "      <td>Raw+rodrigues2021</td>\n",
       "      <td>2784</td>\n",
       "      <td>141.090610</td>\n",
       "      <td>27342.700000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.191148</td>\n",
       "      <td>154.300000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>14.672446</td>\n",
       "      <td>12.261731</td>\n",
       "      <td>1.065593</td>\n",
       "      <td>1.011206</td>\n",
       "      <td>0.917021</td>\n",
       "      <td>0.082979</td>\n",
       "      <td>0.006915</td>\n",
       "      <td>0.005875</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.846824</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.315000</td>\n",
       "      <td>0.048230</td>\n",
       "      <td>0.508547</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>127</th>\n",
       "      <td>Neurokit+rodrigues2021</td>\n",
       "      <td>3459</td>\n",
       "      <td>103.214613</td>\n",
       "      <td>17132.090909</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.677087</td>\n",
       "      <td>121.363636</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>13.994496</td>\n",
       "      <td>12.269024</td>\n",
       "      <td>1.025320</td>\n",
       "      <td>1.011110</td>\n",
       "      <td>0.919997</td>\n",
       "      <td>0.080003</td>\n",
       "      <td>0.005951</td>\n",
       "      <td>0.005585</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.846245</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.314667</td>\n",
       "      <td>0.046923</td>\n",
       "      <td>0.508547</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>koka2022+rodrigues2021</td>\n",
       "      <td>2755</td>\n",
       "      <td>143.779462</td>\n",
       "      <td>27342.700000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.235406</td>\n",
       "      <td>154.300000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>14.704325</td>\n",
       "      <td>12.246915</td>\n",
       "      <td>1.073299</td>\n",
       "      <td>1.010098</td>\n",
       "      <td>0.916255</td>\n",
       "      <td>0.083745</td>\n",
       "      <td>0.007074</td>\n",
       "      <td>0.005933</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.846951</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.315000</td>\n",
       "      <td>0.048463</td>\n",
       "      <td>0.509117</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>Biosppy+rodrigues2021</td>\n",
       "      <td>2838</td>\n",
       "      <td>155.639017</td>\n",
       "      <td>17698.692308</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.327074</td>\n",
       "      <td>126.538462</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>14.629375</td>\n",
       "      <td>12.246915</td>\n",
       "      <td>1.075455</td>\n",
       "      <td>1.010028</td>\n",
       "      <td>0.916304</td>\n",
       "      <td>0.083696</td>\n",
       "      <td>0.006856</td>\n",
       "      <td>0.005867</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.846847</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.315000</td>\n",
       "      <td>0.048172</td>\n",
       "      <td>0.528750</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153</th>\n",
       "      <td>pantompkins1985+rodrigues2021</td>\n",
       "      <td>3177</td>\n",
       "      <td>97.591457</td>\n",
       "      <td>18511.153846</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.319023</td>\n",
       "      <td>128.076923</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>13.939636</td>\n",
       "      <td>12.086693</td>\n",
       "      <td>1.045457</td>\n",
       "      <td>0.996455</td>\n",
       "      <td>0.916916</td>\n",
       "      <td>0.083084</td>\n",
       "      <td>0.011586</td>\n",
       "      <td>0.004869</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.847647</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.315000</td>\n",
       "      <td>0.046873</td>\n",
       "      <td>0.509117</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>hamilton2002+rodrigues2021</td>\n",
       "      <td>2734</td>\n",
       "      <td>131.052724</td>\n",
       "      <td>19425.727273</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.782411</td>\n",
       "      <td>128.272727</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>14.047429</td>\n",
       "      <td>11.992019</td>\n",
       "      <td>1.076740</td>\n",
       "      <td>0.989462</td>\n",
       "      <td>0.914056</td>\n",
       "      <td>0.085944</td>\n",
       "      <td>0.013236</td>\n",
       "      <td>0.005646</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.848983</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.310000</td>\n",
       "      <td>0.048679</td>\n",
       "      <td>0.664680</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>Neurokit+promac_optimized</td>\n",
       "      <td>13670</td>\n",
       "      <td>146.091971</td>\n",
       "      <td>19187.142857</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.166964</td>\n",
       "      <td>136.285714</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>12.542911</td>\n",
       "      <td>11.927526</td>\n",
       "      <td>0.304344</td>\n",
       "      <td>0.981463</td>\n",
       "      <td>0.972951</td>\n",
       "      <td>0.027049</td>\n",
       "      <td>0.004016</td>\n",
       "      <td>0.003559</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.846713</td>\n",
       "      <td>1.486667</td>\n",
       "      <td>0.328966</td>\n",
       "      <td>0.046408</td>\n",
       "      <td>0.514908</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>178</th>\n",
       "      <td>Raw+rodrigues2021+Kubios</td>\n",
       "      <td>4311</td>\n",
       "      <td>89.029765</td>\n",
       "      <td>27342.700000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.752162</td>\n",
       "      <td>154.300000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>14.429797</td>\n",
       "      <td>11.897023</td>\n",
       "      <td>1.101601</td>\n",
       "      <td>0.980355</td>\n",
       "      <td>0.912239</td>\n",
       "      <td>0.087761</td>\n",
       "      <td>0.006819</td>\n",
       "      <td>0.005711</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.847922</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.315000</td>\n",
       "      <td>0.049401</td>\n",
       "      <td>0.700036</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100</th>\n",
       "      <td>koka2022+rodrigues2021+Kubios</td>\n",
       "      <td>4274</td>\n",
       "      <td>91.860287</td>\n",
       "      <td>27342.700000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.787754</td>\n",
       "      <td>154.300000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>14.463236</td>\n",
       "      <td>11.880694</td>\n",
       "      <td>1.110087</td>\n",
       "      <td>0.979097</td>\n",
       "      <td>0.911391</td>\n",
       "      <td>0.088609</td>\n",
       "      <td>0.006976</td>\n",
       "      <td>0.005764</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.848002</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.315000</td>\n",
       "      <td>0.049633</td>\n",
       "      <td>0.700036</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>Biosppy+rodrigues2021+Kubios</td>\n",
       "      <td>4435</td>\n",
       "      <td>102.639533</td>\n",
       "      <td>17698.692308</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.874136</td>\n",
       "      <td>126.538462</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>14.385303</td>\n",
       "      <td>11.880464</td>\n",
       "      <td>1.112518</td>\n",
       "      <td>0.979057</td>\n",
       "      <td>0.911392</td>\n",
       "      <td>0.088608</td>\n",
       "      <td>0.006760</td>\n",
       "      <td>0.005696</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.847859</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.315000</td>\n",
       "      <td>0.049112</td>\n",
       "      <td>0.700036</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>126</th>\n",
       "      <td>Neurokit+rodrigues2021+Kubios</td>\n",
       "      <td>5246</td>\n",
       "      <td>68.304938</td>\n",
       "      <td>15688.214286</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.460363</td>\n",
       "      <td>113.727273</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>13.724646</td>\n",
       "      <td>11.876841</td>\n",
       "      <td>1.071052</td>\n",
       "      <td>0.977643</td>\n",
       "      <td>0.914327</td>\n",
       "      <td>0.085673</td>\n",
       "      <td>0.005819</td>\n",
       "      <td>0.005324</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.847081</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.314667</td>\n",
       "      <td>0.047426</td>\n",
       "      <td>0.707107</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>Biosppy+promac</td>\n",
       "      <td>13789</td>\n",
       "      <td>403.972345</td>\n",
       "      <td>17522.714286</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.266168</td>\n",
       "      <td>132.142857</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>12.482088</td>\n",
       "      <td>11.879088</td>\n",
       "      <td>0.321728</td>\n",
       "      <td>0.977247</td>\n",
       "      <td>0.971563</td>\n",
       "      <td>0.028437</td>\n",
       "      <td>0.006133</td>\n",
       "      <td>0.006092</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.847152</td>\n",
       "      <td>1.485000</td>\n",
       "      <td>0.355185</td>\n",
       "      <td>0.047863</td>\n",
       "      <td>0.565000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>koka2022+promac</td>\n",
       "      <td>13687</td>\n",
       "      <td>421.732135</td>\n",
       "      <td>19603.428571</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.446375</td>\n",
       "      <td>140.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>12.463648</td>\n",
       "      <td>11.875052</td>\n",
       "      <td>0.324022</td>\n",
       "      <td>0.976969</td>\n",
       "      <td>0.971435</td>\n",
       "      <td>0.028565</td>\n",
       "      <td>0.006160</td>\n",
       "      <td>0.006044</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.847131</td>\n",
       "      <td>1.485000</td>\n",
       "      <td>0.354231</td>\n",
       "      <td>0.047838</td>\n",
       "      <td>0.563353</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>177</th>\n",
       "      <td>Raw+promac_optimized</td>\n",
       "      <td>12942</td>\n",
       "      <td>171.981339</td>\n",
       "      <td>15734.428571</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.521216</td>\n",
       "      <td>125.285714</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>12.590936</td>\n",
       "      <td>11.867712</td>\n",
       "      <td>0.317233</td>\n",
       "      <td>0.976529</td>\n",
       "      <td>0.971649</td>\n",
       "      <td>0.028351</td>\n",
       "      <td>0.003889</td>\n",
       "      <td>0.003550</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.847005</td>\n",
       "      <td>1.487500</td>\n",
       "      <td>0.328966</td>\n",
       "      <td>0.047070</td>\n",
       "      <td>0.530330</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>Biosppy+promac_optimized</td>\n",
       "      <td>13191</td>\n",
       "      <td>157.063510</td>\n",
       "      <td>18409.428571</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.340067</td>\n",
       "      <td>133.142857</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>12.579973</td>\n",
       "      <td>11.857713</td>\n",
       "      <td>0.323334</td>\n",
       "      <td>0.975454</td>\n",
       "      <td>0.971043</td>\n",
       "      <td>0.028957</td>\n",
       "      <td>0.003881</td>\n",
       "      <td>0.003502</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.846910</td>\n",
       "      <td>1.487500</td>\n",
       "      <td>0.328966</td>\n",
       "      <td>0.047026</td>\n",
       "      <td>0.512182</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>koka2022+promac_optimized</td>\n",
       "      <td>13028</td>\n",
       "      <td>174.564941</td>\n",
       "      <td>18409.428571</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.523201</td>\n",
       "      <td>133.142857</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>12.590523</td>\n",
       "      <td>11.854273</td>\n",
       "      <td>0.323426</td>\n",
       "      <td>0.975209</td>\n",
       "      <td>0.970946</td>\n",
       "      <td>0.029054</td>\n",
       "      <td>0.003888</td>\n",
       "      <td>0.003511</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.847053</td>\n",
       "      <td>1.487500</td>\n",
       "      <td>0.337500</td>\n",
       "      <td>0.047264</td>\n",
       "      <td>0.530330</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>150</th>\n",
       "      <td>pantompkins1985+promac</td>\n",
       "      <td>13235</td>\n",
       "      <td>396.764501</td>\n",
       "      <td>35679.904762</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.659058</td>\n",
       "      <td>170.285714</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>12.818449</td>\n",
       "      <td>11.830421</td>\n",
       "      <td>0.341269</td>\n",
       "      <td>0.974699</td>\n",
       "      <td>0.970222</td>\n",
       "      <td>0.029778</td>\n",
       "      <td>0.011516</td>\n",
       "      <td>0.005481</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.848738</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.353889</td>\n",
       "      <td>0.048861</td>\n",
       "      <td>0.671751</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>123</th>\n",
       "      <td>Neurokit+promac+Kubios_optimized</td>\n",
       "      <td>14908</td>\n",
       "      <td>271.673086</td>\n",
       "      <td>19187.142857</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.718061</td>\n",
       "      <td>136.285714</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>12.465942</td>\n",
       "      <td>11.840695</td>\n",
       "      <td>0.284803</td>\n",
       "      <td>0.974138</td>\n",
       "      <td>0.974355</td>\n",
       "      <td>0.025645</td>\n",
       "      <td>0.004036</td>\n",
       "      <td>0.003582</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.846921</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.328966</td>\n",
       "      <td>0.045886</td>\n",
       "      <td>0.643467</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>151</th>\n",
       "      <td>pantompkins1985+promac_optimized</td>\n",
       "      <td>14306</td>\n",
       "      <td>172.425381</td>\n",
       "      <td>17671.857143</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.667153</td>\n",
       "      <td>132.714286</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>12.477180</td>\n",
       "      <td>11.820100</td>\n",
       "      <td>0.296317</td>\n",
       "      <td>0.974003</td>\n",
       "      <td>0.973206</td>\n",
       "      <td>0.026794</td>\n",
       "      <td>0.005770</td>\n",
       "      <td>0.004476</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.848662</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.329286</td>\n",
       "      <td>0.046919</td>\n",
       "      <td>0.586032</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>176</th>\n",
       "      <td>Raw+promac</td>\n",
       "      <td>12956</td>\n",
       "      <td>384.722607</td>\n",
       "      <td>28914.904762</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.215402</td>\n",
       "      <td>149.952381</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>12.744782</td>\n",
       "      <td>11.795239</td>\n",
       "      <td>0.368194</td>\n",
       "      <td>0.969822</td>\n",
       "      <td>0.967128</td>\n",
       "      <td>0.032872</td>\n",
       "      <td>0.006884</td>\n",
       "      <td>0.006464</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.847439</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.361923</td>\n",
       "      <td>0.048559</td>\n",
       "      <td>0.525484</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 Name  Exact Matches #    Mean MSE  \\\n",
       "179                 Raw+rodrigues2021             2784  141.090610   \n",
       "127            Neurokit+rodrigues2021             3459  103.214613   \n",
       "101            koka2022+rodrigues2021             2755  143.779462   \n",
       "23              Biosppy+rodrigues2021             2838  155.639017   \n",
       "153     pantompkins1985+rodrigues2021             3177   97.591457   \n",
       "75         hamilton2002+rodrigues2021             2734  131.052724   \n",
       "125         Neurokit+promac_optimized            13670  146.091971   \n",
       "178          Raw+rodrigues2021+Kubios             4311   89.029765   \n",
       "100     koka2022+rodrigues2021+Kubios             4274   91.860287   \n",
       "22       Biosppy+rodrigues2021+Kubios             4435  102.639533   \n",
       "126     Neurokit+rodrigues2021+Kubios             5246   68.304938   \n",
       "20                     Biosppy+promac            13789  403.972345   \n",
       "98                    koka2022+promac            13687  421.732135   \n",
       "177              Raw+promac_optimized            12942  171.981339   \n",
       "21           Biosppy+promac_optimized            13191  157.063510   \n",
       "99          koka2022+promac_optimized            13028  174.564941   \n",
       "150            pantompkins1985+promac            13235  396.764501   \n",
       "123  Neurokit+promac+Kubios_optimized            14908  271.673086   \n",
       "151  pantompkins1985+promac_optimized            14306  172.425381   \n",
       "176                        Raw+promac            12956  384.722607   \n",
       "\n",
       "          Max MSE  Min MSE  Mean MAE     Max MAE  Min MAE  Mean Correct R #  \\\n",
       "179  27342.700000      0.0  2.191148  154.300000      0.0           12.1588   \n",
       "127  17132.090909      0.0  1.677087  121.363636      0.0           12.1588   \n",
       "101  27342.700000      0.0  2.235406  154.300000      0.0           12.1588   \n",
       "23   17698.692308      0.0  2.327074  126.538462      0.0           12.1588   \n",
       "153  18511.153846      0.0  2.319023  128.076923      0.0           12.1588   \n",
       "75   19425.727273      0.0  2.782411  128.272727      0.0           12.1588   \n",
       "125  19187.142857      0.0  2.166964  136.285714      0.0           12.1588   \n",
       "178  27342.700000      0.0  1.752162  154.300000      0.0           12.1588   \n",
       "100  27342.700000      0.0  1.787754  154.300000      0.0           12.1588   \n",
       "22   17698.692308      0.0  1.874136  126.538462      0.0           12.1588   \n",
       "126  15688.214286      0.0  1.460363  113.727273      0.0           12.1588   \n",
       "20   17522.714286      0.0  5.266168  132.142857      0.0           12.1588   \n",
       "98   19603.428571      0.0  5.446375  140.000000      0.0           12.1588   \n",
       "177  15734.428571      0.0  2.521216  125.285714      0.0           12.1588   \n",
       "21   18409.428571      0.0  2.340067  133.142857      0.0           12.1588   \n",
       "99   18409.428571      0.0  2.523201  133.142857      0.0           12.1588   \n",
       "150  35679.904762      0.0  5.659058  170.285714      0.0           12.1588   \n",
       "123  19187.142857      0.0  3.718061  136.285714      0.0           12.1588   \n",
       "151  17671.857143      0.0  2.667153  132.714286      0.0           12.1588   \n",
       "176  28914.904762      0.0  5.215402  149.952381      0.0           12.1588   \n",
       "\n",
       "     Mean Detected R #    Mean TP   Mean FP  Mean Performance  Mean Acc  \\\n",
       "179          14.672446  12.261731  1.065593          1.011206  0.917021   \n",
       "127          13.994496  12.269024  1.025320          1.011110  0.919997   \n",
       "101          14.704325  12.246915  1.073299          1.010098  0.916255   \n",
       "23           14.629375  12.246915  1.075455          1.010028  0.916304   \n",
       "153          13.939636  12.086693  1.045457          0.996455  0.916916   \n",
       "75           14.047429  11.992019  1.076740          0.989462  0.914056   \n",
       "125          12.542911  11.927526  0.304344          0.981463  0.972951   \n",
       "178          14.429797  11.897023  1.101601          0.980355  0.912239   \n",
       "100          14.463236  11.880694  1.110087          0.979097  0.911391   \n",
       "22           14.385303  11.880464  1.112518          0.979057  0.911392   \n",
       "126          13.724646  11.876841  1.071052          0.977643  0.914327   \n",
       "20           12.482088  11.879088  0.321728          0.977247  0.971563   \n",
       "98           12.463648  11.875052  0.324022          0.976969  0.971435   \n",
       "177          12.590936  11.867712  0.317233          0.976529  0.971649   \n",
       "21           12.579973  11.857713  0.323334          0.975454  0.971043   \n",
       "99           12.590523  11.854273  0.323426          0.975209  0.970946   \n",
       "150          12.818449  11.830421  0.341269          0.974699  0.970222   \n",
       "123          12.465942  11.840695  0.284803          0.974138  0.974355   \n",
       "151          12.477180  11.820100  0.296317          0.974003  0.973206   \n",
       "176          12.744782  11.795239  0.368194          0.969822  0.967128   \n",
       "\n",
       "     Mean Err  Mean mean_deviation  Mean std_deviation  Mean mean_ref_ibi  \\\n",
       "179  0.082979             0.006915            0.005875           0.845279   \n",
       "127  0.080003             0.005951            0.005585           0.845279   \n",
       "101  0.083745             0.007074            0.005933           0.845279   \n",
       "23   0.083696             0.006856            0.005867           0.845279   \n",
       "153  0.083084             0.011586            0.004869           0.845279   \n",
       "75   0.085944             0.013236            0.005646           0.845279   \n",
       "125  0.027049             0.004016            0.003559           0.845279   \n",
       "178  0.087761             0.006819            0.005711           0.845279   \n",
       "100  0.088609             0.006976            0.005764           0.845279   \n",
       "22   0.088608             0.006760            0.005696           0.845279   \n",
       "126  0.085673             0.005819            0.005324           0.845279   \n",
       "20   0.028437             0.006133            0.006092           0.845279   \n",
       "98   0.028565             0.006160            0.006044           0.845279   \n",
       "177  0.028351             0.003889            0.003550           0.845279   \n",
       "21   0.028957             0.003881            0.003502           0.845279   \n",
       "99   0.029054             0.003888            0.003511           0.845279   \n",
       "150  0.029778             0.011516            0.005481           0.845279   \n",
       "123  0.025645             0.004036            0.003582           0.845279   \n",
       "151  0.026794             0.005770            0.004476           0.845279   \n",
       "176  0.032872             0.006884            0.006464           0.845279   \n",
       "\n",
       "     Max mean_ref_ibi  Min mean_ref_ibi  Mean std_ref_ibi  Max std_ref_ibi  \\\n",
       "179              1.49              0.31          0.045796         0.526318   \n",
       "127              1.49              0.31          0.045796         0.526318   \n",
       "101              1.49              0.31          0.045796         0.526318   \n",
       "23               1.49              0.31          0.045796         0.526318   \n",
       "153              1.49              0.31          0.045796         0.526318   \n",
       "75               1.49              0.31          0.045796         0.526318   \n",
       "125              1.49              0.31          0.045796         0.526318   \n",
       "178              1.49              0.31          0.045796         0.526318   \n",
       "100              1.49              0.31          0.045796         0.526318   \n",
       "22               1.49              0.31          0.045796         0.526318   \n",
       "126              1.49              0.31          0.045796         0.526318   \n",
       "20               1.49              0.31          0.045796         0.526318   \n",
       "98               1.49              0.31          0.045796         0.526318   \n",
       "177              1.49              0.31          0.045796         0.526318   \n",
       "21               1.49              0.31          0.045796         0.526318   \n",
       "99               1.49              0.31          0.045796         0.526318   \n",
       "150              1.49              0.31          0.045796         0.526318   \n",
       "123              1.49              0.31          0.045796         0.526318   \n",
       "151              1.49              0.31          0.045796         0.526318   \n",
       "176              1.49              0.31          0.045796         0.526318   \n",
       "\n",
       "     Min std_ref_ibi  Mean mean_test_ibi  Max mean_test_ibi  \\\n",
       "179              0.0            0.846824           1.490000   \n",
       "127              0.0            0.846245           1.490000   \n",
       "101              0.0            0.846951           1.490000   \n",
       "23               0.0            0.846847           1.490000   \n",
       "153              0.0            0.847647           1.490000   \n",
       "75               0.0            0.848983           1.490000   \n",
       "125              0.0            0.846713           1.486667   \n",
       "178              0.0            0.847922           1.490000   \n",
       "100              0.0            0.848002           1.490000   \n",
       "22               0.0            0.847859           1.490000   \n",
       "126              0.0            0.847081           1.490000   \n",
       "20               0.0            0.847152           1.485000   \n",
       "98               0.0            0.847131           1.485000   \n",
       "177              0.0            0.847005           1.487500   \n",
       "21               0.0            0.846910           1.487500   \n",
       "99               0.0            0.847053           1.487500   \n",
       "150              0.0            0.848738           1.490000   \n",
       "123              0.0            0.846921           1.490000   \n",
       "151              0.0            0.848662           1.490000   \n",
       "176              0.0            0.847439           1.490000   \n",
       "\n",
       "     Min mean_test_ibi  Mean std_test_ibi  Max std_test_ibi  Min std_test_ibi  \n",
       "179           0.315000           0.048230          0.508547               0.0  \n",
       "127           0.314667           0.046923          0.508547               0.0  \n",
       "101           0.315000           0.048463          0.509117               0.0  \n",
       "23            0.315000           0.048172          0.528750               0.0  \n",
       "153           0.315000           0.046873          0.509117               0.0  \n",
       "75            0.310000           0.048679          0.664680               0.0  \n",
       "125           0.328966           0.046408          0.514908               0.0  \n",
       "178           0.315000           0.049401          0.700036               0.0  \n",
       "100           0.315000           0.049633          0.700036               0.0  \n",
       "22            0.315000           0.049112          0.700036               0.0  \n",
       "126           0.314667           0.047426          0.707107               0.0  \n",
       "20            0.355185           0.047863          0.565000               0.0  \n",
       "98            0.354231           0.047838          0.563353               0.0  \n",
       "177           0.328966           0.047070          0.530330               0.0  \n",
       "21            0.328966           0.047026          0.512182               0.0  \n",
       "99            0.337500           0.047264          0.530330               0.0  \n",
       "150           0.353889           0.048861          0.671751               0.0  \n",
       "123           0.328966           0.045886          0.643467               0.0  \n",
       "151           0.329286           0.046919          0.586032               0.0  \n",
       "176           0.361923           0.048559          0.525484               0.0  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_stats = pd.read_csv(os.path.join(\"./comparison_stats\", \"stats.csv\"))\n",
    "# df_stats.sort_values(by=['Mean Performance'], ascending=False)[[\"Name\", 'Mean Performance', \n",
    "#                    'Mean Acc', 'Mean Err',\n",
    "#                    'Mean Correct R #', 'Mean Detected R #','Mean TP', 'Mean FP']].iloc[:10]\n",
    "df_stats.sort_values(by=['Mean Performance'], ascending=False).iloc[:20]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>Name</th>\n",
       "      <th>Exact Matches #</th>\n",
       "      <th>Mean MSE</th>\n",
       "      <th>Max MSE</th>\n",
       "      <th>Min MSE</th>\n",
       "      <th>Mean MAE</th>\n",
       "      <th>Max MAE</th>\n",
       "      <th>Min MAE</th>\n",
       "      <th>Mean Correct R #</th>\n",
       "      <th>Mean Detected R #</th>\n",
       "      <th>Mean TP</th>\n",
       "      <th>Mean FP</th>\n",
       "      <th>Mean Performance</th>\n",
       "      <th>Mean Acc</th>\n",
       "      <th>Mean Err</th>\n",
       "      <th>Mean mean_deviation</th>\n",
       "      <th>Mean std_deviation</th>\n",
       "      <th>Mean mean_ref_ibi</th>\n",
       "      <th>Max mean_ref_ibi</th>\n",
       "      <th>Min mean_ref_ibi</th>\n",
       "      <th>Mean std_ref_ibi</th>\n",
       "      <th>Max std_ref_ibi</th>\n",
       "      <th>Min std_ref_ibi</th>\n",
       "      <th>Mean mean_test_ibi</th>\n",
       "      <th>Max mean_test_ibi</th>\n",
       "      <th>Min mean_test_ibi</th>\n",
       "      <th>Mean std_test_ibi</th>\n",
       "      <th>Max std_test_ibi</th>\n",
       "      <th>Min std_test_ibi</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>119</th>\n",
       "      <td>Neurokit+Neurokit</td>\n",
       "      <td>13690</td>\n",
       "      <td>814.611749</td>\n",
       "      <td>19542.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9.831041</td>\n",
       "      <td>139.714286</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.935416</td>\n",
       "      <td>11.658273</td>\n",
       "      <td>0.039723</td>\n",
       "      <td>0.960202</td>\n",
       "      <td>0.995934</td>\n",
       "      <td>0.004020</td>\n",
       "      <td>0.003946</td>\n",
       "      <td>0.003433</td>\n",
       "      <td>0.845254</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.846806</td>\n",
       "      <td>1.486667</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.046116</td>\n",
       "      <td>0.516188</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>Neurokit+Neurokit+Kubios</td>\n",
       "      <td>13725</td>\n",
       "      <td>808.714707</td>\n",
       "      <td>19542.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9.890196</td>\n",
       "      <td>139.714286</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.937939</td>\n",
       "      <td>11.552681</td>\n",
       "      <td>0.040503</td>\n",
       "      <td>0.951823</td>\n",
       "      <td>0.995860</td>\n",
       "      <td>0.004094</td>\n",
       "      <td>0.004139</td>\n",
       "      <td>0.003714</td>\n",
       "      <td>0.845254</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.847036</td>\n",
       "      <td>1.486667</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.045106</td>\n",
       "      <td>0.671751</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>171</th>\n",
       "      <td>Raw+Neurokit</td>\n",
       "      <td>13468</td>\n",
       "      <td>839.724399</td>\n",
       "      <td>19579.571429</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10.133815</td>\n",
       "      <td>139.857143</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.952204</td>\n",
       "      <td>11.657263</td>\n",
       "      <td>0.044402</td>\n",
       "      <td>0.959954</td>\n",
       "      <td>0.995665</td>\n",
       "      <td>0.004289</td>\n",
       "      <td>0.003817</td>\n",
       "      <td>0.003438</td>\n",
       "      <td>0.845243</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.846565</td>\n",
       "      <td>1.487500</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.046059</td>\n",
       "      <td>0.514908</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>170</th>\n",
       "      <td>Raw+Neurokit+Kubios</td>\n",
       "      <td>13560</td>\n",
       "      <td>832.627188</td>\n",
       "      <td>19579.571429</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10.161941</td>\n",
       "      <td>139.857143</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.949635</td>\n",
       "      <td>11.552360</td>\n",
       "      <td>0.044585</td>\n",
       "      <td>0.951686</td>\n",
       "      <td>0.995619</td>\n",
       "      <td>0.004335</td>\n",
       "      <td>0.004015</td>\n",
       "      <td>0.003723</td>\n",
       "      <td>0.845243</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.846756</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.045016</td>\n",
       "      <td>0.671751</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>Biosppy+Neurokit</td>\n",
       "      <td>13212</td>\n",
       "      <td>879.663023</td>\n",
       "      <td>16874.142857</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10.633371</td>\n",
       "      <td>129.857143</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.984725</td>\n",
       "      <td>11.647493</td>\n",
       "      <td>0.050319</td>\n",
       "      <td>0.959062</td>\n",
       "      <td>0.995015</td>\n",
       "      <td>0.004985</td>\n",
       "      <td>0.003851</td>\n",
       "      <td>0.003440</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.846493</td>\n",
       "      <td>1.487500</td>\n",
       "      <td>0.314828</td>\n",
       "      <td>0.046229</td>\n",
       "      <td>0.512182</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93</th>\n",
       "      <td>koka2022+Neurokit</td>\n",
       "      <td>13201</td>\n",
       "      <td>877.386602</td>\n",
       "      <td>16874.142857</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10.623835</td>\n",
       "      <td>129.857143</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.984404</td>\n",
       "      <td>11.645704</td>\n",
       "      <td>0.050273</td>\n",
       "      <td>0.958941</td>\n",
       "      <td>0.994989</td>\n",
       "      <td>0.005011</td>\n",
       "      <td>0.003847</td>\n",
       "      <td>0.003448</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.846543</td>\n",
       "      <td>1.487500</td>\n",
       "      <td>0.314828</td>\n",
       "      <td>0.046239</td>\n",
       "      <td>0.512182</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>Biosppy+Neurokit+Kubios</td>\n",
       "      <td>13379</td>\n",
       "      <td>863.837598</td>\n",
       "      <td>16874.142857</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10.553293</td>\n",
       "      <td>129.857143</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.974130</td>\n",
       "      <td>11.543874</td>\n",
       "      <td>0.050686</td>\n",
       "      <td>0.950952</td>\n",
       "      <td>0.994949</td>\n",
       "      <td>0.005051</td>\n",
       "      <td>0.004043</td>\n",
       "      <td>0.003723</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.846697</td>\n",
       "      <td>1.485000</td>\n",
       "      <td>0.314828</td>\n",
       "      <td>0.045324</td>\n",
       "      <td>0.671751</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>koka2022+Neurokit+Kubios</td>\n",
       "      <td>13371</td>\n",
       "      <td>858.943193</td>\n",
       "      <td>16874.142857</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10.515443</td>\n",
       "      <td>129.857143</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.972983</td>\n",
       "      <td>11.542085</td>\n",
       "      <td>0.050732</td>\n",
       "      <td>0.950821</td>\n",
       "      <td>0.994919</td>\n",
       "      <td>0.005081</td>\n",
       "      <td>0.004041</td>\n",
       "      <td>0.003736</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.846677</td>\n",
       "      <td>1.485000</td>\n",
       "      <td>0.314828</td>\n",
       "      <td>0.045282</td>\n",
       "      <td>0.671751</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>pantompkins1985+Neurokit</td>\n",
       "      <td>13806</td>\n",
       "      <td>744.240361</td>\n",
       "      <td>19684.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9.077351</td>\n",
       "      <td>140.285714</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.907435</td>\n",
       "      <td>11.622999</td>\n",
       "      <td>0.051741</td>\n",
       "      <td>0.957870</td>\n",
       "      <td>0.994851</td>\n",
       "      <td>0.005104</td>\n",
       "      <td>0.005768</td>\n",
       "      <td>0.004433</td>\n",
       "      <td>0.845261</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.847708</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.046607</td>\n",
       "      <td>0.537401</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>144</th>\n",
       "      <td>pantompkins1985+Neurokit+Kubios</td>\n",
       "      <td>13863</td>\n",
       "      <td>733.151763</td>\n",
       "      <td>19684.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>9.056879</td>\n",
       "      <td>140.285714</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.907206</td>\n",
       "      <td>11.524609</td>\n",
       "      <td>0.052108</td>\n",
       "      <td>0.950066</td>\n",
       "      <td>0.994821</td>\n",
       "      <td>0.005133</td>\n",
       "      <td>0.005870</td>\n",
       "      <td>0.004623</td>\n",
       "      <td>0.845261</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.847886</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.045712</td>\n",
       "      <td>0.657609</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>Biosppy+nabian2018</td>\n",
       "      <td>7503</td>\n",
       "      <td>895.013980</td>\n",
       "      <td>17712.285714</td>\n",
       "      <td>0.0</td>\n",
       "      <td>11.062525</td>\n",
       "      <td>131.714286</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.310674</td>\n",
       "      <td>11.060869</td>\n",
       "      <td>0.053576</td>\n",
       "      <td>0.916235</td>\n",
       "      <td>0.994401</td>\n",
       "      <td>0.005599</td>\n",
       "      <td>0.012314</td>\n",
       "      <td>0.003324</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.851168</td>\n",
       "      <td>1.487500</td>\n",
       "      <td>0.422381</td>\n",
       "      <td>0.047942</td>\n",
       "      <td>0.671751</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>koka2022+nabian2018</td>\n",
       "      <td>7479</td>\n",
       "      <td>917.231818</td>\n",
       "      <td>18123.428571</td>\n",
       "      <td>0.0</td>\n",
       "      <td>11.308966</td>\n",
       "      <td>128.857143</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.313105</td>\n",
       "      <td>11.062428</td>\n",
       "      <td>0.053713</td>\n",
       "      <td>0.916358</td>\n",
       "      <td>0.994362</td>\n",
       "      <td>0.005638</td>\n",
       "      <td>0.012277</td>\n",
       "      <td>0.003336</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.851106</td>\n",
       "      <td>1.487500</td>\n",
       "      <td>0.422381</td>\n",
       "      <td>0.047880</td>\n",
       "      <td>0.643467</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Biosppy+nabian2018+Kubios</td>\n",
       "      <td>7508</td>\n",
       "      <td>827.289132</td>\n",
       "      <td>21440.500000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10.469002</td>\n",
       "      <td>131.714286</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.305536</td>\n",
       "      <td>10.971332</td>\n",
       "      <td>0.055319</td>\n",
       "      <td>0.909171</td>\n",
       "      <td>0.994299</td>\n",
       "      <td>0.005701</td>\n",
       "      <td>0.012437</td>\n",
       "      <td>0.003564</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.851312</td>\n",
       "      <td>1.487500</td>\n",
       "      <td>0.415000</td>\n",
       "      <td>0.047188</td>\n",
       "      <td>0.692965</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>90</th>\n",
       "      <td>koka2022+nabian2018+Kubios</td>\n",
       "      <td>7514</td>\n",
       "      <td>835.476732</td>\n",
       "      <td>21440.500000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10.543791</td>\n",
       "      <td>131.500000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.304527</td>\n",
       "      <td>10.972203</td>\n",
       "      <td>0.055135</td>\n",
       "      <td>0.909239</td>\n",
       "      <td>0.994253</td>\n",
       "      <td>0.005747</td>\n",
       "      <td>0.012405</td>\n",
       "      <td>0.003578</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.851321</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.415000</td>\n",
       "      <td>0.047121</td>\n",
       "      <td>0.692965</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>Raw+nabian2018</td>\n",
       "      <td>7474</td>\n",
       "      <td>818.258648</td>\n",
       "      <td>21736.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10.398177</td>\n",
       "      <td>147.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.262052</td>\n",
       "      <td>10.892987</td>\n",
       "      <td>0.130132</td>\n",
       "      <td>0.902792</td>\n",
       "      <td>0.986406</td>\n",
       "      <td>0.013594</td>\n",
       "      <td>0.012204</td>\n",
       "      <td>0.003274</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.852592</td>\n",
       "      <td>1.487500</td>\n",
       "      <td>0.422381</td>\n",
       "      <td>0.048274</td>\n",
       "      <td>0.530330</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>168</th>\n",
       "      <td>Raw+nabian2018+Kubios</td>\n",
       "      <td>7489</td>\n",
       "      <td>794.762164</td>\n",
       "      <td>21531.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10.253828</td>\n",
       "      <td>146.666667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.267144</td>\n",
       "      <td>10.805422</td>\n",
       "      <td>0.130453</td>\n",
       "      <td>0.895784</td>\n",
       "      <td>0.986327</td>\n",
       "      <td>0.013673</td>\n",
       "      <td>0.012333</td>\n",
       "      <td>0.003525</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.852635</td>\n",
       "      <td>1.487500</td>\n",
       "      <td>0.416316</td>\n",
       "      <td>0.047208</td>\n",
       "      <td>0.537401</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>143</th>\n",
       "      <td>pantompkins1985+nabian2018</td>\n",
       "      <td>7552</td>\n",
       "      <td>898.636498</td>\n",
       "      <td>18095.181818</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10.543664</td>\n",
       "      <td>129.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.284712</td>\n",
       "      <td>10.899271</td>\n",
       "      <td>0.137104</td>\n",
       "      <td>0.904049</td>\n",
       "      <td>0.986257</td>\n",
       "      <td>0.013743</td>\n",
       "      <td>0.007049</td>\n",
       "      <td>0.004398</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.852966</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.420000</td>\n",
       "      <td>0.047970</td>\n",
       "      <td>0.629325</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>142</th>\n",
       "      <td>pantompkins1985+nabian2018+Kubios</td>\n",
       "      <td>7559</td>\n",
       "      <td>847.820419</td>\n",
       "      <td>18095.181818</td>\n",
       "      <td>0.0</td>\n",
       "      <td>10.125915</td>\n",
       "      <td>129.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.283932</td>\n",
       "      <td>10.815559</td>\n",
       "      <td>0.136003</td>\n",
       "      <td>0.897285</td>\n",
       "      <td>0.986181</td>\n",
       "      <td>0.013819</td>\n",
       "      <td>0.007219</td>\n",
       "      <td>0.004587</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.852822</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.420000</td>\n",
       "      <td>0.046766</td>\n",
       "      <td>0.629325</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>hamilton2002+Neurokit+Kubios</td>\n",
       "      <td>13904</td>\n",
       "      <td>702.641796</td>\n",
       "      <td>21972.833333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8.776168</td>\n",
       "      <td>148.166667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.916976</td>\n",
       "      <td>11.224944</td>\n",
       "      <td>0.191551</td>\n",
       "      <td>0.926722</td>\n",
       "      <td>0.980925</td>\n",
       "      <td>0.019075</td>\n",
       "      <td>0.005180</td>\n",
       "      <td>0.004635</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.850549</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.314828</td>\n",
       "      <td>0.046959</td>\n",
       "      <td>0.551543</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>hamilton2002+Neurokit</td>\n",
       "      <td>13754</td>\n",
       "      <td>715.543581</td>\n",
       "      <td>22128.333333</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8.833694</td>\n",
       "      <td>148.666667</td>\n",
       "      <td>0.0</td>\n",
       "      <td>12.1588</td>\n",
       "      <td>11.925141</td>\n",
       "      <td>11.308013</td>\n",
       "      <td>0.195358</td>\n",
       "      <td>0.933289</td>\n",
       "      <td>0.980816</td>\n",
       "      <td>0.019184</td>\n",
       "      <td>0.005017</td>\n",
       "      <td>0.004396</td>\n",
       "      <td>0.845279</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.045796</td>\n",
       "      <td>0.526318</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.850845</td>\n",
       "      <td>1.490000</td>\n",
       "      <td>0.314828</td>\n",
       "      <td>0.048209</td>\n",
       "      <td>0.544472</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  Name  Exact Matches #    Mean MSE  \\\n",
       "119                  Neurokit+Neurokit            13690  814.611749   \n",
       "118           Neurokit+Neurokit+Kubios            13725  808.714707   \n",
       "171                       Raw+Neurokit            13468  839.724399   \n",
       "170                Raw+Neurokit+Kubios            13560  832.627188   \n",
       "15                    Biosppy+Neurokit            13212  879.663023   \n",
       "93                   koka2022+Neurokit            13201  877.386602   \n",
       "14             Biosppy+Neurokit+Kubios            13379  863.837598   \n",
       "92            koka2022+Neurokit+Kubios            13371  858.943193   \n",
       "145           pantompkins1985+Neurokit            13806  744.240361   \n",
       "144    pantompkins1985+Neurokit+Kubios            13863  733.151763   \n",
       "13                  Biosppy+nabian2018             7503  895.013980   \n",
       "91                 koka2022+nabian2018             7479  917.231818   \n",
       "12           Biosppy+nabian2018+Kubios             7508  827.289132   \n",
       "90          koka2022+nabian2018+Kubios             7514  835.476732   \n",
       "169                     Raw+nabian2018             7474  818.258648   \n",
       "168              Raw+nabian2018+Kubios             7489  794.762164   \n",
       "143         pantompkins1985+nabian2018             7552  898.636498   \n",
       "142  pantompkins1985+nabian2018+Kubios             7559  847.820419   \n",
       "66        hamilton2002+Neurokit+Kubios            13904  702.641796   \n",
       "67               hamilton2002+Neurokit            13754  715.543581   \n",
       "\n",
       "          Max MSE  Min MSE   Mean MAE     Max MAE  Min MAE  Mean Correct R #  \\\n",
       "119  19542.000000      0.0   9.831041  139.714286      0.0           12.1588   \n",
       "118  19542.000000      0.0   9.890196  139.714286      0.0           12.1588   \n",
       "171  19579.571429      0.0  10.133815  139.857143      0.0           12.1588   \n",
       "170  19579.571429      0.0  10.161941  139.857143      0.0           12.1588   \n",
       "15   16874.142857      0.0  10.633371  129.857143      0.0           12.1588   \n",
       "93   16874.142857      0.0  10.623835  129.857143      0.0           12.1588   \n",
       "14   16874.142857      0.0  10.553293  129.857143      0.0           12.1588   \n",
       "92   16874.142857      0.0  10.515443  129.857143      0.0           12.1588   \n",
       "145  19684.000000      0.0   9.077351  140.285714      0.0           12.1588   \n",
       "144  19684.000000      0.0   9.056879  140.285714      0.0           12.1588   \n",
       "13   17712.285714      0.0  11.062525  131.714286      0.0           12.1588   \n",
       "91   18123.428571      0.0  11.308966  128.857143      0.0           12.1588   \n",
       "12   21440.500000      0.0  10.469002  131.714286      0.0           12.1588   \n",
       "90   21440.500000      0.0  10.543791  131.500000      0.0           12.1588   \n",
       "169  21736.333333      0.0  10.398177  147.333333      0.0           12.1588   \n",
       "168  21531.000000      0.0  10.253828  146.666667      0.0           12.1588   \n",
       "143  18095.181818      0.0  10.543664  129.000000      0.0           12.1588   \n",
       "142  18095.181818      0.0  10.125915  129.000000      0.0           12.1588   \n",
       "66   21972.833333      0.0   8.776168  148.166667      0.0           12.1588   \n",
       "67   22128.333333      0.0   8.833694  148.666667      0.0           12.1588   \n",
       "\n",
       "     Mean Detected R #    Mean TP   Mean FP  Mean Performance  Mean Acc  \\\n",
       "119          11.935416  11.658273  0.039723          0.960202  0.995934   \n",
       "118          11.937939  11.552681  0.040503          0.951823  0.995860   \n",
       "171          11.952204  11.657263  0.044402          0.959954  0.995665   \n",
       "170          11.949635  11.552360  0.044585          0.951686  0.995619   \n",
       "15           11.984725  11.647493  0.050319          0.959062  0.995015   \n",
       "93           11.984404  11.645704  0.050273          0.958941  0.994989   \n",
       "14           11.974130  11.543874  0.050686          0.950952  0.994949   \n",
       "92           11.972983  11.542085  0.050732          0.950821  0.994919   \n",
       "145          11.907435  11.622999  0.051741          0.957870  0.994851   \n",
       "144          11.907206  11.524609  0.052108          0.950066  0.994821   \n",
       "13           11.310674  11.060869  0.053576          0.916235  0.994401   \n",
       "91           11.313105  11.062428  0.053713          0.916358  0.994362   \n",
       "12           11.305536  10.971332  0.055319          0.909171  0.994299   \n",
       "90           11.304527  10.972203  0.055135          0.909239  0.994253   \n",
       "169          11.262052  10.892987  0.130132          0.902792  0.986406   \n",
       "168          11.267144  10.805422  0.130453          0.895784  0.986327   \n",
       "143          11.284712  10.899271  0.137104          0.904049  0.986257   \n",
       "142          11.283932  10.815559  0.136003          0.897285  0.986181   \n",
       "66           11.916976  11.224944  0.191551          0.926722  0.980925   \n",
       "67           11.925141  11.308013  0.195358          0.933289  0.980816   \n",
       "\n",
       "     Mean Err  Mean mean_deviation  Mean std_deviation  Mean mean_ref_ibi  \\\n",
       "119  0.004020             0.003946            0.003433           0.845254   \n",
       "118  0.004094             0.004139            0.003714           0.845254   \n",
       "171  0.004289             0.003817            0.003438           0.845243   \n",
       "170  0.004335             0.004015            0.003723           0.845243   \n",
       "15   0.004985             0.003851            0.003440           0.845279   \n",
       "93   0.005011             0.003847            0.003448           0.845279   \n",
       "14   0.005051             0.004043            0.003723           0.845279   \n",
       "92   0.005081             0.004041            0.003736           0.845279   \n",
       "145  0.005104             0.005768            0.004433           0.845261   \n",
       "144  0.005133             0.005870            0.004623           0.845261   \n",
       "13   0.005599             0.012314            0.003324           0.845279   \n",
       "91   0.005638             0.012277            0.003336           0.845279   \n",
       "12   0.005701             0.012437            0.003564           0.845279   \n",
       "90   0.005747             0.012405            0.003578           0.845279   \n",
       "169  0.013594             0.012204            0.003274           0.845279   \n",
       "168  0.013673             0.012333            0.003525           0.845279   \n",
       "143  0.013743             0.007049            0.004398           0.845279   \n",
       "142  0.013819             0.007219            0.004587           0.845279   \n",
       "66   0.019075             0.005180            0.004635           0.845279   \n",
       "67   0.019184             0.005017            0.004396           0.845279   \n",
       "\n",
       "     Max mean_ref_ibi  Min mean_ref_ibi  Mean std_ref_ibi  Max std_ref_ibi  \\\n",
       "119              1.49              0.00          0.045796         0.526318   \n",
       "118              1.49              0.00          0.045796         0.526318   \n",
       "171              1.49              0.00          0.045796         0.526318   \n",
       "170              1.49              0.00          0.045796         0.526318   \n",
       "15               1.49              0.31          0.045796         0.526318   \n",
       "93               1.49              0.31          0.045796         0.526318   \n",
       "14               1.49              0.31          0.045796         0.526318   \n",
       "92               1.49              0.31          0.045796         0.526318   \n",
       "145              1.49              0.00          0.045796         0.526318   \n",
       "144              1.49              0.00          0.045796         0.526318   \n",
       "13               1.49              0.31          0.045796         0.526318   \n",
       "91               1.49              0.31          0.045796         0.526318   \n",
       "12               1.49              0.31          0.045796         0.526318   \n",
       "90               1.49              0.31          0.045796         0.526318   \n",
       "169              1.49              0.31          0.045796         0.526318   \n",
       "168              1.49              0.31          0.045796         0.526318   \n",
       "143              1.49              0.31          0.045796         0.526318   \n",
       "142              1.49              0.31          0.045796         0.526318   \n",
       "66               1.49              0.31          0.045796         0.526318   \n",
       "67               1.49              0.31          0.045796         0.526318   \n",
       "\n",
       "     Min std_ref_ibi  Mean mean_test_ibi  Max mean_test_ibi  \\\n",
       "119              0.0            0.846806           1.486667   \n",
       "118              0.0            0.847036           1.486667   \n",
       "171              0.0            0.846565           1.487500   \n",
       "170              0.0            0.846756           1.490000   \n",
       "15               0.0            0.846493           1.487500   \n",
       "93               0.0            0.846543           1.487500   \n",
       "14               0.0            0.846697           1.485000   \n",
       "92               0.0            0.846677           1.485000   \n",
       "145              0.0            0.847708           1.490000   \n",
       "144              0.0            0.847886           1.490000   \n",
       "13               0.0            0.851168           1.487500   \n",
       "91               0.0            0.851106           1.487500   \n",
       "12               0.0            0.851312           1.487500   \n",
       "90               0.0            0.851321           1.490000   \n",
       "169              0.0            0.852592           1.487500   \n",
       "168              0.0            0.852635           1.487500   \n",
       "143              0.0            0.852966           1.490000   \n",
       "142              0.0            0.852822           1.490000   \n",
       "66               0.0            0.850549           1.490000   \n",
       "67               0.0            0.850845           1.490000   \n",
       "\n",
       "     Min mean_test_ibi  Mean std_test_ibi  Max std_test_ibi  Min std_test_ibi  \n",
       "119           0.000000           0.046116          0.516188               0.0  \n",
       "118           0.000000           0.045106          0.671751               0.0  \n",
       "171           0.000000           0.046059          0.514908               0.0  \n",
       "170           0.000000           0.045016          0.671751               0.0  \n",
       "15            0.314828           0.046229          0.512182               0.0  \n",
       "93            0.314828           0.046239          0.512182               0.0  \n",
       "14            0.314828           0.045324          0.671751               0.0  \n",
       "92            0.314828           0.045282          0.671751               0.0  \n",
       "145           0.000000           0.046607          0.537401               0.0  \n",
       "144           0.000000           0.045712          0.657609               0.0  \n",
       "13            0.422381           0.047942          0.671751               0.0  \n",
       "91            0.422381           0.047880          0.643467               0.0  \n",
       "12            0.415000           0.047188          0.692965               0.0  \n",
       "90            0.415000           0.047121          0.692965               0.0  \n",
       "169           0.422381           0.048274          0.530330               0.0  \n",
       "168           0.416316           0.047208          0.537401               0.0  \n",
       "143           0.420000           0.047970          0.629325               0.0  \n",
       "142           0.420000           0.046766          0.629325               0.0  \n",
       "66            0.314828           0.046959          0.551543               0.0  \n",
       "67            0.314828           0.048209          0.544472               0.0  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_stats.sort_values(by=['Mean Acc'], ascending=False).iloc[:20]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 4.1.2 The Ultimate R Detector"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Detector(object):\n",
    "    def __init__(self, sr = 100, method_denoise = \"Neurokit\", method_rpeak = \"Neurokit\", correct_artifacts = False):\n",
    "        self.sr = sr\n",
    "        self.method_denoise = method_denoise\n",
    "        self.method_rpeak = method_rpeak\n",
    "        self.correct_artifacts = correct_artifacts\n",
    "    \n",
    "    def denoise(self, signal):\n",
    "        return nk2.ecg.ecg_clean(signal, sampling_rate=self.sr, method=self.method_denoise)\n",
    "    \n",
    "    def correct_rpeaks(self, signal, rpeaks, tolerance = 0.05):\n",
    "        return bs.signals.ecg.correct_rpeaks(signal, rpeaks, self.sr, tolerance)[0]\n",
    "    \n",
    "    def detect_rpeaks(self, signal, correct_peaks = False, tolerance = 0.05, return_denoised_signal = False):\n",
    "        # fixed, is_inverted = nk2.ecg.ecg_invert(signal, sampling_rate = self.sr, show=False)\n",
    "        # if is_inverted:\n",
    "        #     signal = fixed\n",
    "        denoised = self.denoise(signal)\n",
    "\n",
    "        rpeaks = nk2.ecg.ecg_peaks(denoised, sampling_rate=self.sr, method=self.method_rpeak, correct_artifacts=self.correct_artifacts)[1][\"ECG_R_Peaks\"]\n",
    "\n",
    "        if correct_peaks:\n",
    "            rpeaks = self.correct_rpeaks(signal, rpeaks, tolerance)\n",
    "        \n",
    "        if return_denoised_signal:\n",
    "            return denoised, rpeaks\n",
    "\n",
    "        return rpeaks\n",
    "    \n",
    "    def detect_fiducials(self, signal):\n",
    "        denoised, rpeaks = self.detect_rpeaks(signal, return_denoised_signal=True)\n",
    "\n",
    "        _, fiducials = nk2.ecg.ecg_delineate(denoised, rpeaks, sampling_rate=self.sr)\n",
    "\n",
    "        fiducials[\"ECG_R_Peaks\"] = rpeaks\n",
    "\n",
    "        return fiducials\n",
    "    \n",
    "    def process(self, signal_12lead):\n",
    "        \"\"\"\n",
    "        Inspired by Biosppy ecg\n",
    "        \"\"\"\n",
    "        # Denoise 12-lead ECG signal\n",
    "        for i in range(12):\n",
    "            signal_12lead[:, i] = self.denoise(signal_12lead[:, i])\n",
    "        \n",
    "        # Detect R peaks using lead I\n",
    "        rpeaks = self.detect_rpeaks(signal_12lead[:, 0])\n",
    "\n",
    "        # Heart beats\n",
    "        templates, rpeaks = bs.signals.ecg.extract_heartbeats(signal = signal_12lead[:, 0], rpeaks = rpeaks, sampling_rate = self.sr, before = 0.2, after = 0.4)\n",
    "\n",
    "        # Heart rate\n",
    "        hr_idx, hr = bs.tools.get_heart_rate(beats = rpeaks, sampling_rate = self.sr, smooth = True, size = 3)\n",
    "\n",
    "        # time vectors\n",
    "        lenght = len(signal)\n",
    "        T = (lenght - 1) / self.sr\n",
    "        ts = np.linspace(0, T, lenght, endpoint=True)\n",
    "        ts_hr = ts[hr_idx]\n",
    "        ts_tmpl = np.linspace(-0.2, 0.4, templates.shape[1], endpoint=False)\n",
    "        \n",
    "        names = (\"ts\", \"filtered\", \"rpeaks\", \"templates_ts\", \"templates\", \"heart_rate_ts\", \"heart_rate\")\n",
    "        \n",
    "        # For P wave features use lead II among (II, V1, aVF)\n",
    "        templatesP, _ = bs.signals.ecg.extract_heartbeats(signal = signal_12lead[:, 1], rpeaks = rpeaks, sampling_rate = self.sr, before = 0.2, after = 0.4)\n",
    "        argsP = (ts, signal_12lead[:, 1], rpeaks, ts_tmpl, templatesP, ts_hr, hr)\n",
    "        processP = bs.utils.ReturnTuple(argsP, names)\n",
    "        P_Peaks, P_Onsets, P_Offsets = bs.signals.ecg.getPPositions(processP)\n",
    "\n",
    "        # For Q wave features use lead I among (I, aVL)\n",
    "        argsQ = (ts, signal_12lead[:, 0], rpeaks, ts_tmpl, templates, ts_hr, hr)\n",
    "        processQ = bs.utils.ReturnTuple(argsQ, names)\n",
    "        Q_Peaks, Q_Onsets = bs.signals.ecg.getQPositions(processQ)\n",
    "\n",
    "        # For S wave features use lead V1 among (V1, V2, V3)\n",
    "        templatesS, _ = bs.signals.ecg.extract_heartbeats(signal = signal_12lead[:, 6], rpeaks = rpeaks, sampling_rate = self.sr, before = 0.2, after = 0.4)\n",
    "        argsS = (ts, signal_12lead[:, 6], rpeaks, ts_tmpl, templatesS, ts_hr, hr)\n",
    "        processS = bs.utils.ReturnTuple(argsS, names)\n",
    "        S_Peaks, S_Offsets = bs.signals.ecg.getSPositions(processS)\n",
    "\n",
    "        # For T wave features use lead V4 among (V4, V5)\n",
    "        templatesT, _ = bs.signals.ecg.extract_heartbeats(signal = signal_12lead[:, 9], rpeaks = rpeaks, sampling_rate = self.sr, before = 0.2, after = 0.4)\n",
    "        argsT = (ts, signal_12lead[:, 9], rpeaks, ts_tmpl, templatesT, ts_hr, hr)\n",
    "        processT = bs.utils.ReturnTuple(argsT, names)\n",
    "        T_Peaks, T_Onsets, T_Offsets = bs.signals.ecg.getTPositions(processT)\n",
    "\n",
    "        fiducials = {\n",
    "            \"R_Peaks\": rpeaks,\n",
    "            \"P_Peaks\": P_Peaks,\n",
    "            \"P_Onsets\": P_Onsets,\n",
    "            \"P_Offsets\": P_Offsets,\n",
    "            \"Q_Peaks\": Q_Peaks,\n",
    "            \"Q_Onsets\": Q_Onsets,\n",
    "            \"S_Peaks\": S_Peaks,\n",
    "            \"S_Offsets\": S_Offsets,\n",
    "            \"T_Peaks\": T_Peaks,\n",
    "            \"T_Onsets\": T_Onsets,\n",
    "            \"T_Offsets\": T_Offsets\n",
    "        }\n",
    "\n",
    "        return templates, fiducials\n",
    "        \n",
    "    def get_stats(self, name, label_filter = None):\n",
    "        stats = {\n",
    "            \"Name\": [],\n",
    "            \"Exact Matches #\": [],\n",
    "            \"Mean MSE\": [],\n",
    "            \"Max MSE\": [],\n",
    "            \"Min MSE\": [],\n",
    "            \"Mean MAE\": [],\n",
    "            \"Max MAE\": [],\n",
    "            \"Min MAE\": [],\n",
    "            \"Mean Correct R #\": [],\n",
    "            \"Mean Detected R #\": [],\n",
    "            \"Mean TP\": [],\n",
    "            \"Mean FP\": [],\n",
    "            \"Mean Performance\": [],\n",
    "            \"Mean Acc\": [],\n",
    "            \"Mean Err\": [],\n",
    "            \"Mean mean_deviation\": [],\n",
    "            \"Mean std_deviation\": [],\n",
    "            \"Mean mean_ref_ibi\": [],\n",
    "            \"Max mean_ref_ibi\": [],\n",
    "            \"Min mean_ref_ibi\": [],\n",
    "            \"Mean std_ref_ibi\": [],\n",
    "            \"Max std_ref_ibi\": [],\n",
    "            \"Min std_ref_ibi\": [],\n",
    "            \"Mean mean_test_ibi\": [],\n",
    "            \"Max mean_test_ibi\": [],\n",
    "            \"Min mean_test_ibi\": [],\n",
    "            \"Mean std_test_ibi\": [],\n",
    "            \"Max std_test_ibi\": [],\n",
    "            \"Min std_test_ibi\": []\n",
    "        }\n",
    "        stats['Name'].append(name)\n",
    "        df = pd.read_csv(os.path.join(comparison_folder, name + \".csv\"))\n",
    "\n",
    "        if label_filter is not None:\n",
    "            df = df[label_filter == df['Label']]\n",
    "        \n",
    "        exactly_matched_rows = df[df[\"Correct R #\"] == df[\"Detected R #\"]]\n",
    "        stats['Exact Matches #'].append(len(exactly_matched_rows))\n",
    "        \n",
    "        stats['Mean MSE'].append(exactly_matched_rows['MSE'].mean())\n",
    "        stats['Max MSE'].append(exactly_matched_rows['MSE'].max())\n",
    "        stats['Min MSE'].append(exactly_matched_rows['MSE'].min())\n",
    "\n",
    "        stats['Mean MAE'].append(exactly_matched_rows['MAE'].mean())\n",
    "        stats['Max MAE'].append(exactly_matched_rows['MAE'].max())\n",
    "        stats['Min MAE'].append(exactly_matched_rows['MAE'].min())\n",
    "\n",
    "        stats['Mean Correct R #'].append(df['Correct R #'].mean())\n",
    "        stats['Mean Detected R #'].append(df['Detected R #'].mean())\n",
    "        stats['Mean TP'].append(df['TP'].mean())\n",
    "        stats['Mean FP'].append(df['FP'].mean())\n",
    "        \n",
    "        stats['Mean Performance'].append(df['Performance'].mean())\n",
    "        stats['Mean Acc'].append(df['Acc'].mean())\n",
    "        stats['Mean Err'].append(df['Err'].mean())\n",
    "\n",
    "        stats['Mean mean_deviation'].append(df['Mean deviation'].mean())\n",
    "        stats['Mean std_deviation'].append(df['Std deviation'].mean())\n",
    "\n",
    "        stats['Mean mean_ref_ibi'].append(df['mean_ref_ibi'].mean())\n",
    "        stats['Max mean_ref_ibi'].append(df['mean_ref_ibi'].max())\n",
    "        stats['Min mean_ref_ibi'].append(df['mean_ref_ibi'].min())\n",
    "\n",
    "        stats['Mean std_ref_ibi'].append(df['std_ref_ibi'].mean())\n",
    "        stats['Max std_ref_ibi'].append(df['std_ref_ibi'].max())\n",
    "        stats['Min std_ref_ibi'].append(df['std_ref_ibi'].min())\n",
    "\n",
    "        stats['Mean mean_test_ibi'].append(df['mean_test_ibi'].mean())\n",
    "        stats['Max mean_test_ibi'].append(df['mean_test_ibi'].max())\n",
    "        stats['Min mean_test_ibi'].append(df['mean_test_ibi'].min())\n",
    "\n",
    "        stats['Mean std_test_ibi'].append(df['std_test_ibi'].mean())\n",
    "        stats['Max std_test_ibi'].append(df['std_test_ibi'].max())\n",
    "        stats['Min std_test_ibi'].append(df['std_test_ibi'].min())\n",
    "        \n",
    "        df_stats = pd.DataFrame.from_dict(stats)\n",
    "        display(df_stats)\n",
    "\n",
    "    def benchmark(self, dataset, save=True, name=\"Detector\"):\n",
    "        info = {\n",
    "            \"MSE\": [],\n",
    "            \"MAE\": [],\n",
    "            \"Correct R #\": [],\n",
    "            \"Detected R #\": [],\n",
    "            \"TP\": [],\n",
    "            \"FP\": [],\n",
    "            \"Performance\": [],\n",
    "            \"Acc\": [],\n",
    "            \"Err\": [],\n",
    "            \"Match\": [],\n",
    "            \"Deviation\": [],\n",
    "            \"Mean deviation\": [],\n",
    "            \"Std deviation\": [],\n",
    "            \"mean_ref_ibi\": [],\n",
    "            \"std_ref_ibi\": [],\n",
    "            \"mean_test_ibi\": [],\n",
    "            \"std_test_ibi\": [],\n",
    "            \"Label\": [],\n",
    "            \"Likelihoods\": []\n",
    "        }\n",
    "        for signal, r_correct, label, likelihood in tqdm(dataset):\n",
    "            signal = signal[:,0]\n",
    "            r_detected = self.detect_rpeaks(signal)\n",
    "\n",
    "            if(len(r_detected) == len(r_correct)):\n",
    "                info['MSE'].append(mean_squared_error(r_correct, r_detected))\n",
    "                info['MAE'].append(mean_absolute_error(r_correct, r_detected))\n",
    "            else:\n",
    "                info['MSE'].append(np.Inf)\n",
    "                info['MAE'].append(np.Inf)            \n",
    "\n",
    "            info['Correct R #'].append(len(r_correct))\n",
    "            info['Detected R #'].append(len(r_detected))\n",
    "            info['Label'].append(str(label))\n",
    "            info['Likelihoods'].append(str(likelihood))\n",
    "            try:\n",
    "                comparison = dict(bs.signals.ecg.compare_segmentation(r_correct, r_detected, sr))\n",
    "                info['TP'].append(comparison['TP'])\n",
    "                info['FP'].append(comparison['FP'])\n",
    "                info['Performance'].append(comparison['performance'])\n",
    "                info['Acc'].append(comparison['acc'])\n",
    "                info['Err'].append(comparison['err'])\n",
    "                info['Match'].append(str(comparison['match']))\n",
    "                info['Deviation'].append(str(comparison['deviation']))\n",
    "                info['Mean deviation'].append(comparison['mean_deviation'])\n",
    "                info['Std deviation'].append(comparison['std_deviation'])\n",
    "                info[\"mean_ref_ibi\"].append(comparison[\"mean_ref_ibi\"])\n",
    "                info[\"std_ref_ibi\"].append(comparison[\"std_ref_ibi\"])\n",
    "                info[\"mean_test_ibi\"].append(comparison[\"mean_test_ibi\"])\n",
    "                info[\"std_test_ibi\"].append(comparison[\"std_test_ibi\"])\n",
    "            except ZeroDivisionError:\n",
    "                info['TP'].append(0)\n",
    "                info['FP'].append(0)\n",
    "                info['Performance'].append(0)\n",
    "                info['Acc'].append(0)\n",
    "                info['Err'].append(0)\n",
    "                info['Match'].append(str(0))\n",
    "                info['Deviation'].append(str(0))\n",
    "                info['Mean deviation'].append(0)\n",
    "                info['Std deviation'].append(0)\n",
    "                info[\"mean_ref_ibi\"].append(0)\n",
    "                info[\"std_ref_ibi\"].append(0)\n",
    "                info[\"mean_test_ibi\"].append(0)\n",
    "                info[\"std_test_ibi\"].append(0)\n",
    "                \n",
    "        df = pd.DataFrame.from_dict(info)\n",
    "        display(df)\n",
    "        if save:\n",
    "            df.to_csv(os.path.join(comparison_folder, name + \".csv\"), index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "detector = Detector()\n",
    "idx = np.random.randint(0, signals.shape[0])\n",
    "signal = signals[idx, :, 0]\n",
    "\n",
    "rpeaks = detector.detect_rpeaks(signal)\n",
    "fiducials = detector.detect_fiducials(signal)\n",
    "signal_peaks = {\n",
    "    'Q_Peaks': fiducials['ECG_Q_Peaks'],\n",
    "    'S_Peaks': fiducials['ECG_S_Peaks'],\n",
    "    'R_Onsets': fiducials['ECG_R_Onsets'],\n",
    "    'R_Offsets': fiducials['ECG_R_Offsets']\n",
    "}\n",
    "plot_ecg(signal, signal_peaks, \"ECG \" + str(labels[idx]))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "torch182",
   "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": "3.9.13"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "ad653feb158c072180ed6e204431a6f2dddd8d6bffea08a55593f1c72c2de124"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}