analysis_descriptive.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "e3642e59-8794-4357-b043-1874eade3553",
   "metadata": {},
   "source": [
    "# Descriptive Analysis of the papers\n",
    "\n",
    "This is a descriptive approach to extract insights on the review in form of quantitative meta analysis and visualization.\n",
    "\n",
    "Steps for a clean xlsx-File:\n",
    "1. Download the [spreadsheet](https://docs.google.com/spreadsheets/d/1uM_VpRDk2YLPrqe4KiA_20HWBWmcSw_f9orz6cuOieQ/edit?pli=1#gid=1960060281) as xlsx.\n",
    "1. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "5298e0cb-88d4-44d3-9ea9-e61bd422b29c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# %pip install -r requirements.txt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "d146ade6-f440-4e18-9c33-c19614973d2f",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "os.chdir(\"/home/jovyan/work/BCI/SLR_EEG_motion_artifact_correction_methods/\")\n",
    "\n",
    "import sys\n",
    "import time\n",
    "from datetime import datetime\n",
    "\n",
    "import math\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "pd.set_option('display.max_rows', 500)\n",
    "pd.set_option('display.max_columns', 500)\n",
    "pd.set_option('display.width', 1000)\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "d6aae1a2-b6c4-454f-a888-4b9ffcc4a8a8",
   "metadata": {},
   "outputs": [],
   "source": [
    "PRISMA_RESULT_PATH = \"Passed_papers/papers_passed.csv\"\n",
    "TABLE_NAME = \"Systematic Literature Review on Motion Artifact Removal of EEG Signals.xlsx\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6f7d9578-52fe-49a6-92c7-8af92596d1a5",
   "metadata": {},
   "source": [
    "## Load the Search Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "76fe1534-5c52-42e0-8598-2f1a4be616d9",
   "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>Reference</th>\n",
       "      <th>Index</th>\n",
       "      <th>PMID</th>\n",
       "      <th>Title</th>\n",
       "      <th>Authors</th>\n",
       "      <th>Citation</th>\n",
       "      <th>First Author</th>\n",
       "      <th>Journal/Book</th>\n",
       "      <th>Publication Year</th>\n",
       "      <th>Create Date</th>\n",
       "      <th>PMCID</th>\n",
       "      <th>NIHMS ID</th>\n",
       "      <th>DOI</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Hossain22</td>\n",
       "      <td>8</td>\n",
       "      <td>35590859</td>\n",
       "      <td>Motion Artifacts Correction from Single-Channe...</td>\n",
       "      <td>Hossain MS, Chowdhury MEH, Reaz MBI, Ali SHM, ...</td>\n",
       "      <td>Sensors (Basel). 2022 Apr 21;22(9):3169. doi: ...</td>\n",
       "      <td>Hossain MS</td>\n",
       "      <td>Sensors (Basel)</td>\n",
       "      <td>2022</td>\n",
       "      <td>2022/05/20</td>\n",
       "      <td>PMC9102309</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10.3390/s22093169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Abu22</td>\n",
       "      <td>10</td>\n",
       "      <td>35459033</td>\n",
       "      <td>Improved Cognitive Vigilance Assessment after ...</td>\n",
       "      <td>Abu Farha N, Al-Shargie F, Tariq U, Al-Nashash H.</td>\n",
       "      <td>Sensors (Basel). 2022 Apr 15;22(8):3051. doi: ...</td>\n",
       "      <td>Abu Farha N</td>\n",
       "      <td>Sensors (Basel)</td>\n",
       "      <td>2022</td>\n",
       "      <td>2022/04/23</td>\n",
       "      <td>PMC9033092</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10.3390/s22083051</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Reference  Index      PMID                                              Title                                            Authors                                           Citation First Author     Journal/Book  Publication Year Create Date       PMCID NIHMS ID                DOI\n",
       "1  Hossain22      8  35590859  Motion Artifacts Correction from Single-Channe...  Hossain MS, Chowdhury MEH, Reaz MBI, Ali SHM, ...  Sensors (Basel). 2022 Apr 21;22(9):3169. doi: ...   Hossain MS  Sensors (Basel)              2022  2022/05/20  PMC9102309      NaN  10.3390/s22093169\n",
       "2      Abu22     10  35459033  Improved Cognitive Vigilance Assessment after ...  Abu Farha N, Al-Shargie F, Tariq U, Al-Nashash H.  Sensors (Basel). 2022 Apr 15;22(8):3051. doi: ...  Abu Farha N  Sensors (Basel)              2022  2022/04/23  PMC9033092      NaN  10.3390/s22083051"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# this file is used as template on google docs\n",
    "df = pd.read_csv(PRISMA_RESULT_PATH)\n",
    "df.index += 1  # increment index such that it fits with the used indices\n",
    "df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1d50efa8-fab9-45bc-a658-fd37f0027db8",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "cd8f07fb-c301-40e2-be46-ecbc1c72540a",
   "metadata": {},
   "source": [
    "## Load the Spreadsheets into separate DataFrames"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9c29b2f3-a23a-48ae-9bc4-dffbe254820b",
   "metadata": {},
   "source": [
    "### papers_passed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "374f1229-6d56-47c1-9d7c-e7c07e1d589a",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 78 entries, Hossain22 to Jung01\n",
      "Data columns (total 4 columns):\n",
      " #   Column            Non-Null Count  Dtype         \n",
      "---  ------            --------------  -----         \n",
      " 0   Title             78 non-null     object        \n",
      " 1   Authors           78 non-null     object        \n",
      " 2   Publication Year  78 non-null     Int64         \n",
      " 3   Create Date       78 non-null     datetime64[ns]\n",
      "dtypes: Int64(1), datetime64[ns](1), object(2)\n",
      "memory usage: 3.1+ KB\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Title</th>\n",
       "      <th>Authors</th>\n",
       "      <th>Publication Year</th>\n",
       "      <th>Create Date</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Reference</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Hossain22</th>\n",
       "      <td>Motion Artifacts Correction from Single-Channe...</td>\n",
       "      <td>Hossain MS, Chowdhury MEH, Reaz MBI, Ali SHM, ...</td>\n",
       "      <td>2022</td>\n",
       "      <td>2022-05-20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Abu22</th>\n",
       "      <td>Improved Cognitive Vigilance Assessment after ...</td>\n",
       "      <td>Abu Farha N, Al-Shargie F, Tariq U, Al-Nashash H.</td>\n",
       "      <td>2022</td>\n",
       "      <td>2022-04-23</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                       Title                                            Authors  Publication Year Create Date\n",
       "Reference                                                                                                                                    \n",
       "Hossain22  Motion Artifacts Correction from Single-Channe...  Hossain MS, Chowdhury MEH, Reaz MBI, Ali SHM, ...              2022  2022-05-20\n",
       "Abu22      Improved Cognitive Vigilance Assessment after ...  Abu Farha N, Al-Shargie F, Tariq U, Al-Nashash H.              2022  2022-04-23"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dtypes = {\n",
    "     'Publication Year': pd.Int64Dtype(),\n",
    "     'Create Date': pd.DatetimeIndex\n",
    "}\n",
    "papers_passed_df = pd.read_excel(TABLE_NAME, engine='openpyxl', sheet_name=\"papers_passed\", dtype=dtypes, \n",
    "                                  usecols=['Reference', 'Title', 'Authors', 'Publication Year', 'Create Date'])\n",
    "papers_passed_df.set_index(\"Reference\", inplace=True)\n",
    "papers_passed_df.info()\n",
    "papers_passed_df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "07c7a3fa-12ba-44f9-a858-1796ed9c81c8",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "7d43555b-139b-4ac0-b2c0-caecd4c3b6bc",
   "metadata": {},
   "source": [
    "### Authors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "deef1df7-4dff-4110-aa9e-aa481dfc9f87",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING: 0 null values\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 471 entries, Hossain22 to Jung01\n",
      "Data columns (total 2 columns):\n",
      " #   Column                         Non-Null Count  Dtype \n",
      "---  ------                         --------------  ----- \n",
      " 0   Author (alle Autoren angeben)  471 non-null    object\n",
      " 1   Institution                    471 non-null    object\n",
      "dtypes: object(2)\n",
      "memory usage: 11.0+ KB\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Author (alle Autoren angeben)</th>\n",
       "      <th>Institution</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Reference</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Hossain22</th>\n",
       "      <td>Hossain MS</td>\n",
       "      <td>University Kebangsaan Malaysia, Malaysia</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Hossain22</th>\n",
       "      <td>Chowdhury MEH</td>\n",
       "      <td>Qatar University, Qatar</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          Author (alle Autoren angeben)                               Institution\n",
       "Reference                                                                        \n",
       "Hossain22                    Hossain MS  University Kebangsaan Malaysia, Malaysia\n",
       "Hossain22                 Chowdhury MEH                   Qatar University, Qatar"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "authors_df = pd.read_excel(TABLE_NAME, engine='openpyxl', sheet_name=\"Authors\", dtype=dtypes)\n",
    "authors_df.set_index(\"Reference\", inplace=True)\n",
    "print(f'WARNING: {sum(authors_df[\"Institution\"].isnull())} null values')\n",
    "# assert sum(authors_df[\"Institution\"].isnull()) == 0  # TODO check this criteria\n",
    "authors_df = authors_df[~authors_df[\"Institution\"].isnull()]\n",
    "authors_df.info()\n",
    "authors_df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "c1c17729",
   "metadata": {},
   "outputs": [],
   "source": [
    "assert len(authors_df.index.unique()) == 77"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "3789de4c-50fe-4b3d-a52a-6859479e4b9f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Hossain22, Abu-Farha22, Flo22, Kumaravel22, Soghoyan21, Karpiel21, Beach21, Dai21, Tost21, Kaur21, Rosanne21, Nahmias21, Saba-Sadiya21, Li21, Anders20, Zhang20, Leach20, Saini20, Mucarquer20, Nordin20, Liu19, Benda19, Val-Calvo19, Saavedra19, Mur19, Blum19, Nordin19, Grosselin19, Li18, Chu18, San-Martin18, Sebek18, Arad18, Symeonidou18, Stone18, Frolich18, Gabard-Durnam18, Tamburro18, Lin18, Delisle-Rodriguez17, Wang17, Jas17, Oliveira17, DelPozo-Banos17, Weiss17, Roy17, Melman16, Oliveira16, Cheveigne16, Costa16, Nathan16, Zou16, Gordon15, Snyder15, Kline15, Raduntz15, Mariani15, Zhang15, Bulea14, Chen14, Yilmaz14, Reis14, Boudet14, Li13, Lawhern13, Nottage13, Yong12, Lau12, Lawhern12, Winkler11, Gwin10, McMenamin10, Mosher09, McMenamin09, Delorme04, Iriarte03, Jung01'"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\", \".join(authors_df.index.unique())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f296677a-9b47-45b2-922c-5fe132b00eb4",
   "metadata": {},
   "source": [
    "### Papers_annotated"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "bb8c8f81-33f2-4de5-98ac-6d2c56d41b24",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "dtypes = {\n",
    "     'Journal/Book': 'category',\n",
    "     'reader': 'category',\n",
    "     'mental strategy': 'category',\n",
    "     'framework': 'category',\n",
    "     'open code': 'category',\n",
    "     'Index': 'int32'\n",
    "}\n",
    "\n",
    "papers_annotated_df = pd.read_excel(TABLE_NAME, engine='openpyxl', sheet_name=\"papers_annotated\", \n",
    "                                    dtype=dtypes, skipfooter=3)\n",
    "# papers_annotated_df = papers_annotated_df[~papers_annotated_df[\"reader\"].isnull()]\n",
    "papers_annotated_df.rename(columns={\n",
    "    \"sample (size, sex, age group, health status)\": \"sample\",\n",
    "    \"open code (yes, no, -)\": \"open code\"}, inplace=True)\n",
    "papers_annotated_df.set_index(\"Reference\", inplace=True)\n",
    "assert papers_annotated_df.shape[0] == 77  # assert 77 papers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "29fbc6d3-68bb-4189-b658-ce5a10a7b525",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 77 entries, Hossain22 to Jung01\n",
      "Data columns (total 16 columns):\n",
      " #   Column                  Non-Null Count  Dtype   \n",
      "---  ------                  --------------  -----   \n",
      " 0   Index                   77 non-null     int32   \n",
      " 1   Title                   77 non-null     object  \n",
      " 2   Authors                 77 non-null     object  \n",
      " 3   Journal/Book            77 non-null     category\n",
      " 4   reader                  77 non-null     category\n",
      " 5   mental strategy         76 non-null     category\n",
      " 6   objective               77 non-null     object  \n",
      " 7   data acquisition        77 non-null     object  \n",
      " 8   sample                  77 non-null     object  \n",
      " 9   #sample                 73 non-null     object  \n",
      " 10  evaluation metrics      76 non-null     object  \n",
      " 11  framework               75 non-null     category\n",
      " 12  open code               76 non-null     object  \n",
      " 13  Innovation / Finding    75 non-null     object  \n",
      " 14  further work suggested  62 non-null     object  \n",
      " 15  Additional information  67 non-null     object  \n",
      "dtypes: category(4), int32(1), object(11)\n",
      "memory usage: 10.6+ KB\n"
     ]
    }
   ],
   "source": [
    "papers_annotated_df.info()  # int: Index. categories: journal, reader, mental strategy, framework"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "6dc6b19f-c616-4e91-9dd8-99718f18ab51",
   "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>Index</th>\n",
       "      <th>Title</th>\n",
       "      <th>Authors</th>\n",
       "      <th>Journal/Book</th>\n",
       "      <th>reader</th>\n",
       "      <th>mental strategy</th>\n",
       "      <th>objective</th>\n",
       "      <th>data acquisition</th>\n",
       "      <th>sample</th>\n",
       "      <th>#sample</th>\n",
       "      <th>evaluation metrics</th>\n",
       "      <th>framework</th>\n",
       "      <th>open code</th>\n",
       "      <th>Innovation / Finding</th>\n",
       "      <th>further work suggested</th>\n",
       "      <th>Additional information</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Reference</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Iriarte03</th>\n",
       "      <td>713</td>\n",
       "      <td>Independent component analysis as a tool to el...</td>\n",
       "      <td>Iriarte J, Urrestarazu E, Valencia M, Alegre M...</td>\n",
       "      <td>J Clin Neurophysiol</td>\n",
       "      <td>mst</td>\n",
       "      <td>-</td>\n",
       "      <td>Objectively test the capability of ICA (JADE a...</td>\n",
       "      <td>64-channel digital EEG using LaMont amplifiers...</td>\n",
       "      <td>Two series of samples of interictal EEG (betwe...</td>\n",
       "      <td>2</td>\n",
       "      <td>Visual</td>\n",
       "      <td>Matlab</td>\n",
       "      <td>-</td>\n",
       "      <td>Demonstrated/Validated utility of ICA (JADE al...</td>\n",
       "      <td>-</td>\n",
       "      <td>Pretty basic ICA evaluation paper; JADE algori...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Jung01</th>\n",
       "      <td>725</td>\n",
       "      <td>Imaging Brain Dynamics Using Independent Compo...</td>\n",
       "      <td>Jung TP, Makeig S, McKeown MJ, Bell AJ, Lee TW...</td>\n",
       "      <td>Proc IEEE Inst Electr Electron Eng</td>\n",
       "      <td>cs</td>\n",
       "      <td>attention strategy</td>\n",
       "      <td>Outlining the assumptions underlying ICA in re...</td>\n",
       "      <td>29-channel EEG system (Electrocap, Inc.) 10-20...</td>\n",
       "      <td>10 subjects (8 males) right-handed with normal...</td>\n",
       "      <td>10</td>\n",
       "      <td>-</td>\n",
       "      <td>NaN</td>\n",
       "      <td>no</td>\n",
       "      <td>There are 4 assumptions underlying ICA, which ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Index                                              Title                                            Authors                        Journal/Book reader     mental strategy                                          objective                                   data acquisition                                             sample #sample evaluation metrics framework open code                               Innovation / Finding further work suggested                             Additional information\n",
       "Reference                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                 \n",
       "Iriarte03    713  Independent component analysis as a tool to el...  Iriarte J, Urrestarazu E, Valencia M, Alegre M...                 J Clin Neurophysiol    mst                   -  Objectively test the capability of ICA (JADE a...  64-channel digital EEG using LaMont amplifiers...  Two series of samples of interictal EEG (betwe...       2             Visual    Matlab         -  Demonstrated/Validated utility of ICA (JADE al...                      -  Pretty basic ICA evaluation paper; JADE algori...\n",
       "Jung01       725  Imaging Brain Dynamics Using Independent Compo...  Jung TP, Makeig S, McKeown MJ, Bell AJ, Lee TW...  Proc IEEE Inst Electr Electron Eng     cs  attention strategy  Outlining the assumptions underlying ICA in re...  29-channel EEG system (Electrocap, Inc.) 10-20...  10 subjects (8 males) right-handed with normal...      10                  -       NaN        no  There are 4 assumptions underlying ICA, which ...                    NaN                                                NaN"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "papers_annotated_df.tail(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "56babf79-30fa-4929-b4a1-44d06e04c7cb",
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Merge with papers_passed\n",
    "# papers_annotated_df = papers_annotated_df.merge(papers_passed_df, how='outer')\n",
    "# papers_annotated_df.tail(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "31bf9666-0f84-48ab-9771-d5afdea2bfe2",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "9b913f78-a035-49ce-821d-2d0783067236",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Pipelines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "3524c483-90ec-437b-9445-0df773cd25eb",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(299, 59)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>is new</th>\n",
       "      <th>EEG (#ch)</th>\n",
       "      <th>electrode type (w,g,d)</th>\n",
       "      <th>MEG</th>\n",
       "      <th>ECG</th>\n",
       "      <th>EMG</th>\n",
       "      <th>Others</th>\n",
       "      <th>filter</th>\n",
       "      <th>MA</th>\n",
       "      <th>AF</th>\n",
       "      <th>Other</th>\n",
       "      <th>Epoch width (s)</th>\n",
       "      <th>Feature Extraction (#)</th>\n",
       "      <th>Fourier Transf.</th>\n",
       "      <th>Wavelet Transf.</th>\n",
       "      <th>Welch</th>\n",
       "      <th>LSP</th>\n",
       "      <th>PCC</th>\n",
       "      <th>CCA</th>\n",
       "      <th>ICA</th>\n",
       "      <th>NMF</th>\n",
       "      <th>PCA</th>\n",
       "      <th>t-SNE</th>\n",
       "      <th>EMD</th>\n",
       "      <th>IVA</th>\n",
       "      <th>CSP</th>\n",
       "      <th>Other.1</th>\n",
       "      <th>Linear Regression</th>\n",
       "      <th>DA</th>\n",
       "      <th>SVM</th>\n",
       "      <th>Spatial sph. spline</th>\n",
       "      <th>GMM</th>\n",
       "      <th>Other.2</th>\n",
       "      <th>Autoencoder</th>\n",
       "      <th>GAN</th>\n",
       "      <th>Other.3</th>\n",
       "      <th>MARA</th>\n",
       "      <th>DSS</th>\n",
       "      <th>DFA</th>\n",
       "      <th>FASTER</th>\n",
       "      <th>ADJUST</th>\n",
       "      <th>HAPPE</th>\n",
       "      <th>ERASE</th>\n",
       "      <th>Other.4</th>\n",
       "      <th>DA.1</th>\n",
       "      <th>KNN</th>\n",
       "      <th>KMC</th>\n",
       "      <th>ED</th>\n",
       "      <th>NB</th>\n",
       "      <th>DT</th>\n",
       "      <th>RF</th>\n",
       "      <th>SVM.1</th>\n",
       "      <th>NN</th>\n",
       "      <th>GMM.1</th>\n",
       "      <th>MEC</th>\n",
       "      <th>Other.5</th>\n",
       "      <th>fit score (0-2)</th>\n",
       "      <th>online score (0-2)</th>\n",
       "      <th>details</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Reference</th>\n",
       "      <th>Pipeline</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Iriarte03</th>\n",
       "      <th>HF_LF_NOTCH</th>\n",
       "      <td>0</td>\n",
       "      <td>64</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1: HF (35 Hz; IIR order 4) + LF (0.3 Hz; IIR o...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Jung01</th>\n",
       "      <th>ICA</th>\n",
       "      <td>0</td>\n",
       "      <td>29</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1: BP 0.1-50 Hz</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>It is averaged across multiple trials and even...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       is new  EEG (#ch) electrode type (w,g,d)  MEG  ECG  EMG Others                                             filter   MA   AF Other Epoch width (s) Feature Extraction (#) Fourier Transf. Wavelet Transf.  Welch  LSP  PCC  CCA  ICA  NMF  PCA t-SNE  EMD  IVA  CSP Other.1 Linear Regression   DA  SVM Spatial sph. spline  GMM Other.2 Autoencoder  GAN Other.3 MARA  DSS  DFA FASTER ADJUST  HAPPE  ERASE Other.4 DA.1  KNN  KMC  ED  NB  DT  RF SVM.1   NN  GMM.1  MEC Other.5  fit score (0-2)  online score (0-2)                                            details\n",
       "Reference Pipeline                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                              \n",
       "Iriarte03 HF_LF_NOTCH       0         64                    NaN  NaN  NaN  NaN    NaN  1: HF (35 Hz; IIR order 4) + LF (0.3 Hz; IIR o...  NaN  NaN   NaN             NaN                    NaN             NaN             NaN    NaN  NaN  NaN  NaN  NaN  NaN  NaN   NaN  NaN  NaN  NaN     NaN               NaN  NaN  NaN                 NaN  NaN     NaN         NaN  NaN     NaN  NaN  NaN  NaN    NaN    NaN    NaN    NaN     NaN  NaN  NaN  NaN NaN NaN NaN NaN   NaN  NaN    NaN  NaN     NaN                1                   2                                                NaN\n",
       "Jung01    ICA               0         29                    NaN  NaN  NaN  NaN    NaN                                    1: BP 0.1-50 Hz  NaN  NaN   NaN             NaN                    NaN             NaN             NaN    NaN  NaN  NaN  NaN    2  NaN  NaN   NaN  NaN  NaN  NaN     NaN               NaN  NaN  NaN                 NaN  NaN     NaN         NaN  NaN     NaN  NaN  NaN  NaN    NaN    NaN    NaN    NaN     NaN  NaN  NaN  NaN NaN NaN NaN NaN   NaN  NaN    NaN  NaN     NaN                0                   1  It is averaged across multiple trials and even..."
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dtypes = {\n",
    "    'is new': pd.Int64Dtype(),\n",
    "    'EEG (#ch)': pd.Int64Dtype(),\n",
    "    # 'Epoch width (s)': 'float64',  # no not possible here\n",
    "    'fit score (0-2)': pd.Int64Dtype(),\n",
    "    'online score (0-2)': pd.Int64Dtype()\n",
    "}\n",
    "pipelines_df = pd.read_excel(TABLE_NAME, engine='openpyxl', sheet_name=\"Pipelines\", header=1, \n",
    "                             na_values=\"-\", dtype=dtypes, skipfooter=1)\n",
    "pipelines_df = pipelines_df[~pipelines_df[\"Pipeline ID (unique per paper)\"].isnull()]\n",
    "pipelines_df.rename(columns={\n",
    "    \"Pipeline ID (unique per paper)\": \"Pipeline\",\n",
    "    \"Filter (HP, LP, BP, BR)\": \"filter\"}, inplace=True)\n",
    "pipelines_df.set_index([\"Reference\", \"Pipeline\"], inplace=True)\n",
    "print(pipelines_df.shape)\n",
    "# TODO check if Bruhn02 is shown and needed? All fields \"Pipeline ID (unique per paper)\" must be provided\n",
    "pipelines_df.tail(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "080d97dc-aeb8-4d57-aabe-385e643a1519",
   "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></th>\n",
       "      <th>fit score (0-2)</th>\n",
       "      <th>online score (0-2)</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Reference</th>\n",
       "      <th>Pipeline</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">Hossain22</th>\n",
       "      <th>WPD(db1)</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>WPD(db2)</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    fit score (0-2)  online score (0-2)\n",
       "Reference Pipeline                                     \n",
       "Hossain22 WPD(db1)                2                   2\n",
       "          WPD(db2)                2                   2"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Show all pipelines that could fit and be online\n",
    "scored_pipelines = pipelines_df[[\"fit score (0-2)\", \"online score (0-2)\"]].dropna()\n",
    "fit_pipelines = scored_pipelines[scored_pipelines[\"fit score (0-2)\"] * scored_pipelines[\"online score (0-2)\"] > 0]\n",
    "fit_pipelines.head(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8114e303-b226-4d65-8cb4-0bca93d9fdab",
   "metadata": {},
   "source": [
    "### Merge statistics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "a10ac019-1009-4961-b6fa-e12fbd634d64",
   "metadata": {},
   "outputs": [],
   "source": [
    "pipelines_annotated_df = pipelines_df.merge(papers_annotated_df, \n",
    "                                  how=\"outer\", left_on=[\"Reference\"], right_on=[\"Reference\"])\n",
    "pipelines_annotated_df[\"year\"] = pipelines_annotated_df.index \n",
    "pipelines_annotated_df[\"year\"] = pipelines_annotated_df[\"year\"].apply(lambda x: 2000+int(x[-2:]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "48925d2c-71c5-452d-afa8-3179e0b8fff5",
   "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>is new</th>\n",
       "      <th>EEG (#ch)</th>\n",
       "      <th>electrode type (w,g,d)</th>\n",
       "      <th>MEG</th>\n",
       "      <th>ECG</th>\n",
       "      <th>EMG</th>\n",
       "      <th>Others</th>\n",
       "      <th>filter</th>\n",
       "      <th>MA</th>\n",
       "      <th>AF</th>\n",
       "      <th>Other</th>\n",
       "      <th>Epoch width (s)</th>\n",
       "      <th>Feature Extraction (#)</th>\n",
       "      <th>Fourier Transf.</th>\n",
       "      <th>Wavelet Transf.</th>\n",
       "      <th>Welch</th>\n",
       "      <th>LSP</th>\n",
       "      <th>PCC</th>\n",
       "      <th>CCA</th>\n",
       "      <th>ICA</th>\n",
       "      <th>NMF</th>\n",
       "      <th>PCA</th>\n",
       "      <th>t-SNE</th>\n",
       "      <th>EMD</th>\n",
       "      <th>IVA</th>\n",
       "      <th>CSP</th>\n",
       "      <th>Other.1</th>\n",
       "      <th>Linear Regression</th>\n",
       "      <th>DA</th>\n",
       "      <th>SVM</th>\n",
       "      <th>Spatial sph. spline</th>\n",
       "      <th>GMM</th>\n",
       "      <th>Other.2</th>\n",
       "      <th>Autoencoder</th>\n",
       "      <th>GAN</th>\n",
       "      <th>Other.3</th>\n",
       "      <th>MARA</th>\n",
       "      <th>DSS</th>\n",
       "      <th>DFA</th>\n",
       "      <th>FASTER</th>\n",
       "      <th>ADJUST</th>\n",
       "      <th>HAPPE</th>\n",
       "      <th>ERASE</th>\n",
       "      <th>Other.4</th>\n",
       "      <th>DA.1</th>\n",
       "      <th>KNN</th>\n",
       "      <th>KMC</th>\n",
       "      <th>ED</th>\n",
       "      <th>NB</th>\n",
       "      <th>DT</th>\n",
       "      <th>RF</th>\n",
       "      <th>SVM.1</th>\n",
       "      <th>NN</th>\n",
       "      <th>GMM.1</th>\n",
       "      <th>MEC</th>\n",
       "      <th>Other.5</th>\n",
       "      <th>fit score (0-2)</th>\n",
       "      <th>online score (0-2)</th>\n",
       "      <th>details</th>\n",
       "      <th>Index</th>\n",
       "      <th>Title</th>\n",
       "      <th>Authors</th>\n",
       "      <th>Journal/Book</th>\n",
       "      <th>reader</th>\n",
       "      <th>mental strategy</th>\n",
       "      <th>objective</th>\n",
       "      <th>data acquisition</th>\n",
       "      <th>sample</th>\n",
       "      <th>#sample</th>\n",
       "      <th>evaluation metrics</th>\n",
       "      <th>framework</th>\n",
       "      <th>open code</th>\n",
       "      <th>Innovation / Finding</th>\n",
       "      <th>further work suggested</th>\n",
       "      <th>Additional information</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Reference</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Frolich18</th>\n",
       "      <td>0</td>\n",
       "      <td>64</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1: BP 2-45Hz (Butterworth)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2: Fourier-ICA</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3: IC_MARC</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>needs STFFT which makes it slower</td>\n",
       "      <td>272</td>\n",
       "      <td>Removal of muscular artifacts in EEG signals: ...</td>\n",
       "      <td>Frølich L, Dowding I.</td>\n",
       "      <td>Brain Inform</td>\n",
       "      <td>mst,cs</td>\n",
       "      <td>motor execution</td>\n",
       "      <td>Comparision of 5 BSS methods during motor exec...</td>\n",
       "      <td>64 equidistant Ag/AgCl electrodes measured at ...</td>\n",
       "      <td>18 subjects in a simulated driving experiment,...</td>\n",
       "      <td>18</td>\n",
       "      <td>ERD peak score (event-related desynchronisatio...</td>\n",
       "      <td>EEGlab, Matlab</td>\n",
       "      <td>-</td>\n",
       "      <td>A high-pass filter with 14 Hz cuttoff increase...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Tamburro18</th>\n",
       "      <td>1</td>\n",
       "      <td>128</td>\n",
       "      <td>g</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1: BP 0.3-100, BR 49-51 (FIR)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4: 5s (sliding 1s overlap) or 2s</td>\n",
       "      <td>5: 14</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3: InfomaxICA</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2: PCA (20)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6: binary SVM RBF kernel</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>Used features: Kurtosis, Maximum epoch varianc...</td>\n",
       "      <td>275</td>\n",
       "      <td>A new ICA-based fingerprint method for the aut...</td>\n",
       "      <td>Tamburro G, Fiedler P, Stone D, Haueisen J, Co...</td>\n",
       "      <td>PeerJ</td>\n",
       "      <td>mst,cs</td>\n",
       "      <td>motor execution</td>\n",
       "      <td>Feature-based classifcation of artifact types ...</td>\n",
       "      <td>RefaExt (Advanced Neuro Technologies) sampled ...</td>\n",
       "      <td>For (1) eyeblinks twelve subjects, (2) eye mov...</td>\n",
       "      <td>12, 19, 19</td>\n",
       "      <td>accuracy, FOR (false omssion rate), sensitivit...</td>\n",
       "      <td>Matlab, EEGlab</td>\n",
       "      <td>-</td>\n",
       "      <td>Findings: (1) New fingerprint method based on ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Very informative paper with a very good introd...</td>\n",
       "      <td>2018</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            is new  EEG (#ch) electrode type (w,g,d)  MEG  ECG  EMG Others                         filter   MA   AF Other                   Epoch width (s) Feature Extraction (#) Fourier Transf. Wavelet Transf.  Welch  LSP  PCC  CCA             ICA  NMF          PCA t-SNE  EMD  IVA  CSP Other.1 Linear Regression   DA                       SVM Spatial sph. spline  GMM     Other.2 Autoencoder  GAN Other.3 MARA  DSS  DFA FASTER ADJUST  HAPPE  ERASE Other.4 DA.1  KNN  KMC  ED  NB  DT  RF SVM.1   NN  GMM.1  MEC Other.5  fit score (0-2)  online score (0-2)                                            details  Index                                              Title                                            Authors  Journal/Book  reader  mental strategy                                          objective                                   data acquisition                                             sample     #sample                                 evaluation metrics       framework open code  \\\n",
       "Reference                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                               \n",
       "Frolich18        0         64                    NaN  NaN  NaN  NaN    NaN     1: BP 2-45Hz (Butterworth)  NaN  NaN   NaN                               NaN                    NaN             NaN             NaN    NaN  NaN  NaN  NaN  2: Fourier-ICA  NaN          NaN   NaN  NaN  NaN  NaN     NaN               NaN  NaN                       NaN                 NaN  NaN  3: IC_MARC         NaN  NaN     NaN  NaN  NaN  NaN    NaN    NaN    NaN    NaN     NaN  NaN  NaN  NaN NaN NaN NaN NaN   NaN  NaN    NaN  NaN     NaN                2                   1                  needs STFFT which makes it slower    272  Removal of muscular artifacts in EEG signals: ...                              Frølich L, Dowding I.  Brain Inform  mst,cs  motor execution  Comparision of 5 BSS methods during motor exec...  64 equidistant Ag/AgCl electrodes measured at ...  18 subjects in a simulated driving experiment,...          18  ERD peak score (event-related desynchronisatio...  EEGlab, Matlab         -   \n",
       "Tamburro18       1        128                      g  NaN  NaN  NaN    NaN  1: BP 0.3-100, BR 49-51 (FIR)  NaN  NaN   NaN  4: 5s (sliding 1s overlap) or 2s                  5: 14             NaN             NaN    NaN  NaN  NaN  NaN   3: InfomaxICA  NaN  2: PCA (20)   NaN  NaN  NaN  NaN     NaN               NaN  NaN  6: binary SVM RBF kernel                 NaN  NaN         NaN         NaN  NaN     NaN  NaN  NaN  NaN    NaN    NaN    NaN    NaN     NaN  NaN  NaN  NaN NaN NaN NaN NaN   NaN  NaN    NaN  NaN     NaN                2                   1  Used features: Kurtosis, Maximum epoch varianc...    275  A new ICA-based fingerprint method for the aut...  Tamburro G, Fiedler P, Stone D, Haueisen J, Co...         PeerJ  mst,cs  motor execution  Feature-based classifcation of artifact types ...  RefaExt (Advanced Neuro Technologies) sampled ...  For (1) eyeblinks twelve subjects, (2) eye mov...  12, 19, 19  accuracy, FOR (false omssion rate), sensitivit...  Matlab, EEGlab         -   \n",
       "\n",
       "                                         Innovation / Finding further work suggested                             Additional information  year  \n",
       "Reference                                                                                                                                      \n",
       "Frolich18   A high-pass filter with 14 Hz cuttoff increase...                    NaN                                                NaN  2018  \n",
       "Tamburro18  Findings: (1) New fingerprint method based on ...                    NaN  Very informative paper with a very good introd...  2018  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pipelines_annotated_df.sample(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a1da43e3-36c3-4ae1-aabd-4d799d05e3fd",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "e3ac4724-3ae8-4455-9276-cdfc03a8bfac",
   "metadata": {},
   "source": [
    "### Pipeline Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "1dc04197-eebc-4096-b13e-eddfa4755d6b",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 787 entries, 0 to 811\n",
      "Data columns (total 6 columns):\n",
      " #   Column     Non-Null Count  Dtype   \n",
      "---  ------     --------------  -----   \n",
      " 0   Reference  787 non-null    category\n",
      " 1   Setup ID   787 non-null    int32   \n",
      " 2   Data ID    787 non-null    int32   \n",
      " 3   Metric     787 non-null    category\n",
      " 4   Pipeline   787 non-null    object  \n",
      " 5   rank       785 non-null    float64 \n",
      "dtypes: category(2), float64(1), int32(2), object(1)\n",
      "memory usage: 31.4+ KB\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Reference</th>\n",
       "      <th>Setup ID</th>\n",
       "      <th>Data ID</th>\n",
       "      <th>Metric</th>\n",
       "      <th>Pipeline</th>\n",
       "      <th>rank</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>730</th>\n",
       "      <td>Chen14</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>SNR</td>\n",
       "      <td>Original EEG</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>273</th>\n",
       "      <td>Saba-Sadiya21</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>f-score</td>\n",
       "      <td>GAAL</td>\n",
       "      <td>8.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100</th>\n",
       "      <td>Soghoyan21</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>ROC-AUC</td>\n",
       "      <td>SVM</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>135</th>\n",
       "      <td>Soghoyan21</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>F1-score</td>\n",
       "      <td>XGB</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>758</th>\n",
       "      <td>Yong12</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>TPR</td>\n",
       "      <td>Ignore</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Reference  Setup ID  Data ID    Metric      Pipeline  rank\n",
       "730         Chen14         2        2       SNR  Original EEG   6.0\n",
       "273  Saba-Sadiya21         1        2   f-score          GAAL   8.5\n",
       "100     Soghoyan21         5        1   ROC-AUC           SVM   1.0\n",
       "135     Soghoyan21         3        1  F1-score           XGB   1.0\n",
       "758         Yong12         2        1       TPR        Ignore  10.0"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dtypes = {\n",
    "     'Reference': 'category',\n",
    "     # 'Setup ID': 'int32',  # only possible if they are non-floats\n",
    "     # 'Data ID': 'int32', \n",
    "     'Metric': 'category',\n",
    "     'Pipeline': 'category',\n",
    "     'rank': 'float64'  # we break ties evenly\n",
    "}\n",
    "\n",
    "results_df = pd.read_excel(TABLE_NAME, engine='openpyxl', sheet_name=\"Results\", \n",
    "                           na_values=\"-\", header=1, dtype=dtypes, \n",
    "                           converters={\"Setup ID\": lambda x: str(x), \"Data ID\": lambda x: str(x)})\n",
    "results_df.rename(columns={\"rank in paper\": \"rank\", \"Pipeline ID (per paper)\": \"Pipeline\"}, inplace=True)\n",
    "# results_df = results_df.loc[results_df.index.dropna()]\n",
    "results_df = results_df.loc[results_df[\"Pipeline\"].dropna().index]  # drop rows with NaN-Pipeline\n",
    "results_df.drop(columns=\"details\", inplace=True)\n",
    "results_df[\"Setup ID\"] = results_df[\"Setup ID\"].astype(np.int32)\n",
    "results_df[\"Data ID\"] = results_df[\"Data ID\"].astype(np.int32)\n",
    "assert results_df[results_df[\"Metric\"].isna()].shape[0] == 0  # should be empty\n",
    "# results_df.dropna(inplace=True)\n",
    "results_df.info()\n",
    "results_df.sample(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "a4e1df8a-a69c-48ec-84ad-473ee11ee7d2",
   "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>Reference</th>\n",
       "      <th>Setup ID</th>\n",
       "      <th>Data ID</th>\n",
       "      <th>Metric</th>\n",
       "      <th>Pipeline</th>\n",
       "      <th>rank</th>\n",
       "      <th>fit score (0-2)</th>\n",
       "      <th>online score (0-2)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Hossain22</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>SNR</td>\n",
       "      <td>WPD(db1)</td>\n",
       "      <td>10.5</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Hossain22</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>SNR</td>\n",
       "      <td>WPD(db2)</td>\n",
       "      <td>10.5</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Hossain22</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>SNR</td>\n",
       "      <td>WPD(db3)</td>\n",
       "      <td>10.5</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Hossain22</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>SNR</td>\n",
       "      <td>WPD(sym4)</td>\n",
       "      <td>10.5</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Hossain22</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>SNR</td>\n",
       "      <td>WPD(sym5)</td>\n",
       "      <td>10.5</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Reference  Setup ID  Data ID Metric   Pipeline  rank  fit score (0-2)  online score (0-2)\n",
       "0  Hossain22         1        1    SNR   WPD(db1)  10.5                2                   2\n",
       "1  Hossain22         1        1    SNR   WPD(db2)  10.5                2                   2\n",
       "2  Hossain22         1        1    SNR   WPD(db3)  10.5                2                   2\n",
       "3  Hossain22         1        1    SNR  WPD(sym4)  10.5                2                   2\n",
       "4  Hossain22         1        1    SNR  WPD(sym5)  10.5                2                   2"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Merge Pipelines with their results\n",
    "fit_results_df = results_df.merge(\n",
    "    right=scored_pipelines, how=\"left\", \n",
    "    left_on=[\"Reference\", \"Pipeline\"], right_on=[\"Reference\", \"Pipeline\"])\n",
    "fit_results_df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e679e624-671f-4187-9b4e-f49182ce6715",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "03cf32ea-5b5c-4ca8-9f8d-41f598b3cb64",
   "metadata": {},
   "source": [
    "## Basic explorative analysis"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "230b4ba8-f44c-4036-be32-97f6933dea65",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Journal, most common over time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "a3f19368-9d0d-46c0-9dcb-0ffad860e209",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# filter the rows in papers_passed, that are really in our set:\n",
    "df = df.loc[df[\"Reference\"].apply(lambda ref: ref in set(authors_df.index))]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "6eeeead5-66d7-4840-b43d-2af8bfc14aeb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Front Neurosci          10\n",
       "Sensors (Basel)          7\n",
       "Front Hum Neurosci       7\n",
       "J Neurosci Methods       6\n",
       "J Healthc Eng            3\n",
       "Psychophysiology         3\n",
       "PLoS One                 3\n",
       "Neuroimage               2\n",
       "Healthc Technol Lett     2\n",
       "Dev Cogn Neurosci        2\n",
       "Name: Journal/Book, dtype: int64"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[\"Journal/Book\"].value_counts()[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "0997b33b-eebc-49f7-9fbf-4d664d7b7e9f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'The trend of the most common Journals over time. Only four Journals contribute more than three publications to this subject of interest.'"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get the journals in order by ascending counts, also in the dataframe\n",
    "top_n = 7\n",
    "joi = df[\"Journal/Book\"].value_counts()[:top_n]\n",
    "other_j = \"other Journal\"\n",
    "df_ = df.copy()\n",
    "df_[\"Journal/Book\"] = df_[\"Journal/Book\"].apply(lambda j: j if j in joi else other_j)\n",
    "journals = list(df_[\"Journal/Book\"].value_counts().index)\n",
    "j_counts = list(df_[\"Journal/Book\"].value_counts())\n",
    "journals.remove(other_j)\n",
    "journals = [other_j] + journals\n",
    "df_[\"Journal/Book\"] = pd.Categorical(df_[\"Journal/Book\"], journals)\n",
    "\n",
    "# Append the journal counts to their names in the dataframe\n",
    "label_dict = {j: f\"{j}: {j_counts[i]}\" for i, j in enumerate(journals)}\n",
    "df_[\"Journal/Book\"] = df_[\"Journal/Book\"].map(label_dict)\n",
    "\n",
    "# Plot the histogram\n",
    "df_[\"Publication Year\"] += 0.5\n",
    "sns.histplot(data=df_, x=\"Publication Year\", hue=\"Journal/Book\", multiple=\"stack\", binwidth=1).set(title=f\"Top {top_n} Journals over time\")\n",
    "\n",
    "plt.xticks(np.arange(2000, 2023, 2))\n",
    "plt.yticks(range(0, 15, 2))\n",
    "plt.xlim(2003, None)\n",
    "plt.savefig(\"Figures/journals_over_time.pdf\", format=\"pdf\", bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\"The trend of the most common Journals over time. Only four Journals contribute more than three publications to this subject of interest.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4e660b9b-3f9f-4f76-a7f4-f0dfe66a72c2",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "22d07a1c-97ec-40a0-b5d0-01c11909c9ff",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Autors, most common"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "117a0628-642b-4d71-85c3-d87fec60de87",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ferris DP      11\n",
       "Hairston WD     7\n",
       "Chen X          4\n",
       "König P         4\n",
       "Makeig S        4\n",
       "Raimondo F      4\n",
       "Zou Y           3\n",
       "Ren H           3\n",
       "Li C            3\n",
       "Cheveigne A     3\n",
       "Name: Author (alle Autoren angeben), dtype: int64"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# There are two persons with the Name X. Chen\n",
    "authors_df[\"Author (alle Autoren angeben)\"].value_counts()[:10] #.value_counts()[:25]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "f6667d8a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1     266\n",
       "2      54\n",
       "3      21\n",
       "4       4\n",
       "11      1\n",
       "7       1\n",
       "Name: Author (alle Autoren angeben), dtype: int64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "authors_df[\"Author (alle Autoren angeben)\"].value_counts().value_counts()[:25]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4c892e79-493f-4d75-8b5f-e443784931e5",
   "metadata": {},
   "source": [
    "### Countries, most common"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "27c2773e-d2d1-4daa-baa4-e99b657cc19a",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "institutions_d = dict()\n",
    "last_author = None\n",
    "last_instances = list()\n",
    "for author, row in authors_df.iterrows():\n",
    "    if author != last_author:\n",
    "        for instance in list(set(last_instances)):\n",
    "            institutions_d[instance] = institutions_d.get(instance, 0) + 1\n",
    "        last_instances = list()\n",
    "        last_author = author\n",
    "    last_instances.append(row[\"Institution\"].split(\",\")[-1].strip())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "2a646ccd-cd5a-4a6e-97f1-4e10bfbde2b5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('USA', 31),\n",
       " ('Germany', 13),\n",
       " ('China', 8),\n",
       " ('Spain', 7),\n",
       " ('France', 6),\n",
       " ('Canada', 5),\n",
       " ('UK', 4),\n",
       " ('Italy', 3),\n",
       " ('India', 3),\n",
       " ('United Arab Emirates', 2),\n",
       " ('Argentina', 2),\n",
       " ('Chile', 2),\n",
       " ('Brazil', 2),\n",
       " ('Denmark', 2),\n",
       " ('Qatar', 1)]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 10 counties involved in the most contributions\n",
    "sorted(institutions_d.items(), key=lambda k: -k[1])[:15]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "9cc9f10e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "33"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(institutions_d)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "7ba5816d-4a95-47a6-9208-412244df5b06",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "USA          170\n",
       "China         45\n",
       "Spain         39\n",
       "France        35\n",
       "Germany       31\n",
       "Canada        16\n",
       "Singapore     13\n",
       "Brazil        13\n",
       "Italy         13\n",
       "Russia        12\n",
       "Name: Institution, dtype: int64"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# duplicate counts for multiple authors\n",
    "authors_df[\"Institution\"].apply(lambda x: x.split(\",\")[-1].strip()).value_counts()[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e549f3f2-4667-41ca-9715-c49877704ea0",
   "metadata": {},
   "source": [
    "### Institutions, most common"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "59fd8d3a-0bca-40e0-8d8e-be77d4acc031",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "institutions_d = dict()\n",
    "last_author = None\n",
    "last_instances = list()\n",
    "for author, row in authors_df.iterrows():\n",
    "    if author != last_author:\n",
    "        for instance in list(set(last_instances)):\n",
    "            institutions_d[instance] = institutions_d.get(instance, 0) + 1\n",
    "        last_instances = list()\n",
    "        last_author = author\n",
    "    last_instances.append(row[\"Institution\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "c50cd0e5-0339-4231-ab8d-8d5f444f1995",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('US Army Research Laboratory, USA', 8),\n",
       " ('University of Michigan, USA', 6),\n",
       " ('University of California San Diego, USA', 4),\n",
       " ('University of Florida, USA', 3),\n",
       " ('CIBER-BBN, Spain', 2),\n",
       " ('University of Maryland, USA', 2),\n",
       " ('University of Southern California, USA', 2),\n",
       " ('Universidad de Valparaiso, Chile', 2),\n",
       " ('Technische Universität Berlin, Germany', 2),\n",
       " ('Harvard University, USA', 2),\n",
       " ('Aberdeen Proving Ground, USA', 2),\n",
       " ('University Medical Center Hamburg-Eppendorf, Germany', 2),\n",
       " ('University of Osnabrück, Germany', 2),\n",
       " ('University of Houston, USA', 2),\n",
       " ('University of British Columbia, Canada', 2)]"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 10 institutions involved in the most contributions\n",
    "sorted(institutions_d.items(), key=lambda k: -k[1])[:15]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "c1ac5de4-6d79-4c44-b316-73749add877a",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Print institutions in form of (county, institution)\n",
    "# sorted({(k.split(\",\")[-1].strip(), k.split(\",\")[0]): v for k, v in institutions_d.items()})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "c3a0c06d-a3d2-45c0-9da9-55c0ec7971bd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "University of Michigan, USA                                                                     21\n",
       "University of California Los Angeles, USA                                                       20\n",
       "University of California San Diego, USA                                                         15\n",
       "Thomas Jefferson University, USA                                                                13\n",
       "US Army Research Laboratory, USA                                                                11\n",
       "National University of Singapore, Singapore                                                     11\n",
       "University of Wisconsin, USA                                                                     9\n",
       "University of Maryland, USA                                                                      8\n",
       "Harvard University, USA                                                                          8\n",
       "University of British Columbia, Canada                                                           7\n",
       "University of Florida, USA                                                                       7\n",
       "China National Digital Switching System Engineering and Technological Research Center, China     7\n",
       "Russian Academy of Sciences, Russia                                                              7\n",
       "Institute de Recerca Sant Joan de Deu, Spain                                                     6\n",
       "University of Houston, USA                                                                       6\n",
       "Name: Institution, dtype: int64"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# with duplicate counts\n",
    "authors_df[\"Institution\"].value_counts()[:15]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "de51abe5-f6b6-476e-a591-0c7f40720c51",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "3e6e3302-ad41-4369-a209-80694994b576",
   "metadata": {},
   "source": [
    "### Framework over year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "bca8153f-cac8-4393-a085-5c5d44697c9e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number of publications using Matlab: \t49\n",
      "number of publications using Python: \t6\n"
     ]
    }
   ],
   "source": [
    "print(f'number of publications using Matlab: \\t{papers_annotated_df[\"framework\"].apply(lambda s: \"matlab\" in s.lower()).sum()}')\n",
    "print(f'number of publications using Python: \\t{papers_annotated_df[\"framework\"].apply(lambda s: \"python\" in s.lower()).sum()}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "6d7bd8cb-94ac-464c-8693-a0d1a27ee357",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Matlab                                                                                            30\n",
       "-                                                                                                 19\n",
       "Python                                                                                             4\n",
       "Matlab, EEGlab                                                                                     3\n",
       "Matlab, EEGLab                                                                                     3\n",
       "Matlab, EEGLAB                                                                                     3\n",
       "EEGlab, Matlab                                                                                     2\n",
       "Matlab, Python                                                                                     2\n",
       "Matlab, EEGLAB, ERPLAB                                                                             1\n",
       "Matlab, Persyst(v12)                                                                               1\n",
       "Matlab, DataSuite                                                                                  1\n",
       "-> SystemPlus(Micromed) for recording \\n\\n-> Spike2 (Cambridge Electronic Design) for analysis     1\n",
       "Matlab, DETECT                                                                                     1\n",
       "Matlab, BrainRecorder software, E-Prime application suite                                          1\n",
       "MATLAB, EEGLAB, ...                                                                                1\n",
       "MATB-II: Microsoft Visual Studio.NET (VS.NET)                                                      1\n",
       "http://www.neurology-kuleuven.be/                                                                  1\n",
       "Name: framework, dtype: int64"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "papers_annotated_df[\"framework\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "63989ce1-0fbf-49fc-a279-5225b1e0731d",
   "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>matlab</th>\n",
       "      <th>python</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>year</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2003</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2004</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2011</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2012</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017</th>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018</th>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021</th>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     matlab python\n",
       "year              \n",
       "2003   True  False\n",
       "2004   True  False\n",
       "2009      1      0\n",
       "2010      2      0\n",
       "2011   True  False\n",
       "2012      3      0\n",
       "2013      2      0\n",
       "2014      2      0\n",
       "2015      4      0\n",
       "2016      5      0\n",
       "2017      4      2\n",
       "2018      7      0\n",
       "2019      4      1\n",
       "2020      5      0\n",
       "2021      4      3\n",
       "2022      3      0"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_ = papers_annotated_df.copy()\n",
    "df_[\"year\"] = df_.index \n",
    "df_[\"year\"] = df_[\"year\"].apply(lambda x: 2000+int(x[-2:]))\n",
    "df_ = df_[~df_[\"framework\"].isnull()]\n",
    "\n",
    "df_[\"matlab\"] = df_[\"framework\"].apply(lambda s: \"matlab\" in str(s).lower())\n",
    "df_[\"python\"] = df_[\"framework\"].apply(lambda s: \"python\" in str(s).lower())\n",
    "# df_.loc[df_[\"matlab\"]==True, \"matlab\"] = True\n",
    "# df_.loc[df_[\"python\"]==True, \"python\"] = True\n",
    "df_.groupby(\"year\").agg({\"matlab\": \"sum\", \"python\": \"sum\"})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "7bc39584-6a2f-4a4b-b7d5-740825de9ed9",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_2 = pd.concat([df_.loc[df_[\"matlab\"]], df_.loc[df_[\"python\"]]])[[\"Index\", \"year\", \"matlab\", \"python\"]]\n",
    "df_2.loc[df_[\"matlab\"], \"framework\"] = \"Matlab\"\n",
    "df_2.loc[df_[\"python\"], \"framework\"] = \"Python\"\n",
    "df_2[\"framework\"] = [\"Matlab\"]*df_[\"matlab\"].sum() + [\"Python\"]*df_[\"python\"].sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "73e3b474-93a1-4e26-8965-d4a790d0c883",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'The trend of programming languages over time depicts a clear dominance of Matlab. Three of five implementations with Python occured in 2022'"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Plot the histogram\n",
    "# df_2[\"year\"] -= 0.5\n",
    "sns.histplot(data=df_2, x=\"year\", hue=\"framework\", multiple=\"stack\", binwidth=0.999).set(title=f\"Top languages over time\")\n",
    "plt.xlim(2008, None)\n",
    "plt.savefig(\"Figures/languages_over_time.pdf\", format=\"pdf\", bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\"The trend of programming languages over time depicts a clear dominance of Matlab. Three of five implementations with Python occured in 2022\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "71658d12-0bf9-4e06-9dab-2cba12b19d46",
   "metadata": {},
   "source": [
    "### Open code over time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "1d6418ef-76fb-4d10-8412-25fd923d379b",
   "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>open code</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>year</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2014</th>\n",
       "      <td>0.250000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016</th>\n",
       "      <td>0.333333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017</th>\n",
       "      <td>0.428571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018</th>\n",
       "      <td>0.090909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019</th>\n",
       "      <td>0.250000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020</th>\n",
       "      <td>0.166667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021</th>\n",
       "      <td>0.300000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022</th>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      open code\n",
       "year           \n",
       "2014   0.250000\n",
       "2015   0.000000\n",
       "2016   0.333333\n",
       "2017   0.428571\n",
       "2018   0.090909\n",
       "2019   0.250000\n",
       "2020   0.166667\n",
       "2021   0.300000\n",
       "2022   0.500000"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_opencode = df_.groupby(\"year\").agg({\"open code\": lambda x: np.sum(x==\"yes\")/len(x) if len(x)>=4 else np.NaN}) # ].value_counts()\n",
    "df_opencode.loc[2014:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "ba3aa99a-5c6e-48f8-a148-45ab42a92314",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'For years with at least four publications, the fraction of papers with open code is depicted. The trend is rising strongly from 2020 (17\\\\%) to 2022 (50\\\\%)'"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sns.lineplot(data=df_opencode, x=\"year\", y=\"open code\", marker=\"o\").set(title=f\"Open Code policy over time\")\n",
    "plt.ylabel(\"Fraction of open code\")\n",
    "# plt.ylim(0,1)\n",
    "plt.show()\n",
    "\"For years with at least four publications, the fraction of papers with open code is depicted. The trend is rising strongly from 2020 (17\\%) to 2022 (50\\%)\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "feeb8b51-97a1-4d6b-93d9-8dd832f17f19",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_ = papers_annotated_df.copy()\n",
    "df_[\"year\"] = df_.index \n",
    "df_[\"year\"] = df_[\"year\"].apply(lambda x: 2000+int(x[-2:]))\n",
    "# df_.groupby(\"year\")[\"open code\"].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "de468d1f-c3d4-41bb-8db4-ffbcc314d197",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAEWCAYAAABhffzLAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAAsTAAALEwEAmpwYAAAeoUlEQVR4nO3deZwU1bn/8c8zwIg4CgKGyzIwqOgVND/AibgvMQIak1y4xuUG1yCYa0TijYn8jEZvTH7qL9eX0USvqGwuuEai0Rij0Si5so0gARFxGWXAZURxAUUGnvtH1Uw64/RMg11dPX2+79erX1NdVX3OUzVdT1efqj7H3B0REQlHWdoBiIhIYSnxi4gERolfRCQwSvwiIoFR4hcRCYwSv4hIYJT4peSYWZWZuZl1LHC9M8zsinj6MDNbWcj6C83MPjaz3dOOQ7adEn/AzOwMM/ubmW00s7fM7EYz65ZSLGZmk8xsmZltMLM6M7vXzPZLI54vyt2fcfe9044jX8zsKTMbnznP3Svc/dW0YpLtp8QfKDP7D+Aq4EKgK3AgMAD4k5mVpxDSr4DzgUlAd2AvYA7w9RRiCZqZdUg7BkmYu+sR2APYBfgYOLHZ/ArgHeCs+PllwH3A3cBHwHPA/8lYvw9wP1APvAZMylh2GXAPMCt+7XKgOks8g4AtwAGtxNw1LqseeB34CVAWL+sA/BJ4F3gVOBdwoGPGa28F3gTWAFcAHbLU09Y27wM8BayPt+mbGctmAFfE00cCdRnLKoHfxvGvA34N7AC8B+yXsd6XgE+A3VqIrSze7tfj/9MsoGu87FHg+83Wfx4YG0//M/CnuL6Vmf/7OO4bgUeADcDXmpXz8/j/82n8vvl1PN+BPTPKuAH4Q7zOX4F/Aq4F3gdeBIbl8t7RI/mHzvjDdDDQmSgRNXH3j4kO3GMyZn8LuJfoLPxOYI6ZdTKzMuAhouTSFzgamGxmozJe+03gLqAb8CBRsmvJ0URJckErMV9PlMB3B44ATgPOjJedDRwPDAOqgROavXYm0ADsGa8zEhhPdtm2uVO8zY8RJejzgDvMrNUmnfgM+vdECbuKaH/d5e6biPbPuIzVTwEed/f6Foo6I34cRbQfKvj7Pr0zfm1jnYOJvsE9bGY7ESX9O+O4TwFuMLMhGWX/G1GC3xmYm1mpu18MPEP0wVLh7t/PsqknEn0w9QQ2Ac8SfXD2JPowvSaOLZf3jiQp7U8ePQr/IEo0b2VZdiXwp3j6MmBexrIyorPmw4ARwBvNXjsFmJ7x2sczlg0GPslS58WZ9bSwvANRIhmcMW8i8FQ8/WfgnIxlI4nP+IFe8Wt3zFh+CvBklrpa2+bDgLeIv2nEy2cDl8XTM2jhjB84iOjMtmML9Y0AVvP3by+LaPZNLGPdJ4B/z3i+N7A53s6dic7WB8TLfg5Mi6dPAp5pVtZNwE8z4p7VxnvmKWB8s3nNz/hvzlh2HrAi4/l+wPqMbc763tEj+UdB73qQovEu0NPMOrp7Q7NlvePljVY3Trj7VjOrI/qa7kAfM1ufsW4HojPDRm9lTG8EOmepc11cbzY9gXKiM+ZGrxOdLRLHs7rZskYDgE7Am2bWOK+s2frNZdtmgNXuvjVLHNlUAq+3sN24+3wz2wAcYWZvEn0reTBLOX34/D7oCPRy9zVm9jBwMtG1m5OBCfF6A4ARzf5XHYHbMp63tj9y9XbG9CctPK/IiKet944kSIk/TM8SnQWPJWqHByBuEjgW+L8Z61ZmLC8D+gFriZpOXnP3QXmI5wngN2ZW7e6LWlj+LtGZ7QDghXhef6L2eojOyCsz1u+fMb2aaFt7tpR4s8i2zQCVZlaWkfz7Ay+1Ud5qoH+WDz2ImqLGEX1Q3ufun2YpZy3RPmjUn+j/0JhgZwM/NbOngR2BJzPq/4u7ZzbhNddWN7357MZ3Nfl778h2UBt/gNz9A+By4HozGx23X1cRtWvX8Y9ngvub2dj4nvjJREl0HrAA+NDMfmxmO5pZBzPb18y+sh3xrCK6MDjbzI40s3Iz62xmJ5vZRe6+hegD6udmtrOZDQAuAG6Pi7gHmGRm/cxsV+CijLLfJGqT/y8z28XMysxsDzM7opWQsm3zfKLmlB/F++xI4BtE7fStWUD04XSlme0Ub9shGctvA8YQJf9ZrZQzG/iBmQ00swrgF8DdGR8mjxB9MPxnPL/xw+n3wF5mdmrjtQoz+4qZ7dNG3JneJrqukA95e+/I9lHiD5S7X010Zv9L4EOipLYaONqji46NfkfURvw+cCrRXSKb42T8DWAo0V0Z7wK3EF2A3R6TiC5U/obojplXiJLhQ/Hy84iS7qtEFx/vBKbFy24G/kh0sfA5ml20JroQXE70beF9oguNrTUtZdvmz4guWB9LtL03AKe5+4utbVjGvtoTeIPow/WkjOV1cdxO680d04g+JJ4m2uefEu2XxnI2EW3714j2T+P8j4iue5xM9K3hLaLmoB1ai7uZXwEnmNn7ZnbdNrzucxJ478g2svjCisjnmNllRBfvxrW1bqlIa5vNbBqw1t1/Ush6JUxq4xdJWdzMNpboVlORxKmpRyRFZvYzYBnw/939tbTjkTCoqUdEJDA64xcRCUy7aOPv2bOnV1VVpR2GiEi7UlNT866779Z8frtI/FVVVSxa1NLvekREJBsze72l+WrqEREJjBK/iEhglPhFRALTLtr4W7J582bq6ur49NNs/VmVjs6dO9OvXz86deqUdigiUgLabeKvq6tj5513pqqqiozudkuOu7Nu3Trq6uoYOHBg2uGISAlot009n376KT169CjppA9gZvTo0SOIbzYiUhjtNvEDJZ/0G4WynSJSGO068YuIyLZT4k9BbW0t++67b9phiEiglPhFSkzfyv6YWV4ffSv7t12xtBvt9q6ellxzzTVMmxYNyjR+/HgmT55MbW0to0ePZsSIESxevJi99tqLWbNm0aVLF2pqarjgggv4+OOP6dmzJzNmzKB3794ceeSRjBgxgieffJL169dz6623cthhh32uvquvvprbbruNsrIyjj32WK688kqWLFnCOeecw8aNG9ljjz2YNm0au+66KzU1NZx11ll06dKFQw89tKmMLVu2cNFFF/HUU0+xadMmzj33XCZOnFiwfSalZ23dak666X/yWubdEw/Oa3mSrpI546+pqWH69OnMnz+fefPmcfPNN7N48WIAVq5cyYQJE1i6dCm77LILN9xwA5s3b+a8887jvvvua0rKF198cVN5DQ0NLFiwgGuvvZbLL7/8c/X94Q9/YM6cOcyfP5/nn3+eH/3oRwCcdtppXHXVVSxdupT99tuv6bVnnnkm1113Hc8+++w/lHPrrbfStWtXFi5cyMKFC7n55pt57TV1yy4iySmZxD937lzGjBnDTjvtREVFBWPHjuWZZ6LhSysrKznkkGhs63HjxjF37lxWrlzJsmXLOOaYYxg6dChXXHEFdXV1TeWNHTsWgP3335/a2trP1ff4449z5pln0qVLFwC6d+/OBx98wPr16zniiGgc79NPP52nn376c/NPPfXUpnIee+wxZs2axdChQxkxYgTr1q1j1apV+d9BIiKxkmnqaW1Amea3Q5oZ7s6QIUM+dwbeaIcdonGoO3ToQENDQ4v15XqbZWvrujvXX389o0aNyqksEZEvqmTO+A8//HDmzJnDxo0b2bBhAw888EBTu/wbb7zRlOBnz57NoYceyt577019fX3T/M2bN7N8+fKc6xs5ciTTpk1j48aNALz33nt07dqVXXfdtembxm233cYRRxxBt27d6Nq1K3PnzgXgjjvuaCpn1KhR3HjjjWzevBmAl156iQ0bNnzBvSEikl3JnPEPHz6cM844gwMOOACILu4OGzaM2tpa9tlnH2bOnMnEiRMZNGgQ3/ve9ygvL+e+++5j0qRJfPDBBzQ0NDB58mSGDBmSU32jR49myZIlVFdXU15eznHHHccvfvELZs6c2XRxd/fdd2f69OkATJ8+venibubZ/fjx46mtrWX48OG4O7vtthtz5szJ+/4REWnULsbcra6u9uYDsaxYsYJ99tmnzdfW1tZy/PHHs2zZsqTCK4hct1fEzBK5q6c95Ar5R2ZW4+7VzeeXTFOPiIjkpuQTf1VVVbs/2xcRyaeST/wiIvKPlPhFRAKjxC8iEhglfhGRwJRk4s9374TqmVBESkliP+Ays2nA8cA77r5vPK87cDdQBdQCJ7r7+/muO9+9E6pnQhEpJUme8c8ARjebdxHwhLsPAp6In5eExl8In3322QwZMoSRI0fyySefsGTJEg488EC+/OUvM2bMGN5/P++fcyIi2ySxxO/uTwPvNZv9LWBmPD0T+Jek6k/DqlWrOPfcc1m+fDndunXj/vvvz9pNs4hIWgrdxt/L3d8EiP9+qcD1J2rgwIEMHToUiLpzfuWVV1rspllEJE1Fe3HXzCaY2SIzW1RfX592ODlp7MoZou6c169fn14wIoFKYujJUrvJo9C9c75tZr3d/U0z6w28k21Fd58KTIWok7ZCBZhPmd00H3bYYU3dNItIcpIYehJK6yaPQif+B4HTgSvjv79LopI+/Srz+k/q069yu1+brZtmEZG0JHk752zgSKCnmdUBPyVK+PeY2XeBN4BvJ1H3mtVvJFFsq5p3BvfDH/6waXrevHkFj0dEJJvEEr+7n5Jl0dFJ1SkiIm0r2ou7IiKSDCV+EZHAKPGLiARGiV9EJDBK/CIigSnJxF9V2Tevv9irquyb9iaJiORNoX/AVRCv163Fp389b+XZmQ/nrSwRkbSVZOJPwyWXXELPnj05//zzAbj44ovp1asXmzZt4p577mHTpk2MGTOGyy+/nA0bNnDiiSdSV1fHli1buOSSSzjppJNS3gIRCUVJNvWk4bvf/S4zZ0Y9Tm/dupW77rqLXr16sWrVKhYsWMCSJUuoqanh6aef5tFHH6VPnz48//zzLFu2jNGjmw9bICKSHCX+PKmqqqJHjx4sXryYxx57jGHDhrFw4cKm6eHDh/Piiy+yatUq9ttvPx5//HF+/OMf88wzz9C1a9e0wxeRgKipJ4/Gjx/PjBkzeOuttzjrrLN44oknmDJlChMnTvzcujU1NTzyyCNMmTKFkSNHcumll6YQsYiESIk/j8aMGcOll17K5s2bufPOO+nYsSOXXHIJ3/nOd6ioqGDNmjV06tSJhoYGunfvzrhx46ioqGDGjBlphy4iASnJxD+gX5+83okzoF+fnNYrLy/nqKOOolu3bnTo0IGRI0eyYsUKDjroIAAqKiq4/fbbefnll7nwwgspKyujU6dO3HjjjXmLVUSkLSWZ+GtXr0ml3q1btzJv3jzuvffepnnnn39+050+jfbYYw9GjRpV6PBERABd3M2bF154gT333JOjjz6aQYMGpR2OiEhWJXnGn4bBgwfz6quvph2GiEib2vUZv3u7HIp3m4WynSJSGO028Xfu3Jl169aVfFJ0d9atW0fnzp3TDkVESkS7berp168fdXV11NfXpx1K4jp37ky/fv3SDkNESkS7TfydOnVi4MCBaYchItLutNumHhER2T5K/CIigVHiFxEJjBK/iEhglPhFRAKjxC8iEhglfhGRwCjxi4gERolfRCQwqSR+M/uBmS03s2VmNtvM1BGNiEiBFDzxm1lfYBJQ7e77Ah2Akwsdh4hIqNJq6ukI7GhmHYEuwNqU4hARCU7BE7+7rwF+CbwBvAl84O6PNV/PzCaY2SIzWxRCD5wiRa2sI2aW90ffyv5pb1mQCt47p5ntCnwLGAisB+41s3Hufnvmeu4+FZgKUF1dXdqd7osUu60NnHTT/+S92LsnHpz3MqVtaTT1fA14zd3r3X0z8FtA/30RkQJJI/G/ARxoZl3MzICjgRUpxCEiEqQ02vjnA/cBzwF/i2OYWug4RERClcoIXO7+U+CnadQtIhI6/XJXRCQwSvwiIoFR4hcRCYwSv4hIYJT4RUQCo8QvIhIYJX4RkcAo8YuIBEaJX0QkMEr8IiKBUeIXEQmMEr+ISGCU+EVEAqPEL5KSvpX9ExnOUKQtqXTLLCKwtm61hjOUVOiMX0QkMEr8IiKBUeIXEQmMEr+ISGCU+EVEAqPELyISGCV+EZHAKPGLiARGiV9EJDA5JX4zOySXeSIiUvxyPeO/Psd5IiJS5Frtq8fMDgIOBnYzswsyFu0CdEgyMBERSUZbnbSVAxXxejtnzP8QOCGpoEREJDmtJn53/wvwFzOb4e6vFygmERFJUK7dMu9gZlOBqszXuPtXt6dSM+sG3ALsCzhwlrs/uz1liYjItsk18d8L/DdRst6Sh3p/BTzq7ieYWTnQJQ9liohIDnJN/A3ufmM+KjSzXYDDgTMA3P0z4LN8lC0iIm3L9XbOh8zs382st5l1b3xsZ527A/XAdDNbbGa3mNlOzVcyswlmtsjMFtXX129nVRKapIYz7FjeWUMkJqGso/ZrCnI94z89/nthxjwnSuLbU+dw4Dx3n29mvwIuAi7JXMndpwJTAaqrq3076pEAJTmcYb7L1RCJwNYG7dcU5JT43X1gHuusA+rcfX78/D6ixC8iIgWQU+I3s9Namu/us7a1Qnd/y8xWm9ne7r4SOBp4YVvLERGR7ZNrU89XMqY7EyXr54BtTvyx84A74jt6XgXO3M5yRERkG+Xa1HNe5nMz6wrctr2VuvsSoHp7Xy8iIttve7tl3ggMymcgIiJSGLm28T9EdBcPRJ2z7QPck1RQIiKSnFzb+H+ZMd0AvO7udQnEIyIiCcupqSfurO1Foh46d0W/tBURabdyHYHrRGAB8G3gRGC+malbZhGRdijXpp6Lga+4+zsAZrYb8DjRj69ERKQdyfWunrLGpB9btw2vFRGRIpLrGf+jZvZHYHb8/CTgkWRCEhGRJLU15u6eQC93v9DMxgKHAgY8C9xRgPhERCTP2mquuRb4CMDdf+vuF7j7D4jO9q9NNjQREUlCW4m/yt2XNp/p7ouIhmEUEZF2pq3E37mVZTvmMxARESmMthL/QjM7u/lMM/suUJNMSCIikqS27uqZDDxgZt/h74m+GigHxiQYlwSgb2V/1tatTjsMkeC0mvjd/W3gYDM7Ctg3nv2wu/858cik5CUxTKKG3RNpW6798T8JPJlwLCIiUgD69a2ISGCU+EVEAqPELyISGCV+EZHAKPGLiARGiV9EJDBK/CIigVHiFxEJjBK/iEhglPhFRAKjxC8iEhglfhGRwKSW+M2sg5ktNrPfpxWDiEiI0jzjPx9YkWL9IiJBSiXxm1k/4OvALWnULyISsrTO+K8FfgRszbaCmU0ws0Vmtqi+vr5ggYmItKisI2aW10ffyv6pbEpOA7Hkk5kdD7zj7jVmdmS29dx9KjAVoLq62gsTnYhIFlsbSmbEuDTO+A8BvmlmtcBdwFfN7PYU4hARCVLBE7+7T3H3fu5eBZwM/NndxxU6DhGRUOk+fhGRwBS8jT+Tuz8FPJVmDCIiodEZv4hIYJT4RUQCo8QvIhIYJX4RkcAo8YuIBEaJX0QkMEr8IiKBUeIXEQmMEr+ISGCU+EVEAqPELyISGCV+EZHAKPGLiAQm1d45Jf/6VvZnbd3qvJbZodMObNm8Ka9lClDWMbURmEqa9mublPhLzNq61YkMD5fvMhvLDdrWBnz61/NerJ35cN7LbFe0X9ukph4RkcAo8YuIBEaJX0QkMEr8IiKBUeIXEQmMEr+ISGCU+EVEAqPELyISGCV+EZHAKPGLiARGiV9EJDBK/CIigVHiFxEJjBK/iEhgCp74zazSzJ40sxVmttzMzi90DCIiIUujP/4G4D/c/Tkz2xmoMbM/ufsLKcQiIhKcgp/xu/ub7v5cPP0RsALoW+g4RERCleoIXGZWBQwD5rewbAIwAaB///45l9lehh7s06+SNavfyGuZkqAkhvMra0cD4Gk4QyyBfdChU3ley8tVau88M6sA7gcmu/uHzZe7+1RgKkB1dbXnWm57GXow9IOo3UlgOL92NZSfhjPES+g9kMpdPWbWiSjp3+Huv00jBhGRUKVxV48BtwIr3P2aQtcvIhK6NM74DwFOBb5qZkvix3EpxCEiEqSCt/G7+1zACl2viIhE9MtdEZHAKPGLiARGiV9EJDBK/CIigVHiFxEJjBK/iEhglPhFRAKjxC8iEhglfhGRwCjxi4gERolfRCQwSvwiIoFR4hcRCUw7Gvut9ERDE7QDSQ27p+H8EpHEEIFJDRPZnmItJdpDKWo3Q9klOOxeqQxlV0za0xCB7SnWUqKmHhGRwCjxi4gERolfRCQwSvwiIoFR4hcRCYwSv4hIYJT4RUQCo8QvIhIYJX4RkcAo8YuIBEaJX0QkMEr8IiKBUeIXEQmMEr+ISGBSSfxmNtrMVprZy2Z2URoxiIiEquCJ38w6AL8BjgUGA6eY2eBCxyEiEqo0zvgPAF5291fd/TPgLuBbKcQhIhIkc/fCVmh2AjDa3cfHz08FRrj795utNwGYED/dF1hW0EC3X0/g3bSDyJFiTYZiTYZi3XYD3H235jPTGHqxpYFmP/fp4+5TgakAZrbI3auTDiwfFGsyFGsyFGsyij3WNJp66oDKjOf9gLUpxCEiEqQ0Ev9CYJCZDTSzcuBk4MEU4hARCVLBm3rcvcHMvg/8EegATHP35W28bGrykeWNYk2GYk2GYk1GUcda8Iu7IiKSLv1yV0QkMEr8IiKBSavLhkoze9LMVpjZcjM7P57f3cz+ZGar4r+7ZrxmStzFw0ozG5Ux/xQz+5uZLTWzR82sZ5qxmlmPeP2PzezXzcraP471ZTO7zsxaurU19VjNrIuZPWxmL8blXJnPOPMZa7MyHzSzvP/eI8/vgXIzm2pmL8X791+LONZiO7aOMbOaOKYaM/tqRlnFdmy1GGshjq2cuHvBH0BvYHg8vTPwElH3DVcDF8XzLwKuiqcHA88DOwADgVeILgx3BN4BesbrXQ1clnKsOwGHAucAv25W1gLgIKLfMvwBOLYYYwW6AEfF0+XAM8Uaa0Z5Y4E7gWVF8H5t7T1wOXBFPF3W+N4ttliL9NgaBvSJp/cF1hTxsdVirIU4tnLankJXmGWn/g44BlgJ9M7Y0Svj6SnAlIz1/xj/kzsB9cCA+B/+38CENGPNWO+MZgdSb+DFjOenADcVY6wtlPMr4OxijRWoAObGB2LeE3+eY10N7JR0jHl4vxbtsRXPN2Ad0clg0R5bzWNtYVnix1ZLj9Tb+M2siujTcT7Qy93fBIj/filerS/RAdOoDujr7puB7wF/I/oR2GDg1pRjzaYvUdyN6uJ5ifiCsWaW0w34BvBE/qNsqqOKLxbrz4D/AjYmFWOjLxJrvC8BfmZmz5nZvWbWqxhjbQfH1r8Ci919E8V/bGXGmllONxI+trJJNfGbWQVwPzDZ3T9sbdUW5rmZdSJ6cw4D+gBLib4d5N02xJq1iBbmJXIvbR5ibSynIzAbuM7dX81XfM3q+EKxmtlQYE93fyDfsbVQ1xfdrx2Jfqn+V3cfDjwL/DKPITbJw34t2mPLzIYAVwETG2e1sFpRHFstxNo4P/FjqzWpJf74jXU/cIe7/zae/baZ9Y6X9yZqY4Ts3TwMBXD3Vzz63nQPcHDKsWZTF8fdKJGuKvIUa6OpwCp3vzbfccax5CPWg4D9zayWqLlnLzN7qkhjXUf0raTxQ+peYHiRxjoUiu/YMrN+RPvvNHd/JZ5dlMdWllgbJXpstSWtu3qM6GvjCne/JmPRg8Dp8fTpRO1ojfNPNrMdzGwgMIjoYs4aYLCZNfY+dwywIuVYWxR/DfzIzA6MyzytrdekFWtc1hVAV2ByPmPMKD9f+/VGd+/j7lVEFylfcvcjizRWBx4CGuM7GnihGGOlCI+tuGnkYaLrfX9tXLkYj61sscbLEj22clLoiwrxBY1Dib6KLQWWxI/jgB5E7V2r4r/dM15zMdHdPCvJuApOdDfCirish4AeRRBrLfAe8DHR2cjgeH41UffSrwC/Jv7ldLHFSnTG5PF+bSxnfDHG2qzMKpK5qyef74EBwNNxWU8A/Ys41qI6toCfABsy1l0CfKkYj61ssVKAYyuXh7psEBEJTOp39YiISGEp8YuIBEaJX0QkMEr8IiKBUeIXEQmMEr+ISGCU+EUKxMw6pB2DCCjxi7TIzH7W2Od6/PznZjbJzC40s4UW9VF/ecbyOXG/68vNbELG/I/N7D/NbD5R9xIiqVPiF2nZrcQ/xTezMuBk4G2i7kIOIOrLZn8zOzxe/yx335/oF6STzKxHPH8nol8Tj3D3uQWMXySrjmkHIFKM3L3WzNaZ2TCgF7AY+AowMp6GaByAQURdMEwyszHx/Mp4/jpgC1HHXiJFQ4lfJLtbiAYo+SdgGlGnav/P3W/KXMnMjgS+Bhzk7hvj3kE7x4s/dfctBYpXJCdq6hHJ7gFgNNGZ/h/jx1lxn+yYWV8z+xJRT4vvx0n/n4ED0wpYJBc64xfJwt0/M7MngfXxWftjZrYP8GzUSy8fA+OAR4FzzGwpUe+x89KKWSQX6p1TJIv4ou5zwLfdfVXa8Yjki5p6RFpgZoOBl4EnlPSl1OiMX0QkMDrjFxEJjBK/iEhglPhFRAKjxC8iEhglfhGRwPwvSIJlbB+kkpQAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'For years with at least four publications, the fraction of papers with open code is depicted.'"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Plot the histogram\n",
    "df_ = papers_annotated_df.copy()\n",
    "df_[\"year\"] = df_.index \n",
    "df_[\"year\"] = df_[\"year\"].apply(lambda x: 2000+int(x[-2:])) - 0.5\n",
    "\n",
    "df_[\"open code\"] = df_[\"open code\"].apply(lambda row: [\"yes\" if str(row).lower().strip() == \"yes\" else \"no\"][0])\n",
    "sns.histplot(data=df_, x=\"year\", hue=\"open code\", multiple=\"stack\", binwidth=0.999).set(title=f\"Open Code policy over time\")\n",
    "plt.xlim(2008, None)\n",
    "plt.savefig(\"Figures/opencode_over_time.pdf\", format=\"pdf\", bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\"For years with at least four publications, the fraction of papers with open code is depicted.\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "be42c831-8bb6-4fe7-9b91-ef23126b78f6",
   "metadata": {},
   "source": [
    "### Fraction of fit use-cases and online scores over year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "472488ea-569e-4f11-a7fd-bb30bf2c7086",
   "metadata": {},
   "outputs": [],
   "source": [
    "pipelines_annotated_df_proposed = pipelines_annotated_df.loc[pipelines_annotated_df[\"is new\"]==1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "b83d3f8e-7f11-4977-ac68-07fb5f2ff52a",
   "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>fit score (0-2)</th>\n",
       "      <th>online score (0-2)</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>year</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2011</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2012</th>\n",
       "      <td>0.800000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.142857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017</th>\n",
       "      <td>0.583333</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018</th>\n",
       "      <td>0.846154</td>\n",
       "      <td>0.230769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2019</th>\n",
       "      <td>0.125000</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020</th>\n",
       "      <td>0.166667</td>\n",
       "      <td>0.166667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2021</th>\n",
       "      <td>0.888889</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2022</th>\n",
       "      <td>0.791667</td>\n",
       "      <td>0.750000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      fit score (0-2)  online score (0-2)\n",
       "year                                     \n",
       "2011         0.000000            0.000000\n",
       "2012         0.800000            1.000000\n",
       "2013         0.000000            0.000000\n",
       "2014              NaN                 NaN\n",
       "2015         0.500000            0.500000\n",
       "2016         0.000000            0.142857\n",
       "2017         0.583333            0.000000\n",
       "2018         0.846154            0.230769\n",
       "2019         0.125000            0.500000\n",
       "2020         0.166667            0.166667\n",
       "2021         0.888889            0.000000\n",
       "2022         0.791667            0.750000"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def share_of_twos(x):\n",
    "    if len(x) >= 4: \n",
    "        return np.sum(x==2) / len(x)\n",
    "    return np.NaN\n",
    "\n",
    "df_3 = pipelines_annotated_df_proposed.groupby(\"year\").agg({\"fit score (0-2)\": share_of_twos, \"online score (0-2)\": share_of_twos})\n",
    "df_3.loc[2011:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "e9b1fefc-846a-468d-92ba-43e5965b3bd0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'The fraction of proposed pipelines that fit an use case with BCI and are online capable is rising slightly. Only years with four or more proposed pipelines are taken into account.'"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sns.lineplot(data=df_3, x=\"year\", y=\"fit score (0-2)\", marker=\"o\", label=\"Fraction of fit pipelines\")\n",
    "sns.lineplot(data=df_3, x=\"year\", y=\"online score (0-2)\", marker=\"D\", linestyle=\"--\", label=\"Fraction of online pipelines\")\n",
    "plt.ylabel(\"Fraction of pipelines\")\n",
    "# plt.ylim(0,1)\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\"The fraction of proposed pipelines that fit an use case with BCI and are online capable is rising slightly. Only years with four or more proposed pipelines are taken into account.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "9d3107c3-b882-40de-b8e6-a2d6d10bb8f6",
   "metadata": {},
   "outputs": [],
   "source": [
    "def number_of_fulls(x):\n",
    "    return np.sum(x==2)\n",
    "\n",
    "def number_of_non_fulls(x):\n",
    "    return len(x) - np.sum(x==2)\n",
    "\n",
    "# df_3 = pipelines_annotated_df_proposed.groupby(\"year\").agg({\"fit score (0-2)\": number_of_fulls, \n",
    "#                                                             \"online score (0-2)\": number_of_fulls})\n",
    "# df_3.columns = [\"fit\", \"online\"]\n",
    "df_3 = pipelines_annotated_df_proposed.groupby(\"year\").agg({\"fit score (0-2)\": number_of_fulls, \n",
    "                                                            \"online score (0-2)\": number_of_fulls,\n",
    "                                                           \"is new\": \"sum\"})\n",
    "df_3.columns = [\"fit\", \"online\", \"total\"]\n",
    "# df_3 = pipelines_annotated_df_proposed.groupby(\"year\").agg({\"fit score (0-2)\": [number_of_fulls, number_of_non_fulls], \n",
    "#                                                             \"online score (0-2)\": [number_of_fulls, number_of_non_fulls]})\n",
    "# df_3.columns = [\"fit\", \"not fit / unknown\", \"online\", \"not online / unknown\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "52141935-db96-46c9-95bb-7b87ebedce55",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "df_3[\"year\"] = df_3.index\n",
    "\n",
    "df_4 = pd.melt(df_3, id_vars=\"year\", var_name=\"score\", value_name=\"sum\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "f9db1ac3-8136-4b6a-af65-ccd560e105da",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'The number of totally proposed pipelines per year since 2011 with fraction of pipelines that are fit for an use case with BCI and are online capable. In 2021, none of the proposed pipelines were online capable.'"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "g = sns.catplot(data=df_4[df_4[\"year\"] >= 2011],\n",
    "            x=\"year\", y=\"sum\", hue=\"score\", kind=\"bar\",\n",
    "                palette=[\"C0\", \"C1\", \"grey\"], legend=None\n",
    "           ).set(title=f\"Fit and online pipelines over time\")\n",
    "plt.ylabel(\"Number of pipelines\")\n",
    "# plt.ylim(0,1)\n",
    "g.fig.set_figwidth(8)\n",
    "plt.legend()\n",
    "\n",
    "plt.savefig(\"Figures/fractions_over_time.pdf\", format=\"pdf\", bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\"The number of totally proposed pipelines per year since 2011 with fraction of pipelines that are fit for an use case with BCI and are online capable. In 2021, none of the proposed pipelines were online capable.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "79122570-7c10-470b-9bf2-91423f946d51",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "79d6ffcb-8c51-4243-ba46-3f4bd607a919",
   "metadata": {},
   "source": [
    "### Evaluation metric"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "dfb8ce7c-c083-407e-91fd-320352a65200",
   "metadata": {},
   "outputs": [],
   "source": [
    "evametrics_d = dict()\n",
    "for metrics in list(papers_annotated_df[\"evaluation metrics\"].apply(lambda x: [m.split(\"(\")[0].strip().lower() for m in str(x).split(\",\")]).values):\n",
    "    for m in metrics:\n",
    "        evametrics_d[m] = evametrics_d.get(m, 0) + 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "05916030-f25c-48c1-9320-7e609f016d6e",
   "metadata": {},
   "outputs": [],
   "source": [
    "# sorted(evametrics_d.items(), key=lambda x: x[0])  # TODO check for the items, remove \"\" and \"-\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "4e7860a8-ef96-4837-8d52-632fc8350ea6",
   "metadata": {},
   "outputs": [],
   "source": [
    "if \"\" in evametrics_d: evametrics_d.pop(\"\")\n",
    "if \"-\" in evametrics_d: evametrics_d.pop(\"-\")\n",
    "if \"nan\" in evametrics_d: evametrics_d.pop(\"nan\")\n",
    "evametrics_d[\"rmse\"] += evametrics_d.pop(\"rrmse\", 0)\n",
    "evametrics_d[\"rmse\"] += evametrics_d.pop(\"mse\", 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "d0556797-fb23-4f6a-aef9-05933f9df794",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[('accuracy', 22), ('visual', 14), ('snr', 11), ('rmse', 7), ('sensitivity', 5), ('pds', 5), ('specificity', 4), ('f1-score', 4), ('correlation coefficient', 4)]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'Number of totally used evaluation metrics in publications. It can be seen that the majority of used metrics are used infrequently. The accuracy, visual proof and the signal-to-noise ratio (SNR) are the most comment evaluation metrics.'"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get the journals in order by ascending counts, also in the dataframe\n",
    "top_n = 9\n",
    "metric_map = {\n",
    "    \"snr\": \"SNR\",\n",
    "    \"rmse\": \"RMSE\",\n",
    "    \"f1-score\": \"F1-score\",\n",
    "    \"auc\": \"AUC\",\n",
    "    \"pds\": \"power density spectrum\"\n",
    "}\n",
    "\n",
    "metrics_oi = sorted(evametrics_d.items(), key=lambda x: -x[1])[:top_n]\n",
    "metrics_oi.append((\"other metrics\", sum([count for m, count in sorted(evametrics_d.items(), key=lambda x: -x[1])[top_n:]])))\n",
    "metrics_oi = [[f\"{metric_map.get(p, p)}: {v}\", v] for p,v in metrics_oi]\n",
    "\n",
    "# label_dict = {j: f\"{j}: {j_counts[i]}\" for i, j in enumerate(journals)}\n",
    "print(sorted(evametrics_d.items(), key=lambda x: -x[1])[:top_n])\n",
    "plt.pie(np.array(metrics_oi)[:,1], labels=np.array(metrics_oi)[:,0])\n",
    "plt.title(f\"The {top_n} most common evaluation metrics\")\n",
    "plt.savefig(\"Figures/mostcommon_metrics.pdf\", format=\"pdf\", bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\"Number of totally used evaluation metrics in publications. It can be seen that the majority of used metrics are used infrequently. The accuracy, visual proof and the signal-to-noise ratio (SNR) are the most comment evaluation metrics.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a7d090e2-8204-4811-81de-eb1fd25fe5c3",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "bac39d76-6461-4c79-9d46-2214ad300a4f",
   "metadata": {},
   "source": [
    "### Distribution of the sample size and number of channels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "a137eec2-2424-4346-a692-14c421ff137a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([  1.  ,   7.  ,  10.  ,  18.25, 200.  ]), 16.41304347826087)"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "samples = list()\n",
    "for sizes in papers_annotated_df[\"#sample\"].apply(lambda x: [s.strip() for s in str(x).split(\",\")]):\n",
    "    for s in sizes:\n",
    "        if s != \"nan\":\n",
    "            samples.append(int(s))\n",
    "np.quantile(samples, [0,0.25,0.50,0.75,1.00]), np.mean(samples)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "e9e87199-e9c2-4a8c-9ef4-20fbd35dfd47",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.07608695652173914"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean([s>=30 for s in samples])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "76c8e21a-a92f-4021-a4a9-50ee6f8ce065",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([  2.,  19.,  64.,  97., 256.]), 69.28767123287672)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n_channels = list()\n",
    "for sizes in pipelines_annotated_df.groupby(\"Reference\").agg({\"EEG (#ch)\": \"unique\"})[\"EEG (#ch)\"].values:\n",
    "    for s in list(sizes):\n",
    "        if not pd.isnull(s):\n",
    "            # print(s, sizes)\n",
    "            n_channels.append(int(s))\n",
    "np.quantile(n_channels, [0,0.25,0.50,0.75,1.00]), np.mean(n_channels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "7cbaa204-822a-4411-a5a9-593c5322472a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'Number of subjects over all used datasets and number of used EEG channels per publication. Half of the datasets consist of less than ten participants.'"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(2,1, figsize=(8,4))\n",
    "sns.boxplot(x=samples, ax=ax[0])\n",
    "ax[0].set_ylabel(\"sample size\")\n",
    "ax[0].set_xticks(range(0, 201, 10))\n",
    "sns.boxplot(x=n_channels, ax=ax[1])\n",
    "ax[1].set_ylabel(\"#channels\")\n",
    "ax[1].set_xticks(range(0, 261, 20))\n",
    "ax[0].set_title(\"Number of subjects and channels\")\n",
    "plt.savefig(\"Figures/subjects_channels_boxplot.pdf\", format=\"pdf\", bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\"Number of subjects over all used datasets and number of used EEG channels per publication. Half of the datasets consist of less than ten participants.\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "44bc8681",
   "metadata": {},
   "source": [
    "### Most common BCI-paradigms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "4b4694f7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('motor execution', 28),\n",
       " ('attention strategy', 25),\n",
       " ('-', 20),\n",
       " ('motor imagery', 8),\n",
       " ('ssvep', 2),\n",
       " ('nan', 1),\n",
       " ('mental arithmetic', 1),\n",
       " ('p300', 1),\n",
       " ('motor exectution', 1),\n",
       " ('aep', 1),\n",
       " ('vep', 1),\n",
       " ('motor imagery tasks', 1),\n",
       " ('motor tapping study', 1),\n",
       " ('nine different cognitive tasks', 1),\n",
       " ('...)', 1)]"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bcimetrics_d = dict()\n",
    "for metrics in list(papers_annotated_df[\"mental strategy\"].astype(\"object\").apply(\n",
    "    lambda x: [str(m).split(\"(\")[0].strip().lower() for m in str(x).split(\",\")]).values):\n",
    "    if str(metrics) == \"nan\":\n",
    "        continue\n",
    "    # print(metrics)\n",
    "    for m in metrics:\n",
    "        bcimetrics_d[m] = bcimetrics_d.get(m, 0) + 1\n",
    "\n",
    "sorted(bcimetrics_d.items(), key=lambda x: -x[1])  # TODO check for the items, remove \"\" and \"-\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "63c8e17c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('motor execution', 28),\n",
       " ('attention strategy', 25),\n",
       " ('-', 20),\n",
       " ('other', 10),\n",
       " ('motor imagery', 8),\n",
       " ('ssvep', 2)]"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "BCI_PARADIGM_MIN_COUNT = 2\n",
    "\n",
    "bcimetrics_d[\"other\"] = 0\n",
    "for key, count in bcimetrics_d.items():\n",
    "    if count < BCI_PARADIGM_MIN_COUNT:\n",
    "        bcimetrics_d[\"other\"] += bcimetrics_d.get(key, 0)\n",
    "bcimetrics_d = {key: count for key, count in bcimetrics_d.items() if count >= BCI_PARADIGM_MIN_COUNT}\n",
    "sorted(bcimetrics_d.items(), key=lambda x: -x[1])  # TODO check for the items, remove \"\" and \"-\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "2f59316b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[('motor execution', 28), ('attention strategy', 25), ('-', 20), ('other', 10), ('motor imagery', 8), ('ssvep', 2)]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'Number of used BCI paradigms in publications.'"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get the journals in order by ascending counts, also in the dataframe\n",
    "top_n = 9\n",
    "metric_map = {\n",
    "    \"-\": \"no paradigm\",\n",
    "    \"ssvep\": \"SSVEP\",\n",
    "    \"other\": \"other paradigm\",\n",
    "}\n",
    "\n",
    "metrics_oi = sorted(bcimetrics_d.items(), key=lambda x: -x[1])[:top_n]\n",
    "# metrics_oi.append((\"other pardigm\", sum([count for m, count in sorted(bcimetrics_d.items(), key=lambda x: -x[1])[top_n:]])))\n",
    "metrics_oi = [[f\"{metric_map.get(p, p)}: {v}\", v] for p,v in metrics_oi]\n",
    "\n",
    "# label_dict = {j: f\"{j}: {j_counts[i]}\" for i, j in enumerate(journals)}\n",
    "print(sorted(bcimetrics_d.items(), key=lambda x: -x[1])[:top_n])\n",
    "plt.pie(np.array(metrics_oi)[:,1], labels=np.array(metrics_oi)[:,0])\n",
    "plt.title(f\"The most common BCI paradigms\")\n",
    "plt.savefig(\"Figures/mostcommon_bciparadigms.pdf\", format=\"pdf\", bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\"Number of used BCI paradigms in publications.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d001e82c",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "c4ed91f6-341b-4204-ac88-69e9f70185cd",
   "metadata": {},
   "source": [
    "### Distribution of the Filter's cutoff-frequencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "541e2a3d-2689-466e-b85c-a80595f50d26",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1: LP 200Hz (cutoff-f=200-300Hz) + Notch (50Hz) + BP 8-35Hz (5th-order Butterworth; gain=1.5) , 200\n",
      "could not convert string to float: 'order Butterworth; gain=1.5)'\n",
      "1: LP 200Hz (+Powerline-BP) , 200\n",
      "could not convert string to float: ')'\n",
      "1: LP 1000Hz (+Powerline-BP) , 1000\n",
      "could not convert string to float: ')'\n",
      "1: LP 200Hz (+Powerline-BP) , )\n",
      "could not convert string to float: ')'\n",
      "1: LP 1000Hz (+Powerline-BP) , )\n",
      "could not convert string to float: ')'\n"
     ]
    }
   ],
   "source": [
    "def extract_highpass_cutoff(x):\n",
    "    x = x.split(\"\\t\")[-1]\n",
    "    values = list()\n",
    "    if \"HP\" in x:\n",
    "        values.append(\n",
    "            float(x.split(\"HP\")[-1].split(\"Hz\")[0].split(\"(\")[0].strip())\n",
    "        )\n",
    "    if \"BP\" in x:\n",
    "        try:\n",
    "            values.append(\n",
    "                float(x.split(\"BP\")[-1].split(\"-\")[0].split(\"Hz\")[0].split(\"(\")[0].strip())\n",
    "            )\n",
    "        except Exception as e:\n",
    "            print(x, \",\", x.split(\"BP\")[-1].split(\"-\")[0].split(\"Hz\")[0].split(\"(\")[0].strip())\n",
    "            print(e)\n",
    "            time.sleep(1)\n",
    "    return values\n",
    "\n",
    "def extract_lowpass_cutoff(x):\n",
    "    x = x.split(\"\\t\")[-1]\n",
    "    values = list()\n",
    "    try:\n",
    "        if \"LP\" in x:\n",
    "            values.append(\n",
    "                float(x.split(\"LP\")[-1].split(\"Hz\")[0].split(\"(\")[0].split(\",\")[0].strip())\n",
    "            )\n",
    "        if \"BP\" in x:\n",
    "            values.append(\n",
    "                float(x.split(\"BP\")[-1].split(\"-\")[-1].split(\"Hz\")[0].split(\"(\")[0].split(\",\")[0].strip())\n",
    "            )\n",
    "    except Exception as e:\n",
    "        print(x, \",\", x.split(\"LP\")[-1].split(\"Hz\")[0].split(\"(\")[0].split(\",\")[0].strip())\n",
    "        print(e)\n",
    "        time.sleep(0.1)\n",
    "    return values\n",
    "\n",
    "pipelines_annotated_df[\"Ref+Filter\"] = pipelines_annotated_df.index + \"\\t\" + pipelines_annotated_df[\"filter\"]\n",
    "lowpass_coffs = list(map(lambda x: extract_lowpass_cutoff(str(x)), pipelines_annotated_df[\"Ref+Filter\"].unique()))\n",
    "highpass_coffs = list(map(lambda x: extract_highpass_cutoff(str(x)), pipelines_annotated_df[\"Ref+Filter\"].unique()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "af568a54-b2ba-4cce-84a2-677b9e6f4930",
   "metadata": {},
   "outputs": [],
   "source": [
    "lps = [v for sublist in lowpass_coffs for v in sublist]\n",
    "hps = [v for sublist in highpass_coffs for v in sublist]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "81933c5b-895c-49a6-9dde-6658ba326055",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'Cutoff-frequencies of all filters, categorised into low-passes and high-passes and plotted on a logarithmic scale.'"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(2,1, figsize=(8,4))\n",
    "ax[0].set_title(\"Cutoff frequencies of the filters\")\n",
    "plot = sns.boxplot(x=lps, ax=ax[0]).set(xscale=\"log\")\n",
    "ax[0].set_ylabel(\"Low pass\")\n",
    "ax[0].set_xticks([0.1, 1, 10, 100, 1000])\n",
    "ax[0].set_xlim(0.04, 1200)\n",
    "\n",
    "sns.boxplot(x=hps, ax=ax[1]).set(xscale=\"log\")\n",
    "ax[1].set_ylabel(\"High pass\")\n",
    "ax[1].set_xticks([0.1, 1, 10, 100, 1000])\n",
    "ax[1].set_xlim(0.04, 1200)\n",
    "ax[1].set_xlabel(\"Cutoff frequency (Hz)\")\n",
    "\n",
    "plt.savefig(\"Figures/filters_boxplot.pdf\", format=\"pdf\", bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\"Cutoff-frequencies of all filters, categorised into low-passes and high-passes and plotted on a logarithmic scale.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "74274361-85e9-47b6-b76a-5ba5e7d4ef1c",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "72ecbc31-b482-4d76-9a27-0660a76b46fe",
   "metadata": {},
   "source": [
    "## Analysis for the Methods"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8009ee9e-3230-4923-b49e-812070b8e608",
   "metadata": {},
   "source": [
    "### Decomposition methods"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7e461983-6403-4087-b909-5cd6cd10f9d7",
   "metadata": {},
   "source": [
    "#### Decomposition methods (number of pipelines)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "46ce9b9e-e925-4ac2-9823-221beeff576a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Fourier Transf.     29\n",
       "Wavelet Transf.     48\n",
       "CCA                 23\n",
       "ICA                112\n",
       "PCA                 42\n",
       "EMD                 24\n",
       "Other               51\n",
       "dtype: int64"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distr = (~pipelines_annotated_df[['Fourier Transf.', 'Wavelet Transf.', 'Welch', 'LSP', 'PCC', 'CCA', 'ICA', 'NMF', 'PCA', 't-SNE', 'EMD', 'IVA', 'CSP', 'Other.1']].isnull()).sum(axis=0)\n",
    "distr[\"Other\"] = distr[\"Other.1\"]\n",
    "\n",
    "distr[\"Other\"] += distr[\"Welch\"]\n",
    "distr[\"Other\"] += distr[\"LSP\"]\n",
    "distr[\"Other\"] += distr[\"PCC\"]\n",
    "distr[\"Other\"] += distr[\"NMF\"]\n",
    "distr[\"Other\"] += distr[\"t-SNE\"]\n",
    "distr[\"Other\"] += distr[\"IVA\"]\n",
    "distr[\"Other\"] += distr[\"CSP\"]\n",
    "[distr.pop(key) for key in ['Welch', 'LSP', 'PCC', 'NMF', 't-SNE', 'IVA', 'CSP', 'Other.1']]\n",
    "distr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "cb8fc67d-44c0-425e-82ca-e6a5272e2a26",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'Number of totally used decomposition methods.'"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels = [f\"{p}: {v}\" for p,v in distr.items()]\n",
    "plt.pie(distr, labels=labels)\n",
    "plt.title(f\"Most common decomposition methods\")\n",
    "plt.savefig(\"Figures/mostcommon_decompositions.pdf\", format=\"pdf\", bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\"Number of totally used decomposition methods.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6ed4afb9-42bb-46a6-8bea-deb4c9d8931b",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "758b8b0d-d8f6-4d05-9eff-7b3ac9dc6ea8",
   "metadata": {},
   "source": [
    "#### Decomposition methods (total number of publications)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "d9522c4e-4948-4e31-9c84-932a64e1d42c",
   "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>total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Fourier Transf.</th>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wavelet Transf.</th>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CCA</th>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ICA</th>\n",
       "      <td>36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PCA</th>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 total\n",
       "Fourier Transf.     10\n",
       "Wavelet Transf.     12\n",
       "CCA                  9\n",
       "ICA                 36\n",
       "PCA                 18"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distr = (pipelines_annotated_df.groupby(\"Reference\").agg(\"count\")[\n",
    "    ['Fourier Transf.', 'Wavelet Transf.', 'Welch', 'LSP', 'PCC', 'CCA', 'ICA', 'NMF', 'PCA', 't-SNE', 'EMD', 'IVA', 'CSP', 'Other.1']\n",
    "        ]>0).sum(axis=0)\n",
    "distr[\"Other\"] = distr[\"Other.1\"]\n",
    "\n",
    "distr[\"Other\"] += distr[\"Welch\"]\n",
    "distr[\"Other\"] += distr[\"LSP\"]\n",
    "distr[\"Other\"] += distr[\"PCC\"]\n",
    "distr[\"Other\"] += distr[\"NMF\"]\n",
    "distr[\"Other\"] += distr[\"t-SNE\"]\n",
    "distr[\"Other\"] += distr[\"IVA\"]\n",
    "distr[\"Other\"] += distr[\"CSP\"]\n",
    "[distr.pop(key) for key in ['Welch', 'LSP', 'PCC', 'NMF', 't-SNE', 'IVA', 'CSP', 'Other.1']]\n",
    "\n",
    "# store in df_decomp\n",
    "df_decomp = pd.DataFrame(data=distr, columns=[\"total\"])\n",
    "df_decomp.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "684278fb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'Number of totally used decomposition methods.'"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels = [f\"{p}: {v}\" for p,v in distr.items()]\n",
    "plt.pie(distr, labels=labels)\n",
    "plt.title(f\"Most common decomposition methods\")\n",
    "# plt.savefig(\"Figures/mostcommon_ICAflavors.pdf\", format=\"pdf\", bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\"Number of totally used decomposition methods.\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "23ab8b6b-ee71-45ea-8a37-d5d1a0ee0786",
   "metadata": {},
   "source": [
    "#### Decomposition methods (proposed number of publications)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "41339ad8-b4cc-44b8-91ed-e3463dc22aa3",
   "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>total</th>\n",
       "      <th>proposed</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Fourier Transf.</th>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wavelet Transf.</th>\n",
       "      <td>12</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CCA</th>\n",
       "      <td>9</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ICA</th>\n",
       "      <td>36</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PCA</th>\n",
       "      <td>18</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 total  proposed\n",
       "Fourier Transf.     10         7\n",
       "Wavelet Transf.     12         9\n",
       "CCA                  9         6\n",
       "ICA                 36        17\n",
       "PCA                 18         9"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distr = (pipelines_annotated_df[pipelines_annotated_df[\"is new\"] > 0].groupby(\"Reference\").agg(\"count\")[\n",
    "    ['Fourier Transf.', 'Wavelet Transf.', 'Welch', 'LSP', 'PCC', 'CCA', 'ICA', 'NMF', 'PCA', 't-SNE', 'EMD', 'IVA', 'CSP', 'Other.1']\n",
    "        ]>0).sum(axis=0)\n",
    "distr[\"Other\"] = distr[\"Other.1\"]\n",
    "\n",
    "distr[\"Other\"] += distr[\"Welch\"]\n",
    "distr[\"Other\"] += distr[\"LSP\"]\n",
    "distr[\"Other\"] += distr[\"PCC\"]\n",
    "distr[\"Other\"] += distr[\"NMF\"]\n",
    "distr[\"Other\"] += distr[\"t-SNE\"]\n",
    "distr[\"Other\"] += distr[\"IVA\"]\n",
    "distr[\"Other\"] += distr[\"CSP\"]\n",
    "[distr.pop(key) for key in ['Welch', 'LSP', 'PCC', 'NMF', 't-SNE', 'IVA', 'CSP', 'Other.1']]\n",
    "\n",
    "# store in df_decomp\n",
    "df_decomp[\"proposed\"] = distr\n",
    "df_decomp.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "580927a9",
   "metadata": {},
   "source": [
    "#### Decomposition methods (number since 2020)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "dc6c2165",
   "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>total</th>\n",
       "      <th>proposed</th>\n",
       "      <th>since 2020</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Fourier Transf.</th>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wavelet Transf.</th>\n",
       "      <td>12</td>\n",
       "      <td>9</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CCA</th>\n",
       "      <td>9</td>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ICA</th>\n",
       "      <td>36</td>\n",
       "      <td>17</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PCA</th>\n",
       "      <td>18</td>\n",
       "      <td>9</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 total  proposed  since 2020\n",
       "Fourier Transf.     10         7           1\n",
       "Wavelet Transf.     12         9           3\n",
       "CCA                  9         6           4\n",
       "ICA                 36        17          11\n",
       "PCA                 18         9           5"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distr = (pipelines_annotated_df[pipelines_annotated_df[\"year\"] >= 2020].groupby(\"Reference\").agg(\"count\")[\n",
    "    ['Fourier Transf.', 'Wavelet Transf.', 'Welch', 'LSP', 'PCC', 'CCA', 'ICA', 'NMF', 'PCA', 't-SNE', 'EMD', 'IVA', 'CSP', 'Other.1']\n",
    "        ]>0).sum(axis=0)\n",
    "distr[\"Other\"] = distr[\"Other.1\"]\n",
    "\n",
    "distr[\"Other\"] += distr[\"Welch\"]\n",
    "distr[\"Other\"] += distr[\"LSP\"]\n",
    "distr[\"Other\"] += distr[\"PCC\"]\n",
    "distr[\"Other\"] += distr[\"NMF\"]\n",
    "distr[\"Other\"] += distr[\"t-SNE\"]\n",
    "distr[\"Other\"] += distr[\"IVA\"]\n",
    "distr[\"Other\"] += distr[\"CSP\"]\n",
    "[distr.pop(key) for key in ['Welch', 'LSP', 'PCC', 'NMF', 't-SNE', 'IVA', 'CSP', 'Other.1']]\n",
    "\n",
    "# store in df_decomp\n",
    "df_decomp[\"since 2020\"] = distr\n",
    "df_decomp.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b5870f25",
   "metadata": {},
   "source": [
    "#### Decomposition methods (number of publications with fit pipelines of the type)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "499903f2",
   "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>total</th>\n",
       "      <th>proposed</th>\n",
       "      <th>since 2020</th>\n",
       "      <th>fit</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Fourier Transf.</th>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wavelet Transf.</th>\n",
       "      <td>12</td>\n",
       "      <td>9</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CCA</th>\n",
       "      <td>9</td>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ICA</th>\n",
       "      <td>36</td>\n",
       "      <td>17</td>\n",
       "      <td>11</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PCA</th>\n",
       "      <td>18</td>\n",
       "      <td>9</td>\n",
       "      <td>5</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 total  proposed  since 2020  fit\n",
       "Fourier Transf.     10         7           1    5\n",
       "Wavelet Transf.     12         9           3    6\n",
       "CCA                  9         6           4    5\n",
       "ICA                 36        17          11   17\n",
       "PCA                 18         9           5    9"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distr = (pipelines_annotated_df[pipelines_annotated_df[\"fit score (0-2)\"] == 2].groupby(\"Reference\").agg(\"count\")[\n",
    "    ['Fourier Transf.', 'Wavelet Transf.', 'Welch', 'LSP', 'PCC', 'CCA', 'ICA', 'NMF', 'PCA', 't-SNE', 'EMD', 'IVA', 'CSP', 'Other.1']\n",
    "        ]>0).sum(axis=0)\n",
    "distr[\"Other\"] = distr[\"Other.1\"]\n",
    "\n",
    "distr[\"Other\"] += distr[\"Welch\"]\n",
    "distr[\"Other\"] += distr[\"LSP\"]\n",
    "distr[\"Other\"] += distr[\"PCC\"]\n",
    "distr[\"Other\"] += distr[\"NMF\"]\n",
    "distr[\"Other\"] += distr[\"t-SNE\"]\n",
    "distr[\"Other\"] += distr[\"IVA\"]\n",
    "distr[\"Other\"] += distr[\"CSP\"]\n",
    "[distr.pop(key) for key in ['Welch', 'LSP', 'PCC', 'NMF', 't-SNE', 'IVA', 'CSP', 'Other.1']]\n",
    "\n",
    "# store in df_decomp\n",
    "df_decomp[\"fit\"] = distr\n",
    "df_decomp.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a0eb08e6",
   "metadata": {},
   "source": [
    "#### Decomposition methods (number of publications with online capable pipelines of the type)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "89283606",
   "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>total</th>\n",
       "      <th>proposed</th>\n",
       "      <th>since 2020</th>\n",
       "      <th>fit</th>\n",
       "      <th>online</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Fourier Transf.</th>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Wavelet Transf.</th>\n",
       "      <td>12</td>\n",
       "      <td>9</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CCA</th>\n",
       "      <td>9</td>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ICA</th>\n",
       "      <td>36</td>\n",
       "      <td>17</td>\n",
       "      <td>11</td>\n",
       "      <td>17</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PCA</th>\n",
       "      <td>18</td>\n",
       "      <td>9</td>\n",
       "      <td>5</td>\n",
       "      <td>9</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 total  proposed  since 2020  fit  online\n",
       "Fourier Transf.     10         7           1    5       6\n",
       "Wavelet Transf.     12         9           3    6       5\n",
       "CCA                  9         6           4    5       2\n",
       "ICA                 36        17          11   17      10\n",
       "PCA                 18         9           5    9       4"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distr = (pipelines_annotated_df[pipelines_annotated_df[\"online score (0-2)\"] == 2].groupby(\"Reference\").agg(\"count\")[\n",
    "    ['Fourier Transf.', 'Wavelet Transf.', 'Welch', 'LSP', 'PCC', 'CCA', 'ICA', 'NMF', 'PCA', 't-SNE', 'EMD', 'IVA', 'CSP', 'Other.1']\n",
    "        ]>0).sum(axis=0)\n",
    "distr[\"Other\"] = distr[\"Other.1\"]\n",
    "\n",
    "distr[\"Other\"] += distr[\"Welch\"]\n",
    "distr[\"Other\"] += distr[\"LSP\"]\n",
    "distr[\"Other\"] += distr[\"PCC\"]\n",
    "distr[\"Other\"] += distr[\"NMF\"]\n",
    "distr[\"Other\"] += distr[\"t-SNE\"]\n",
    "distr[\"Other\"] += distr[\"IVA\"]\n",
    "distr[\"Other\"] += distr[\"CSP\"]\n",
    "[distr.pop(key) for key in ['Welch', 'LSP', 'PCC', 'NMF', 't-SNE', 'IVA', 'CSP', 'Other.1']]\n",
    "\n",
    "# store in df_decomp\n",
    "df_decomp[\"online\"] = distr\n",
    "df_decomp.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b7f2db3d",
   "metadata": {},
   "source": [
    "### Plot the stacked barplot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "9e7408cf",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fourier Transf.: \tsample_count=1, \treference_count=10, \tp_value=0.013\n",
      "Wavelet Transf.: \tsample_count=3, \treference_count=12, \tp_value=0.439\n",
      "CCA: \tsample_count=4, \treference_count=9, \tp_value=0.985\n",
      "ICA: \tsample_count=11, \treference_count=36, \tp_value=0.888\n",
      "PCA: \tsample_count=5, \treference_count=18, \tp_value=0.68\n",
      "EMD: \tsample_count=4, \treference_count=8, \tp_value=0.996\n",
      "Other: \tsample_count=3, \treference_count=27, \tp_value=0.0\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.013, 0.439, 0.985, 0.888, 0.68, 0.996, 0.0]"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def statistic(x, y, axis=0):\n",
    "    return np.mean(x, axis=axis) - np.mean(y, axis=axis)\n",
    "\n",
    "def bootstrap_test_multiple(references, samples, alternative=\"less\", n_resamples=1_000, verbose=0):\n",
    "    significances = []\n",
    "    for key in references.index:\n",
    "        total_bools = np.array([1] * references[key] + [0] * (references.sum() - references[key]))\n",
    "        sample_bools = [1] * samples[key] + [0] * (samples.sum() - samples[key])\n",
    "        diffs = np.zeros(n_resamples)\n",
    "        for idx in range(n_resamples):\n",
    "            sampled_orig = np.random.choice(total_bools, size=len(total_bools), replace=True)\n",
    "            diff = statistic(sampled_orig, sample_bools)\n",
    "            if diff > 0:\n",
    "                diffs[idx] = 1\n",
    "        if alternative == \"less\":\n",
    "            pvalue = np.round(1.0 -np.mean(diffs), 6)\n",
    "        else:\n",
    "            pvalue = np.round(np.mean(diffs), 6)\n",
    "        significances.append(pvalue)\n",
    "\n",
    "        if verbose >= 1:\n",
    "            print(f\"{key}: \\tsample_count={samples[key]}, \\treference_count={references[key]}, \\tp_value={pvalue}\")\n",
    "    return significances\n",
    "\n",
    "bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"since 2020\"], alternative=\"less\", n_resamples=1_000, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "7e70b5f5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# calculate the p-values to all cases\n",
    "p_values_proposed = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"proposed\"], alternative=\"greater\", n_resamples=10_000)\n",
    "p_values_2020 = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"since 2020\"], alternative=\"greater\", n_resamples=10_000)\n",
    "p_values_fit = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"fit\"], alternative=\"greater\", n_resamples=10_000)\n",
    "p_values_online = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"online\"], alternative=\"greater\", n_resamples=10_000)\n",
    "\n",
    "p_values_proposed_less = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"proposed\"], alternative=\"less\", n_resamples=10_000)\n",
    "p_values_2020_less = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"since 2020\"], alternative=\"less\", n_resamples=10_000)\n",
    "p_values_fit_less = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"fit\"], alternative=\"less\", n_resamples=10_000)\n",
    "p_values_online_less = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"online\"], alternative=\"less\", n_resamples=10_000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "1d77eca6",
   "metadata": {},
   "outputs": [],
   "source": [
    "# annotate the significances\n",
    "ALPHA = 0.05\n",
    "sig_char = \">\"\n",
    "sig_neg_char = \"<\"\n",
    "insig_char = \" \"\n",
    "\n",
    "significances_proposed = [sig_neg_char if p_values_proposed_less[idx] <= ALPHA else char \n",
    "                          for idx, char in enumerate([sig_char if p_values_proposed[idx] <= ALPHA else insig_char for idx in range(len(df_decomp))])]\n",
    "significances_2020 = [sig_neg_char if p_values_2020_less[idx] <= ALPHA else char \n",
    "                      for idx, char in enumerate([sig_char if p_values_2020[idx] <= ALPHA else insig_char for idx in range(len(df_decomp))])]\n",
    "significances_fit = [sig_neg_char if p_values_fit_less[idx] <= ALPHA else char \n",
    "                     for idx, char in enumerate([sig_char if p_values_fit[idx] <= ALPHA else insig_char for idx in range(len(df_decomp))])]\n",
    "significances_online = [sig_neg_char if p_values_online_less[idx] <= ALPHA else char \n",
    "                        for idx, char in enumerate([sig_char if p_values_online[idx] <= ALPHA else insig_char for idx in range(len(df_decomp))])]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "d759b8b0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAHiCAYAAAAH0tCqAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAAsTAAALEwEAmpwYAAA/4ElEQVR4nO3deXwV1fnH8c9DWMK+yFIQKIhoISGEfd+0oIIioGy1FVTEvbYKiLVFRP2pbRWrbaUqFioKKBaligsi4AIKQcOOoBAkguyLrLKc3x8zGS4hywVy7w3J9/165ZW5Z7Znbib3ueecmTPmnENERASgSKwDEBGR/ENJQUREAkoKIiISUFIQEZGAkoKIiASUFEREJKCkIGExswlm9kiM9m1m9m8z22VmC2MRQ35mZvvM7IIc5q8ws87Ri+jsmNlgM/s0j7bV2czS82JbhYWSwjnKzNLMbIuZlQ4pG2Jmc2MYVqS0B7oCNZ1zLTPP9D9EjvkfjvvMbL2fRC6KfqjR55wr45xbB1knb+dcgnNubl7v18zmmpkzs8aZyt/0yzuHsY06/rJF8zo+OTNKCue2osDdsQ7idJlZ3Gmu8nMgzTm3P4dlFjjnygDlgV8CB4HFZpZ4hmFKeNYA12e8MLPzgNbAtphFJGdFSeHc9hdgmJlVyDwjq29g/je7If70YDP7zMzGmtluM1tnZm398o1mttXMBmXabGUzm2VmP5rZPDP7eci2f+HP22lmX5tZv5B5E8zsOTObaWb7gS5ZxFvDzGb4639jZjf75TcBLwJt/FrAQzm9Ic65Y865b51ztwPzgNEh+2htZvP9410S+k3WzCr5tYtNfjPVmyHzbvZj2unHWCNknjOz281srf++PGxm9cxsgZntNbPXzKy4v2xnM0s3sz+Y2Xa/tnddyLbKm9l/zGybmW0wsz+aWRF/3oX+e77HX3dqphguNLOhwHXACP+9+p8/P83MfulPlzCzp/3j3ORPl8gU373+33+zmd2Q0/sNvAL0D0n0A4HpwE8h8RUxs5Fm9q2Z7fDfk0r+7I/937v9mNuErPdX/2+x3syuCCnP8lzx55X0z7ddZrYSaBEarJndZ2bf+3+rr83s0lyOr/BxzunnHPwB0vC+Ef8XeMQvGwLM9afrAA4oGrLOXGCIPz0YOArcAMQBjwDfAf8ASgDdgB+BMv7yE/zXHf35fwM+9eeVBjb62yoKNAW2Awkh6+4B2uF9EYnP4njmAf8E4oFkvG+al4bE+mkO70WW84EbgS3+9PnADqC7H0NX/3UVf/47wFSgIlAM6OSXX+IfS1P/uJ8FPg7ZhwNmAOWABOAwMBu4AK/WshIY5C/b2X/Pn/K31QnYD1zsz/8P8BZQ1v/7rQFu8udNBh7IeP+A9pliuDDkvX4kq3PFnx4DfA5UBaoA84GHM8U3xn8PugMHgIrZvO9z8c65D4Ar/LKFQBsgHejsl/3O32dN/7j/BUzO4TwdDBwBbsY7N28DNgEWxrnyOPAJUAmoBSwH0v15F+OdpzVC9l0v1v/L+e0n5gHo5wz/cCeSQiLeB24VTj8prA2Z18hfvlpI2Q4g2Z+eAEwJmVcGOOb/4/UHPskU37+AB0PW/U8Ox1LL31bZkLLHgAkhsZ5JUrgcOOJP3we8nGn++8AgoDpwnCw+/IDxwJ8zHfcRoI7/2gHtQuYvBu4Lef0k8LQ/3RnvQ7d0yPzXgD/5H36HgYYh824J+Xv+B3ger18lc4ynkxS+BbqHzLsMr2kuI76Dmc6ZrUDrbN73uf4592u8pHUxsMafF5oUVuF/aPuvq/vvYVGyTwrfhLwu5S/zszDOlXXA5SHzhnIiKVzoH88vgWKx/h/Orz9qPjrHOeeWA28DI89g9S0h0wf97WUuKxPyemPIfvcBO4EaeG3+rfxmmd1mthuvGeNnWa2bhRrATufcjyFlG/C+3Z+N8/0Y8WPsmynG9ngfULX8/e/KJrYNGS/8496RKbbM71lO7+Eud3LfyAZ/H5WB4qH74uT3YARgwELzria6MbuDzsVJxxOy/ww7nHNHQ14fyBR/Vv6LV6O6C3g5i/k/B6aHvO+r8D7Yq+WwzR8yJpxzB/zJMuR+rtTg5HMt9G/3DV6tZTSw1cymhDYFikdJoWB4EK+qHfpBlfHBUyqkLPRD+kzUypgwszJ4VfRNeP+E85xzFUJ+yjjnbgtZN6fheDcBlcysbEhZbeD7s4y3N15TAn6ML2eKsbRz7nF/XiXLom/Gjy2076Q0cN5ZxFbRQq4YwzvOTXhNVEdC90XIe+Cc+8E5d7NzrgZeDeKfZnZhFtvPbdjjk44nZP9nzP/QfhevmSerpLARr3kp9L2Pd859H0a8meV2rmwm5Dz154XG+qpzrj3ee+CAJ05z/wWekkIB4H8Dmgr8NqRsG94/yq/NLM7/ZlnvLHfV3cza+x2nDwNfOOc24tVULjKz35hZMf+nhZk1CDP+jXht24+ZWbyZJQE34XVinhb/WOua2bN4zSEZHdOTgKvM7DJ/mXi/Y7Wmc24z3ofaP82soh9/R3+9V4EbzCzZ75D9P/+40043thAPmVlxM+sAXAm87pw7hteU9KiZlTWvE/8eP27MrK+Z1fTX34X3gXYsi21vwevPyM5k4I9mVsXMKgOjMvZxlv6A1w+TlsW8cXjH9XMAf99X+/O24TXd5RRzIIxz5TXgfv/vWBOv9oK/34vN7BL/73gIrxaX1XtYqCkpFBxj8Dp8Q90MDMdr7kjA+2c6G6/i1Up2As3wmojwq/LdgAF43+R+wPsGVuI0tj0Qr315E97VKw8652adxvptzGwfsBevrbsc0MI5t8yPcSNwNd6H1za8b6/DOfE/8Bu8b+qr8dqdf+evNxuvzf8NvG+h9fzjPFM/4H2ob8L7ILvVObfan3cXXg1vHfAp3vv9kj+vBfCFf4wzgLudc+uz2P54oKHfVPNmFvMfAVKApcAy4Eu/7Kw45zY557K74exvfswfmNmPeJ3Orfz1DgCPAp/5MbcOY3c5nSsP4TUZrcfrAA+tuZTA64jejvd3qIp3PkiIjN58EYkw8y6BneScq5nLoiIxo5qCiIgElBRERCSg5iMREQmopiAiIgElBRERCZzTw9VWrlzZ1alTJ9ZhiIicUxYvXrzdOVclq3nndFKoU6cOKSkpsQ5DROScYmYbspun5iMREQkoKYiISEBJQUREAud0n4KIZO3IkSOkp6dz6NChWIciMRQfH0/NmjUpVqxY2OsoKYgUQOnp6ZQtW5Y6depgZrEOR2LAOceOHTtIT0+nbt26Ya+n5iORAujQoUOcd955SgiFmJlx3nnnnXZtUUlBpIBSQpAzOQeUFEQkIuLi4khOTg5+0tLSznqbmzZt4tprrz3j9Vu1akVycjK1a9emSpUqeRpbZtu2baNVq1Y0adKETz75JPcVgBtvvJGqVauSmJh4UvnOnTvp2rUr9evXp2vXruzaldWTY/PGOT0gXvPmzZ1uXhM51apVq2jQ4MSD7+qMfCdPt5/2eI9clylTpgz79u3Ls30ePXqUokVPrxv02LFjxMXFnVI+YcIEUlJS+Pvf/37W+8jOlClTePfdd5k4cWLY63z88ceUKVOG66+/nuXLlwflI0aMoFKlSowcOZLHH3+cXbt28cQT4T1JNPO5AGBmi51zzbNaXjUFEYma1NRUWrduTVJSEr179w6+8Xbu3DkYnWD79u1kDF8zYcIE+vbty1VXXUW3bt1IS0sLvkUfO3aM4cOH06JFC5KSkvjXv/4FwNy5c+nSpQu/+tWvaNSoUa4xjR49mqFDh9KtWzeuv/560tLS6NChA02bNqVp06bMnz8/2G7nzp259tpr+cUvfsF1111HxpfqkSNH0rBhQ5KSkhg2bBipqamMGDGCmTNnkpyczMGDB8N6fzp27EilSpVOKX/rrbcYNGgQAIMGDeLNN98Ma3tnQlcfiUhEHDx4kOTkZADq1q3L9OnTuf7663n22Wfp1KkTo0aN4qGHHuLpp5/OcTsLFixg6dKlVKpU6aRmnvHjx1O+fHkWLVrE4cOHadeuHd26dQNg4cKFLF++POyrbhYvXsynn35KyZIlOXDgALNmzSI+Pp61a9cycODAIGF99dVXrFixgho1atCuXTs+++wzGjZsyPTp01m9ejVmxu7du6lQoQJjxozJsjaSkpLCuHHjePHFF8N7I4EtW7ZQvXp1AKpXr87WrVvDXvd0KSmISESULFmS1NTU4PWePXvYvXs3nTp1ArxvvH379s11O127ds3y2/MHH3zA0qVLmTZtWrD9tWvXUrx4cVq2bHlal2H27NmTkiVLAt49HnfeeSepqanExcWxZs2aYLmWLVtSs6b3NNWMvojWrVsTHx/PkCFD6NGjB1deeWWO+2revPlpJYRoU/ORiMRc0aJFOX78OMApl1CWLl06y3Wcczz77LOkpqaSmprK+vXrg5pCdutkJ3T5sWPHUq1aNZYsWUJKSgo//fRTMK9EiRLBdFxcXNAHsXDhQq655hrefPNNLr/88tPadziqVavG5s2bAdi8eTNVq1bN831kUFIQkagoX748FStWDK7Eefnll4NaQ506dVi8eDFA8M0/N5dddhnPPfccR44cAWDNmjXs37//rOPcs2cP1atXp0iRIrz88sscO3Ysx+X37dvHnj176N69O08//fRJtaO80rNnz6DDeuLEiVx99dV5vo8MSgoiEjUTJ05k+PDhJCUlkZqayqhRowAYNmwYzz33HG3btmX79u1hbWvIkCE0bNiQpk2bkpiYyC233MLRo0fPOsbbb7+diRMn0rp1a9asWZNrrePHH3/kyiuvJCkpiU6dOjF27NhTlhk3bhzjxo0DvD6FIUOGZLmtgQMH0qZNG77++mtq1qzJ+PHjAa8je9asWdSvX59Zs2YxcuTIszzK7EXsklQzewm4EtjqnEv0y/4CXAX8BHwL3OCc2+3Pux+4CTgG/NY5935u+9AlqSJZy+oyRCmc8tMlqROAzI1rs4BE51wSsAa43w+wITAASPDX+aeZnXpxsYiIRFTEkoJz7mNgZ6ayD5xzGfW7z4Ga/vTVwBTn3GHn3HrgG6BlpGITEZGsxbJP4UbgXX/6fGBjyLx0v0xERKIoJvcpmNkDwFHglYyiLBbLsrPDzIYCQwFq16591rGkjwxvTJJIq/l4h1iHICIS/ZqCmQ3C64C+zp3o5U4HaoUsVhPYlNX6zrnnnXPNnXPNq1SpEtlgRUQKmagmBTO7HLgP6OmcOxAyawYwwMxKmFldoD6wMJqxiYhIBJOCmU0GFgAXm1m6md0E/B0oC8wys1QzGwfgnFsBvAasBN4D7nDO5XzHiIjkW7///e9PGtPosssuO+na/HvvvZennnoqz/Y3ePDgXG96mzBhAps2ndoAcccdd5CcnEzDhg0pWbJkMJx2uDfRna6BAweSlJSU5f0MWfn73//OhRdeiJmddA/HK6+8QlJSEklJSbRt25YlS5bkSXwR61Nwzg3Monh8Dss/CjwaqXhECrXR5fN4e3tynN22bVtef/11fve733H8+HG2b9/O3r17g/nz58/PdSC8vDZhwgQSExOpUaPGSeX/+Mc/AEhLS+PKK6885Y7k7IbfPhM//PAD8+fPZ8OGDWGv065dO6688ko6d+58UnndunWZN28eFStW5N1332Xo0KF88cUXZx2j7mgWkTzXrl27YMjpFStWkJiYSNmyZdm1axeHDx9m1apVNGnShDFjxtCiRQsSExMZOnQozjlWrVpFy5YnrkhPS0sjKSkJ8EYz7dSpE82aNeOyyy4LxgMKldUy06ZNIyUlheuuuy6soayzGn67V69eNGvWjISEBJ5//vlg2TJlyvDAAw/QuHFjWrduzZYtWwB4/fXXSUxMpHHjxnTs2BGAbt26sXXrVpKTk8N+8E6TJk2CocRDtW3blooVKwLQunVr0tPTw9pebpQURCTP1ahRg6JFi/Ldd98xf/582rRpQ6tWrViwYAEpKSkkJSVRvHhx7rzzThYtWsTy5cs5ePAgb7/9Ng0aNOCnn35i3bp1AEydOpV+/fpx5MgR7rrrLqZNm8bixYu58cYbeeCBB07ab3bLXHvttTRv3pxXXnmF1NTUYETUnCxcuJBHH32UlStXAvDSSy+xePFiUlJSeOaZZ9ixYwcA+/fvp3Xr1ixZsoSOHTvywgsvADBmzBjef/99lixZwowZMwCYMWMG9erVIzU1lQ4dTr7iMGOY8TMxfvx4rrjiijNeP5SGzhaRiMioLcyfP5977rmH77//nvnz51O+fHnatm0LwJw5c/jzn//MgQMH2LlzJwkJCVx11VX069eP1157jZEjRzJ16lSmTp3K119/zfLly+natSvgNetkPGMgQzjLhCvz8NvPPPMM06dPB2Djxo2sXbuW8847j+LFiwfDZTdr1oxZs2YFxz948GD69etHnz59ct3fmQ6kN2fOHMaPH8+nn356RutnpqQgIhHRtm1b5s+fz7Jly0hMTKRWrVo8+eSTlCtXjhtvvJFDhw5x++23k5KSQq1atRg9enQwbHb//v3p27cvffr0wcyoX78+y5YtIyEhgQULFmS7T+dcrsuEK3QgvLlz5/Lhhx+yYMECSpUqRefOnYNYixUrhpl3q1XGcNrgDYL3xRdf8M4775CcnByR0VOXLl3KkCFDePfddznvvPPyZJtqPhKRiGjXrh1vv/02lSpVIi4ujkqVKrF7924WLFhAmzZtgg/VypUrs2/fvpOu9qlXrx5xcXE8/PDD9O/fH4CLL76Ybdu2BR/4R44cYcWKFSftM6dlypYty48//nhGx7Jnzx4qVqxIqVKlWL16NZ9//nmu63z77be0atWKMWPGULlyZTZu3JjrOqfju+++o0+fPrz88stcdNFFebZdJQURiYhGjRqxfft2WrdufVJZ+fLlqVy5MhUqVODmm2+mUaNG9OrVixYtWpy0fv/+/Zk0aRL9+vUDoHjx4kybNo377ruPxo0bk5ycHHRmZ8hpmcGDB3Prrbee1jOTM1x++eUcPXqUpKQk/vSnP510TNkZPnw4jRo1IjExkY4dO9K4ceOT5mceQju7PoVnnnmGmjVrkp6eTlJSUrDOmDFj2LFjB7fffjvJyck0b57loKenLWJDZ0dDXgydrWEupCDS0NmSIT8NnS0iIucYJQUREQkoKYiISEBJQUREAkoKIiISUFIQEZGAkoKIRMQPP/zAgAEDqFevHg0bNqR79+6sWbMGgDVr1tC9e3cuvPBCGjRoQL9+/YKB5ADuvvtuzj//fI4fPx7Wvu677z4SExNJTExk6tSpETmewkLDXIgUAo0mNsrT7S0btCzH+c45evfuzaBBg5gyZQrgje2zZcsWateuTY8ePXjqqae46qqrAG/8nm3btlGtWjWOHz/O9OnTqVWrFh9//PEpQ0Zn9s477/Dll1+SmprK4cOH6dSpE1dccQXlypXLk2MtbFRTEJE8N2fOHIoVK8att94alCUnJ9OhQwdeffVV2rRpEyQEgC5dupCYmBism5iYyG233cbkyZNz3dfKlSvp1KkTRYsWpXTp0jRu3Jj33nsv7w+qkFBSEJE8t3z5cpo1a3ba8wAmT57MwIED6d27N2+//TZHjhwBvGGnR40adcryjRs35t133+XAgQNs376dOXPm5Pk4Q4WJmo9EJN/46aefmDlzJmPHjqVs2bK0atWKDz74gB49etCzZ0969ux5yjrdunVj0aJFtG3blipVqtCmTRuKFtVH25lSTUFE8lxCQgKLFy8+7Xnvvfcee/bsoVGjRtSpU4dPP/00rCakBx54gNTUVGbNmoVzjvr1659V/IWZkoKI5LlLLrmEw4cPB08hA1i0aBHz5s3jV7/6FfPnz+edd94J5r333nssW7aMyZMn8+KLL5KWlkZaWhrr16/ngw8+4MCBA9nu69ixY8FT0JYuXcrSpUvp1q1b5A6ugFNSEJE8Z2ZMnz6dWbNmUa9ePRISEhg9ejQ1atSgZMmSvP322zz77LPUr1+fhg0bMmHCBMqVK8f7779Pjx49gu2ULl2a9u3b87///S/bPoUjR47QoUMHGjZsyNChQ5k0aZKaj86Chs7W0NlSAGnobMmgobNFROSMKSmIiEhASUFERAJKCiIiElBSEBGRgJKCiIgElBREJCLKlCkTTOflUNkbNmygWbNmJCcnk5CQwLhx44J5zjkeeOABLrroIho0aMAzzzyTtwdVCOgOD5FCYNUv8vaehQarV4W97KFDh/J0qOzq1aszf/58SpQowb59+0hMTKRnz57UqFGDCRMmsHHjRlavXk2RIkXYunXr2RxmoaSkICIRld1Q2Rkyhsru378/kydPzjUpFC9ePJg+fPjwSbWL5557jldffZUiRbxGkKpVq+bRURQeaj4SkYjK66GyATZu3EhSUhK1atXivvvuo0aNGgB8++23TJ06lebNm3PFFVewdu3avD+gAk5JQURiJmOo7F69elGuXLlgqGyAnj17MmbMmCzXq1WrFkuXLuWbb75h4sSJQf/E4cOHiY+PJyUlhZtvvpkbb7wxasdSUCgpiEhERWKo7Aw1atQgISGBTz7xxjCrWbMm11xzDQC9e/dm6dKlZ38AhYySgohEVF4PlZ2ens7BgwcB2LVrF5999hkXX3wxAL169eKjjz4CYN68eVx00UURPLKCSUlBRCIqr4fKXrVqFa1ataJx48Z06tSJYcOG0ahRIwBGjhzJG2+8QaNGjbj//vt58cUXo3acBYWGztbQ2VIAaehsyaChs0VE5IwpKYiISEBJQUREAkoKIiISUFIQEZGAkoKIiASUFEQkIuLi4khOTiYxMZG+ffsGN6T98MMPDBgwgHr16tGwYUO6d+/OmjVrgvXGjh1LfHw8e/bsCWs/N954I1WrViUxMfGk8tTUVFq3bk1ycjLNmzdn4cKFeXdwBZhGSRUpBP5x60d5ur07xl2S6zIlS5YkNTUVgOuuu45x48bx+9//nt69ezNo0CCmTJkCeB/eW7ZsCe4+njx5Mi1atGD69OkMHjw41/0MHjyYO++8k+uvv/6k8hEjRvDggw9yxRVXMHPmTEaMGMHcuXNP6zgLI9UURCTiOnTowDfffMOcOXMoVqwYt956azAvOTmZDh28mze//fZb9u3bxyOPPBL2GEgdO3akUqVKp5SbGXv37gVgz549wUiqkjMlBRGJqKNHj/Luu+/SqFGjsIfR7tChA19//XXwkJyUlBSGDBlyWvt9+umnGT58OLVq1WLYsGE89thjZ3UchYWSgohExMGDB4P2/Nq1a3PTTTflus6UKVMYMGAARYoUoU+fPrz++usANG/e/LTHMXruuecYO3YsGzduZOzYsWHtX9SnICIREtqnkCEhIYFp06ZlufzSpUtZu3YtXbt2BbxnLVxwwQXccccdZ7T/iRMn8re//Q2Avn37nnZNo7BSTUFEouaSSy7h8OHDvPDCC0HZokWLmDdvHpMnT2b06NHBMNqbNm3i+++/Z8OGDWe0rxo1ajBv3jwAPvroI+rXr58nx1DQKSmISNSYGdOnT2fWrFnUq1ePhIQERo8eTY0aNZgyZQq9e/c+afnevXszZcqUHPsUBg4cSJs2bfj666+pWbMm48ePB+CFF17g3nvvpXHjxvzhD3/g+eefj/jxFQQaOltDZ0sBpKGzJYOGzhYRkTOmpCAiIgElBRERCSgpiIhIQElBREQCSgoiIhJQUhCRiMgYOjvj5/HHHwegc+fO1K5dm9DL4Xv16kWZMmUASEtLo2TJkjRp0oQGDRrQsmVLJk6cGNY+x44dS0JCAomJiQwcOJBDhw7l/YEVcBrmQqQQeLL/lXm6vXunvp3rMlkNc5GhQoUKfPbZZ7Rv357du3ezefPmk+bXq1ePr776CoB169bRp08fjh8/zg033JDt/r7//nueeeYZVq5cScmSJenXrx9TpkwJa/htOSFiNQUze8nMtprZ8pCySmY2y8zW+r8rhsy738y+MbOvzeyySMUlIrE3YMCA4HkK//3vf+nTp0+2y15wwQU89dRTPPPMM7lu9+jRoxw8eJCjR49y4MABDZd9BiLZfDQBuDxT2UhgtnOuPjDbf42ZNQQGAAn+Ov80s7gIxiYiEZYxSmrGz9SpU4N5l156KR9//DHHjh1jypQp9O/fP8dtNW3alNWrVwPZD6N9/vnnM2zYMGrXrk316tUpX7483bp1y9uDKgQilhSccx8DOzMVXw1kNA5OBHqFlE9xzh12zq0HvgFaRio2EYm8jOajjJ/QD/64uDjat2/P1KlTOXjwIHXq1MlxW6H9D9kNo71r1y7eeust1q9fz6ZNm9i/fz+TJk3Ks+MpLKLd0VzNObcZwP9d1S8/H9gYsly6XyYiBdSAAQO466676NevX67LfvXVV7mO5fThhx9St25dqlSpQrFixejTpw/z58/Pq3ALjfxy9ZFlUZblSH1mNtTMUswsZdu2bREOS0QipUOHDtx///0MHDgwx+XS0tIYNmwYd911V47L1a5dm88//5wDBw7gnGP27NkaFPAMRPvqoy1mVt05t9nMqgNb/fJ0oFbIcjWBTVltwDn3PPA8eKOkRjJYETlzGX0KGS6//PLgslTwhtEeNmxYlut+++23NGnShEOHDlG2bFnuuuuu4MqjlJQUxo0bd0oTUqtWrbj22mtp2rQpRYsWpUmTJgwdOjTvD6yAi+jQ2WZWB3jbOZfov/4LsMM597iZjQQqOedGmFkC8CpeP0INvE7o+s65YzltX0Nni2RNQ2dLhtMdOjtiNQUzmwx0BiqbWTrwIPA48JqZ3QR8B/QFcM6tMLPXgJXAUeCO3BKCiIjkvYglBedcdg2Fl2az/KPAo5GKR0REcpdfOppFRCQfUFIQEZGAkoKIiASUFEREJKCkICIRkZ6eztVXX039+vWpV68ed999Nz/99BMAqampzJw5M1h29OjR/PWvf41IHE899RQNGzYkKSmJSy+9lA0bNgAwZ86ck8Zmio+P580334xIDOcSDZ0tUgjk9f04ud1X45yjT58+3Hbbbbz11lscO3aMoUOH8sADD/CXv/yF1NRUUlJS6N69e57Ec+zYMeLish5Ds0mTJqSkpFCqVCmee+45RowYwdSpU+nSpUswtPfOnTu58MILNYAeqimISAR89NFHxMfHB3chx8XFMXbsWF566SX27t3LqFGjmDp16kmjp65cuZLOnTtzwQUXnDRM9qRJk2jZsiXJycnccsstHDvm3cJUpkwZRo0aRatWrViwYEG2sXTp0oVSpUoB0Lp1a9LT009ZZtq0aVxxxRXBcoWZkoKI5LkVK1bQrFmzk8rKlStH7dq1SUtLY8yYMfTv3/+k0VNXr17N+++/z8KFC3nooYc4cuQIq1atYurUqXz22WekpqYSFxfHK6+8AsD+/ftJTEzkiy++oH379gwZMoTcRjgYP348V1xxxSnlU6ZMyXUMpsJCzUcikuecc5idOs5lduUAPXr0oESJEpQoUYKqVauyZcsWZs+ezeLFi2nRogXgjadUtao3uHJcXBzXXHNNsH5Ww2mHmjRpEikpKcybN++k8s2bN7Ns2TIuu0zP9gIlBRGJgISEBN54442Tyvbu3cvGjRupV68eixcvPmWdEiVKBNNxcXEcPXoU5xyDBg3iscceO2X5+Pj4bPsRMvvwww959NFHmTdv3kn7AXjttdfo3bs3xYoVC2tbBZ2aj0Qkz1166aUcOHCA//znP4DXEXzvvfcyePBgSpUqRdmyZfnxxx/D2s60adPYutUbUHnnzp3B1UPh+uqrr7jllluYMWNGUMsINXnyZDUdhVBSEJE8Z2ZMnz6d119/nfr163PRRRcRHx/P//3f/wFe5+/KlStPeUxnZg0bNuSRRx6hW7duJCUl0bVrVzZv3pzlstn1KQwfPpx9+/bRt29fkpOT6dmzZzAvLS2NjRs30qlTp7M84oIjokNnR5qGzhbJmobOlgynO3S2agoiIhJQUhARkYCSgoiIBJQURAqoc7m/UPLGmZwDSgoiBVB8fDw7duxQYijEnHPs2LGD+Pj401pPN6+JFEA1a9YkPT2dbdu2xToUiaH4+Hhq1qx5WusU+qQwdf0TsQ4BgHvRJamSd4oVK0bdunVjHYacg9R8JCIiASUFEREJKCmIiEhASUFERAJKCiIiElBSEBGRgJKCiIgElBRERCSgpCAiIgElBRERCSgpiIhIQElBREQCSgoiIhJQUhARkYCSgoiIBJQUREQkoKQgIiIBJQUREQkoKYiISEBJQUREAkoKIiISUFIQEZGAkoKIiASUFEREJKCkICIiASUFEREJFI11ALEWX/GeWIcgIpJvqKYgIiIBJQUREQkoKYiISEBJQUREAkoKIiISUFIQEZGAkoKIiASUFEREJKCkICIiASUFEREJKCmIiEhASUFERAJKCiIiEohJUjCz35vZCjNbbmaTzSzezCqZ2SwzW+v/rhiL2ERECrOoJwUzOx/4LdDcOZcIxAEDgJHAbOdcfWC2/1pERKIoVs1HRYGSZlYUKAVsAq4GJvrzJwK9YhOaiEjhFfWk4Jz7Hvgr8B2wGdjjnPsAqOac2+wvsxmoGu3YREQKu1g0H1XEqxXUBWoApc3s16ex/lAzSzGzlG3btkUqTBGRQikWzUe/BNY757Y5544A/wXaAlvMrDqA/3trVis75553zjV3zjWvUqVK1IIWESkMck0KZvZnMytnZsXMbLaZbT+db/ZZ+A5obWalzMyAS4FVwAxgkL/MIOCts9iHiIicgXBqCt2cc3uBK4F04CJg+Jnu0Dn3BTAN+BJY5sfwPPA40NXM1gJd/dciIhJFRcNYppj/uzsw2Tm30/uCf+accw8CD2YqPoxXaxARkRgJJyn8z8xWAweB282sCnAosmGJiEgs5Np85JwbCbTBu9nsCLAf7+ohEREpYMKpKQA0AOr4N5tl+E8E4hERkRjKNSmY2ctAPSAVOOYXO5QUREQKnHBqCs2Bhs45F+lgREQktsK5JHU58LNIByIiIrEXTk2hMrDSzBbiXTYKgHOuZ8SiEhGRmAgnKYyOdBAiIpI/5JoUnHPzzKwa0MIvWuicy3JcIhERObeFM/ZRP2Ah0BfoB3xhZtdGOjAREYm+cJqPHgBaZNQO/DuaP8Qbv0hERAqQcK4+KpKpuWhHmOuJiMg5Jpyawntm9j4w2X/dH5gZuZBERCRWwuloHm5m1wDtAAOed85Nj3hkIiISdWGNfeScewN4I8KxiIhIjGWbFMzsU+dcezP7EW+so2AW4Jxz5SIenYiIRFW2ScE5197/XTZ64YiISCyFc5/Cy+GUiYjIuS+cS0sTQl/4z1RoFplwREQklrJNCmZ2v9+fkGRme/2fH4EtwFtRi1BERKIm26TgnHvM70/4i3OunP9T1jl3nnPu/ijGKCIiURLOfQr3m1lFoD4QH1L+cSQDExGR6AvncZxDgLuBmniP5GwNLAAuiWhkIiISdeF0NN+NN2z2BudcF6AJsC2iUYmISEyEkxQOOecOAZhZCefcauDiyIYlIiKxEM4wF+lmVgF4E5hlZruATZEMSkREYiOcjube/uRoM5sDlAfei2hUIiISE+Hc0dzazMqC92hOYA5ev4KIiBQw4fQpPAfsC3m93y8TEZECJpw+BXPOBaOkOueO+0NdSAGTPvKTWIcAQM3HO8Q6BJFCK5yawjoz+62ZFfN/7gbWRTowERGJvnCSwq1AW+B7IB1oBQyNZFAiIhIb4Vx9tBUYEIVYREQkxnJ68toI59yfzexZTn7yGgDOud9GNDIREYm6nGoKq/zfKdEIREREYi+nx3H+z/89MXrhiIhILOXUfPQ/smg2yuCc6xmRiEREJGZyaj76a9SiEBGRfCGn5qN5GdNmVhz4BV7N4Wvn3E9RiE1ERKIsnIfs9ADGAd8CBtQ1s1ucc+9GOjgREYmucIareBLo4pz7BsDM6gHvAAUiKVwy945Yh+BblfsiIiIRFs4dzVszEoJvHbA1QvGIiEgM5XT1UR9/coWZzQRew+tT6AssikJsIiISZTk1H10VMr0F6ORPbwMqRiwiERGJmZyuProhmoGIiEjshXP10b/JeuyjGyMSkYiIxEw4Vx+9HTIdD/QGNkUmHBERiaVwhs5+I/S1mU0GPoxYRCIiEjPhXJKaWX2gdl4HIiIisRdOn8KPeH0K5v/+AbgvwnGJiEgMhNN8VDYagYiISOyF09GccSNbe7yawifOuTcjGZSIiMRGrn0KZvZP4FZgGbAcuNXM/hHpwEREJPrCqSl0AhKdcw7AzCbiJQgRESlgwrn66GtOvtqoFrA0MuGIiEgshVNTOA9YZWYL/dctgAVmNgP0WE4RkYIknKQwKuJRiIhIvhDOJanzcltGREQKhjO5o1lERAooJQUREQlkmxTMbLb/+4m83qmZVTCzaWa22sxWmVkbM6tkZrPMbK3/Ww/yERGJspxqCtXNrBPQ08yamFnT0J+z3O/fgPecc78AGuM9tX4kMNs5Vx+Y7b8WEZEoyqmjeRTeB3NN4KlM8xxwyZns0MzKAR2BwQDOuZ+An8zsaqCzv9hEYC4aeE9EJKpyehznNGCamf3JOfdwHu7zArznPP/bzBoDi4G7gWrOuc3+vjebWdU83KeIiIQh145m59zDZtbTzP7q/1x5lvssCjQFnnPONQH2cxpNRWY21MxSzCxl27ZtZxmKiIiECmdAvMfwvsmv9H/u9svOVDqQ7pz7wn89DS9JbDGz6v4+qwNbs1rZOfe8c665c655lSpVziIMERHJLJxLUnsAXZ1zLznnXgIu98vOiHPuB2CjmV3sF12Kl2xmAIP8skHAW2e6DxEROTNhPU8BqADs9KfL58F+7wJeMbPiwDrgBrwE9ZqZ3QR8B/TNg/2IyFlKH/lJrEOg5uMdYh1CoRFOUngM+MrM5uA9krMjcP/Z7NQ5lwo0z2LWpWezXREROTvhjH002czm4o2OasB9fhOQiIgUMGE1H/mXis6IcCwiIhJjGvtIREQCSgoiIhLIMSmYWREzWx6tYEREJLZyTArOuePAEjOrndNyIiJSMITT0VwdWOE/o3l/RqGezSwiUvCEkxQeingUIiKSL4T1jGYz+zlQ3zn3oZmVAuIiH5qIiERbOAPi3Yw3aN2//KLzgTcjGJOIiMRIOJek3gG0A/YCOOfWAnrWgYhIARROUjjsPx0NADMrivfkNRERKWDCSQrzzOwPQEkz6wq8DvwvsmGJiEgshJMURuI9PnMZcAswE/hjJIMSEZHYCOfqo+NmNhH4Aq/Z6GvnnJqPREQKoFyTgpn1AMYB3+INnV3XzG5xzr0b6eAkuqaufyLWIQBwL3qgikishHPz2pNAF+fcNwBmVg94B1BSEBEpYMLpU9iakRB864CtEYpHRERiKNuagpn18SdXmNlM4DW8PoW+wKIoxCYiIlGWU/PRVSHTW4BO/vQ2oGLEIhIRkZjJNik4526IZiAiIhJ74Vx9VBe4C6gTuryGzhYRKXjCufroTWA83l3MxyMajYiIxFQ4SeGQc+6ZiEciIiIxF05S+JuZPQh8ABzOKHTOfRmxqEREJCbCSQqNgN8Al3Ci+cj5r0VEpAAJJyn0Bi4IHT5bREQKpnDuaF4CVIhwHCIikg+EU1OoBqw2s0Wc3KegS1JFRAqYcJLCgxGPQkRE8oVwnqcwLxqBiIhI7IVzR/OPnHgmc3GgGLDfOVcukoGJiEj0hVNTKBv62sx6AS0jFZCIiMROOFcfncQ59ya6R0FEpEAKp/moT8jLIkBzTjQniYhIARLO1Uehz1U4CqQBV0ckGhERialw+hT0XIVCIr7iPbEOQURiLKfHcY7KYT3nnHs4AvGIiEgM5VRT2J9FWWngJuA8QElBRKSAyelxnE9mTJtZWeBu4AZgCvBkduuJiMi5K8c+BTOrBNwDXAdMBJo653ZFIzARyR+mrn8i1iFwLx1iHUKhkVOfwl+APsDzQCPn3L6oRSUiIjGR081r9wI1gD8Cm8xsr//zo5ntjU54IiISTTn1KZz23c4iInJu0we/iIgElBRERCSgpCAiIgElBRERCSgpiIhIQElBREQCSgoiIhJQUhARkYCSgoiIBJQUREQkoKQgIiIBJQUREQkoKYiISCDHh+yIiMRXvCfWIUgUqaYgIiIBJQUREQnELCmYWZyZfWVmb/uvK5nZLDNb6/+uGKvYREQKq1jWFO4GVoW8HgnMds7VB2b7r0VEJIpikhTMrCbQA3gxpPhqYKI/PRHoFeWwREQKvVjVFJ4GRgDHQ8qqOec2A/i/q8YgLhGRQi3qScHMrgS2OucWn+H6Q80sxcxStm3blsfRiYgUbrGoKbQDeppZGjAFuMTMJgFbzKw6gP97a1YrO+eed841d841r1KlSrRiFhEpFKKeFJxz9zvnajrn6gADgI+cc78GZgCD/MUGAW9FOzYRkcIuP93R/DjwmpndBHwH9I1xPIXOJXPviHUIvlW5LyIiERHTpOCcmwvM9ad3AJfGMh4RkcJOdzSLiEhASUFERAJKCiIiElBSEBGRgJKCiIgElBRERCSgpCAiIgElBRERCSgpiIhIID8NcyEikq+lj/wk1iFQ8/EOEd2+agoiIhJQUhARkYCSgoiIBJQUREQkoKQgIiIBJQUREQkoKYiISEBJQUREAkoKIiIS0B3NIpKjS+beEesQgFWxDqDQUE1BREQCSgoiIhJQUhARkYCSgoiIBJQUREQkoKQgIiKBQn9Jar/788dbsCzWAchJCsPDVESyopqCiIgElBREoiR9zw9s378r1mGI5EhJQSSCDh/9if+t+ojrpt5Lu38N4Nud38U6JJEc5Y8GdZECZtW2b5m65B3+u3IWxYrE0bPBL3mgy200rHphrEMTyZGSgkgEdHvpBiqWLM8jXX9Hj4s7E1ckLtYhiYRFzUciEfCnLndQpVRFhr/7Z+6Z+Rgfr1/EcXc81mGJ5EpJQSQChrbsz+wh/+GVfn+laJGi3Dz9j7R+ri9rtqfFOjSRHKn5SCSCmtdsRPOajXjo0t/y1qoPsVgHJJILJQWRKChTohTXJfeMdRiSD6Xv+YH4oiWoXLpirEMBlBREsjR1/ROxDoF70R3N+U1enRdHjx1j+aYtLFy3kbVbt3Nb5zZcUKVSWOtG+rxQUhARiZLNu/eycP1GFm/4nrgiRWhSuwZXNm5AjQrlYh1aQElBRCRKnvzgE0oVL0afpokk1axOkSL5r5dJVx+JiETJVY0bUDa+BK+lLGXKoiWs+WEbx52LdVgnUU1BJAK+27Gb2udViHUYksfiK95zVutf1tr7WffDCuavnsmE+XMpVaIsd/R4nOoV6+RNkGdJSUEkAv7+0XyqlStDy7q1aPrz8yldonisQ5J85IKfJXDBzxK4tu0dpHzzEZaPLlZWUhCJgAeuvIQvN3zP5+u+4+2lq0k8vxqt6tbiwmqVKWL55wNAYiu+eCnaN7wy1mGcRH0KIhFQvmQ8XX5Rj+GXd+KOLm3Yuncfz3+8kPXbdsY6NJEcqaYgEiGbd+8lZcP3fLXhe447R6eLL+Bn5cvGOiyRHCkpiETAUx98wuY9e7moWhV6NU0goUY14oqoYi75n5KCSAQknv8zbmzfnAqlSsY6FJHToqQgEgHdEurHOgSRM6L6rIiIBJQUREQkoOYjkSyc7Z2rIucq1RRERCSgpCAiIgE1H4mIhOmSuXfEOgRgVUS3rpqCiIgElBRERCSgpCAiIoFC36ewbP13sQ4h3+h3f/44HZbFOgCRQkw1BRERCSgpiIhIIOpJwcxqmdkcM1tlZivM7G6/vJKZzTKztf7vitGOTUSksItFTeEocK9zrgHQGrjDzBoCI4HZzrn6wGz/tYiIRFHUk4JzbrNz7kt/+ke8OzHOB64GJvqLTQR6RTs2EZHCLqZ9CmZWB2gCfAFUc85tBi9xAFVjGJqISKEUs6RgZmWAN4DfOef2nsZ6Q80sxcxStm3bFrkARUQKoZgkBTMrhpcQXnHO/dcv3mJm1f351YGtWa3rnHveOdfcOde8SpUq0QlYRKSQiMXVRwaMB1Y5554KmTUDGORPDwLeinZsIiKFXSxuYW0H/AZYZmapftkfgMeB18zsJuA7oG8MYhMBCsdomCJZiXpScM59Clg2sy+NZiwiInIy3dEsIiIBJQUREQkoKYiISEBJQUREAkoKIiISUFIQEZGAkoKIiASUFEREJKCkICIigfzxpHbJF5at/y7WIUg+1O/+2H9MLIt1AIWIagoiIhJQUhARkYCSgoiIBJQUREQkoKQgIiIBJQUREQkoKYiISEBJQUREAkoKIiISiP2tiiIi54jCcHe3agoiIhJQUhARkYCSgoiIBJQUREQkoKQgIiIBJQUREQnE/voqEcnX9PClEwrDe6GagoiIBJQUREQkUOibj+ocejXWIQCQFusARERQTUFEREIoKYiISEBJQUREAkoKIiISUFIQEZGAkoKIiAQK/SWpIlkpDA9TEcmKagoiIhJQUhARkYCSgoiIBGLfcCr5hob8yDsH1h2g1AWlYh2GyGlTUhCJgHWPrKNEjRJU7FiRCm0rULSM/tXk3KAzVSQCLn7yYnYv2M2uebvY8toWyjYtS6WOlSjdsDRWxGIdnki2lBREIqBYxWJU6V6FKt2rcGDdATb9exNpf02j7si6lP5F6ViHJ2coPzSxpkV4+0oKIhFyaOMhdn+2m92f78Ydd1S+vDIlapaIdVgiOVJSEImAb0Z9w6GNhyiTWIbq11WnXJNyWFE1G0n+p6QgEgHlmpWj9t21KX5e8ViHInJalBREIqDq1VVjHUKeKQzt6HKCbl4TEZGAkoKIiATUfCSShWXrv4t1CFIAHd70NSVqXBzrMHKkpCAiEiU/TBpOscq1KZPUldIJXYgrWS7WIZ1CSUFEJErOv+3f7F85h31L3mfX3H9Tqn4byiR1I75OY8zyR2t+/ohCRKQQKFr2PMq3upYaN/2Tn/3qCY7s2MjW1/7E4Y0rYh1aQDUFEZEo+mlbGvuXf8T+lXNxx49TrmUfilX5eazDCigpiIhEyaZ//5Yj29KIr9OEir+8hVIXtsLi8tfHcP6KRkSkACt1URvKXPMniparEutQsqWkICISJRXaDYx1CLlSR7OIiASUFESiZMPu42zdfzzWYYjkKN81H5nZ5cDfgDjgRefc4zEOSQqhvBoEzh09woG1n7Nv6Qcc2rCEagP/j/haiWGtm5YnEYicnnyVFMwsDvgH0BVIBxaZ2Qzn3MrYRiZyen7alsa+JR+wf+VcrEgcpRp0pGKXGyhe9YJYhyaSo3yVFICWwDfOuXUAZjYFuBpQUpBzyuaX7qRIyXJU6norpS5uhxWJi3VIImHJb30K5wMbQ16n+2Ui55SKXW4irlQFdrz7DDtmPs3B9V/hnPoTJP8z51ysYwiYWV/gMufcEP/1b4CWzrm7QpYZCgz1X14MfB31QE9VGdge6yDyCb0XJ1QGDgJVgIrAMWANcCiWQcWIzosT8sN78XPnXJY3S+S35qN0oFbI65rAptAFnHPPA89HM6jcmFmKc655rOPID/RenBD6XphZWWAg8IlzblVsI4s+nRcn5Pf3Ir8lhUVAfTOrC3wPDAB+FduQRM6ec+5H8tmXGZGs5Kuk4Jw7amZ3Au/jXZL6knMu/wwfKCJSwOWrpADgnJsJzIx1HKdJ3wBP0Htxgt6LE/RenJCv34t81dEsIiKxld8uSRURkRhSUsiGmVUws9tzWaaOmeXaEe4vtzzvoitczGyumUXsag0ze9HMGkZw+6+Y2ddmttzMXjKzYn65mdkzZvaNmS01s6Z+eS0zm2Nmq8xshZndHbKtSmY2y8zW+r8rRirus2Fmv/Xj32VmI/2yXpF8n/MzM5tgZtf60xE9386WkkL2KgA5JgWgDgX46ih/2JECzzk3JMJDqbwC/AJoBJQEhvjlVwD1/Z+hwHN++VHgXudcA6A1cEfIh8hIYLZzrj4w23+dH90OdHfOVQwZv6wXkG8/DKMlCufbWVFSyN7jQD0zSzWzv/g/y81smZn1D1mmg7/M7/0awSdm9qX/0zaG8efIj3W1mU30v6VOM7NSZpZmZqPM7FOgr5kN9I95uZk9EbL+PjN70j/O2WZWxS9PNrPP/W1Oz/gm639zXOmXT/HLSvvfnBeZ2VdmdrVfXtLMpvjLTsX7IM2LYy5tZu+Y2RL/ePr75UFNxD+uR/1lPjezan55Nf94lvg/bf3yX5vZQv8c+FdWidQ5N9P5gIV499+AN4TLf/xZnwMVzKy6c26zc+5Lf90fgVWcuLP/amCiPz0R74M2XzGzccAFwAz//+Lv/vvVE/iL/17Vi22UZ8/M7vHPo+Vm9jv/f2qVmb3g1/A+MLNTzt0wz7cqZvaG/7+xyMzaRe3AnHP6yeIHrxaw3J++BpiFd5lsNeA7oDrQGXg7ZJ1SQLw/XR9Iybyt/PLjx+SAdv7rl4BheINzjvDLavjHWgXvSrWPgF7+PAdc50+PAv7uTy8FOvnTY4Cn/elNQAl/uoL/+/+AX2eU4d3tWxq4B+9yZIAkvG/OzfPgmK8BXgh5Xd7/PTdj+/5xXeVP/xn4oz89FfidPx0HlAcaAP8Divnl/wSuz2H/xYAvgQ7+67eB9iHzZ2c+Tv/v9B1Qzn+9O9P8XbE+l7I51jS8O3cHh5wbE4BrYx1bHh1fM2CZf76WAVYATfxzNdlf5rWQ8zs49jDPt1czzg2gNrAqWsemmkJ42gOTnXPHnHNbgHlAiyyWKwa8YGbLgNfJ/1Xljc65z/zpSXjHCd4HIHjHONc5t805dxSvGaSjP+94yHKTgPZmVh7vA3+eXz4xZPmlwCtm9mu8fxyAbsBIM0vF+0eJx/sH6OhvE+fcUn/dvLAM+KWZPWFmHZxze7JY5ie8D2uAxXgfygCX4Dfv+OfBHuBSvA+HRf4xXIr3DTk7/wQ+ds594r+2LJYJLgc0szLAG3jJaG/uhydR1B6Y7pzb75zbB/wX6ACsd86l+suEnj/Zye58+yXwd/+8mgGUM++u+IjLd/cp5FNZ/fNm5ffAFqAxXtNcfh/jJvP1yBmv9/u/wz3urLaVWQ+8D/uewJ/MLMHf/jXOuZPGrzKzcLZ32pxza8ysGdAdeMzMPnDOjcm02BHnfz3DG6sop/8RAyY65+7Pbd9m9iBejeuWkOJsh3UxrzP6DeAV59x/Q5bZktHEZGbVga257VsiIrv/jcMh08fIvekzu/OtCNDGOXfwzEM8M6opZO9HICMzfwz0N7M4v+28I17bcOgy4DUpbHbecJi/wWtmyM9qm1kbf3og8Gmm+V8Ancysst9WPhCvlgTeuXOtP/0r4FP/2/MuM+vgl/8GmGdmRYBazrk5wAi8pqIyeHeu32V+FjCzJv56HwPX+WWJeE1IZ83MagAHnHOTgL8CTU9j9dnAbf524sysnF92rZlV9csrmdnPs9jvEOAyYKA7eajUGcD15mkN7PE/7A0Yj9dk8FSmzc0ABvnTg4C3TuMYYi3z/8u57GOgl3n9cKWB3sAnuaxzOj4A7sx4YWbJebjtHCkpZMM5twP4zLxLSdvgNWEswWtXH+Gc+8EvO+p3Ev0er3lgkJl9DlzEiW/c+dUqvHiXApU4cfULAM65zcD9wBy8Y//SOZfxIbQfSDCzxXhNKxnfuAfhdSYuBZL98jhgkt+s9hUw1jm3G3gYr8ltqf8+P+xv4zmgjL+NEXgJOC80Ahb6VfIHgEdOY927gS7+MSwGEpx3BckfgQ/8WGfh9TVlNg6vL2qB38k6yi+fCawDvgFe4MTVbu3wEuol/vKpZtbdn/c40NXM1uI9jOpcejLhFGC4eRcVnNMdzc67EGAC3rn5BfAisCsPd/FboLl5F1usBG7Nw23nSHc0F1JmVgevkzy8Z0Oeuv4+51yZvI1KRGJNNQUREQmopiAiIgHVFEREJKCkICIiASUFEREJKCmIiEhASUFERAJKCiIiEvh/ZzvuxXHvNJMAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the results\n",
    "fig = df_decomp.T.plot(kind='bar', stacked=True, figsize=(6,8))\n",
    "text_init_height = -1.1\n",
    "\n",
    "# Print the significance levels\n",
    "cumsum = text_init_height\n",
    "for idx, val in enumerate(df_decomp[\"proposed\"]):\n",
    "    cumsum += val/2\n",
    "    plt.text(0.9725, cumsum, significances_proposed[idx], fontweight=\"bold\")\n",
    "    cumsum += val/2\n",
    "cumsum = text_init_height\n",
    "for idx, val in enumerate(df_decomp[\"since 2020\"]):\n",
    "    cumsum += val/2\n",
    "    plt.text(1.9725, cumsum, significances_2020[idx], fontweight=\"bold\")\n",
    "    cumsum += val/2\n",
    "cumsum = text_init_height\n",
    "for idx, val in enumerate(df_decomp[\"fit\"]):\n",
    "    cumsum += val/2\n",
    "    plt.text(2.9725, cumsum, significances_fit[idx], fontweight=\"bold\")\n",
    "    cumsum += val/2\n",
    "cumsum = text_init_height\n",
    "for idx, val in enumerate(df_decomp[\"online\"]):\n",
    "    cumsum += val/2\n",
    "    plt.text(3.9725, cumsum, significances_online[idx], fontweight=\"bold\")\n",
    "    cumsum += val/2\n",
    "\n",
    "# plot layout\n",
    "plt.title(\"Number of Decomposition Methods\")\n",
    "plt.ylabel(\"Number of publications\")\n",
    "plt.xticks(rotation=0)\n",
    "plt.legend(labels, loc=\"upper right\")\n",
    "plt.savefig(\"Figures/barplot_decompositions.pdf\", format=\"pdf\", bbox_inches=\"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "db782f02",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "407bc0a4-90cc-45fa-94eb-018944992abe",
   "metadata": {},
   "source": [
    "#### ICA flavors methods (total)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "720baff1-0469-4f51-9d1b-36e259328a19",
   "metadata": {},
   "outputs": [],
   "source": [
    "ica_map = {\n",
    "    \"Fast ICA\": \"FastICA\",\n",
    "    \"ICA-\": \"wICA\",\n",
    "    \"w-ICA\": \"wICA\"\n",
    "}\n",
    "    \n",
    "\n",
    "def extract_flavor(x):\n",
    "    x = str(x).split(\"\\t\")[-1]\n",
    "    if str(x) == \"nan\":\n",
    "        return list()\n",
    "    flavors = list()\n",
    "    for m in str(x).split(\",\"):\n",
    "        for method in m.split(\"+\"):\n",
    "            r = method.split(\":\")[-1].strip()\n",
    "            if not r.isnumeric() and r != \"ICA\":\n",
    "                flavors.append(\n",
    "                    ica_map.get(r, r)\n",
    "                )\n",
    "            else:\n",
    "                flavors.append(\"unspecified\")\n",
    "    return flavors\n",
    "\n",
    "pipelines_annotated_df[\"Ref+ICA\"] = pipelines_annotated_df.index + \"\\t\" + pipelines_annotated_df[\"ICA\"]\n",
    "flavors = list(map(extract_flavor, pipelines_annotated_df[\"Ref+ICA\"].unique()))\n",
    "flavors = [f for sublist in flavors for f in sublist]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "b8d691a2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array(['AMICA', 'CCA', 'CUDAICA', 'EAWICA', 'ERICA', 'FastICA',\n",
       "        'Fourier-ICA', 'ICA-W', 'IMICA', 'Infomax', 'InfomaxICA', 'JADE',\n",
       "        'MCCA', 'SCICA', 'TDSEP', 'extended Infomax', 'unspecified',\n",
       "        'wICA'], dtype='<U16'),\n",
       " array([ 8,  1,  1,  1,  1,  8,  1,  1,  1,  2,  7,  2,  1,  1,  3,  3, 13,\n",
       "         9]))"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.unique(flavors, return_counts=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "12270a3a-5bf9-4cb4-9978-f91791aaf618",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'Most common ICA methods.'"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "threshold = 3\n",
    "labels, counts = np.unique(flavors, return_counts=True)\n",
    "labels = [f\"{f}: {counts[i]}\" for i, f in enumerate(labels)]\n",
    "\n",
    "labels.append(\"Other\")\n",
    "counts = list(counts)\n",
    "counts.append(0)\n",
    "droplist = []\n",
    "for i, c in enumerate(counts):\n",
    "    if c < threshold:\n",
    "        counts[-1] += c\n",
    "        droplist.append(i)\n",
    "labels = [lab for i, lab in enumerate(labels) if i not in droplist]\n",
    "counts = [c for i, c in enumerate(counts) if i not in droplist]\n",
    "labels[-1] = f\"Other: {counts[-1]}\"\n",
    "\n",
    "plt.pie(sorted(counts), \n",
    "        labels=[lab for _, lab in sorted(zip(counts, labels))])\n",
    "plt.title(f\"Most common ICA flavors\")\n",
    "plt.savefig(\"Figures/mostcommon_ICAflavors.pdf\", format=\"pdf\", bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\"Most common ICA methods.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d133e114",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "efb1cb86-2306-49bf-a1dd-4459380caa51",
   "metadata": {},
   "source": [
    "#### Artefact detection and correction methods (total number of publications using them)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "5564fbc3-2f3a-48e3-9e54-7a80aed5577d",
   "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>total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Linear Regression</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DA</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SVM</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Spatial sph. spline</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Other</th>\n",
       "      <td>23</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     total\n",
       "Linear Regression        3\n",
       "DA                       3\n",
       "SVM                      4\n",
       "Spatial sph. spline      4\n",
       "Other                   23"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distr = (pipelines_annotated_df.groupby(\"Reference\").agg(\"count\")[\n",
    "    ['Linear Regression', 'DA', 'SVM', 'Spatial sph. spline', 'GMM', 'Other.2', 'Autoencoder', 'GAN', 'Other.3']\n",
    "        ]>0).sum(axis=0)\n",
    "distr[\"Other\"] = distr[\"Other.2\"] + distr[\"Other.3\"]\n",
    "# distr[\"CSP\"] = distr[\"CSP.1\"]\n",
    "\n",
    "distr[\"Other\"] += distr[\"GMM\"]\n",
    "distr[\"Other\"] += distr[\"Autoencoder\"]\n",
    "distr[\"Other\"] += distr[\"GAN\"]\n",
    "[distr.pop(key) for key in distr.keys() if key in ['GMM', 'Autoencoder', 'GAN', 'Other.2', 'Other.3']]\n",
    "\n",
    "# store in df_decomp\n",
    "df_decomp = pd.DataFrame(data=distr, columns=[\"total\"])\n",
    "df_decomp.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "c132af61-e702-4edd-9ff6-58345e49307a",
   "metadata": {},
   "outputs": [],
   "source": [
    "colors = {\n",
    "    'Linear Regression': \"C1\",\n",
    "     'SVM': \"C2\",\n",
    "     'Spatial sph. spline': \"C3\",\n",
    "     'CSP': \"C4\",\n",
    "     'DA': \"C5\",\n",
    "     'Other': \"C6\",\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "b2a54f0a-f3d8-4801-8ae0-6ab5feb88bfe",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'Number of publications pipelines with selected artifact methods, methods occuring only in one publications are ignored.'"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels = [f\"{p}: {v}\" for p,v in distr.items()]\n",
    "plt.pie(distr, labels=labels, colors=[colors[lab.split(\":\")[0]] for lab in labels])\n",
    "plt.title(f\"Most common artifact detection\\nmethods by number of papers\")\n",
    "plt.savefig(\"Figures/mostcommon_detection_by_paper.pdf\", format=\"pdf\", bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\"Number of publications pipelines with selected artifact methods, methods occuring only in one publications are ignored.\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b4f5ca1b-1628-4bf8-ac2c-b6d11c2fe8a8",
   "metadata": {},
   "source": [
    "#### Artefact detection and correction methods (proposed number of publications using them)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "ae599a42-2d0f-4d32-9195-4e12c93f5088",
   "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>total</th>\n",
       "      <th>proposed</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Linear Regression</th>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DA</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SVM</th>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Spatial sph. spline</th>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Other</th>\n",
       "      <td>23</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     total  proposed\n",
       "Linear Regression        3         2\n",
       "DA                       3         1\n",
       "SVM                      4         3\n",
       "Spatial sph. spline      4         3\n",
       "Other                   23        19"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distr = (pipelines_annotated_df[pipelines_annotated_df[\"is new\"] > 0].groupby(\"Reference\").agg(\"count\")[\n",
    "    ['Linear Regression', 'DA', 'SVM', 'Spatial sph. spline', 'GMM', 'Other.2', 'Autoencoder', 'GAN', 'Other.3']\n",
    "        ]>0).sum(axis=0)\n",
    "distr[\"Other\"] = distr[\"Other.2\"] + distr[\"Other.3\"]\n",
    "# distr[\"CSP\"] = distr[\"CSP.1\"]\n",
    "\n",
    "distr[\"Other\"] += distr[\"GMM\"]\n",
    "distr[\"Other\"] += distr[\"Autoencoder\"]\n",
    "distr[\"Other\"] += distr[\"GAN\"]\n",
    "[distr.pop(key) for key in distr.keys() if key in ['GMM', 'Autoencoder', 'GAN', 'CSP.1', 'Other.2', 'Other.3']]\n",
    "\n",
    "# store in df_decomp\n",
    "df_decomp[\"proposed\"] = distr\n",
    "df_decomp.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "51a8c185",
   "metadata": {},
   "source": [
    "#### Artefact detection and correction methods (number of publications since 2020 using them)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "2000b7f6",
   "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>total</th>\n",
       "      <th>proposed</th>\n",
       "      <th>since 2020</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Linear Regression</th>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DA</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SVM</th>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Spatial sph. spline</th>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Other</th>\n",
       "      <td>23</td>\n",
       "      <td>19</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     total  proposed  since 2020\n",
       "Linear Regression        3         2           1\n",
       "DA                       3         1           0\n",
       "SVM                      4         3           2\n",
       "Spatial sph. spline      4         3           3\n",
       "Other                   23        19           9"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distr = (pipelines_annotated_df[pipelines_annotated_df[\"year\"] >= 2020].groupby(\"Reference\").agg(\"count\")[\n",
    "    ['Linear Regression', 'DA', 'SVM', 'Spatial sph. spline', 'GMM', 'Other.2', 'Autoencoder', 'GAN', 'Other.3']\n",
    "        ]>0).sum(axis=0)\n",
    "distr[\"Other\"] = distr[\"Other.2\"] + distr[\"Other.3\"]\n",
    "# distr[\"CSP\"] = distr[\"CSP.1\"]\n",
    "\n",
    "distr[\"Other\"] += distr[\"GMM\"]\n",
    "distr[\"Other\"] += distr[\"Autoencoder\"]\n",
    "distr[\"Other\"] += distr[\"GAN\"]\n",
    "[distr.pop(key) for key in distr.keys() if key in ['GMM', 'Autoencoder', 'GAN', 'CSP.1', 'Other.2', 'Other.3']]\n",
    "\n",
    "# store in df_decomp\n",
    "df_decomp[\"since 2020\"] = distr\n",
    "df_decomp.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4b70e844",
   "metadata": {},
   "source": [
    "#### Artefact detection and correction methods (number publications of fit pipelines using them)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "db7cedfe",
   "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>total</th>\n",
       "      <th>proposed</th>\n",
       "      <th>since 2020</th>\n",
       "      <th>fit</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Linear Regression</th>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DA</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SVM</th>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Spatial sph. spline</th>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Other</th>\n",
       "      <td>23</td>\n",
       "      <td>19</td>\n",
       "      <td>9</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     total  proposed  since 2020  fit\n",
       "Linear Regression        3         2           1    1\n",
       "DA                       3         1           0    2\n",
       "SVM                      4         3           2    2\n",
       "Spatial sph. spline      4         3           3    0\n",
       "Other                   23        19           9   13"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distr = (pipelines_annotated_df[pipelines_annotated_df[\"fit score (0-2)\"] == 2].groupby(\"Reference\").agg(\"count\")[\n",
    "    ['Linear Regression', 'DA', 'SVM', 'Spatial sph. spline', 'GMM', 'Other.2', 'Autoencoder', 'GAN', 'Other.3']\n",
    "        ]>0).sum(axis=0)\n",
    "distr[\"Other\"] = distr[\"Other.2\"] + distr[\"Other.3\"]\n",
    "# distr[\"CSP\"] = distr[\"CSP.1\"]\n",
    "\n",
    "distr[\"Other\"] += distr[\"GMM\"]\n",
    "distr[\"Other\"] += distr[\"Autoencoder\"]\n",
    "distr[\"Other\"] += distr[\"GAN\"]\n",
    "[distr.pop(key) for key in distr.keys() if key in ['GMM', 'Autoencoder', 'GAN', 'CSP.1', 'Other.2', 'Other.3']]\n",
    "\n",
    "# store in df_decomp\n",
    "df_decomp[\"fit\"] = distr\n",
    "df_decomp.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f03c7b19",
   "metadata": {},
   "source": [
    "#### Artefact detection and correction methods (number publications of online pipelines using them)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "ae5b4f86",
   "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>total</th>\n",
       "      <th>proposed</th>\n",
       "      <th>since 2020</th>\n",
       "      <th>fit</th>\n",
       "      <th>online</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Linear Regression</th>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DA</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SVM</th>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Spatial sph. spline</th>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Other</th>\n",
       "      <td>23</td>\n",
       "      <td>19</td>\n",
       "      <td>9</td>\n",
       "      <td>13</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     total  proposed  since 2020  fit  online\n",
       "Linear Regression        3         2           1    1       0\n",
       "DA                       3         1           0    2       3\n",
       "SVM                      4         3           2    2       1\n",
       "Spatial sph. spline      4         3           3    0       1\n",
       "Other                   23        19           9   13       5"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distr = (pipelines_annotated_df[pipelines_annotated_df[\"online score (0-2)\"] == 2].groupby(\"Reference\").agg(\"count\")[\n",
    "    ['Linear Regression', 'DA', 'SVM', 'Spatial sph. spline', 'GMM', 'Other.2', 'Autoencoder', 'GAN', 'Other.3']\n",
    "        ]>0).sum(axis=0)\n",
    "distr[\"Other\"] = distr[\"Other.2\"] + distr[\"Other.3\"]\n",
    "# distr[\"CSP\"] = distr[\"CSP.1\"]\n",
    "\n",
    "distr[\"Other\"] += distr[\"GMM\"]\n",
    "distr[\"Other\"] += distr[\"Autoencoder\"]\n",
    "distr[\"Other\"] += distr[\"GAN\"]\n",
    "[distr.pop(key) for key in distr.keys() if key in ['GMM', 'Autoencoder', 'GAN', 'CSP.1', 'Other.2', 'Other.3']]\n",
    "\n",
    "# store in df_decomp\n",
    "df_decomp[\"online\"] = distr\n",
    "df_decomp.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eea886bf",
   "metadata": {},
   "source": [
    "### Plot the stacked barplot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "8eed9bd5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# calculate the p-values to all cases\n",
    "p_values_proposed = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"proposed\"], alternative=\"greater\", n_resamples=10_000)\n",
    "p_values_2020 = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"since 2020\"], alternative=\"greater\", n_resamples=10_000)\n",
    "p_values_fit = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"fit\"], alternative=\"greater\", n_resamples=10_000)\n",
    "p_values_online = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"online\"], alternative=\"greater\", n_resamples=10_000)\n",
    "\n",
    "p_values_proposed_less = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"proposed\"], alternative=\"less\", n_resamples=10_000)\n",
    "p_values_2020_less = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"since 2020\"], alternative=\"less\", n_resamples=10_000)\n",
    "p_values_fit_less = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"fit\"], alternative=\"less\", n_resamples=10_000)\n",
    "p_values_online_less = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"online\"], alternative=\"less\", n_resamples=10_000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "3de59785",
   "metadata": {},
   "outputs": [],
   "source": [
    "# annotate the significances\n",
    "ALPHA = 0.05\n",
    "sig_char = \">\"\n",
    "sig_neg_char = \"<\"\n",
    "insig_char = \" \"\n",
    "\n",
    "significances_proposed = [sig_neg_char if p_values_proposed_less[idx] <= ALPHA else char \n",
    "                          for idx, char in enumerate([sig_char if p_values_proposed[idx] <= ALPHA else insig_char for idx in range(len(df_decomp))])]\n",
    "significances_2020 = [sig_neg_char if p_values_2020_less[idx] <= ALPHA else char \n",
    "                      for idx, char in enumerate([sig_char if p_values_2020[idx] <= ALPHA else insig_char for idx in range(len(df_decomp))])]\n",
    "significances_fit = [sig_neg_char if p_values_fit_less[idx] <= ALPHA else char \n",
    "                     for idx, char in enumerate([sig_char if p_values_fit[idx] <= ALPHA else insig_char for idx in range(len(df_decomp))])]\n",
    "significances_online = [sig_neg_char if p_values_online_less[idx] <= ALPHA else char \n",
    "                        for idx, char in enumerate([sig_char if p_values_online[idx] <= ALPHA else insig_char for idx in range(len(df_decomp))])]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "7c59a88b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the results\n",
    "fig = df_decomp.T.plot(kind='bar', stacked=True, figsize=(6,6))\n",
    "\n",
    "# Print the significance levels\n",
    "cumsum = -0.4\n",
    "for idx, val in enumerate(df_decomp[\"proposed\"]):\n",
    "    cumsum += val/2\n",
    "    plt.text(0.9725, cumsum, significances_proposed[idx], fontweight=\"bold\")\n",
    "    cumsum += val/2\n",
    "cumsum = -0.85\n",
    "for idx, val in enumerate(df_decomp[\"since 2020\"]):\n",
    "    cumsum += val/2\n",
    "    plt.text(1.9725, cumsum, significances_2020[idx], fontweight=\"bold\")\n",
    "    cumsum += val/2\n",
    "cumsum = -0.4\n",
    "for idx, val in enumerate(df_decomp[\"fit\"]):\n",
    "    if val > 0:\n",
    "        cumsum += val/2\n",
    "        plt.text(2.9725, cumsum, significances_fit[idx], fontweight=\"bold\")\n",
    "        cumsum += val/2\n",
    "cumsum = -0.4\n",
    "for idx, val in enumerate(df_decomp[\"online\"]):\n",
    "    if val > 0:\n",
    "        cumsum += val/2\n",
    "        plt.text(3.9725, cumsum, significances_online[idx], fontweight=\"bold\")\n",
    "        cumsum += val/2\n",
    "\n",
    "# plot layout\n",
    "plt.title(\"Number of Artifact Detection Methods\")\n",
    "plt.ylabel(\"Number of publications\")\n",
    "plt.xticks(rotation=0)\n",
    "plt.legend(labels, loc=\"upper right\")\n",
    "plt.savefig(\"Figures/barplot_detections.pdf\", format=\"pdf\", bbox_inches=\"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c4d5afed-91d0-4e6f-b3f6-5f70a79bdcda",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "f96ae714-8366-4c81-b79b-67b4cbb73ccc",
   "metadata": {},
   "source": [
    "#### Special methods (total number of publications using them)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "bd527cff-3336-4c78-8474-a5e6e4dcad1e",
   "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>total</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MARA</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FASTER</th>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ADJUST</th>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Other</th>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        total\n",
       "MARA        4\n",
       "FASTER      8\n",
       "ADJUST      8\n",
       "Other      14"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distr = (pipelines_annotated_df.groupby(\"Reference\").agg(\"count\")[\n",
    "    ['MARA', 'DSS', 'DFA', 'FASTER', 'ADJUST', 'HAPPE', 'ERASE', 'Other.4']\n",
    "        ]>0).sum(axis=0)\n",
    "distr[\"Other\"] = distr[\"Other.4\"]\n",
    "\n",
    "distr[\"Other\"] += distr[\"DSS\"]\n",
    "distr[\"Other\"] += distr[\"DFA\"]\n",
    "distr[\"Other\"] += distr[\"HAPPE\"]\n",
    "distr[\"Other\"] += distr[\"ERASE\"]\n",
    "[distr.pop(key) for key in distr.keys() if key in ['DSS', 'DFA', 'HAPPE', 'ERASE', 'Other.4']]\n",
    "\n",
    "# store in df_decomp\n",
    "df_decomp = pd.DataFrame(data=distr, columns=[\"total\"])\n",
    "df_decomp.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "d4598d79-96f1-480a-8d5e-6cd054135c5a",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "'Number of publications pipelines with selected special methods an subpipelines, methods occuring only in one publications are ignored.'"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "labels = [f\"{p}: {v}\" for p,v in distr.items()]\n",
    "plt.pie(distr, labels=labels)\n",
    "plt.title(f\"Most common special EEG methods and subpipelines\")\n",
    "plt.savefig(\"Figures/barplot_specials.pdf\", format=\"pdf\", bbox_inches=\"tight\")\n",
    "plt.show()\n",
    "\"Number of publications pipelines with selected special methods an subpipelines, methods occuring only in one publications are ignored.\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a9bd6043",
   "metadata": {},
   "source": [
    "#### Artefact detection and correction methods (proposed number of publications using them)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "4e928bbe",
   "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>total</th>\n",
       "      <th>proposed</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MARA</th>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FASTER</th>\n",
       "      <td>8</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ADJUST</th>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Other</th>\n",
       "      <td>14</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        total  proposed\n",
       "MARA        4         2\n",
       "FASTER      8         3\n",
       "ADJUST      8         2\n",
       "Other      14        10"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distr = (pipelines_annotated_df[pipelines_annotated_df[\"is new\"] > 0].groupby(\"Reference\").agg(\"count\")[\n",
    "    ['MARA', 'DSS', 'DFA', 'FASTER', 'ADJUST', 'HAPPE', 'ERASE', 'Other.4']\n",
    "        ]>0).sum(axis=0)\n",
    "distr[\"Other\"] = distr[\"Other.4\"]\n",
    "\n",
    "distr[\"Other\"] += distr[\"DSS\"]\n",
    "distr[\"Other\"] += distr[\"DFA\"]\n",
    "distr[\"Other\"] += distr[\"HAPPE\"]\n",
    "distr[\"Other\"] += distr[\"ERASE\"]\n",
    "[distr.pop(key) for key in distr.keys() if key in ['DSS', 'DFA', 'HAPPE', 'ERASE', 'Other.4']]\n",
    "\n",
    "# store in df_decomp\n",
    "df_decomp[\"proposed\"] = distr\n",
    "df_decomp.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ca2633ea",
   "metadata": {},
   "source": [
    "#### Artefact detection and correction methods (number of publications since 2020 using them)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "eeab1f31",
   "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>total</th>\n",
       "      <th>proposed</th>\n",
       "      <th>since 2020</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MARA</th>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FASTER</th>\n",
       "      <td>8</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ADJUST</th>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Other</th>\n",
       "      <td>14</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        total  proposed  since 2020\n",
       "MARA        4         2           2\n",
       "FASTER      8         3           3\n",
       "ADJUST      8         2           4\n",
       "Other      14        10           6"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distr = (pipelines_annotated_df[pipelines_annotated_df[\"year\"] >= 2020].groupby(\"Reference\").agg(\"count\")[\n",
    "    ['MARA', 'DSS', 'DFA', 'FASTER', 'ADJUST', 'HAPPE', 'ERASE', 'Other.4']\n",
    "        ]>0).sum(axis=0)\n",
    "distr[\"Other\"] = distr[\"Other.4\"]\n",
    "\n",
    "distr[\"Other\"] += distr[\"DSS\"]\n",
    "distr[\"Other\"] += distr[\"DFA\"]\n",
    "distr[\"Other\"] += distr[\"HAPPE\"]\n",
    "distr[\"Other\"] += distr[\"ERASE\"]\n",
    "[distr.pop(key) for key in distr.keys() if key in ['DSS', 'DFA', 'HAPPE', 'ERASE', 'Other.4']]\n",
    "\n",
    "# store in df_decomp\n",
    "df_decomp[\"since 2020\"] = distr\n",
    "df_decomp.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3ba6ed52",
   "metadata": {},
   "source": [
    "#### Artefact detection and correction methods (number publications of fit pipelines using them)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "09e56c55",
   "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>total</th>\n",
       "      <th>proposed</th>\n",
       "      <th>since 2020</th>\n",
       "      <th>fit</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MARA</th>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FASTER</th>\n",
       "      <td>8</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ADJUST</th>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Other</th>\n",
       "      <td>14</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        total  proposed  since 2020  fit\n",
       "MARA        4         2           2    0\n",
       "FASTER      8         3           3    3\n",
       "ADJUST      8         2           4    3\n",
       "Other      14        10           6    6"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distr = (pipelines_annotated_df[pipelines_annotated_df[\"fit score (0-2)\"] == 2].groupby(\"Reference\").agg(\"count\")[\n",
    "    ['MARA', 'DSS', 'DFA', 'FASTER', 'ADJUST', 'HAPPE', 'ERASE', 'Other.4']\n",
    "        ]>0).sum(axis=0)\n",
    "distr[\"Other\"] = distr[\"Other.4\"]\n",
    "\n",
    "distr[\"Other\"] += distr[\"DSS\"]\n",
    "distr[\"Other\"] += distr[\"DFA\"]\n",
    "distr[\"Other\"] += distr[\"HAPPE\"]\n",
    "distr[\"Other\"] += distr[\"ERASE\"]\n",
    "[distr.pop(key) for key in distr.keys() if key in ['DSS', 'DFA', 'HAPPE', 'ERASE', 'Other.4']]\n",
    "\n",
    "# store in df_decomp\n",
    "df_decomp[\"fit\"] = distr\n",
    "df_decomp.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bbb129b8",
   "metadata": {},
   "source": [
    "#### Artefact detection and correction methods (number publications of online pipelines using them)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "id": "80127458",
   "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>total</th>\n",
       "      <th>proposed</th>\n",
       "      <th>since 2020</th>\n",
       "      <th>fit</th>\n",
       "      <th>online</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>MARA</th>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FASTER</th>\n",
       "      <td>8</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ADJUST</th>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Other</th>\n",
       "      <td>14</td>\n",
       "      <td>10</td>\n",
       "      <td>6</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        total  proposed  since 2020  fit  online\n",
       "MARA        4         2           2    0       0\n",
       "FASTER      8         3           3    3       2\n",
       "ADJUST      8         2           4    3       1\n",
       "Other      14        10           6    6       2"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distr = (pipelines_annotated_df[pipelines_annotated_df[\"online score (0-2)\"] == 2].groupby(\"Reference\").agg(\"count\")[\n",
    "    ['MARA', 'DSS', 'DFA', 'FASTER', 'ADJUST', 'HAPPE', 'ERASE', 'Other.4']\n",
    "        ]>0).sum(axis=0)\n",
    "distr[\"Other\"] = distr[\"Other.4\"]\n",
    "\n",
    "distr[\"Other\"] += distr[\"DSS\"]\n",
    "distr[\"Other\"] += distr[\"DFA\"]\n",
    "distr[\"Other\"] += distr[\"HAPPE\"]\n",
    "distr[\"Other\"] += distr[\"ERASE\"]\n",
    "[distr.pop(key) for key in distr.keys() if key in ['DSS', 'DFA', 'HAPPE', 'ERASE', 'Other.4']]\n",
    "\n",
    "# store in df_decomp\n",
    "df_decomp[\"online\"] = distr\n",
    "df_decomp.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b467727e",
   "metadata": {},
   "source": [
    "### Plot the stacked barplot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "id": "0d72f1a5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# calculate the p-values to all cases\n",
    "p_values_proposed = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"proposed\"], alternative=\"greater\", n_resamples=10_000)\n",
    "p_values_2020 = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"since 2020\"], alternative=\"greater\", n_resamples=10_000)\n",
    "p_values_fit = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"fit\"], alternative=\"greater\", n_resamples=10_000)\n",
    "p_values_online = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"online\"], alternative=\"greater\", n_resamples=10_000)\n",
    "\n",
    "p_values_proposed_less = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"proposed\"], alternative=\"less\", n_resamples=10_000)\n",
    "p_values_2020_less = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"since 2020\"], alternative=\"less\", n_resamples=10_000)\n",
    "p_values_fit_less = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"fit\"], alternative=\"less\", n_resamples=10_000)\n",
    "p_values_online_less = bootstrap_test_multiple(df_decomp[\"total\"], df_decomp[\"online\"], alternative=\"less\", n_resamples=10_000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "53cb830c",
   "metadata": {},
   "outputs": [],
   "source": [
    "# annotate the significances\n",
    "ALPHA = 0.05\n",
    "sig_char = \">\"\n",
    "sig_neg_char = \"<\"\n",
    "insig_char = \" \"\n",
    "\n",
    "significances_proposed = [sig_neg_char if p_values_proposed_less[idx] <= ALPHA else char \n",
    "                          for idx, char in enumerate([sig_char if p_values_proposed[idx] <= ALPHA else insig_char for idx in range(len(df_decomp))])]\n",
    "significances_2020 = [sig_neg_char if p_values_2020_less[idx] <= ALPHA else char \n",
    "                      for idx, char in enumerate([sig_char if p_values_2020[idx] <= ALPHA else insig_char for idx in range(len(df_decomp))])]\n",
    "significances_fit = [sig_neg_char if p_values_fit_less[idx] <= ALPHA else char \n",
    "                     for idx, char in enumerate([sig_char if p_values_fit[idx] <= ALPHA else insig_char for idx in range(len(df_decomp))])]\n",
    "significances_online = [sig_neg_char if p_values_online_less[idx] <= ALPHA else char \n",
    "                        for idx, char in enumerate([sig_char if p_values_online[idx] <= ALPHA else insig_char for idx in range(len(df_decomp))])]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "1d3b839a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the results\n",
    "fig = df_decomp.T.plot(kind='bar', stacked=True, figsize=(6,5))\n",
    "\n",
    "# Print the significance levels\n",
    "cumsum = -0.4\n",
    "for idx, val in enumerate(df_decomp[\"proposed\"]):\n",
    "    cumsum += val/2\n",
    "    plt.text(0.9725, cumsum, significances_proposed[idx], fontweight=\"bold\")\n",
    "    cumsum += val/2\n",
    "cumsum = -0.85\n",
    "for idx, val in enumerate(df_decomp[\"since 2020\"]):\n",
    "    cumsum += val/2\n",
    "    plt.text(1.9725, cumsum, significances_2020[idx], fontweight=\"bold\")\n",
    "    cumsum += val/2\n",
    "cumsum = -0.4\n",
    "for idx, val in enumerate(df_decomp[\"fit\"]):\n",
    "    if val > 0:\n",
    "        cumsum += val/2\n",
    "        plt.text(2.9725, cumsum, significances_fit[idx], fontweight=\"bold\")\n",
    "        cumsum += val/2\n",
    "cumsum = -0.4\n",
    "for idx, val in enumerate(df_decomp[\"online\"]):\n",
    "    if val > 0:\n",
    "        cumsum += val/2\n",
    "        plt.text(3.9725, cumsum, significances_online[idx], fontweight=\"bold\")\n",
    "        cumsum += val/2\n",
    "\n",
    "# plot layout\n",
    "plt.title(\"Number of Special Methods and Subpipelines\")\n",
    "plt.ylabel(\"Number of publications\")\n",
    "plt.xticks(rotation=0)\n",
    "plt.legend(labels, loc=\"upper right\")\n",
    "plt.savefig(\"Figures/barplot_specials.pdf\", format=\"pdf\", bbox_inches=\"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "48540003-ec9d-4f04-9943-2ebcde3407d3",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "35a3b48c",
   "metadata": {},
   "source": [
    "## Find publications that compare ICA with another method group"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "id": "cf2649e0",
   "metadata": {},
   "outputs": [],
   "source": [
    "ica_refs = set(pipelines_annotated_df[~pipelines_annotated_df[\"ICA\"].isnull()].groupby(\"Reference\").agg(\"count\").index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "id": "aff80b1b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "set()"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ica_refs.intersection(\n",
    "    set(pipelines_annotated_df[~pipelines_annotated_df[\"EMD\"].isnull()].groupby(\"Reference\").agg(\"count\").index)\n",
    "    ).difference({\"Chen14\", \"Dai21\", \"Saini20\"})  # remove some as there are no comparisions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "id": "f23cae23",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Chen14', 'Dai21'}"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ica_refs.intersection(\n",
    "    set(pipelines_annotated_df[~pipelines_annotated_df[\"CCA\"].isnull()].groupby(\"Reference\").agg(\"count\").index)\n",
    ").difference({\"Nordin20\"})  # remove some as there are no comparisions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "id": "446d85f0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Arad18', 'Gordon15', 'Rosanne21'}"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ica_refs.intersection(\n",
    "    set(pipelines_annotated_df[~pipelines_annotated_df[\"PCA\"].isnull()].groupby(\"Reference\").agg(\"count\").index)\n",
    ").difference({\"Nordin20\", \"Anders20\", 'Gabard-Durnam18', \"Flo22\", 'Sebek18', 'Tamburro18', \"McMenamin10\", \"Winkler11\"})  # remove some as there are no comparisions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "id": "689d26b0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "set()"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ica_refs.intersection(\n",
    "    set(pipelines_annotated_df[~pipelines_annotated_df[\"CSP\"].isnull()].groupby(\"Reference\").agg(\"count\").index)\n",
    ").difference({\"Winkler11\"})  # remove some as there are no comparisions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b1187417",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}