templates_for_plots/template_model_comparison.ipynb

{
 "metadata": {
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.0"
  },
  "orig_nbformat": 4,
  "kernelspec": {
   "name": "python3",
   "display_name": "Python 3.9.0 64-bit"
  },
  "interpreter": {
   "hash": "e2bfb1b1dd0bcdebdb315279aa118b1f834444d4ba3ba6d660e9f6ce7703f6a2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2,
 "cells": [
  {
   "cell_type": "markdown",
   "source": [
    "## Load Data\r\n",
    "\r\n",
    "Preparation of data takes place in separate scripts. In the following, the aggregated data frames are read in from parquet files.\r\n",
    "\r\n",
    "For this template script to work, the data frames must have the expected structure:"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "source": [
    "# Expected data frame structure (empty example data frame)\r\n",
    "import pandas as pd\r\n",
    "expected_df_structure = pd.DataFrame(columns=[\r\n",
    "    'ID', 'PARAMETER', 'RUNTIME', 'MODEL', 'SCALED', 'LIBRARY', 'ENDOGENOUS', 'EXOGENOUS', 'FIRST_FORECAST', 'ALARM_TYPE', \r\n",
    "    'TP', 'FN', 'FP', 'TN', \r\n",
    "    'FPR', 'TPR', 'FNR', 'TNR', 'ACC', 'F1S', \r\n",
    "    'N_CHUNKS', 'N_ITERATIONS'])"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "source": [
    "import pandas as pd\r\n",
    "import pyarrow as pa\r\n",
    "arimax_overview = pd.read_parquet('../data/arima-data_week14/arima-x_forecast_overview_first2000_2021-07-20T13-09-54.parquet', engine='pyarrow')\r\n",
    "# rnn_overview = pd.read_parquet('../data/INSERT.parquet', engine='pyarrow')"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "source": [
    "# Sort by ID, but sort alarm type low in front of high\r\n",
    "arimax_overview['TEMPORARY_SORTING_COLUMN'] = arimax_overview.ID.replace( ['_H', '_L'], ['', ''], regex=True)\r\n",
    "arimax_overview = arimax_overview.sort_values(by=['TEMPORARY_SORTING_COLUMN', 'ALARM_TYPE'], ascending = [True, False]).drop(['TEMPORARY_SORTING_COLUMN'], axis = 1).reset_index(drop=True)\r\n",
    "display(arimax_overview)"
   ],
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "                ID PARAMETER RUNTIME   MODEL   LIBRARY ENDOGENOUS EXOGENOUS  \\\n",
       "0   BP_A_01_12_d_L        BP    None   ARIMA     darts     Median       n/a   \n",
       "1   BP_A_01_12_d_H        BP    None   ARIMA     darts     Median       n/a   \n",
       "2   BP_A_01_12_p_L        BP    None   ARIMA  pmdarima     Median       n/a   \n",
       "3   BP_A_01_12_p_H        BP    None   ARIMA  pmdarima     Median       n/a   \n",
       "4   BP_A_01_30_d_L        BP    None   ARIMA     darts     Median       n/a   \n",
       "..             ...       ...     ...     ...       ...        ...       ...   \n",
       "67  O2_A_03_12_p_H        O2    None  ARIMAX  pmdarima        Max    Median   \n",
       "68  O2_A_03_30_d_L        O2    None  ARIMAX     darts        Min    Median   \n",
       "69  O2_A_03_30_d_H        O2    None  ARIMAX     darts        Max    Median   \n",
       "70  O2_A_03_30_p_L        O2    None  ARIMAX  pmdarima        Min    Median   \n",
       "71  O2_A_03_30_p_H        O2    None  ARIMAX  pmdarima        Max    Median   \n",
       "\n",
       "   FIRST_FORECAST ALARM_TYPE   TP  ...   FP     TN     FPR     TPR     FNR  \\\n",
       "0              12        Low   30  ...   49   4639  0.0105  0.2027  0.7973   \n",
       "1              12       High  150  ...  103   4451  0.0226  0.5319  0.4681   \n",
       "2              12        Low   30  ...   49   4639  0.0105  0.2027  0.7973   \n",
       "3              12       High  150  ...  103   4451  0.0226  0.5319  0.4681   \n",
       "4              30        Low    2  ...    6   1125  0.0053  0.0588  0.9412   \n",
       "..            ...        ...  ...  ...  ...    ...     ...     ...     ...   \n",
       "67             12       High   36  ...  304  25557  0.0118  1.0000  0.0000   \n",
       "68             30        Low   55  ...  250  13689  0.0179  0.1910  0.8090   \n",
       "69             30       High   12  ...  156  14059  0.0110  1.0000  0.0000   \n",
       "70             30        Low   55  ...  250  13689  0.0179  0.1910  0.8090   \n",
       "71             30       High   12  ...  156  14059  0.0110  1.0000  0.0000   \n",
       "\n",
       "       TNR     ACC     F1S  N_CHUNKS  N_ITERATIONS  \n",
       "0   0.9895  0.9655  0.2643     405.0        4836.0  \n",
       "1   0.9774  0.9514  0.5607     405.0        4836.0  \n",
       "2   0.9895  0.9655  0.2643     405.0        4836.0  \n",
       "3   0.9774  0.9514  0.5607     405.0        4836.0  \n",
       "4   0.9947  0.9674  0.0952      88.0        1165.0  \n",
       "..     ...     ...     ...       ...           ...  \n",
       "67  0.9882  0.9883  0.1915     923.0       25897.0  \n",
       "68  0.9821  0.9661  0.1855     426.0       14227.0  \n",
       "69  0.9890  0.9890  0.1333     426.0       14227.0  \n",
       "70  0.9821  0.9661  0.1855     426.0       14227.0  \n",
       "71  0.9890  0.9890  0.1333     426.0       14227.0  \n",
       "\n",
       "[72 rows x 21 columns]"
      ],
      "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>ID</th>\n",
       "      <th>PARAMETER</th>\n",
       "      <th>RUNTIME</th>\n",
       "      <th>MODEL</th>\n",
       "      <th>LIBRARY</th>\n",
       "      <th>ENDOGENOUS</th>\n",
       "      <th>EXOGENOUS</th>\n",
       "      <th>FIRST_FORECAST</th>\n",
       "      <th>ALARM_TYPE</th>\n",
       "      <th>TP</th>\n",
       "      <th>...</th>\n",
       "      <th>FP</th>\n",
       "      <th>TN</th>\n",
       "      <th>FPR</th>\n",
       "      <th>TPR</th>\n",
       "      <th>FNR</th>\n",
       "      <th>TNR</th>\n",
       "      <th>ACC</th>\n",
       "      <th>F1S</th>\n",
       "      <th>N_CHUNKS</th>\n",
       "      <th>N_ITERATIONS</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>BP_A_01_12_d_L</td>\n",
       "      <td>BP</td>\n",
       "      <td>None</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>darts</td>\n",
       "      <td>Median</td>\n",
       "      <td>n/a</td>\n",
       "      <td>12</td>\n",
       "      <td>Low</td>\n",
       "      <td>30</td>\n",
       "      <td>...</td>\n",
       "      <td>49</td>\n",
       "      <td>4639</td>\n",
       "      <td>0.0105</td>\n",
       "      <td>0.2027</td>\n",
       "      <td>0.7973</td>\n",
       "      <td>0.9895</td>\n",
       "      <td>0.9655</td>\n",
       "      <td>0.2643</td>\n",
       "      <td>405.0</td>\n",
       "      <td>4836.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>BP_A_01_12_d_H</td>\n",
       "      <td>BP</td>\n",
       "      <td>None</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>darts</td>\n",
       "      <td>Median</td>\n",
       "      <td>n/a</td>\n",
       "      <td>12</td>\n",
       "      <td>High</td>\n",
       "      <td>150</td>\n",
       "      <td>...</td>\n",
       "      <td>103</td>\n",
       "      <td>4451</td>\n",
       "      <td>0.0226</td>\n",
       "      <td>0.5319</td>\n",
       "      <td>0.4681</td>\n",
       "      <td>0.9774</td>\n",
       "      <td>0.9514</td>\n",
       "      <td>0.5607</td>\n",
       "      <td>405.0</td>\n",
       "      <td>4836.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>BP_A_01_12_p_L</td>\n",
       "      <td>BP</td>\n",
       "      <td>None</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Median</td>\n",
       "      <td>n/a</td>\n",
       "      <td>12</td>\n",
       "      <td>Low</td>\n",
       "      <td>30</td>\n",
       "      <td>...</td>\n",
       "      <td>49</td>\n",
       "      <td>4639</td>\n",
       "      <td>0.0105</td>\n",
       "      <td>0.2027</td>\n",
       "      <td>0.7973</td>\n",
       "      <td>0.9895</td>\n",
       "      <td>0.9655</td>\n",
       "      <td>0.2643</td>\n",
       "      <td>405.0</td>\n",
       "      <td>4836.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>BP_A_01_12_p_H</td>\n",
       "      <td>BP</td>\n",
       "      <td>None</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Median</td>\n",
       "      <td>n/a</td>\n",
       "      <td>12</td>\n",
       "      <td>High</td>\n",
       "      <td>150</td>\n",
       "      <td>...</td>\n",
       "      <td>103</td>\n",
       "      <td>4451</td>\n",
       "      <td>0.0226</td>\n",
       "      <td>0.5319</td>\n",
       "      <td>0.4681</td>\n",
       "      <td>0.9774</td>\n",
       "      <td>0.9514</td>\n",
       "      <td>0.5607</td>\n",
       "      <td>405.0</td>\n",
       "      <td>4836.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>BP_A_01_30_d_L</td>\n",
       "      <td>BP</td>\n",
       "      <td>None</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>darts</td>\n",
       "      <td>Median</td>\n",
       "      <td>n/a</td>\n",
       "      <td>30</td>\n",
       "      <td>Low</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>6</td>\n",
       "      <td>1125</td>\n",
       "      <td>0.0053</td>\n",
       "      <td>0.0588</td>\n",
       "      <td>0.9412</td>\n",
       "      <td>0.9947</td>\n",
       "      <td>0.9674</td>\n",
       "      <td>0.0952</td>\n",
       "      <td>88.0</td>\n",
       "      <td>1165.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>O2_A_03_12_p_H</td>\n",
       "      <td>O2</td>\n",
       "      <td>None</td>\n",
       "      <td>ARIMAX</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Max</td>\n",
       "      <td>Median</td>\n",
       "      <td>12</td>\n",
       "      <td>High</td>\n",
       "      <td>36</td>\n",
       "      <td>...</td>\n",
       "      <td>304</td>\n",
       "      <td>25557</td>\n",
       "      <td>0.0118</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.9882</td>\n",
       "      <td>0.9883</td>\n",
       "      <td>0.1915</td>\n",
       "      <td>923.0</td>\n",
       "      <td>25897.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>O2_A_03_30_d_L</td>\n",
       "      <td>O2</td>\n",
       "      <td>None</td>\n",
       "      <td>ARIMAX</td>\n",
       "      <td>darts</td>\n",
       "      <td>Min</td>\n",
       "      <td>Median</td>\n",
       "      <td>30</td>\n",
       "      <td>Low</td>\n",
       "      <td>55</td>\n",
       "      <td>...</td>\n",
       "      <td>250</td>\n",
       "      <td>13689</td>\n",
       "      <td>0.0179</td>\n",
       "      <td>0.1910</td>\n",
       "      <td>0.8090</td>\n",
       "      <td>0.9821</td>\n",
       "      <td>0.9661</td>\n",
       "      <td>0.1855</td>\n",
       "      <td>426.0</td>\n",
       "      <td>14227.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>O2_A_03_30_d_H</td>\n",
       "      <td>O2</td>\n",
       "      <td>None</td>\n",
       "      <td>ARIMAX</td>\n",
       "      <td>darts</td>\n",
       "      <td>Max</td>\n",
       "      <td>Median</td>\n",
       "      <td>30</td>\n",
       "      <td>High</td>\n",
       "      <td>12</td>\n",
       "      <td>...</td>\n",
       "      <td>156</td>\n",
       "      <td>14059</td>\n",
       "      <td>0.0110</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.9890</td>\n",
       "      <td>0.9890</td>\n",
       "      <td>0.1333</td>\n",
       "      <td>426.0</td>\n",
       "      <td>14227.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>O2_A_03_30_p_L</td>\n",
       "      <td>O2</td>\n",
       "      <td>None</td>\n",
       "      <td>ARIMAX</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Min</td>\n",
       "      <td>Median</td>\n",
       "      <td>30</td>\n",
       "      <td>Low</td>\n",
       "      <td>55</td>\n",
       "      <td>...</td>\n",
       "      <td>250</td>\n",
       "      <td>13689</td>\n",
       "      <td>0.0179</td>\n",
       "      <td>0.1910</td>\n",
       "      <td>0.8090</td>\n",
       "      <td>0.9821</td>\n",
       "      <td>0.9661</td>\n",
       "      <td>0.1855</td>\n",
       "      <td>426.0</td>\n",
       "      <td>14227.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>O2_A_03_30_p_H</td>\n",
       "      <td>O2</td>\n",
       "      <td>None</td>\n",
       "      <td>ARIMAX</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Max</td>\n",
       "      <td>Median</td>\n",
       "      <td>30</td>\n",
       "      <td>High</td>\n",
       "      <td>12</td>\n",
       "      <td>...</td>\n",
       "      <td>156</td>\n",
       "      <td>14059</td>\n",
       "      <td>0.0110</td>\n",
       "      <td>1.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.9890</td>\n",
       "      <td>0.9890</td>\n",
       "      <td>0.1333</td>\n",
       "      <td>426.0</td>\n",
       "      <td>14227.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>72 rows × 21 columns</p>\n",
       "</div>"
      ]
     },
     "metadata": {}
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "source": [
    "# Model Overview (removes parameters, focuses on model configurations)\r\n",
    "arimax_overview.replace(['HR_', 'BP_', 'O2_'], ['', '', ''], regex=True)[\r\n",
    "    ['ID', 'MODEL', 'LIBRARY', 'ENDOGENOUS', 'EXOGENOUS', 'FIRST_FORECAST', 'ALARM_TYPE']\r\n",
    "    ].drop_duplicates().reset_index(drop=True)"
   ],
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "             ID   MODEL   LIBRARY ENDOGENOUS EXOGENOUS FIRST_FORECAST  \\\n",
       "0   A_01_12_d_L   ARIMA     darts     Median       n/a             12   \n",
       "1   A_01_12_d_H   ARIMA     darts     Median       n/a             12   \n",
       "2   A_01_12_p_L   ARIMA  pmdarima     Median       n/a             12   \n",
       "3   A_01_12_p_H   ARIMA  pmdarima     Median       n/a             12   \n",
       "4   A_01_30_d_L   ARIMA     darts     Median       n/a             30   \n",
       "5   A_01_30_d_H   ARIMA     darts     Median       n/a             30   \n",
       "6   A_01_30_p_L   ARIMA  pmdarima     Median       n/a             30   \n",
       "7   A_01_30_p_H   ARIMA  pmdarima     Median       n/a             30   \n",
       "8   A_02_12_d_L   ARIMA     darts        Min       n/a             12   \n",
       "9   A_02_12_d_H   ARIMA     darts        Max       n/a             12   \n",
       "10  A_02_12_p_L   ARIMA  pmdarima        Min       n/a             12   \n",
       "11  A_02_12_p_H   ARIMA  pmdarima        Max       n/a             12   \n",
       "12  A_02_30_d_L   ARIMA     darts        Min       n/a             30   \n",
       "13  A_02_30_d_H   ARIMA     darts        Max       n/a             30   \n",
       "14  A_02_30_p_L   ARIMA  pmdarima        Min       n/a             30   \n",
       "15  A_02_30_p_H   ARIMA  pmdarima        Max       n/a             30   \n",
       "16  A_03_12_d_L  ARIMAX     darts        Min    Median             12   \n",
       "17  A_03_12_d_H  ARIMAX     darts        Max    Median             12   \n",
       "18  A_03_12_p_L  ARIMAX  pmdarima        Min    Median             12   \n",
       "19  A_03_12_p_H  ARIMAX  pmdarima        Max    Median             12   \n",
       "20  A_03_30_d_L  ARIMAX     darts        Min    Median             30   \n",
       "21  A_03_30_d_H  ARIMAX     darts        Max    Median             30   \n",
       "22  A_03_30_p_L  ARIMAX  pmdarima        Min    Median             30   \n",
       "23  A_03_30_p_H  ARIMAX  pmdarima        Max    Median             30   \n",
       "\n",
       "   ALARM_TYPE  \n",
       "0         Low  \n",
       "1        High  \n",
       "2         Low  \n",
       "3        High  \n",
       "4         Low  \n",
       "5        High  \n",
       "6         Low  \n",
       "7        High  \n",
       "8         Low  \n",
       "9        High  \n",
       "10        Low  \n",
       "11       High  \n",
       "12        Low  \n",
       "13       High  \n",
       "14        Low  \n",
       "15       High  \n",
       "16        Low  \n",
       "17       High  \n",
       "18        Low  \n",
       "19       High  \n",
       "20        Low  \n",
       "21       High  \n",
       "22        Low  \n",
       "23       High  "
      ],
      "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>ID</th>\n",
       "      <th>MODEL</th>\n",
       "      <th>LIBRARY</th>\n",
       "      <th>ENDOGENOUS</th>\n",
       "      <th>EXOGENOUS</th>\n",
       "      <th>FIRST_FORECAST</th>\n",
       "      <th>ALARM_TYPE</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A_01_12_d_L</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>darts</td>\n",
       "      <td>Median</td>\n",
       "      <td>n/a</td>\n",
       "      <td>12</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A_01_12_d_H</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>darts</td>\n",
       "      <td>Median</td>\n",
       "      <td>n/a</td>\n",
       "      <td>12</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A_01_12_p_L</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Median</td>\n",
       "      <td>n/a</td>\n",
       "      <td>12</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A_01_12_p_H</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Median</td>\n",
       "      <td>n/a</td>\n",
       "      <td>12</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>A_01_30_d_L</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>darts</td>\n",
       "      <td>Median</td>\n",
       "      <td>n/a</td>\n",
       "      <td>30</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>A_01_30_d_H</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>darts</td>\n",
       "      <td>Median</td>\n",
       "      <td>n/a</td>\n",
       "      <td>30</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>A_01_30_p_L</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Median</td>\n",
       "      <td>n/a</td>\n",
       "      <td>30</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>A_01_30_p_H</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Median</td>\n",
       "      <td>n/a</td>\n",
       "      <td>30</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>A_02_12_d_L</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>darts</td>\n",
       "      <td>Min</td>\n",
       "      <td>n/a</td>\n",
       "      <td>12</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>A_02_12_d_H</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>darts</td>\n",
       "      <td>Max</td>\n",
       "      <td>n/a</td>\n",
       "      <td>12</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>A_02_12_p_L</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Min</td>\n",
       "      <td>n/a</td>\n",
       "      <td>12</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>A_02_12_p_H</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Max</td>\n",
       "      <td>n/a</td>\n",
       "      <td>12</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>A_02_30_d_L</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>darts</td>\n",
       "      <td>Min</td>\n",
       "      <td>n/a</td>\n",
       "      <td>30</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>A_02_30_d_H</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>darts</td>\n",
       "      <td>Max</td>\n",
       "      <td>n/a</td>\n",
       "      <td>30</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>A_02_30_p_L</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Min</td>\n",
       "      <td>n/a</td>\n",
       "      <td>30</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>A_02_30_p_H</td>\n",
       "      <td>ARIMA</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Max</td>\n",
       "      <td>n/a</td>\n",
       "      <td>30</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>A_03_12_d_L</td>\n",
       "      <td>ARIMAX</td>\n",
       "      <td>darts</td>\n",
       "      <td>Min</td>\n",
       "      <td>Median</td>\n",
       "      <td>12</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>A_03_12_d_H</td>\n",
       "      <td>ARIMAX</td>\n",
       "      <td>darts</td>\n",
       "      <td>Max</td>\n",
       "      <td>Median</td>\n",
       "      <td>12</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>A_03_12_p_L</td>\n",
       "      <td>ARIMAX</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Min</td>\n",
       "      <td>Median</td>\n",
       "      <td>12</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>A_03_12_p_H</td>\n",
       "      <td>ARIMAX</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Max</td>\n",
       "      <td>Median</td>\n",
       "      <td>12</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>A_03_30_d_L</td>\n",
       "      <td>ARIMAX</td>\n",
       "      <td>darts</td>\n",
       "      <td>Min</td>\n",
       "      <td>Median</td>\n",
       "      <td>30</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>A_03_30_d_H</td>\n",
       "      <td>ARIMAX</td>\n",
       "      <td>darts</td>\n",
       "      <td>Max</td>\n",
       "      <td>Median</td>\n",
       "      <td>30</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>A_03_30_p_L</td>\n",
       "      <td>ARIMAX</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Min</td>\n",
       "      <td>Median</td>\n",
       "      <td>30</td>\n",
       "      <td>Low</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>A_03_30_p_H</td>\n",
       "      <td>ARIMAX</td>\n",
       "      <td>pmdarima</td>\n",
       "      <td>Max</td>\n",
       "      <td>Median</td>\n",
       "      <td>30</td>\n",
       "      <td>High</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {},
     "execution_count": 8
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "source": [
    "# Results for selected parameter\r\n",
    "selected_parameter = 'HR'\r\n",
    "\r\n",
    "results_selected_parameter = arimax_overview[arimax_overview.PARAMETER == selected_parameter][\r\n",
    "    ['ID', 'ALARM_TYPE', 'TP', 'FN', 'FP', 'TN', 'FPR', 'TPR', 'FNR', 'TNR', 'ACC', 'F1S', 'N_CHUNKS', 'N_ITERATIONS']\r\n",
    "    ].reset_index(drop=True)\r\n",
    "\r\n",
    "# Print best FPR & F1S - Multiple lines are displayed if there are identical scores (e.g. darts and pmdarima)\r\n",
    "print(\"Best (lowest) FPR - Low Alarm:\")\r\n",
    "print(results_selected_parameter[results_selected_parameter.FPR == results_selected_parameter[results_selected_parameter.ALARM_TYPE == 'Low'].FPR.min()][['ID','FPR']])\r\n",
    "print(\"----------\")\r\n",
    "print(\"Best (lowest) FPR - High Alarm:\")\r\n",
    "print(results_selected_parameter[results_selected_parameter.FPR == results_selected_parameter[results_selected_parameter.ALARM_TYPE == 'High'].FPR.min()][['ID','FPR']])\r\n",
    "print(\"----------\")\r\n",
    "print(\"Best (highest) F1S - Low Alarm:\")\r\n",
    "print(results_selected_parameter[results_selected_parameter.F1S == results_selected_parameter[results_selected_parameter.ALARM_TYPE == 'Low'].F1S.max()][['ID','F1S']])\r\n",
    "print(\"----------\")\r\n",
    "print(\"Best (highest) F1S - High Alarm:\")\r\n",
    "print(results_selected_parameter[results_selected_parameter.F1S == results_selected_parameter[results_selected_parameter.ALARM_TYPE == 'High'].F1S.max()][['ID','F1S']])\r\n",
    "print(\"----------\")"
   ],
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Best (lowest) FPR - Low Alarm:\n",
      "               ID     FPR\n",
      "4  HR_A_01_30_d_L  0.0055\n",
      "6  HR_A_01_30_p_L  0.0055\n",
      "----------\n",
      "Best (lowest) FPR - High Alarm:\n",
      "               ID     FPR\n",
      "5  HR_A_01_30_d_H  0.0041\n",
      "7  HR_A_01_30_p_H  0.0041\n",
      "----------\n",
      "Best (highest) F1S - Low Alarm:\n",
      "                ID   F1S\n",
      "12  HR_A_02_30_d_L  0.44\n",
      "14  HR_A_02_30_p_L  0.44\n",
      "----------\n",
      "Best (highest) F1S - High Alarm:\n",
      "               ID     F1S\n",
      "5  HR_A_01_30_d_H  0.4064\n",
      "7  HR_A_01_30_p_H  0.4064\n",
      "----------\n"
     ]
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Model Comparison FPR & F1S"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "source": [
    "import matplotlib.pyplot as plt\r\n",
    "import seaborn as sns\r\n",
    "\r\n",
    "overview = arimax_overview\r\n",
    "# overview = arimax_overview[arimax_overview.LIBRARY == 'darts']\r\n",
    "\r\n",
    "sns.set_style(\"whitegrid\")\r\n",
    "\r\n",
    "plotdata = overview.replace( ['_H', '_L'], ['', ''], regex=True)\r\n",
    "\r\n",
    "fig, axs = plt.subplots(\r\n",
    "    2,\r\n",
    "    3,\r\n",
    "    figsize = (20, 8),\r\n",
    "    dpi = 72\r\n",
    "    )\r\n",
    "fig.subplots_adjust(hspace=0.6, wspace=0.4)\r\n",
    "\r\n",
    "fpr_ylimits = [0,max(overview.FPR)]\r\n",
    "f1s_ylimits = [0,max(overview.F1S)]\r\n",
    "\r\n",
    "for i, parameter in enumerate(['BP', 'HR', 'O2']):\r\n",
    "\r\n",
    "    plot1 = sns.barplot(\r\n",
    "        ax = axs[0,i],\r\n",
    "        data = plotdata[plotdata.PARAMETER == parameter],\r\n",
    "        x='ID',\r\n",
    "        y='FPR',\r\n",
    "        hue='ALARM_TYPE',\r\n",
    "        palette = sns.color_palette(\"colorblind\"),\r\n",
    "        ci=None,\r\n",
    "        hue_order=['Low','High'])\r\n",
    "    axs[0,i].set_title(str(parameter), fontweight='bold', color= 'black', fontsize=14)\r\n",
    "    axs[0,i].set_ylim(fpr_ylimits)\r\n",
    "    axs[0,i].tick_params(axis=\"x\", rotation=90)\r\n",
    "    axs[0,i].legend(bbox_to_anchor=(1.02, 1), loc='upper left', borderaxespad=0)\r\n",
    "\r\n",
    "    best_fpr_low = plotdata[plotdata.FPR == plotdata[(plotdata.PARAMETER == parameter) & (plotdata.ALARM_TYPE == 'Low')].FPR.min()]['FPR'].unique()\r\n",
    "    best_fpr_high = plotdata[plotdata.FPR == plotdata[(plotdata.PARAMETER == parameter) & (plotdata.ALARM_TYPE == 'High')].FPR.min()]['FPR'].unique()\r\n",
    "\r\n",
    "    for bar in plot1.patches:\r\n",
    "        if bar.get_height() == best_fpr_low or bar.get_height() == best_fpr_high :\r\n",
    "            plot1.annotate(\"↑\",\r\n",
    "            (bar.get_x() + bar.get_width() / 2, bar.get_height()),\r\n",
    "            ha='center', va='center', size=15, xytext=(0, -10), color='white', fontsize=16,\r\n",
    "            textcoords='offset points')\r\n",
    "\r\n",
    "    plot2 = sns.barplot(\r\n",
    "        ax = axs[1,i],\r\n",
    "        data = plotdata[plotdata.PARAMETER == parameter],\r\n",
    "        x='ID',\r\n",
    "        y='F1S',\r\n",
    "        hue='ALARM_TYPE',\r\n",
    "        palette = sns.color_palette(\"colorblind\"),\r\n",
    "        ci=None,\r\n",
    "        hue_order=['Low','High'])\r\n",
    "    axs[1,i].set_ylim(f1s_ylimits)\r\n",
    "    axs[1,i].tick_params(axis=\"x\", rotation=90)\r\n",
    "    axs[1,i].legend(bbox_to_anchor=(1.02, 1), loc='upper left', borderaxespad=0)\r\n",
    "\r\n",
    "    best_f1s_low = plotdata[plotdata.F1S == plotdata[(plotdata.PARAMETER == parameter) & (plotdata.ALARM_TYPE == 'Low')].F1S.max()]['F1S'].unique()\r\n",
    "    best_f1s_high = plotdata[plotdata.F1S == plotdata[(plotdata.PARAMETER == parameter) & (plotdata.ALARM_TYPE == 'High')].F1S.max()]['F1S'].unique()\r\n",
    "\r\n",
    "    for bar in plot2.patches:\r\n",
    "        if bar.get_height() == best_f1s_low or bar.get_height() == best_f1s_high :\r\n",
    "            plot2.annotate(\"↑\",\r\n",
    "            (bar.get_x() + bar.get_width() / 2, bar.get_height()),\r\n",
    "            ha='center', va='center', size=15, xytext=(0, -10), color='white', fontsize=16,\r\n",
    "            textcoords='offset points')\r\n",
    "\r\n",
    "plt.show(fig)"
   ],
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 1440x576 with 6 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {}
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## darts vs pmdarima: FPR & F1S"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "source": [
    "import matplotlib.pyplot as plt\r\n",
    "import seaborn as sns\r\n",
    "\r\n",
    "overview = arimax_overview\r\n",
    "\r\n",
    "sns.set_style(\"whitegrid\")\r\n",
    "\r\n",
    "plotdata = overview.replace( ['_d', '_p'], ['', ''], regex=True)\r\n",
    "\r\n",
    "fig, axs = plt.subplots(\r\n",
    "    2,\r\n",
    "    3,\r\n",
    "    figsize = (20, 8),\r\n",
    "    dpi = 72\r\n",
    "    )\r\n",
    "fig.subplots_adjust(hspace=0.6, wspace=0.4)\r\n",
    "\r\n",
    "fpr_ylimits = [0,max(overview.FPR)]\r\n",
    "f1s_ylimits = [0,max(overview.F1S)]\r\n",
    "\r\n",
    "for i, parameter in enumerate(['BP', 'HR', 'O2']):\r\n",
    "\r\n",
    "    plot1 = sns.barplot(\r\n",
    "        ax = axs[0,i],\r\n",
    "        data = plotdata[plotdata.PARAMETER == parameter],\r\n",
    "        x='ID',\r\n",
    "        y='FPR',\r\n",
    "        hue='LIBRARY',\r\n",
    "        palette = sns.color_palette(\"colorblind\"),\r\n",
    "        ci=None)\r\n",
    "    axs[0,i].set_title(str(parameter), fontweight='bold', color= 'black', fontsize=14)\r\n",
    "    axs[0,i].set_ylim(fpr_ylimits)\r\n",
    "    axs[0,i].tick_params(axis=\"x\", rotation=90)\r\n",
    "    axs[0,i].legend(bbox_to_anchor=(1.02, 1), loc='upper left', borderaxespad=0)\r\n",
    "\r\n",
    "    plot2 = sns.barplot(\r\n",
    "        ax = axs[1,i],\r\n",
    "        data = plotdata[plotdata.PARAMETER == parameter],\r\n",
    "        x='ID',\r\n",
    "        y='F1S',\r\n",
    "        hue='LIBRARY',\r\n",
    "        palette = sns.color_palette(\"colorblind\"),\r\n",
    "        ci=None)\r\n",
    "    axs[1,i].set_ylim(f1s_ylimits)\r\n",
    "    axs[1,i].tick_params(axis=\"x\", rotation=90)\r\n",
    "    axs[1,i].legend(bbox_to_anchor=(1.02, 1), loc='upper left', borderaxespad=0)\r\n",
    "\r\n",
    "plt.show(fig)"
   ],
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 1440x576 with 6 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {}
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## darts vs pmdarima: RUNTIME"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "source": [
    "# import matplotlib.pyplot as plt\r\n",
    "# import seaborn as sns\r\n",
    "\r\n",
    "# overview = arimax_overview\r\n",
    "\r\n",
    "# sns.set_style(\"whitegrid\")\r\n",
    "\r\n",
    "# plotdata = overview.replace( ['_d', '_p'], ['', ''], regex=True)\r\n",
    "\r\n",
    "# fig, axs = plt.subplots(\r\n",
    "#     1,\r\n",
    "#     3,\r\n",
    "#     figsize = (20, 4),\r\n",
    "#     dpi = 72\r\n",
    "#     )\r\n",
    "# fig.subplots_adjust(hspace = 0.8)\r\n",
    "\r\n",
    "# runtime_ylimits = [0,max(overview[overview.RUNTIME.notnull()]['RUNTIME'])]\r\n",
    "\r\n",
    "# for i, parameter in enumerate(['BP', 'HR', 'O2']):\r\n",
    "\r\n",
    "#     sns.barplot(\r\n",
    "#         ax = axs[i],\r\n",
    "#         data = plotdata[plotdata.PARAMETER == parameter],\r\n",
    "#         x='ID',\r\n",
    "#         y='RUNTIME',\r\n",
    "#         hue='LIBRARY',\r\n",
    "#         palette = sns.color_palette(\"colorblind\"),\r\n",
    "#         ci=None)\r\n",
    "#     axs[i].set_title(str(parameter), fontweight='bold', color= 'black', fontsize=14)\r\n",
    "#     axs[i].set_ylim(runtime_ylimits)\r\n",
    "#     axs[i].tick_params(axis=\"x\", rotation=90)\r\n",
    "\r\n",
    "# plt.show(fig)"
   ],
   "outputs": [],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## FIRST FORECAST 12 vs 30: FPR & F1S"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "source": [
    "import matplotlib.pyplot as plt\r\n",
    "import seaborn as sns\r\n",
    "\r\n",
    "overview = arimax_overview\r\n",
    "# overview = arimax_overview[arimax_overview.LIBRARY == 'darts']\r\n",
    "\r\n",
    "sns.set_style(\"whitegrid\")\r\n",
    "\r\n",
    "plotdata = overview.replace( ['_12', '_30'], ['', ''], regex=True)\r\n",
    "\r\n",
    "fig, axs = plt.subplots(\r\n",
    "    2,\r\n",
    "    3,\r\n",
    "    figsize = (20, 8),\r\n",
    "    dpi = 72\r\n",
    "    )\r\n",
    "fig.subplots_adjust(hspace=0.6, wspace=0.4)\r\n",
    "\r\n",
    "fpr_ylimits = [0,max(overview.FPR)]\r\n",
    "f1s_ylimits = [0,max(overview.F1S)]\r\n",
    "\r\n",
    "for i, parameter in enumerate(['BP', 'HR', 'O2']):\r\n",
    "\r\n",
    "    plot1 = sns.barplot(\r\n",
    "        ax = axs[0,i],\r\n",
    "        data = plotdata[plotdata.PARAMETER == parameter],\r\n",
    "        x='ID',\r\n",
    "        y='FPR',\r\n",
    "        hue='FIRST_FORECAST',\r\n",
    "        palette = sns.color_palette(\"colorblind\"),\r\n",
    "        ci=None)\r\n",
    "    axs[0,i].set_title(str(parameter), fontweight='bold', color= 'black', fontsize=14)\r\n",
    "    axs[0,i].set_ylim(fpr_ylimits)\r\n",
    "    axs[0,i].tick_params(axis=\"x\", rotation=90)\r\n",
    "    axs[0,i].legend(bbox_to_anchor=(1.02, 1), loc='upper left', borderaxespad=0)\r\n",
    "\r\n",
    "    plot2 = sns.barplot(\r\n",
    "        ax = axs[1,i],\r\n",
    "        data = plotdata[plotdata.PARAMETER == parameter],\r\n",
    "        x='ID',\r\n",
    "        y='F1S',\r\n",
    "        hue='FIRST_FORECAST',\r\n",
    "        palette = sns.color_palette(\"colorblind\"),\r\n",
    "        ci=None)\r\n",
    "    axs[1,i].set_ylim(f1s_ylimits)\r\n",
    "    axs[1,i].tick_params(axis=\"x\", rotation=90)\r\n",
    "    axs[1,i].legend(bbox_to_anchor=(1.02, 1), loc='upper left', borderaxespad=0)\r\n",
    "\r\n",
    "plt.show(fig)"
   ],
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 1440x576 with 6 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {}
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "markdown",
   "source": [
    "## Confusion Matrix Stacked Plot"
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "source": [
    "import pandas as pd\r\n",
    "import matplotlib.pyplot as plt\r\n",
    "import seaborn as sns\r\n",
    "\r\n",
    "for i, parameter in enumerate(['BP', 'HR', 'O2']):\r\n",
    "\r\n",
    "    sns.set_style(\"white\")\r\n",
    "    sns.set_context({\"figure.figsize\": (10, 4)})\r\n",
    "\r\n",
    "    plotdata = overview[overview.PARAMETER == parameter][['ID','TP','FN','FP']]\r\n",
    "    plotdata['Series1'] = plotdata.TP + plotdata.FN + plotdata.FP\r\n",
    "    plotdata['Series2'] = plotdata.TP + plotdata.FN\r\n",
    "    plotdata['Series3'] = plotdata.TP\r\n",
    "\r\n",
    "    fp = sns.barplot(\r\n",
    "        data=plotdata,\r\n",
    "        x='ID',\r\n",
    "        y='Series1',\r\n",
    "        color = sns.color_palette(\"colorblind\")[1],\r\n",
    "        ci=None).tick_params(axis=\"x\", rotation=90)\r\n",
    "\r\n",
    "    fn = sns.barplot(\r\n",
    "        data=plotdata,\r\n",
    "        x='ID',\r\n",
    "        y='Series2',\r\n",
    "        color = sns.color_palette(\"colorblind\")[0],\r\n",
    "        ci=None).set_title(str(parameter), fontweight='bold', color= 'black', fontsize=14)\r\n",
    "    \r\n",
    "    tp = sns.barplot(\r\n",
    "        data=plotdata,\r\n",
    "        x='ID',\r\n",
    "        y='Series3',\r\n",
    "        color = sns.color_palette(\"colorblind\")[2],\r\n",
    "        ci=None).set(ylabel=\"Count\")\r\n",
    "    \r\n",
    "    leg_fp = plt.Rectangle((0,0),1,1,fc=sns.color_palette(\"colorblind\")[1], edgecolor = 'none')\r\n",
    "    leg_fn = plt.Rectangle((0,0),1,1,fc=sns.color_palette(\"colorblind\")[0], edgecolor = 'none')\r\n",
    "    lef_tp = plt.Rectangle((0,0),1,1,fc=sns.color_palette(\"colorblind\")[2], edgecolor = 'none')\r\n",
    "    l = plt.legend([leg_fp, leg_fn, lef_tp], ['FP', 'FN', 'TP'], loc=1, ncol = 3, prop={'size':12}, bbox_to_anchor=(1, 1.12))\r\n",
    "    l.draw_frame(False)\r\n",
    "\r\n",
    "    plt.show()\r\n"
   ],
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 720x288 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {}
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 720x288 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {}
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "<Figure size 720x288 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {}
    }
   ],
   "metadata": {}
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "source": [
    "arimax_overview['TP_FP_RATIO'] = round(arimax_overview.TP / arimax_overview.FP, 2)\r\n",
    "\r\n",
    "display(arimax_overview[arimax_overview.PARAMETER == 'BP'][['ID','TP','FN','FP', 'TN', 'TP_FP_RATIO']].reset_index(drop=True))\r\n",
    "display(arimax_overview[arimax_overview.PARAMETER == 'HR'][['ID','TP','FN','FP', 'TN', 'TP_FP_RATIO']].reset_index(drop=True))\r\n",
    "display(arimax_overview[arimax_overview.PARAMETER == 'O2'][['ID','TP','FN','FP', 'TN', 'TP_FP_RATIO']].reset_index(drop=True))"
   ],
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "                ID   TP   FN   FP    TN  TP_FP_RATIO\n",
       "0   BP_A_01_12_d_L   30  118   49  4639         0.61\n",
       "1   BP_A_01_12_d_H  150  132  103  4451         1.46\n",
       "2   BP_A_01_12_p_L   30  118   49  4639         0.61\n",
       "3   BP_A_01_12_p_H  150  132  103  4451         1.46\n",
       "4   BP_A_01_30_d_L    2   32    6  1125         0.33\n",
       "5   BP_A_01_30_d_H   18   29    6  1112         3.00\n",
       "6   BP_A_01_30_p_L    2   32    6  1125         0.33\n",
       "7   BP_A_01_30_p_H   18   29    6  1112         3.00\n",
       "8   BP_A_02_12_d_L   33  136   55  4612         0.60\n",
       "9   BP_A_02_12_d_H  155  138  108  4435         1.44\n",
       "10  BP_A_02_12_p_L   33  136   55  4612         0.60\n",
       "11  BP_A_02_12_p_H  155  138  108  4435         1.44\n",
       "12  BP_A_02_30_d_L    4   32    7  1122         0.57\n",
       "13  BP_A_02_30_d_H   19   30    6  1110         3.17\n",
       "14  BP_A_02_30_p_L    4   32    7  1122         0.57\n",
       "15  BP_A_02_30_p_H   19   30    6  1110         3.17\n",
       "16  BP_A_03_12_d_L   33  136   61  4606         0.54\n",
       "17  BP_A_03_12_d_H  153  140  108  4435         1.42\n",
       "18  BP_A_03_12_p_L   33  136   61  4606         0.54\n",
       "19  BP_A_03_12_p_H  153  140  108  4435         1.42\n",
       "20  BP_A_03_30_d_L    4   32    6  1123         0.67\n",
       "21  BP_A_03_30_d_H   18   31    6  1110         3.00\n",
       "22  BP_A_03_30_p_L    4   32    6  1123         0.67\n",
       "23  BP_A_03_30_p_H   18   31    6  1110         3.00"
      ],
      "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>ID</th>\n",
       "      <th>TP</th>\n",
       "      <th>FN</th>\n",
       "      <th>FP</th>\n",
       "      <th>TN</th>\n",
       "      <th>TP_FP_RATIO</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>BP_A_01_12_d_L</td>\n",
       "      <td>30</td>\n",
       "      <td>118</td>\n",
       "      <td>49</td>\n",
       "      <td>4639</td>\n",
       "      <td>0.61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>BP_A_01_12_d_H</td>\n",
       "      <td>150</td>\n",
       "      <td>132</td>\n",
       "      <td>103</td>\n",
       "      <td>4451</td>\n",
       "      <td>1.46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>BP_A_01_12_p_L</td>\n",
       "      <td>30</td>\n",
       "      <td>118</td>\n",
       "      <td>49</td>\n",
       "      <td>4639</td>\n",
       "      <td>0.61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>BP_A_01_12_p_H</td>\n",
       "      <td>150</td>\n",
       "      <td>132</td>\n",
       "      <td>103</td>\n",
       "      <td>4451</td>\n",
       "      <td>1.46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>BP_A_01_30_d_L</td>\n",
       "      <td>2</td>\n",
       "      <td>32</td>\n",
       "      <td>6</td>\n",
       "      <td>1125</td>\n",
       "      <td>0.33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>BP_A_01_30_d_H</td>\n",
       "      <td>18</td>\n",
       "      <td>29</td>\n",
       "      <td>6</td>\n",
       "      <td>1112</td>\n",
       "      <td>3.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>BP_A_01_30_p_L</td>\n",
       "      <td>2</td>\n",
       "      <td>32</td>\n",
       "      <td>6</td>\n",
       "      <td>1125</td>\n",
       "      <td>0.33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>BP_A_01_30_p_H</td>\n",
       "      <td>18</td>\n",
       "      <td>29</td>\n",
       "      <td>6</td>\n",
       "      <td>1112</td>\n",
       "      <td>3.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>BP_A_02_12_d_L</td>\n",
       "      <td>33</td>\n",
       "      <td>136</td>\n",
       "      <td>55</td>\n",
       "      <td>4612</td>\n",
       "      <td>0.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>BP_A_02_12_d_H</td>\n",
       "      <td>155</td>\n",
       "      <td>138</td>\n",
       "      <td>108</td>\n",
       "      <td>4435</td>\n",
       "      <td>1.44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>BP_A_02_12_p_L</td>\n",
       "      <td>33</td>\n",
       "      <td>136</td>\n",
       "      <td>55</td>\n",
       "      <td>4612</td>\n",
       "      <td>0.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>BP_A_02_12_p_H</td>\n",
       "      <td>155</td>\n",
       "      <td>138</td>\n",
       "      <td>108</td>\n",
       "      <td>4435</td>\n",
       "      <td>1.44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>BP_A_02_30_d_L</td>\n",
       "      <td>4</td>\n",
       "      <td>32</td>\n",
       "      <td>7</td>\n",
       "      <td>1122</td>\n",
       "      <td>0.57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>BP_A_02_30_d_H</td>\n",
       "      <td>19</td>\n",
       "      <td>30</td>\n",
       "      <td>6</td>\n",
       "      <td>1110</td>\n",
       "      <td>3.17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>BP_A_02_30_p_L</td>\n",
       "      <td>4</td>\n",
       "      <td>32</td>\n",
       "      <td>7</td>\n",
       "      <td>1122</td>\n",
       "      <td>0.57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>BP_A_02_30_p_H</td>\n",
       "      <td>19</td>\n",
       "      <td>30</td>\n",
       "      <td>6</td>\n",
       "      <td>1110</td>\n",
       "      <td>3.17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>BP_A_03_12_d_L</td>\n",
       "      <td>33</td>\n",
       "      <td>136</td>\n",
       "      <td>61</td>\n",
       "      <td>4606</td>\n",
       "      <td>0.54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>BP_A_03_12_d_H</td>\n",
       "      <td>153</td>\n",
       "      <td>140</td>\n",
       "      <td>108</td>\n",
       "      <td>4435</td>\n",
       "      <td>1.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>BP_A_03_12_p_L</td>\n",
       "      <td>33</td>\n",
       "      <td>136</td>\n",
       "      <td>61</td>\n",
       "      <td>4606</td>\n",
       "      <td>0.54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>BP_A_03_12_p_H</td>\n",
       "      <td>153</td>\n",
       "      <td>140</td>\n",
       "      <td>108</td>\n",
       "      <td>4435</td>\n",
       "      <td>1.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>BP_A_03_30_d_L</td>\n",
       "      <td>4</td>\n",
       "      <td>32</td>\n",
       "      <td>6</td>\n",
       "      <td>1123</td>\n",
       "      <td>0.67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>BP_A_03_30_d_H</td>\n",
       "      <td>18</td>\n",
       "      <td>31</td>\n",
       "      <td>6</td>\n",
       "      <td>1110</td>\n",
       "      <td>3.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>BP_A_03_30_p_L</td>\n",
       "      <td>4</td>\n",
       "      <td>32</td>\n",
       "      <td>6</td>\n",
       "      <td>1123</td>\n",
       "      <td>0.67</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>BP_A_03_30_p_H</td>\n",
       "      <td>18</td>\n",
       "      <td>31</td>\n",
       "      <td>6</td>\n",
       "      <td>1110</td>\n",
       "      <td>3.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {}
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "                ID   TP   FN   FP     TN  TP_FP_RATIO\n",
       "0   HR_A_01_12_d_L  197  192  341  47614         0.58\n",
       "1   HR_A_01_12_d_H  231  453  265  47395         0.87\n",
       "2   HR_A_01_12_p_L  197  192  341  47614         0.58\n",
       "3   HR_A_01_12_p_H  231  453  265  47395         0.87\n",
       "4   HR_A_01_30_d_L  108  122  166  29829         0.65\n",
       "5   HR_A_01_30_d_H  140  287  122  29676         1.15\n",
       "6   HR_A_01_30_p_L  108  122  166  29829         0.65\n",
       "7   HR_A_01_30_p_H  140  287  122  29676         1.15\n",
       "8   HR_A_02_12_d_L  211  215  348  47567         0.61\n",
       "9   HR_A_02_12_d_H  264  514  296  47267         0.89\n",
       "10  HR_A_02_12_p_L  211  215  348  47567         0.61\n",
       "11  HR_A_02_12_p_H  264  514  296  47267         0.89\n",
       "12  HR_A_02_30_d_L  121  136  172  29794         0.70\n",
       "13  HR_A_02_30_d_H  149  347  147  29580         1.01\n",
       "14  HR_A_02_30_p_L  121  136  172  29794         0.70\n",
       "15  HR_A_02_30_p_H  149  347  147  29580         1.01\n",
       "16  HR_A_03_12_d_L  210  216  358  47560         0.59\n",
       "17  HR_A_03_12_d_H  268  510  301  47265         0.89\n",
       "18  HR_A_03_12_p_L  210  216  358  47560         0.59\n",
       "19  HR_A_03_12_p_H  268  510  301  47265         0.89\n",
       "20  HR_A_03_30_d_L  122  135  176  29792         0.69\n",
       "21  HR_A_03_30_d_H  160  336  159  29570         1.01\n",
       "22  HR_A_03_30_p_L  122  135  176  29792         0.69\n",
       "23  HR_A_03_30_p_H  160  336  159  29570         1.01"
      ],
      "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>ID</th>\n",
       "      <th>TP</th>\n",
       "      <th>FN</th>\n",
       "      <th>FP</th>\n",
       "      <th>TN</th>\n",
       "      <th>TP_FP_RATIO</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>HR_A_01_12_d_L</td>\n",
       "      <td>197</td>\n",
       "      <td>192</td>\n",
       "      <td>341</td>\n",
       "      <td>47614</td>\n",
       "      <td>0.58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>HR_A_01_12_d_H</td>\n",
       "      <td>231</td>\n",
       "      <td>453</td>\n",
       "      <td>265</td>\n",
       "      <td>47395</td>\n",
       "      <td>0.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>HR_A_01_12_p_L</td>\n",
       "      <td>197</td>\n",
       "      <td>192</td>\n",
       "      <td>341</td>\n",
       "      <td>47614</td>\n",
       "      <td>0.58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>HR_A_01_12_p_H</td>\n",
       "      <td>231</td>\n",
       "      <td>453</td>\n",
       "      <td>265</td>\n",
       "      <td>47395</td>\n",
       "      <td>0.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>HR_A_01_30_d_L</td>\n",
       "      <td>108</td>\n",
       "      <td>122</td>\n",
       "      <td>166</td>\n",
       "      <td>29829</td>\n",
       "      <td>0.65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>HR_A_01_30_d_H</td>\n",
       "      <td>140</td>\n",
       "      <td>287</td>\n",
       "      <td>122</td>\n",
       "      <td>29676</td>\n",
       "      <td>1.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>HR_A_01_30_p_L</td>\n",
       "      <td>108</td>\n",
       "      <td>122</td>\n",
       "      <td>166</td>\n",
       "      <td>29829</td>\n",
       "      <td>0.65</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>HR_A_01_30_p_H</td>\n",
       "      <td>140</td>\n",
       "      <td>287</td>\n",
       "      <td>122</td>\n",
       "      <td>29676</td>\n",
       "      <td>1.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>HR_A_02_12_d_L</td>\n",
       "      <td>211</td>\n",
       "      <td>215</td>\n",
       "      <td>348</td>\n",
       "      <td>47567</td>\n",
       "      <td>0.61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>HR_A_02_12_d_H</td>\n",
       "      <td>264</td>\n",
       "      <td>514</td>\n",
       "      <td>296</td>\n",
       "      <td>47267</td>\n",
       "      <td>0.89</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>HR_A_02_12_p_L</td>\n",
       "      <td>211</td>\n",
       "      <td>215</td>\n",
       "      <td>348</td>\n",
       "      <td>47567</td>\n",
       "      <td>0.61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>HR_A_02_12_p_H</td>\n",
       "      <td>264</td>\n",
       "      <td>514</td>\n",
       "      <td>296</td>\n",
       "      <td>47267</td>\n",
       "      <td>0.89</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>HR_A_02_30_d_L</td>\n",
       "      <td>121</td>\n",
       "      <td>136</td>\n",
       "      <td>172</td>\n",
       "      <td>29794</td>\n",
       "      <td>0.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>HR_A_02_30_d_H</td>\n",
       "      <td>149</td>\n",
       "      <td>347</td>\n",
       "      <td>147</td>\n",
       "      <td>29580</td>\n",
       "      <td>1.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>HR_A_02_30_p_L</td>\n",
       "      <td>121</td>\n",
       "      <td>136</td>\n",
       "      <td>172</td>\n",
       "      <td>29794</td>\n",
       "      <td>0.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>HR_A_02_30_p_H</td>\n",
       "      <td>149</td>\n",
       "      <td>347</td>\n",
       "      <td>147</td>\n",
       "      <td>29580</td>\n",
       "      <td>1.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>HR_A_03_12_d_L</td>\n",
       "      <td>210</td>\n",
       "      <td>216</td>\n",
       "      <td>358</td>\n",
       "      <td>47560</td>\n",
       "      <td>0.59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>HR_A_03_12_d_H</td>\n",
       "      <td>268</td>\n",
       "      <td>510</td>\n",
       "      <td>301</td>\n",
       "      <td>47265</td>\n",
       "      <td>0.89</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>HR_A_03_12_p_L</td>\n",
       "      <td>210</td>\n",
       "      <td>216</td>\n",
       "      <td>358</td>\n",
       "      <td>47560</td>\n",
       "      <td>0.59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>HR_A_03_12_p_H</td>\n",
       "      <td>268</td>\n",
       "      <td>510</td>\n",
       "      <td>301</td>\n",
       "      <td>47265</td>\n",
       "      <td>0.89</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>HR_A_03_30_d_L</td>\n",
       "      <td>122</td>\n",
       "      <td>135</td>\n",
       "      <td>176</td>\n",
       "      <td>29792</td>\n",
       "      <td>0.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>HR_A_03_30_d_H</td>\n",
       "      <td>160</td>\n",
       "      <td>336</td>\n",
       "      <td>159</td>\n",
       "      <td>29570</td>\n",
       "      <td>1.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>HR_A_03_30_p_L</td>\n",
       "      <td>122</td>\n",
       "      <td>135</td>\n",
       "      <td>176</td>\n",
       "      <td>29792</td>\n",
       "      <td>0.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>HR_A_03_30_p_H</td>\n",
       "      <td>160</td>\n",
       "      <td>336</td>\n",
       "      <td>159</td>\n",
       "      <td>29570</td>\n",
       "      <td>1.01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {}
    },
    {
     "output_type": "display_data",
     "data": {
      "text/plain": [
       "                ID   TP   FN    FP     TN  TP_FP_RATIO\n",
       "0   O2_A_01_12_d_L  138  343  1163  24253         0.12\n",
       "1   O2_A_01_12_d_H   36    0   234  25627         0.15\n",
       "2   O2_A_01_12_p_L  138  343  1163  24253         0.12\n",
       "3   O2_A_01_12_p_H   36    0   234  25627         0.15\n",
       "4   O2_A_01_30_d_L   48  193   223  13763         0.22\n",
       "5   O2_A_01_30_d_H   12    0    62  14153         0.19\n",
       "6   O2_A_01_30_p_L   48  193   223  13763         0.22\n",
       "7   O2_A_01_30_p_H   12    0    62  14153         0.19\n",
       "8   O2_A_02_12_d_L  158  400  1169  24168         0.14\n",
       "9   O2_A_02_12_d_H   29    7   253  25606         0.11\n",
       "10  O2_A_02_12_p_L  158  400  1169  24168         0.14\n",
       "11  O2_A_02_12_p_H   29    7   253  25606         0.11\n",
       "12  O2_A_02_30_d_L   55  233   246  13693         0.22\n",
       "13  O2_A_02_30_d_H   12    0    81  14134         0.15\n",
       "14  O2_A_02_30_p_L   55  233   246  13693         0.22\n",
       "15  O2_A_02_30_p_H   12    0    81  14134         0.15\n",
       "16  O2_A_03_12_d_L  151  407  1205  24134         0.13\n",
       "17  O2_A_03_12_d_H   36    0   304  25557         0.12\n",
       "18  O2_A_03_12_p_L  151  407  1205  24134         0.13\n",
       "19  O2_A_03_12_p_H   36    0   304  25557         0.12\n",
       "20  O2_A_03_30_d_L   55  233   250  13689         0.22\n",
       "21  O2_A_03_30_d_H   12    0   156  14059         0.08\n",
       "22  O2_A_03_30_p_L   55  233   250  13689         0.22\n",
       "23  O2_A_03_30_p_H   12    0   156  14059         0.08"
      ],
      "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>ID</th>\n",
       "      <th>TP</th>\n",
       "      <th>FN</th>\n",
       "      <th>FP</th>\n",
       "      <th>TN</th>\n",
       "      <th>TP_FP_RATIO</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>O2_A_01_12_d_L</td>\n",
       "      <td>138</td>\n",
       "      <td>343</td>\n",
       "      <td>1163</td>\n",
       "      <td>24253</td>\n",
       "      <td>0.12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>O2_A_01_12_d_H</td>\n",
       "      <td>36</td>\n",
       "      <td>0</td>\n",
       "      <td>234</td>\n",
       "      <td>25627</td>\n",
       "      <td>0.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>O2_A_01_12_p_L</td>\n",
       "      <td>138</td>\n",
       "      <td>343</td>\n",
       "      <td>1163</td>\n",
       "      <td>24253</td>\n",
       "      <td>0.12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>O2_A_01_12_p_H</td>\n",
       "      <td>36</td>\n",
       "      <td>0</td>\n",
       "      <td>234</td>\n",
       "      <td>25627</td>\n",
       "      <td>0.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>O2_A_01_30_d_L</td>\n",
       "      <td>48</td>\n",
       "      <td>193</td>\n",
       "      <td>223</td>\n",
       "      <td>13763</td>\n",
       "      <td>0.22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>O2_A_01_30_d_H</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>62</td>\n",
       "      <td>14153</td>\n",
       "      <td>0.19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>O2_A_01_30_p_L</td>\n",
       "      <td>48</td>\n",
       "      <td>193</td>\n",
       "      <td>223</td>\n",
       "      <td>13763</td>\n",
       "      <td>0.22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>O2_A_01_30_p_H</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>62</td>\n",
       "      <td>14153</td>\n",
       "      <td>0.19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>O2_A_02_12_d_L</td>\n",
       "      <td>158</td>\n",
       "      <td>400</td>\n",
       "      <td>1169</td>\n",
       "      <td>24168</td>\n",
       "      <td>0.14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>O2_A_02_12_d_H</td>\n",
       "      <td>29</td>\n",
       "      <td>7</td>\n",
       "      <td>253</td>\n",
       "      <td>25606</td>\n",
       "      <td>0.11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>O2_A_02_12_p_L</td>\n",
       "      <td>158</td>\n",
       "      <td>400</td>\n",
       "      <td>1169</td>\n",
       "      <td>24168</td>\n",
       "      <td>0.14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>O2_A_02_12_p_H</td>\n",
       "      <td>29</td>\n",
       "      <td>7</td>\n",
       "      <td>253</td>\n",
       "      <td>25606</td>\n",
       "      <td>0.11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>O2_A_02_30_d_L</td>\n",
       "      <td>55</td>\n",
       "      <td>233</td>\n",
       "      <td>246</td>\n",
       "      <td>13693</td>\n",
       "      <td>0.22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>O2_A_02_30_d_H</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>81</td>\n",
       "      <td>14134</td>\n",
       "      <td>0.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>O2_A_02_30_p_L</td>\n",
       "      <td>55</td>\n",
       "      <td>233</td>\n",
       "      <td>246</td>\n",
       "      <td>13693</td>\n",
       "      <td>0.22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>O2_A_02_30_p_H</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>81</td>\n",
       "      <td>14134</td>\n",
       "      <td>0.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>O2_A_03_12_d_L</td>\n",
       "      <td>151</td>\n",
       "      <td>407</td>\n",
       "      <td>1205</td>\n",
       "      <td>24134</td>\n",
       "      <td>0.13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>O2_A_03_12_d_H</td>\n",
       "      <td>36</td>\n",
       "      <td>0</td>\n",
       "      <td>304</td>\n",
       "      <td>25557</td>\n",
       "      <td>0.12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>O2_A_03_12_p_L</td>\n",
       "      <td>151</td>\n",
       "      <td>407</td>\n",
       "      <td>1205</td>\n",
       "      <td>24134</td>\n",
       "      <td>0.13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>O2_A_03_12_p_H</td>\n",
       "      <td>36</td>\n",
       "      <td>0</td>\n",
       "      <td>304</td>\n",
       "      <td>25557</td>\n",
       "      <td>0.12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>O2_A_03_30_d_L</td>\n",
       "      <td>55</td>\n",
       "      <td>233</td>\n",
       "      <td>250</td>\n",
       "      <td>13689</td>\n",
       "      <td>0.22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>O2_A_03_30_d_H</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>156</td>\n",
       "      <td>14059</td>\n",
       "      <td>0.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>O2_A_03_30_p_L</td>\n",
       "      <td>55</td>\n",
       "      <td>233</td>\n",
       "      <td>250</td>\n",
       "      <td>13689</td>\n",
       "      <td>0.22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>O2_A_03_30_p_H</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>156</td>\n",
       "      <td>14059</td>\n",
       "      <td>0.08</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "metadata": {}
    }
   ],
   "metadata": {}
  }
 ]
}