notebooks/mimic.sample.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn.utils import shuffle\n",
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Study population"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_dir = '../datafiles/mimic/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "metadata_df = pd.read_csv(data_dir + 'mimic-cxr-2.0.0-metadata.csv')\n",
    "diagnosis_df = pd.read_csv(data_dir + 'mimic-cxr-2.0.0-chexpert.csv')\n",
    "admissions_df = pd.read_csv(data_dir + 'admissions.csv')\n",
    "patients_df = pd.read_csv(data_dir + 'patients.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>dicom_id</th>\n",
       "      <th>subject_id</th>\n",
       "      <th>study_id</th>\n",
       "      <th>PerformedProcedureStepDescription</th>\n",
       "      <th>ViewPosition</th>\n",
       "      <th>Rows</th>\n",
       "      <th>Columns</th>\n",
       "      <th>StudyDate</th>\n",
       "      <th>StudyTime</th>\n",
       "      <th>ProcedureCodeSequence_CodeMeaning</th>\n",
       "      <th>ViewCodeSequence_CodeMeaning</th>\n",
       "      <th>PatientOrientationCodeSequence_CodeMeaning</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>02aa804e-bde0afdd-112c0b34-7bc16630-4e384014</td>\n",
       "      <td>10000032</td>\n",
       "      <td>50414267</td>\n",
       "      <td>CHEST (PA AND LAT)</td>\n",
       "      <td>PA</td>\n",
       "      <td>3056</td>\n",
       "      <td>2544</td>\n",
       "      <td>21800506</td>\n",
       "      <td>213014.531</td>\n",
       "      <td>CHEST (PA AND LAT)</td>\n",
       "      <td>postero-anterior</td>\n",
       "      <td>Erect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>174413ec-4ec4c1f7-34ea26b7-c5f994f8-79ef1962</td>\n",
       "      <td>10000032</td>\n",
       "      <td>50414267</td>\n",
       "      <td>CHEST (PA AND LAT)</td>\n",
       "      <td>LATERAL</td>\n",
       "      <td>3056</td>\n",
       "      <td>2544</td>\n",
       "      <td>21800506</td>\n",
       "      <td>213014.531</td>\n",
       "      <td>CHEST (PA AND LAT)</td>\n",
       "      <td>lateral</td>\n",
       "      <td>Erect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2a2277a9-b0ded155-c0de8eb9-c124d10e-82c5caab</td>\n",
       "      <td>10000032</td>\n",
       "      <td>53189527</td>\n",
       "      <td>CHEST (PA AND LAT)</td>\n",
       "      <td>PA</td>\n",
       "      <td>3056</td>\n",
       "      <td>2544</td>\n",
       "      <td>21800626</td>\n",
       "      <td>165500.312</td>\n",
       "      <td>CHEST (PA AND LAT)</td>\n",
       "      <td>postero-anterior</td>\n",
       "      <td>Erect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>e084de3b-be89b11e-20fe3f9f-9c8d8dfe-4cfd202c</td>\n",
       "      <td>10000032</td>\n",
       "      <td>53189527</td>\n",
       "      <td>CHEST (PA AND LAT)</td>\n",
       "      <td>LATERAL</td>\n",
       "      <td>3056</td>\n",
       "      <td>2544</td>\n",
       "      <td>21800626</td>\n",
       "      <td>165500.312</td>\n",
       "      <td>CHEST (PA AND LAT)</td>\n",
       "      <td>lateral</td>\n",
       "      <td>Erect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>68b5c4b1-227d0485-9cc38c3f-7b84ab51-4b472714</td>\n",
       "      <td>10000032</td>\n",
       "      <td>53911762</td>\n",
       "      <td>CHEST (PORTABLE AP)</td>\n",
       "      <td>AP</td>\n",
       "      <td>2705</td>\n",
       "      <td>2539</td>\n",
       "      <td>21800723</td>\n",
       "      <td>80556.875</td>\n",
       "      <td>CHEST (PORTABLE AP)</td>\n",
       "      <td>antero-posterior</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                       dicom_id  subject_id  study_id  \\\n",
       "0  02aa804e-bde0afdd-112c0b34-7bc16630-4e384014    10000032  50414267   \n",
       "1  174413ec-4ec4c1f7-34ea26b7-c5f994f8-79ef1962    10000032  50414267   \n",
       "2  2a2277a9-b0ded155-c0de8eb9-c124d10e-82c5caab    10000032  53189527   \n",
       "3  e084de3b-be89b11e-20fe3f9f-9c8d8dfe-4cfd202c    10000032  53189527   \n",
       "4  68b5c4b1-227d0485-9cc38c3f-7b84ab51-4b472714    10000032  53911762   \n",
       "\n",
       "  PerformedProcedureStepDescription ViewPosition  Rows  Columns  StudyDate  \\\n",
       "0                CHEST (PA AND LAT)           PA  3056     2544   21800506   \n",
       "1                CHEST (PA AND LAT)      LATERAL  3056     2544   21800506   \n",
       "2                CHEST (PA AND LAT)           PA  3056     2544   21800626   \n",
       "3                CHEST (PA AND LAT)      LATERAL  3056     2544   21800626   \n",
       "4               CHEST (PORTABLE AP)           AP  2705     2539   21800723   \n",
       "\n",
       "    StudyTime ProcedureCodeSequence_CodeMeaning ViewCodeSequence_CodeMeaning  \\\n",
       "0  213014.531                CHEST (PA AND LAT)             postero-anterior   \n",
       "1  213014.531                CHEST (PA AND LAT)                      lateral   \n",
       "2  165500.312                CHEST (PA AND LAT)             postero-anterior   \n",
       "3  165500.312                CHEST (PA AND LAT)                      lateral   \n",
       "4   80556.875               CHEST (PORTABLE AP)             antero-posterior   \n",
       "\n",
       "  PatientOrientationCodeSequence_CodeMeaning  \n",
       "0                                      Erect  \n",
       "1                                      Erect  \n",
       "2                                      Erect  \n",
       "3                                      Erect  \n",
       "4                                        NaN  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metadata_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of images: 377110\n",
      "Number of patients: 65379\n"
     ]
    }
   ],
   "source": [
    "print(\"Number of images: \" + str(len(metadata_df)))\n",
    "print(\"Number of patients: \" + str(metadata_df.subject_id.nunique()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>subject_id</th>\n",
       "      <th>hadm_id</th>\n",
       "      <th>admittime</th>\n",
       "      <th>dischtime</th>\n",
       "      <th>deathtime</th>\n",
       "      <th>admission_type</th>\n",
       "      <th>admission_location</th>\n",
       "      <th>discharge_location</th>\n",
       "      <th>insurance</th>\n",
       "      <th>language</th>\n",
       "      <th>marital_status</th>\n",
       "      <th>ethnicity</th>\n",
       "      <th>edregtime</th>\n",
       "      <th>edouttime</th>\n",
       "      <th>hospital_expire_flag</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>14679932</td>\n",
       "      <td>21038362</td>\n",
       "      <td>2139-09-26 14:16:00</td>\n",
       "      <td>2139-09-28 11:30:00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ELECTIVE</td>\n",
       "      <td>NaN</td>\n",
       "      <td>HOME</td>\n",
       "      <td>Other</td>\n",
       "      <td>ENGLISH</td>\n",
       "      <td>SINGLE</td>\n",
       "      <td>UNKNOWN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>15585972</td>\n",
       "      <td>24941086</td>\n",
       "      <td>2123-10-07 23:56:00</td>\n",
       "      <td>2123-10-12 11:22:00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ELECTIVE</td>\n",
       "      <td>NaN</td>\n",
       "      <td>HOME</td>\n",
       "      <td>Other</td>\n",
       "      <td>ENGLISH</td>\n",
       "      <td>NaN</td>\n",
       "      <td>WHITE</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>11989120</td>\n",
       "      <td>21965160</td>\n",
       "      <td>2147-01-14 09:00:00</td>\n",
       "      <td>2147-01-17 14:25:00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ELECTIVE</td>\n",
       "      <td>NaN</td>\n",
       "      <td>HOME</td>\n",
       "      <td>Other</td>\n",
       "      <td>ENGLISH</td>\n",
       "      <td>NaN</td>\n",
       "      <td>UNKNOWN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>17817079</td>\n",
       "      <td>24709883</td>\n",
       "      <td>2165-12-27 17:33:00</td>\n",
       "      <td>2165-12-31 21:18:00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ELECTIVE</td>\n",
       "      <td>NaN</td>\n",
       "      <td>HOME</td>\n",
       "      <td>Other</td>\n",
       "      <td>ENGLISH</td>\n",
       "      <td>NaN</td>\n",
       "      <td>OTHER</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>15078341</td>\n",
       "      <td>23272159</td>\n",
       "      <td>2122-08-28 08:48:00</td>\n",
       "      <td>2122-08-30 12:32:00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ELECTIVE</td>\n",
       "      <td>NaN</td>\n",
       "      <td>HOME</td>\n",
       "      <td>Other</td>\n",
       "      <td>ENGLISH</td>\n",
       "      <td>NaN</td>\n",
       "      <td>BLACK/AFRICAN AMERICAN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   subject_id   hadm_id            admittime            dischtime deathtime  \\\n",
       "0    14679932  21038362  2139-09-26 14:16:00  2139-09-28 11:30:00       NaN   \n",
       "1    15585972  24941086  2123-10-07 23:56:00  2123-10-12 11:22:00       NaN   \n",
       "2    11989120  21965160  2147-01-14 09:00:00  2147-01-17 14:25:00       NaN   \n",
       "3    17817079  24709883  2165-12-27 17:33:00  2165-12-31 21:18:00       NaN   \n",
       "4    15078341  23272159  2122-08-28 08:48:00  2122-08-30 12:32:00       NaN   \n",
       "\n",
       "  admission_type admission_location discharge_location insurance language  \\\n",
       "0       ELECTIVE                NaN               HOME     Other  ENGLISH   \n",
       "1       ELECTIVE                NaN               HOME     Other  ENGLISH   \n",
       "2       ELECTIVE                NaN               HOME     Other  ENGLISH   \n",
       "3       ELECTIVE                NaN               HOME     Other  ENGLISH   \n",
       "4       ELECTIVE                NaN               HOME     Other  ENGLISH   \n",
       "\n",
       "  marital_status               ethnicity edregtime edouttime  \\\n",
       "0         SINGLE                 UNKNOWN       NaN       NaN   \n",
       "1            NaN                   WHITE       NaN       NaN   \n",
       "2            NaN                 UNKNOWN       NaN       NaN   \n",
       "3            NaN                   OTHER       NaN       NaN   \n",
       "4            NaN  BLACK/AFRICAN AMERICAN       NaN       NaN   \n",
       "\n",
       "   hospital_expire_flag  \n",
       "0                     0  \n",
       "1                     0  \n",
       "2                     0  \n",
       "3                     0  \n",
       "4                     0  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "admissions_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "WHITE                            337630\n",
       "BLACK/AFRICAN AMERICAN            80293\n",
       "HISPANIC/LATINO                   29823\n",
       "OTHER                             26813\n",
       "ASIAN                             24506\n",
       "UNKNOWN                           19400\n",
       "UNABLE TO OBTAIN                   3740\n",
       "AMERICAN INDIAN/ALASKA NATIVE      1535\n",
       "Name: ethnicity, dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "admissions_df.ethnicity.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "OTHER_WHITE                                                   2489\n",
       "UNKNOWN_WHITE                                                 1131\n",
       "BLACK/AFRICAN AMERICAN_OTHER                                   560\n",
       "UNABLE TO OBTAIN_WHITE                                         308\n",
       "HISPANIC/LATINO_OTHER                                          307\n",
       "HISPANIC/LATINO_WHITE                                          204\n",
       "HISPANIC/LATINO_UNKNOWN                                        173\n",
       "BLACK/AFRICAN AMERICAN_WHITE                                   168\n",
       "BLACK/AFRICAN AMERICAN_UNKNOWN                                 156\n",
       "OTHER_UNKNOWN                                                  130\n",
       "BLACK/AFRICAN AMERICAN_HISPANIC/LATINO                         111\n",
       "ASIAN_OTHER                                                     98\n",
       "UNABLE TO OBTAIN_UNKNOWN                                        49\n",
       "ASIAN_WHITE                                                     46\n",
       "BLACK/AFRICAN AMERICAN_UNABLE TO OBTAIN                         44\n",
       "ASIAN_UNKNOWN                                                   41\n",
       "OTHER_UNABLE TO OBTAIN                                          29\n",
       "AMERICAN INDIAN/ALASKA NATIVE_WHITE                             28\n",
       "OTHER_UNKNOWN_WHITE                                             21\n",
       "HISPANIC/LATINO_UNABLE TO OBTAIN                                21\n",
       "AMERICAN INDIAN/ALASKA NATIVE_OTHER                             19\n",
       "HISPANIC/LATINO_OTHER_WHITE                                     19\n",
       "OTHER_UNABLE TO OBTAIN_WHITE                                    19\n",
       "ASIAN_UNABLE TO OBTAIN                                          18\n",
       "HISPANIC/LATINO_OTHER_UNKNOWN                                   16\n",
       "BLACK/AFRICAN AMERICAN_OTHER_WHITE                              11\n",
       "ASIAN_BLACK/AFRICAN AMERICAN                                    11\n",
       "AMERICAN INDIAN/ALASKA NATIVE_BLACK/AFRICAN AMERICAN            11\n",
       "UNABLE TO OBTAIN_UNKNOWN_WHITE                                   9\n",
       "HISPANIC/LATINO_UNKNOWN_WHITE                                    8\n",
       "AMERICAN INDIAN/ALASKA NATIVE_UNABLE TO OBTAIN                   8\n",
       "BLACK/AFRICAN AMERICAN_HISPANIC/LATINO_OTHER                     8\n",
       "ASIAN_HISPANIC/LATINO                                            7\n",
       "AMERICAN INDIAN/ALASKA NATIVE_HISPANIC/LATINO                    6\n",
       "AMERICAN INDIAN/ALASKA NATIVE_UNKNOWN                            5\n",
       "AMERICAN INDIAN/ALASKA NATIVE_BLACK/AFRICAN AMERICAN_OTHER       4\n",
       "AMERICAN INDIAN/ALASKA NATIVE_ASIAN                              4\n",
       "BLACK/AFRICAN AMERICAN_OTHER_UNKNOWN                             4\n",
       "ASIAN_UNKNOWN_WHITE                                              3\n",
       "HISPANIC/LATINO_UNABLE TO OBTAIN_UNKNOWN                         3\n",
       "ASIAN_UNABLE TO OBTAIN_UNKNOWN                                   3\n",
       "AMERICAN INDIAN/ALASKA NATIVE_OTHER_WHITE                        2\n",
       "BLACK/AFRICAN AMERICAN_UNKNOWN_WHITE                             2\n",
       "BLACK/AFRICAN AMERICAN_UNABLE TO OBTAIN_UNKNOWN                  2\n",
       "BLACK/AFRICAN AMERICAN_HISPANIC/LATINO_WHITE                     2\n",
       "ASIAN_BLACK/AFRICAN AMERICAN_OTHER                               1\n",
       "AMERICAN INDIAN/ALASKA NATIVE_UNKNOWN_WHITE                      1\n",
       "ASIAN_OTHER_UNKNOWN                                              1\n",
       "ASIAN_BLACK/AFRICAN AMERICAN_UNKNOWN                             1\n",
       "OTHER_UNABLE TO OBTAIN_UNKNOWN_WHITE                             1\n",
       "BLACK/AFRICAN AMERICAN_HISPANIC/LATINO_OTHER_WHITE               1\n",
       "ASIAN_OTHER_WHITE                                                1\n",
       "ASIAN_BLACK/AFRICAN AMERICAN_HISPANIC/LATINO_OTHER               1\n",
       "ASIAN_HISPANIC/LATINO_WHITE                                      1\n",
       "BLACK/AFRICAN AMERICAN_OTHER_UNABLE TO OBTAIN                    1\n",
       "ASIAN_HISPANIC/LATINO_OTHER                                      1\n",
       "BLACK/AFRICAN AMERICAN_HISPANIC/LATINO_OTHER_UNKNOWN             1\n",
       "AMERICAN INDIAN/ALASKA NATIVE_ASIAN_OTHER_WHITE                  1\n",
       "BLACK/AFRICAN AMERICAN_OTHER_UNKNOWN_WHITE                       1\n",
       "ASIAN_BLACK/AFRICAN AMERICAN_WHITE                               1\n",
       "Name: ethnicity, dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# remove patients who have inconsistent documented race information\n",
    "# credit to github.com/robintibor\n",
    "ethnicity_df = admissions_df.loc[:,['subject_id', 'ethnicity']].drop_duplicates()\n",
    "\n",
    "v = ethnicity_df.subject_id.value_counts()\n",
    "subject_id_more_than_once = v.index[v.gt(1)]\n",
    "\n",
    "ambiguous_ethnicity_df = ethnicity_df[ethnicity_df.subject_id.isin(subject_id_more_than_once)]\n",
    "inconsistent_race = ambiguous_ethnicity_df.subject_id.unique()\n",
    "\n",
    "grouped = ambiguous_ethnicity_df.groupby('subject_id')\n",
    "grouped.aggregate(lambda x: \"_\".join(sorted(x))).ethnicity.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "merge_df = pd.merge(metadata_df,diagnosis_df,on=['subject_id', 'study_id'])\n",
    "merge_df = pd.merge(merge_df,ethnicity_df,on='subject_id')\n",
    "merge_df = merge_df[~merge_df.subject_id.isin(inconsistent_race)]\n",
    "merge_df = merge_df.rename(columns={\"ethnicity\": \"race\"})\n",
    "merge_df = merge_df[merge_df.race.isin(['ASIAN','BLACK/AFRICAN AMERICAN','WHITE'])]\n",
    "merge_df = merge_df[merge_df.ViewPosition.isin(['AP','PA'])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cxr = pd.merge(merge_df,patients_df, on='subject_id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total images after inclusion/exclusion criteria: 183207\n",
      "Total patients after inclusion/exclusion criteria: 43209\n"
     ]
    }
   ],
   "source": [
    "print(\"Total images after inclusion/exclusion criteria: \" + str(len(df_cxr)))\n",
    "print(\"Total patients after inclusion/exclusion criteria: \" + str(df_cxr.subject_id.nunique()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cxr = df_cxr.rename(columns={'gender': 'sex'})\n",
    "df_cxr = df_cxr.rename(columns={'anchor_age': 'age'})\n",
    "\n",
    "study_year = np.floor(df_cxr['StudyDate'] / 10000)\n",
    "delta_years = study_year - df_cxr['anchor_year']\n",
    "df_cxr['age'] = df_cxr['age'] + delta_years"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "white = 'White'\n",
    "asian = 'Asian'\n",
    "black = 'Black'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['White', 'Black', 'Asian'], dtype=object)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mask = (df_cxr.race.str.contains(\"BLACK\", na=False))\n",
    "df_cxr.loc[mask, \"race\"] = black\n",
    "\n",
    "mask = (df_cxr.race.str.contains(\"WHITE\", na=False))\n",
    "df_cxr.loc[mask, \"race\"] = white\n",
    "\n",
    "mask = (df_cxr.race.str.contains(\"ASIAN\", na=False))\n",
    "df_cxr.loc[mask, \"race\"] = asian\n",
    "\n",
    "df_cxr['race'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cxr = df_cxr[df_cxr.race.isin([asian,black,white])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cxr['race_label'] = df_cxr['race']\n",
    "\n",
    "df_cxr.loc[df_cxr['race_label'] == white, 'race_label'] = 0\n",
    "df_cxr.loc[df_cxr['race_label'] == asian, 'race_label'] = 1\n",
    "df_cxr.loc[df_cxr['race_label'] == black, 'race_label'] = 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cxr.loc[df_cxr['sex'] == 'F', 'sex'] = 'Female'\n",
    "df_cxr.loc[df_cxr['sex'] == 'M', 'sex'] = 'Male'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cxr['sex_label'] = df_cxr['sex']\n",
    "\n",
    "df_cxr.loc[df_cxr['sex_label'] == 'Male', 'sex_label'] = 0\n",
    "df_cxr.loc[df_cxr['sex_label'] == 'Female', 'sex_label'] = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Female', 'Male'], dtype=object)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_cxr['sex'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>dicom_id</th>\n",
       "      <th>subject_id</th>\n",
       "      <th>study_id</th>\n",
       "      <th>PerformedProcedureStepDescription</th>\n",
       "      <th>ViewPosition</th>\n",
       "      <th>Rows</th>\n",
       "      <th>Columns</th>\n",
       "      <th>StudyDate</th>\n",
       "      <th>StudyTime</th>\n",
       "      <th>ProcedureCodeSequence_CodeMeaning</th>\n",
       "      <th>...</th>\n",
       "      <th>Pneumothorax</th>\n",
       "      <th>Support Devices</th>\n",
       "      <th>race</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>anchor_year</th>\n",
       "      <th>anchor_year_group</th>\n",
       "      <th>dod</th>\n",
       "      <th>race_label</th>\n",
       "      <th>sex_label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>02aa804e-bde0afdd-112c0b34-7bc16630-4e384014</td>\n",
       "      <td>10000032</td>\n",
       "      <td>50414267</td>\n",
       "      <td>CHEST (PA AND LAT)</td>\n",
       "      <td>PA</td>\n",
       "      <td>3056</td>\n",
       "      <td>2544</td>\n",
       "      <td>21800506</td>\n",
       "      <td>213014.531</td>\n",
       "      <td>CHEST (PA AND LAT)</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>52.0</td>\n",
       "      <td>2180</td>\n",
       "      <td>2014 - 2016</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2a2277a9-b0ded155-c0de8eb9-c124d10e-82c5caab</td>\n",
       "      <td>10000032</td>\n",
       "      <td>53189527</td>\n",
       "      <td>CHEST (PA AND LAT)</td>\n",
       "      <td>PA</td>\n",
       "      <td>3056</td>\n",
       "      <td>2544</td>\n",
       "      <td>21800626</td>\n",
       "      <td>165500.312</td>\n",
       "      <td>CHEST (PA AND LAT)</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>52.0</td>\n",
       "      <td>2180</td>\n",
       "      <td>2014 - 2016</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>68b5c4b1-227d0485-9cc38c3f-7b84ab51-4b472714</td>\n",
       "      <td>10000032</td>\n",
       "      <td>53911762</td>\n",
       "      <td>CHEST (PORTABLE AP)</td>\n",
       "      <td>AP</td>\n",
       "      <td>2705</td>\n",
       "      <td>2539</td>\n",
       "      <td>21800723</td>\n",
       "      <td>80556.875</td>\n",
       "      <td>CHEST (PORTABLE AP)</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>52.0</td>\n",
       "      <td>2180</td>\n",
       "      <td>2014 - 2016</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>fffabebf-74fd3a1f-673b6b41-96ec0ac9-2ab69818</td>\n",
       "      <td>10000032</td>\n",
       "      <td>53911762</td>\n",
       "      <td>CHEST (PORTABLE AP)</td>\n",
       "      <td>AP</td>\n",
       "      <td>2906</td>\n",
       "      <td>2258</td>\n",
       "      <td>21800723</td>\n",
       "      <td>80556.875</td>\n",
       "      <td>CHEST (PORTABLE AP)</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>52.0</td>\n",
       "      <td>2180</td>\n",
       "      <td>2014 - 2016</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ea030e7a-2e3b1346-bc518786-7a8fd698-f673b44c</td>\n",
       "      <td>10000032</td>\n",
       "      <td>56699142</td>\n",
       "      <td>CHEST (PORTABLE AP)</td>\n",
       "      <td>AP</td>\n",
       "      <td>3056</td>\n",
       "      <td>2544</td>\n",
       "      <td>21800805</td>\n",
       "      <td>234424.765</td>\n",
       "      <td>CHEST (PORTABLE AP)</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>52.0</td>\n",
       "      <td>2180</td>\n",
       "      <td>2014 - 2016</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 34 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                       dicom_id  subject_id  study_id  \\\n",
       "0  02aa804e-bde0afdd-112c0b34-7bc16630-4e384014    10000032  50414267   \n",
       "1  2a2277a9-b0ded155-c0de8eb9-c124d10e-82c5caab    10000032  53189527   \n",
       "2  68b5c4b1-227d0485-9cc38c3f-7b84ab51-4b472714    10000032  53911762   \n",
       "3  fffabebf-74fd3a1f-673b6b41-96ec0ac9-2ab69818    10000032  53911762   \n",
       "4  ea030e7a-2e3b1346-bc518786-7a8fd698-f673b44c    10000032  56699142   \n",
       "\n",
       "  PerformedProcedureStepDescription ViewPosition  Rows  Columns  StudyDate  \\\n",
       "0                CHEST (PA AND LAT)           PA  3056     2544   21800506   \n",
       "1                CHEST (PA AND LAT)           PA  3056     2544   21800626   \n",
       "2               CHEST (PORTABLE AP)           AP  2705     2539   21800723   \n",
       "3               CHEST (PORTABLE AP)           AP  2906     2258   21800723   \n",
       "4               CHEST (PORTABLE AP)           AP  3056     2544   21800805   \n",
       "\n",
       "    StudyTime ProcedureCodeSequence_CodeMeaning  ... Pneumothorax  \\\n",
       "0  213014.531                CHEST (PA AND LAT)  ...          NaN   \n",
       "1  165500.312                CHEST (PA AND LAT)  ...          NaN   \n",
       "2   80556.875               CHEST (PORTABLE AP)  ...          NaN   \n",
       "3   80556.875               CHEST (PORTABLE AP)  ...          NaN   \n",
       "4  234424.765               CHEST (PORTABLE AP)  ...          NaN   \n",
       "\n",
       "  Support Devices   race     sex   age  anchor_year  anchor_year_group  dod  \\\n",
       "0             NaN  White  Female  52.0         2180        2014 - 2016  NaN   \n",
       "1             NaN  White  Female  52.0         2180        2014 - 2016  NaN   \n",
       "2             NaN  White  Female  52.0         2180        2014 - 2016  NaN   \n",
       "3             NaN  White  Female  52.0         2180        2014 - 2016  NaN   \n",
       "4             NaN  White  Female  52.0         2180        2014 - 2016  NaN   \n",
       "\n",
       "   race_label  sex_label  \n",
       "0           0          1  \n",
       "1           0          1  \n",
       "2           0          1  \n",
       "3           0          1  \n",
       "4           0          1  \n",
       "\n",
       "[5 rows x 34 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_cxr.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "labels = [\n",
    "    'No Finding',\n",
    "    'Enlarged Cardiomediastinum',\n",
    "    'Cardiomegaly',\n",
    "    'Lung Opacity',\n",
    "    'Lung Lesion',\n",
    "    'Edema',\n",
    "    'Consolidation',\n",
    "    'Pneumonia',\n",
    "    'Atelectasis',\n",
    "    'Pneumothorax',\n",
    "    'Pleural Effusion',\n",
    "    'Pleural Other',\n",
    "    'Fracture',\n",
    "    'Support Devices']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cxr['disease'] = df_cxr[labels[0]]\n",
    "df_cxr.loc[df_cxr[labels[0]] == 1, 'disease'] = labels[0]\n",
    "df_cxr.loc[df_cxr[labels[10]] == 1, 'disease'] = labels[10]\n",
    "df_cxr.loc[df_cxr['disease'].isna(), 'disease'] = 'Other'\n",
    "\n",
    "df_cxr['disease_label'] = df_cxr['disease']\n",
    "df_cxr.loc[df_cxr['disease_label'] == labels[0], 'disease_label'] = 0\n",
    "df_cxr.loc[df_cxr['disease_label'] == labels[10], 'disease_label'] = 1\n",
    "df_cxr.loc[df_cxr['disease_label'] == 'Other', 'disease_label'] = 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cxr.subject_id = df_cxr.subject_id.astype(str)\n",
    "df_cxr.study_id = df_cxr.study_id.astype(str)\n",
    "df_cxr.insert(2, \"path\", \"\")\n",
    "df_cxr.path = df_cxr.subject_id.str[0:2]\n",
    "df_cxr.path = \"p\" + df_cxr.path\n",
    "df_cxr.path = df_cxr.path + \"/p\" + df_cxr.subject_id + \"/s\" + df_cxr.study_id + \"/\" + df_cxr.dicom_id + \".jpg\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>dicom_id</th>\n",
       "      <th>subject_id</th>\n",
       "      <th>path</th>\n",
       "      <th>study_id</th>\n",
       "      <th>PerformedProcedureStepDescription</th>\n",
       "      <th>ViewPosition</th>\n",
       "      <th>Rows</th>\n",
       "      <th>Columns</th>\n",
       "      <th>StudyDate</th>\n",
       "      <th>StudyTime</th>\n",
       "      <th>...</th>\n",
       "      <th>race</th>\n",
       "      <th>sex</th>\n",
       "      <th>age</th>\n",
       "      <th>anchor_year</th>\n",
       "      <th>anchor_year_group</th>\n",
       "      <th>dod</th>\n",
       "      <th>race_label</th>\n",
       "      <th>sex_label</th>\n",
       "      <th>disease</th>\n",
       "      <th>disease_label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>02aa804e-bde0afdd-112c0b34-7bc16630-4e384014</td>\n",
       "      <td>10000032</td>\n",
       "      <td>p10/p10000032/s50414267/02aa804e-bde0afdd-112c...</td>\n",
       "      <td>50414267</td>\n",
       "      <td>CHEST (PA AND LAT)</td>\n",
       "      <td>PA</td>\n",
       "      <td>3056</td>\n",
       "      <td>2544</td>\n",
       "      <td>21800506</td>\n",
       "      <td>213014.531</td>\n",
       "      <td>...</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>52.0</td>\n",
       "      <td>2180</td>\n",
       "      <td>2014 - 2016</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>No Finding</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2a2277a9-b0ded155-c0de8eb9-c124d10e-82c5caab</td>\n",
       "      <td>10000032</td>\n",
       "      <td>p10/p10000032/s53189527/2a2277a9-b0ded155-c0de...</td>\n",
       "      <td>53189527</td>\n",
       "      <td>CHEST (PA AND LAT)</td>\n",
       "      <td>PA</td>\n",
       "      <td>3056</td>\n",
       "      <td>2544</td>\n",
       "      <td>21800626</td>\n",
       "      <td>165500.312</td>\n",
       "      <td>...</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>52.0</td>\n",
       "      <td>2180</td>\n",
       "      <td>2014 - 2016</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>No Finding</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>68b5c4b1-227d0485-9cc38c3f-7b84ab51-4b472714</td>\n",
       "      <td>10000032</td>\n",
       "      <td>p10/p10000032/s53911762/68b5c4b1-227d0485-9cc3...</td>\n",
       "      <td>53911762</td>\n",
       "      <td>CHEST (PORTABLE AP)</td>\n",
       "      <td>AP</td>\n",
       "      <td>2705</td>\n",
       "      <td>2539</td>\n",
       "      <td>21800723</td>\n",
       "      <td>80556.875</td>\n",
       "      <td>...</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>52.0</td>\n",
       "      <td>2180</td>\n",
       "      <td>2014 - 2016</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>No Finding</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>fffabebf-74fd3a1f-673b6b41-96ec0ac9-2ab69818</td>\n",
       "      <td>10000032</td>\n",
       "      <td>p10/p10000032/s53911762/fffabebf-74fd3a1f-673b...</td>\n",
       "      <td>53911762</td>\n",
       "      <td>CHEST (PORTABLE AP)</td>\n",
       "      <td>AP</td>\n",
       "      <td>2906</td>\n",
       "      <td>2258</td>\n",
       "      <td>21800723</td>\n",
       "      <td>80556.875</td>\n",
       "      <td>...</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>52.0</td>\n",
       "      <td>2180</td>\n",
       "      <td>2014 - 2016</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>No Finding</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ea030e7a-2e3b1346-bc518786-7a8fd698-f673b44c</td>\n",
       "      <td>10000032</td>\n",
       "      <td>p10/p10000032/s56699142/ea030e7a-2e3b1346-bc51...</td>\n",
       "      <td>56699142</td>\n",
       "      <td>CHEST (PORTABLE AP)</td>\n",
       "      <td>AP</td>\n",
       "      <td>3056</td>\n",
       "      <td>2544</td>\n",
       "      <td>21800805</td>\n",
       "      <td>234424.765</td>\n",
       "      <td>...</td>\n",
       "      <td>White</td>\n",
       "      <td>Female</td>\n",
       "      <td>52.0</td>\n",
       "      <td>2180</td>\n",
       "      <td>2014 - 2016</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>No Finding</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 37 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                       dicom_id subject_id  \\\n",
       "0  02aa804e-bde0afdd-112c0b34-7bc16630-4e384014   10000032   \n",
       "1  2a2277a9-b0ded155-c0de8eb9-c124d10e-82c5caab   10000032   \n",
       "2  68b5c4b1-227d0485-9cc38c3f-7b84ab51-4b472714   10000032   \n",
       "3  fffabebf-74fd3a1f-673b6b41-96ec0ac9-2ab69818   10000032   \n",
       "4  ea030e7a-2e3b1346-bc518786-7a8fd698-f673b44c   10000032   \n",
       "\n",
       "                                                path  study_id  \\\n",
       "0  p10/p10000032/s50414267/02aa804e-bde0afdd-112c...  50414267   \n",
       "1  p10/p10000032/s53189527/2a2277a9-b0ded155-c0de...  53189527   \n",
       "2  p10/p10000032/s53911762/68b5c4b1-227d0485-9cc3...  53911762   \n",
       "3  p10/p10000032/s53911762/fffabebf-74fd3a1f-673b...  53911762   \n",
       "4  p10/p10000032/s56699142/ea030e7a-2e3b1346-bc51...  56699142   \n",
       "\n",
       "  PerformedProcedureStepDescription ViewPosition  Rows  Columns  StudyDate  \\\n",
       "0                CHEST (PA AND LAT)           PA  3056     2544   21800506   \n",
       "1                CHEST (PA AND LAT)           PA  3056     2544   21800626   \n",
       "2               CHEST (PORTABLE AP)           AP  2705     2539   21800723   \n",
       "3               CHEST (PORTABLE AP)           AP  2906     2258   21800723   \n",
       "4               CHEST (PORTABLE AP)           AP  3056     2544   21800805   \n",
       "\n",
       "    StudyTime  ...   race     sex   age  anchor_year  anchor_year_group  dod  \\\n",
       "0  213014.531  ...  White  Female  52.0         2180        2014 - 2016  NaN   \n",
       "1  165500.312  ...  White  Female  52.0         2180        2014 - 2016  NaN   \n",
       "2   80556.875  ...  White  Female  52.0         2180        2014 - 2016  NaN   \n",
       "3   80556.875  ...  White  Female  52.0         2180        2014 - 2016  NaN   \n",
       "4  234424.765  ...  White  Female  52.0         2180        2014 - 2016  NaN   \n",
       "\n",
       "   race_label  sex_label     disease  disease_label  \n",
       "0           0          1  No Finding              0  \n",
       "1           0          1  No Finding              0  \n",
       "2           0          1  No Finding              0  \n",
       "3           0          1  No Finding              0  \n",
       "4           0          1  No Finding              0  \n",
       "\n",
       "[5 rows x 37 columns]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_cxr.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pre-processing of imaging dataset (this may take a few hours, needed only once)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "img_data_dir = '<path_to_imaging_data>'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████| 183207/183207 [29:48<00:00, 102.46it/s]\n"
     ]
    }
   ],
   "source": [
    "from skimage.io import imread\n",
    "from skimage.io import imsave\n",
    "from skimage.transform import resize\n",
    "\n",
    "df_cxr['path_preproc'] = df_cxr['path']\n",
    "\n",
    "preproc_dir = 'preproc_224x224/'\n",
    "out_dir = img_data_dir\n",
    "\n",
    "if not os.path.exists(out_dir + preproc_dir):\n",
    "    os.makedirs(out_dir + preproc_dir)\n",
    "\n",
    "for idx, p in enumerate(tqdm(df_cxr['path'])):\n",
    "\n",
    "    split =  p.split(\"/\")\n",
    "    preproc_filename = split[2] + '_'+ split[3]\n",
    "    df_cxr.loc[idx, 'path_preproc'] = preproc_dir + preproc_filename\n",
    "    out_path = out_dir + preproc_dir + preproc_filename\n",
    "    \n",
    "    if not os.path.exists(out_path):\n",
    "        image = imread(img_data_dir + p)\n",
    "        image = resize(image, output_shape=(224, 224), preserve_range=True)\n",
    "        imsave(out_path, image.astype(np.uint8))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random splitting into train, validation, test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cxr.insert(5, \"split\",\"none\", True)\n",
    "unique_sub_id = df_cxr.subject_id.unique()\n",
    "\n",
    "train_percent, valid_percent, test_percent = 0.60, 0.10, 0.30\n",
    "\n",
    "unique_sub_id = shuffle(unique_sub_id)\n",
    "value1 = (round(len(unique_sub_id)*train_percent))\n",
    "value2 = (round(len(unique_sub_id)*valid_percent))\n",
    "value3 = value1 + value2\n",
    "value4 = (round(len(unique_sub_id)*test_percent))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Patients in training set: 25925\n"
     ]
    }
   ],
   "source": [
    "print(\"Patients in training set: \" + str(value1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Patients in validation set: 4321\n"
     ]
    }
   ],
   "source": [
    "print(\"Patients in validation set: \" + str(value2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Patients in testing set: 12963\n"
     ]
    }
   ],
   "source": [
    "print(\"Patients in testing set: \" + str(value4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cxr = shuffle(df_cxr)\n",
    "\n",
    "train_sub_id = unique_sub_id[:value1]\n",
    "validate_sub_id = unique_sub_id[value1:value3]\n",
    "test_sub_id = unique_sub_id[value3:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cxr.loc[df_cxr.subject_id.isin(train_sub_id), \"split\"]=\"train\"\n",
    "df_cxr.loc[df_cxr.subject_id.isin(validate_sub_id), \"split\"]=\"validate\"\n",
    "df_cxr.loc[df_cxr.subject_id.isin(test_sub_id), \"split\"]=\"test\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "train       0.601942\n",
       "test        0.301637\n",
       "validate    0.096421\n",
       "Name: split, dtype: float64"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_cxr.split.value_counts(normalize=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "train       110280\n",
       "test         55262\n",
       "validate     17665\n",
       "Name: split, dtype: int64"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_cxr.split.value_counts(normalize=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save full sample"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cxr.to_csv(data_dir + 'mimic.sample.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Creating subsets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cxr = pd.read_csv(data_dir + 'mimic.sample.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train = df_cxr[df_cxr.split==\"train\"]\n",
    "df_val = df_cxr[df_cxr.split==\"validate\"]\n",
    "df_test = df_cxr[df_cxr.split==\"test\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train.to_csv(data_dir + 'mimic.sample.train.csv')\n",
    "df_val.to_csv(data_dir + 'mimic.sample.val.csv')\n",
    "df_test.to_csv(data_dir + 'mimic.sample.test.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train_white = df_train[df_train['race'] == white]\n",
    "df_val_white = df_val[df_val['race'] == white]\n",
    "df_train_white.to_csv(data_dir + 'mimic.sample.train.white.csv')\n",
    "df_val_white.to_csv(data_dir + 'mimic.sample.val.white.csv')\n",
    "\n",
    "df_cxr_white = df_cxr[df_cxr['race'] == white]\n",
    "df_cxr_asian = df_cxr[df_cxr['race'] == asian]\n",
    "df_cxr_black = df_cxr[df_cxr['race'] == black]\n",
    "\n",
    "df_test_white = df_test[df_test['race'] == white]\n",
    "df_test_asian = df_test[df_test['race'] == asian]\n",
    "df_test_black = df_test[df_test['race'] == black]\n",
    "\n",
    "df_train_asian = df_train[df_train['race'] == asian]\n",
    "df_train_black = df_train[df_train['race'] == black]\n",
    "\n",
    "df_val_asian = df_val[df_val['race'] == asian]\n",
    "df_val_black = df_val[df_val['race'] == black]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train_male = df_train[df_train['sex'] == 'Male']\n",
    "df_val_male = df_val[df_val['sex'] == 'Male']\n",
    "df_train_male.to_csv(data_dir + 'mimic.sample.train.male.csv')\n",
    "df_val_male.to_csv(data_dir + 'mimic.sample.val.male.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWgAAAFqCAYAAAA3LcDeAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAo60lEQVR4nO3deXiV1bn38e8dhjDJ4ERQmWpwHiGgVBDQgn2r9G3fovWgWNsetUeFgla0aBWn1iqtisopIFq10nqUTuAAismxImUQUBHQoCICgoigDGHM/f7x7MSdkIQN2cPKzu9zXftK9jPtO2Tzy9rrWc96zN0REZHw5GS6ABERqZoCWkQkUApoEZFAKaBFRAKlgBYRCZQCWkQkUApoEZFApTWgzexsM/unma02MzezyyutNzMbbWZrzKzEzIrM7MRK27Qxs6fM7MvY4ykza53On0NEJB3S3YJuASwGfg6UVLF+JHA9MBToDnwGvGxmB8VtMxnoCnw79ugKPJXCmkVEMsIydSWhmW0BrnX3P8aeG7AGeNjd744ta0oU0r9w9/FmdjywBOjl7rNi2/QC/gUc5+7vpf8nERFJjZD6oDsDecCMsgXuXgK8BnwztqgnsAV4I26/WcDWuG1ERLJCw0wXECcv9nVdpeXrgCPjtlnvcc1+d3cz+yxu/wrM7ErgSoATTjih27vvvpvUokUk/caOHcvy5ctr3Ka4uBiALl267PN4+fn5DBs2LCm1VcMOZKeQAjol3H0CMAGgoKBAM0OJZIFEwrRsm7Fjx6a6nJQJKaDXxr62BVbGLW8bt24tcJiZWVkrOtZ3fXjcNhmRyF/0VatWAXDUUUfVuF0a/ppLQBJ574DeP/VRSH3QHxGFbP+yBWbWBOjN133Os4lGgvSM268n0JyK/dJBKikpoaSkqsErIvum90/9k9YWtJm1APJjT3OADmZ2GvCFu680sweAUWa2DHgfuIXopOBkAHdfamYvAeNjfcsA44FpmR7BUV8+cknyJdra1fun/kl3F0cBUBj3/PbY4wngcuBeoCnwCNAGmAMMcPfNcfsMBh4Cpsee/xO4NqVVS9LpY3349DvKvLQGtLsXUcPZzFi/8ujYo7ptNgKXJrk0CZQ+0odPv6PUCekkodQj+lgfPv2OMi+kk4QiIhJHLegsl8zhf6B+RJF0UkCL+hBFAqWAznIa/idSd6kPWkQkUApoEZFAKaBFRAKlgBYRCZQCWkQkUApoEZFAKaBFRAKlgBYRCZQCWkQkUApoEZFAKaBFRAKlgBYRCZQCWkQkUApoEZFAKaBFRAKlgBYRCZQCWkQkUApoEZFAKaBFRAKlgBYRCZQCWkQkUApoEZFAKaBFRAKlgBYRCZQCWkQkUApoEZFAKaBFRAKlgBYRCZQCWkQkUApoEZFAKaBFRAKlgBYRCZQCWkQkUApoEZFAKaBFRAKlgBYRCZQCWkQkUApoEZFAKaBFRAKlgBYRCZQCWkQkUApoEZFAKaBFRAKlgBYRCZQCWkQkUEEFtJk1MLM7zewjM9se+3qXmTWM28bMbLSZrTGzEjMrMrMTM1m3iEgqBBXQwI3ANcAw4Djg57Hnv4zbZiRwPTAU6A58BrxsZgelt1QRkdRquO9N0uqbwFR3nxp7vsLM/gmcAVHrGRgO3OPuU2LLfkQU0oOB8WmvWEQkRUIL6NeBq83sOHdfZmYnAOcAv4mt7wzkATPKdnD3EjN7jSjcFdASlLFjx7J8+fKkHKu4uBiAYcOG1fpY+fn5STmOpFZoAf1b4CBgiZntIarvbncfF1ufF/u6rtJ+64AjqzqgmV0JXAnQoUOHpBcsUpPly5ez8J0llDY7uNbHsp0OwJsfrK3VcXK2fVHrWiQ9QgvoHwKXEXVXvAucBjxoZh+5+6QDOaC7TwAmABQUFHiS6hRJWGmzg9l+wgWZLqNckyXTMl2CJCi0gL4PGOPuf4k9f8fMOhKdJJwElDUd2gIr4/ZrG7dORCQrhDaKoxmwp9KyPXxd50dEQdy/bKWZNQF6A2+ko0ARkXQJrQU9FbjJzD4i6uI4HbgOeBLA3d3MHgBGmdky4H3gFmALMDkjFYuIpEhoAT0UuBMYBxwOfApMBO6I2+ZeoCnwCNAGmAMMcPfN6S0185I1QiCZowNAIwTqglWrViXtd6T3T+oEFdCxkB0ee1S3jQOjY496bfny5by/eAEdWlTuFdo/jXdFPUjbV8yrdU0rtzSo9TEk9UpKSpLy3gG9f1IpqICW/dehxR5uKdiS6TLK3TW/RaZLkASF9t4BvX8qU0DvQ6gXGkD0MfXQpBxJREKkgN6HEC80gOhigxZNGkGjWh9KRAKlgE5AaBcaQOxig9J6d15UpF4JbRy0iIjEKKBFRAKlgBYRCZT6oEUkGOu25bCzuDgpI52y4QIaBbSIBGP7HmObb2Hh6oW1P1hp9CUpx9pU+0McCAW0iISlNZT2Lc10FRXkFGWmN1h90CIigVJAi4gESgEtIhIoBbSISKAU0CIigVJAi4gESgEtIhIoBbSISKAU0CIigdKVhJJ0od6QVDcjlbpGAS1JV1JSwsJ3F0LrJBwsWfMpbKp1JSJpp4CW1Ggd1nwKmZpLQaQ29K4VEQmUAlpEJFAKaBGRQCmgRUQCpYAWEQmUAlpEJFAaZieSQqtWrSJn25c0WTIt06WUy9m2gR3m0CjTlci+qAUtIhIotaBFUuioo45i3Y6GbD/hgkyXUq7Jkmnklm4GSjJdiuyDWtAiIoFSC3ofQuxDBPUjitQHakGLiARKLeh9CLEPEdSPKFIfqAUtIhIoBbSISKAU0CIigVIfdB22Y8cOPt7egLvmt8h0KeU+3tyAXbYDmmW6EqlJiO8dgB17DLZkuopwqAUtIhIotaDrsNzcXNo3KuGWgnCaHHfNb8Enu3LZxa5MlyI1CPG9A3BFYSu2tbBMlxEMtaBFRAKlgBYRCZQCWkQkUOqDFpFg7Cw12AQ5RYG1HTfBKl+V9pcN7F9BRETK7LMFbWbHA/8B9AE6AU2B9cAC4EVgirvvSGGNIlJPNM5xdrc2SvuWZrqUCnKKcjjqyKPS/7rVrTCzrmb2CrAQOAt4AxgDjAKeABy4G1hjZjeaWW4a6hURqTdqakH/DbgXuNDdN1a3kZn1BEYAvyAKbBERSYKaArqLu+/c1wHcfTYw28waJ68sERGptotjX+FsZhXu5ZFImCfCzNqZ2RNmtt7MtpvZEjPrE7fezGy0ma0xsxIzKzKzE5Px2iIiIUloFIeZDTOzH8Q9nwSUmNl7ZnZssooxs9bALMCA84HjgaHAZ3GbjQSujy3vHlv3spkdlKw6RERCkOgwu2FEIzcws7OBi4DBwCLgd0msZyTwqbtf5u5z3f0jd5/p7ktjr23AcOAed5/i7ouBHwEHxeoREckaiV6ociTwUez7gcCz7v4/ZvYO8K8k1vM94CUzewboB6wBHgUecXcHOgN5wIyyHdy9xMxeA74JjE9iLXKAduzYATsDu9hgU2YuNBCpjUT/B30FHB77vj8wM/b9LqBJEuv5BnA18CFwHvAgcA9wTWx9Xuzrukr7rYtbV4GZXWlm881s/vr165NYqohIaiXagp4BTDSzBUA+0QUqACfydcs6GXKA+e7+y9jzhWbWhSigHz6QA7r7BGACQEFBgSelSqlRbm4uu5rtCupig0xdaCBSG4m2oK8BXgcOAwa5+xex5V2BPyexnk+BJZWWLQU6xL5fG/vattI2bePWiYhkhUQu9W4IXEp0Ym5N/Dp3vy3J9cwCKo8KOQb4OPb9R0RB3B+YF6uvCdAbuCHJtYiIZNQ+W9Duvhu4D2i0r22T4H7gTDO72czyzexCohEkj8RqceAB4EYz+39mdhLwR6K7mE1OQ30iImmTaB/0v4FufN2STQl3n2dm3wN+DfwKWBn7Oi5us3uJJmx6BGgDzAEGuPvmVNYmIpJuiQb0RGCMmXUA3gS2xq909wXJKsjdnweer2G9A6NjDxGRrJVoQJd1H/y+inUONEhOOSIiUibRgO6c0ipERGQvCQW0u6e071lERPaW8D0JY8PtehCNSa4wtai7P5nkukRE6r2EAtrMjgOmEnV1GLAntu8uYAeQ1QGds+0LmiyZVuvj2PavAPAmLWt9rJxtX0CTdIx8lNoK7f2j907dkWgL+gGi0RunEV0ochrQCvhv4JYU1BWM/Pz8pB2ruDgaCdjl6CqnDdlPeaxatQp2b0rCsSRVwnz/6L1TVyQa0N2BPu6+1cxKgYbuvsDMRgIPAaekrMIMGzZsWNKPNXbs2KQdb/uKT5NyLEmNUN8/eu/UDYnOxWHAttj364mmHwVYRTR5koiIJFmiLejFwKlE04DOJbrUeg9wBbA8RbWJiNRriQb03UDz2Pe3EF3pVwh8TnR3FRERSbJEx0FPj/v+Q+B4MzsY2Bi79FpERJJsv+5JZGaHmtkZZpbr7l8onEVEUifRu3ofZGb/Q3QH7TeInSQ0sz+Y2ejUlSciUn8l2gf9W6JQ7kp0Z5Uy04j6p0cntyxJ1MotDbhrfotaHWPdtujvdNtmtb9F1cotDbDcWh9GREg8oL8LfN/dF5lZfLfGUqIbvUoGJOsiiJ3FxQA06dSl1sc6Bli1ahVb2FLrY4nUd4kGdBtgQxXLDyK67FsyIFkXQaTiApr1q3UHdZHaSvQk4TyiVnSZslb0VUR90iIikmSJtqBHAdPN7MTYPtfFvu8BnJ2q4kRE6rNEx0G/YWbfBH4BfACcCywAerr7OymsT0RSJBknmCG5J5l37LFaHyObJDwfdCyIf5TCWkQkTZI5y14yTzI3LS7WCeY4ic4HvQh4FPiTu29KZUEiknohz7K3cPXCWh8nWyR6kvB5YCTwqZn92czOTWFNIiJC4n3QN5vZLcC3gR8Dz5vZp8DjwB/dfWUKaxSR+mQT5BTt1ywUVSvrKal9Nzts4utJltNof/qgHXgReDE2UdJVwG3ArWY2E7jf3V9KTZkiUh8k9w40Ud94lyNr3zfOkcmtLVEJB3QZMzsT+AnwQ2ANUSu6HfCcmT3q7sOTWqGI1Buh9o1nSqInCQ8HLiPq3jga+CcwyN1fjtvmKeBlYHjyy5Q6Z1NgH1M3kZGPqCK1kWgLehXRnVMmAU+4++dVbPMu0RWHUs8F+TE1Qx9RRWoj0YA+193/VdMG7v4V0K/2JUldp4+pIsmR0GfQfYWziIgkX7UBbWavmFmvfR3AzFqb2c1mNjS5pYmI1G81dXE8BfzZzLYRTcw/n2jUxnai6UdPAHoRjY3+O9GFLCIikiTVBrS7P2Fmk4ELgf8gGlrXqmw1sASYDpzu7u+lulARkfqmxpOE7r4LmBx7YGatgKbAhtg6ERFJkf26UMXdvwS+TFEtIiISJwlXEoiISCoooEVEAqWAFhEJlAJaRCRQ+xXQZlZgZj80s+ax583NbL9nxBMRkX1LdDa7tsA/iO7i7UAX4EPg90QXrvw8VQWKiNRXibag7wfWAYcA2+KWPwsMSHZRIiKyH7PZEc1ot9Gswm3RPwA6JL0qERFJuAXdFNhZxfLDiLo4REQkyRIN6NeAy+Oeu5k1AG4EZia7KBERSbyLYyTwv2bWHcgFfgecSDR50lkpqk1EpF5LdML+JcDJwBvADKAJ0QnC0939g9SVJyJSfyU8htnd1wK3pbAWERGJk+g46LOrWeVEJwk/cPcvklaViIgk3IIuIgpjgLJxdvHPS83sn8AQd9+avPJEROqvREdxnA8sBS4F8mOPS4F3gR/EHqcB9ySzODP7pZm5mT0ct8zMbLSZrTGzEjMrMrMTk/m6IiIhSLQFfRfwc3ePH1L3oZmtB37r7t3MbA/wEJCUm8ea2ZnAlcDblVaNBK4nGvb3HnAr8LKZHevum5Px2iIiIUi0BX0CsLqK5atj6wDeAfKSUVTs1lpPE90HcWPccgOGA/e4+xR3Xwz8CDgIGJyM1xYRCUWiAb0EuNnMcssWxL4fFVsH0B5Ym6S6JgDPuXthpeWdif4IzChb4O4lRBfSfDNJry0iEoREuziuBqYCq81scWzZSUApcEHs+TeAcbUtyMyu4Os+7srKWujrKi1fBxxZzfGuJOoqoUMHTRsiInVHQgHt7nPMrDNRaB4bWzwZmOzuW2LbPFnbYszsWODXQK9k3TXc3ScQtcgpKCjwfWwuIhKM/blQZSswvvJyM/uWu7+SpHp6AocC78bNmtcAONvMfkZ0eTlAW2Bl3H5tSV73iohIEA7olldmdqSZ3WJmHwLTk1jP34kuKT8t7jEf+Evs+/eJgrh/XC1NgN5El6GLiGSNhFvQsdnr/i/wn0QB+TbwB6I5OZLC3TcBmyq97lbgi9iIDczsAWCUmS0jCuxbgC1EXS4iIlljnwEd6xf+T+AyYCtREPYnumpwSU37psi9RPNTPwK0AeYAAzQGWkSyTY0BbWb/IhqtMQW4yN3/N7b8xjTUBoC796303IHRsYeISNbaVwu6J1FLdYK7v5uGekREJGZfJwm7E4X462a20MxGmFlSrhYUEZGa1RjQ7r7Q3a8B2gG/B74LfBLb73wza5P6EkVE6qdE76iy3d2fcvd+wPHAfcAIYK2ZvZjKAkVE6qv9Hgft7svd/SaiuTcuouq7fYuISC0lPA66MnffA/wj9hARkSQ7oCsJRUQk9RTQIiKBUkCLiARKAS0iEigFtIhIoBTQIiKBUkCLiARKAS0iEigFtIhIoBTQIiKBUkCLiARKAS0iEigFtIhIoBTQIiKBUkCLiARKAS0iEigFtIhIoBTQIiKBUkCLiARKAS0iEigFtIhIoBTQIiKBUkCLiARKAS0iEigFtIhIoBTQIiKBUkCLiARKAS0iEigFtIhIoBTQIiKBUkCLiARKAS0iEqiGmS4gW4wdO5bly5fXuM2iRYsAOPvssznttNOq3S4/P59hw4YlsToRqYvUghYRCZRa0Emyrxbv2WefXeH5okWLeO2111JZkojUcWpBi4gESgEtIhIoBbSISKAU0CIigVJAi4gESgEtIhIoBbSISKAU0CIigVJAi4gEKqiANrNfmtk8M/vKzNab2VQzO6nSNmZmo81sjZmVmFmRmZ2YqZpFRFIlqIAG+gLjgG8C5wC7gVfM7OC4bUYC1wNDge7AZ8DLZnZQeksVEUmtoObicPfz4p+b2RDgS+AsYKqZGTAcuMfdp8S2+RFRSA8Gxqe1YBGRFAoqoKtwEFErf2PseWcgD5hRtoG7l5jZa0StbgW01DmJTFULmq62Pgqti6OyB4FFwOzY87zY13WVtlsXt64CM7vSzOab2fz169enpEgRkVQItgVtZr8HegG93H3PgR7H3ScAEwAKCgo8SeWJJE0irV1NV1s/BdmCNrP7gf8AznH3D+NWrY19bVtpl7Zx60REskJwAW1mD/J1OC+rtPojoiDuH7d9E6A38EbaihQRSYOgujjM7BFgCPA9YKOZlfUrb3H3Le7uZvYAMMrMlgHvA7cAW4DJGShZRCRlggpo4OrY15mVlt8OjI59fy/QFHgEaAPMAQa4++Z0FFjXJDJCoLi4GEisLzRZIwQSHbmQaG0auZB8+h1lXlAB7e6WwDZOFNajU11PfdG0adNMl1CtkGuTiH5HqRNUQEvyhdpiCbUu+Zp+R5kX3ElCERGJKKDTpPKYVY1hlf2h90/9pIAWEQmURefc6oeCggKfP39+pssQkVran9FJXbp02efx0jDCZJ8DIKqik4QikpWyYXSJWtAiIql3QC1o9UGLiARKAS0iEigFtIhIoBTQIiKBUkCLiARKAS0iEigFtIhIoBTQIiKBUkCLiARKAS0iEigFtIhIoBTQIiKBUkCLiARKAS0iEigFtIhIoLJqPug333yzcU5Ozn81aNDgx+7eikpzsG7YsKFju3btMlSdiNRXO3fu/LjSIjezL/fs2fN4aWnpf3fr1m1nVftl1R1VGjZsOLFly5ZnHXHEEVsbN268waziHNlLlizpePzxx2eoOhGpxz6Pf+Lu7Ny5s9GaNWuGfvXVV12BH1W1U7Z1cfTq2LHjl7m5ubsqh7OISCjMjNzc3F0dO3b8EuhV3XbZFtANcnJysqfPRkSyWiyvGlS7Po21iIjIflBA11GdOnVizJgxtd5G0u/yyy/nggsuyHQZQtTV8NxzzyXteMn+P5dVJwmr89P/Gnrsho1fNnZ3GjVqlPLXa3vowTzyQGK/pPHjx3PdddexceNGGjduDMDOnTtp3bo13/jGN1i8eHH5tsuXL6dLly688sorCR173rx5NG/evPy5mfHss88yaNCg/fhp0u+XI67hyw1r0/Z6rQ7J4zf3P7Lf+y1YsIDu3btz5plnMmvWrIT3e/DBB6nro6euvf5a1m1Yl7bXa3tIWx7+3cP7tc/ll1/OE088Uf78kEMO4cwzz2TMmDEcd9xxyS4xJepFQG/Y+GXjT9p/q3HaXvDTooQ37devH9u2bWPu3Ln06hWdK5gzZw6tWrWiuLiY9evXc9hhhwFQWFhIbm4uZ511VkLHLtuvrvlyw1puyn8/ba93z/ID2+/RRx/l6quv5sknn2Tp0qUkOkKoVatWB/aCAVm3YR1ruq1J3wu+eWC7fetb3+Kpp54CYM2aNdxwww18//vfZ+nSpUksLnXUxZFhxxxzDEcccQSFhYXlywoLCzn33HMpKCigqKiowvKePXvSpEkTALZv385VV11Fy5YtOeqoo7jvvvsqHDv+41anTp0AuPDCCzGz8ucAU6dOpVu3bjRp0oTOnTtz8803s3NnlcMyJaakpITJkydz5ZVXMmjQICZNmlRh/R133EHHjh3Jzc0lLy+Pyy67rHxd5S6Ol156id69e9OmTRsOPvhgzjvvvAoBsmLFCsyMKVOm0L9/f5o1a8YJJ5zAyy+/nPoftI4r+/fPy8uja9eujBgxgmXLllFSUlLl9jfddBPHHnssTZs2pVOnTowcOZLt27dX2OaFF17gjDPOoGnTphxyyCEMHDhwr23K/OlPf6Jly5Y8/fTTB/RXWQEdgH79+u0V0H379qVv374VlhcVFdGvX7/y5/fffz8nn3wyCxYs4MYbb2TkyJHMnj27yteYN28eABMnTuTTTz8tfz59+nQuueQSrr32Wt59910ee+wxnnvuOUaNGpWKHzVrPPfcc3Ts2JGTTz6ZIUOG8OSTT7Jr1y4ApkyZwpgxYxg3bhzFxcVMmzaNHj16VHusrVu3Mnz4cObOnUtRURGtWrVi4MCBe/2RvPnmmxk2bBhvvfUW3bt35+KLL2bLli0p/TmzyebNm3nmmWc4+eSTadq0aZXbNG/enMcee4ylS5cybtw4/vKXv3D33XeXr3/ppZf47ne/S//+/XnzzTcpLCykT58+lJaW7nWsBx98kKFDhzJt2jQuueSSLw+k5nrRxRG6fv36ce2117Jjxw7cndmzZzNx4kQ6dOjAz3/+cwCWLVvGp59+yjnnnFO+34ABA7j22msBGDp0KGPHjmXmzJn07Nlzr9co6+5o3bo1eXl55cvvvvtubrjhBn784x8DcPTRR/Pb3/6WSy+9lPvuuw+NJ6/apEmTGDJkCAB9+vShWbNm/OMf/2DQoEF8/PHHtGvXjgEDBtCoUSM6dOhAQUFBtcf6wQ9+UOH5448/TsuWLSt0ewGMGDGCgQMHAvDrX/+aJ598kkWLFlXYRip66aWXaNGiBRD9IWzfvj0vvPBCtdv/6le/Kv++U6dOjBo1ijFjxnDnnXcCcOeddzJo0CDuuuuu8u1OOeWUKo8zYcIEXn31VU4//fQDrl8t6ACcc845bN++ndmzZzN79mwOO+ww8vPzOeuss/jggw9Yu3YthYWFNGvWjDPOOKN8v8pvjCOOOILPPvtsv177zTff5O6776ZFixblj8GDB7N161bWrk3fibq6ZPny5bz++usMHjwYiE6+XnLJJeXdHBdeeCHbt2+nc+fO/PSnP+XZZ59lx44d1R7vgw8+YPDgwRx99NG0bNmStm3bUlpaysqVKytsF//7PuKIIwD2+/dd35x99tksWrSIRYsWMXfuXM4991wGDBjAJ598UuX2zz33HL169SIvL48WLVowYsSICr+HhQsXcu6559b4mg8++CAPPfQQr7/+eq3CGRTQQejcuTMdO3akqKiIoqIi+vTpA0Qft7p161a+vFevXhVGoVQekWJmVX7UqklpaSm33XZb+Zt40aJFvP322xQXF9fZk4yp9uijj7Jnzx46dOhAw4YNadiwIffccw8zZszgk08+oX379rz33nuMHz+eli1bcv3119OtWze2bt1a5fEuuOAC1q9fz/jx45kzZw4LFy6kYcOGe3VxxP++yz7Z7O/vu75p1qwZ+fn55Ofn0717dx599FG++uorJkyYsNe2//73v7n44os577zzmDp1KgsXLuSuu+4q77pKVK9evTAz/vznP9e6fnVxBKKsH9rdK5xQ6tu3L6+++ipFRUVcd911tXqNRo0asWfPngrLunbtyrJly8jPz6/VseuL3bt388QTT/Cb3/xmr7HMQ4YM4fHHH+fWW2+lSZMmnH/++Zx//vncdNNN5OXlMWvWLAYMGFBhnw0bNrBs2TLGjRtXfn5hwYIF7N69O20/U31iZuTk5LBt27a91s2aNYsjjzyyQjfHxx9XnOPo9NNPZ+bMmVxxxRXVvka3bt247rrr6N+/P2ZW4Xj7SwEdiH79+jF58mQAHnvssfLlffr04aKLLmLz5s0VThAeiE6dOjFz5kz69OlDbm4ubdq04dZbb+WCCy6gY8eOXHTRRTRs2JDFixczd+5c7r333lq9XjZ6/vnn+fzzz7niiis45JBDKqy7+OKL+cMf/kD79u3Zs2cPZ5xxBi1atOCZZ56hUaNGdOnSZa/jtWnThkMPPZSJEyfSvn17Vq9ezQ033EDDhvqvmQw7duwo76rbuHEjDz/8MFu2bCnvy493zDHHsHr1ap5++ml69uzJ9OnT92oF33zzzQwcOJD8/HwGDx6MuzNjxgyuuuoqmjVrVr5d9+7dmTFjBgMGDMDMuOWWWw6ofnVxBKJfv37s3LmTww8/vEJrtlevXpSUlNCyZUu6detWq9f43e9+R2FhIe3bty/vGzvvvPN4/vnnKSwspEePHvTo0YN77rmHDh061Oq1stWkSZPo16/fXuEMUd/zihUraN68OZMmTaJ3796cdNJJTJkyhb/+9a907tx5r31ycnJ45plnePvttznppJO45ppruPPOO8nNzU3Hj5P1XnnlFdq1a0e7du0444wzmDdvHs8++yx9+/bda9uBAwdyww03MHz4cE455RRefvll7rjjjgrbfOc73+Fvf/sbL774Iqeffjp9+vShsLCQnJy9o7RHjx7MmDGDMWPGMHLkyAOa5zir5oN+6623Vpx66qmfV14edyVh49CuJJS91ZUrCeuzunAlYWCqvdTmrbfeOvTUU0/tVNW6ehHQZZYsWdLthBNOSGdJIiJwgAGtLg4RkUApoEVEAqWAFhEJlAJaRCRQ2RbQpaWlpZo8QkTqhFheVXs5aFYFtJmtLSkpaZLpOkREElFSUtLEzKodU5pVAb179+7bV6xY0Xjr1q1N1ZIWkVCVlpba1q1bm65YsaLx7t27b69uu6waBw2wYMGC8xo2bHibu+dR6Q/Qhg0bOrZrd0AX9IiIHLCdO3d+XGlRqZmt3b179+1du3adXt1+WRfQNSkoKPD58+dnugwRqX8O6BN9VnVxiIhkkzob0GZ2tZl9ZGbbzexNM+ud6ZpERJKpTga0mf0QeBD4NXA68AbwoplpCjYRyRp1MqCB64A/uvtEd1/q7kOBT4H/ynBdIiJJU+cC2swaA92AGZVWzQC+mf6KRERSoy7etuFQoAFQeTLadcC3Km9sZlcCV8aebjGz91JbXp11KFDtVK0i+6D3T81ecvdv7+9OdTGg94u7TwD2vkOkVGBm8929INN1SN2k909q1LkuDqK/0nuAtpWWtwXSdxsOEZEUq3MB7e47ie5O0L/Sqv5EozlERLJCXe3i+D3wlJnNBWYBPwOOAP6Q0arqNnUDSW3o/ZMCdfZSbzO7GhgJtAMWAyPc/bXMViUikjx1NqBFRLJdneuDlvQxsxVm9otM1yHZycxGm9niTNcRMgV0FjGzP5qZm9mkKtb9NrZuWiZqk8yKe29UfpyW6dqkegro7PMJcJGZNS9bYGYNgcuAlRmrSkLwCtE5m/iHWrABU0Bnn7eBYuCiuGXnA9uBorIFZtbdzGaY2edm9pWZvW5mPWs6sJm1MrMJZvaZmW02s/81M12cUHfscPe1lR67zWxgbEbI7bEZIu+OTakAlHd13RprhW82s0/M7Idm1trM/mJmW8ys2MwGxO3TwMwmxY5XEls/0sxqzBwz+7GZLYnV8r6ZjdjXPtms3v7gWW4S8JO45z8BHgfizwgfBDwF9AZ6AIuAF8zskKoOaGYGPA8cCVxANIvga8CrZqbb1NRRZnYe8DTwMHAi0XtlENFMkfGGA3OBrsD/AE8Ak4EXgNOI3gt/MrOye4LmAKuJGgrHAzcDo4Af11DLFbHXvTW2z/XAjcDVtfoh6zJ31yNLHsAfgWlAG6AE6ALkATuADmXrq9nXiGYEvDRu2QrgF7HvzwG2AE0r7bcIGJnpn12PhN4bu2O/w7LHi0TB+qtK234vtr5slNcK4M9x61sQ/bEfG7esU2xZQQ013AO8Evd8NLA47vlKYEilfYYDSzL975epR129UEVq4O4bzexvRK2hTUCRu6+MGsERMzscuBPoR3SZfAOgKVGQV6Ub0AxYH38coAlwdJJ/BEmN1/h64jCI/oi/D/QwsxvjlucQvRfyiP5oQ9R1BoC7bzGzbcA7cfuUTV52eNkCM/sZ8J9Ax9jxGgGV781Xtu1hQHtgvJn9d9yqhhzg7aKygQI6ez1G9DF0C9FHxsqeIArmEUQtpB3ATKBxFdtC9J92HVGXSGVf1bJWSY9t7r48fkGsf/d24Nkqtl8f9/2uSuu80rKy7rOc2HF/CDwA/IJoCoavgGuA71dTW1l368/QlA3lFNDZayawk2gayL9Xsb4XMMzdnwcws7ZEZ/Wrs4Ao0Evd/cPklioZtAA4rnJwJ0EvYI67P1y2wMyq/aTl7uvMbA1wtLs/meRa6iwFdJZydzezU4j6EXdUscn7wKVmNgdoDtxLFOjVeYVo3pN/mNlIYBnRR+BvE/Ur/iupP4Ckyx3ANDP7mOjk327gJKCHu4+sxXHfBy43s/8DLAcuBvoAG2vY5zbgITPbRHTysRHRSckj3f03tailztIojizm7pvdvbruh58Qnex5E/gLUZfIihqO5cB3gFeBicB7RP+hjwXWJK9qSSd3n040DLMf0SiNucBN1H7M/Hii98dkYB7RScTf7aOWR4nel0OAt4B/EfWZf1TLWuoszcUhIhIotaBFRAKlgBYRCZQCWkQkUApoEZFAKaBFRAKlgBYRCZQCWkQkUApoqbfM7DAzGxeb73iHma0zs5lm1j/TtYmALvWW+m0K0Qx9PyW6HPlwosuRq5wTWyTddCWh1Etm1ppoXoj+7v5KNds0JpqS9RLgYOBd4BZ3nx67gcEMoqkw+8fmPmkBLARmuPs1afgxJMupi0Pqq7JJ678bdxeQyh4nalEPJppA6AlgqpmdGpub5EdEdxMpu/P5WKIJp3QndEkKtaCl3jKzHxBN/NSMqOU7C3jW3efEpsYsBjq5+8q4ff4OrHH3q2PPvwc8QzQb4EiiWeDeSufPIdlLAS31Wqz13BvoSTR1ak+i++cVE83GtrXSLrnAq+5+XtwxHgcuJ7r1131pKFvqCZ0klHrN3bcDL8ced5jZo0T3yhtCdJeQ7ux9N5GSsm9iAd8d2APkp6FkqUcU0CIVLSH6f7GU6ARgnrsX1rD9fUSt6v7AdDN73t3/mfoypT5QF4fUS2Z2CNF9+B4juiHqZqAAeAh4x937m9mfiLo/rie6NdTBQF/gQ3f/a+xuIf8Aesf6rUcR3YX6FHdfm+YfSbKQAlrqJTPLJbrFUn+irolcYDUwFbjL3b8ws0ZE/dGXAUcBXxDdceR2ojuOvAOMc/c7YsfMAQqBbcB3XP+5pJYU0CIigdI4aBGRQCmgRUQCpYAWEQmUAlpEJFAKaBGRQCmgRUQCpYAWEQmUAlpEJFAKaBGRQP1/XjTVkJBfZVIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 425.125x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fontsize = 14\n",
    "plt.rc('xtick', labelsize=fontsize) \n",
    "plt.rc('ytick', labelsize=fontsize)\n",
    "\n",
    "fig = sns.catplot(x='sex', y='age', hue='race', order=['Male', 'Female'], hue_order=['White', 'Asian', 'Black'], kind=\"box\", data=df_cxr)\n",
    "fig._legend.remove()\n",
    "plt.xlabel('Sex', fontsize=fontsize)\n",
    "plt.ylabel('Age (years)', fontsize=fontsize)\n",
    "plt.ylim([0, 100])\n",
    "plt.legend(fontsize=14, loc='lower center', ncol=3)\n",
    "plt.show()\n",
    "# fig.savefig(\"MIMIC-age-sex-race.png\", bbox_inches='tight', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 425.125x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = sns.catplot(x='disease', y='age', hue='race', order=['Other', 'Pleural Effusion', 'No Finding'], hue_order=['White', 'Asian', 'Black'], kind=\"box\", data=df_cxr)\n",
    "fig._legend.remove()\n",
    "plt.xlabel('Disease', fontsize=fontsize)\n",
    "plt.ylabel('Age (years)', fontsize=fontsize)\n",
    "plt.ylim([0, 100])\n",
    "plt.legend(fontsize=14, loc='lower center', ncol=3)\n",
    "plt.show()\n",
    "# fig.savefig(\"MIMIC-age-pathology-race.png\", bbox_inches='tight', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432.125x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = sns.catplot(x='disease', hue='sex', order=['Other', 'Pleural Effusion', 'No Finding'], hue_order=['Male', 'Female'], data=df_cxr, kind='count')\n",
    "fig._legend.remove()\n",
    "plt.xlabel('Disease', fontsize=fontsize)\n",
    "plt.ylabel('Count', fontsize=fontsize)\n",
    "plt.legend(fontsize=14)\n",
    "plt.show()\n",
    "# fig.savefig(\"MIMIC-sex-pathology.png\", bbox_inches='tight', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 425.125x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = sns.catplot(x='disease', hue='race', order=['Other', 'Pleural Effusion', 'No Finding'], hue_order=['White', 'Asian', 'Black'], data=df_cxr, kind='count')\n",
    "fig._legend.remove()\n",
    "plt.xlabel('Disease', fontsize=fontsize)\n",
    "plt.ylabel('Count', fontsize=fontsize)\n",
    "plt.legend(fontsize=14)\n",
    "plt.show()\n",
    "# fig.savefig(\"MIMIC-race-pathology.png\", bbox_inches='tight', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = sns.catplot(x='disease', y='age', order=['Other', 'Pleural Effusion', 'No Finding'], kind=\"box\", data=df_cxr)\n",
    "plt.xlabel('Disease', fontsize=fontsize)\n",
    "plt.ylabel('Age (years)', fontsize=fontsize)\n",
    "plt.ylim([0, 100])\n",
    "plt.show()\n",
    "# fig.savefig(\"MIMIC-age-pathology.png\", bbox_inches='tight', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = sns.catplot(x='sex', y='age', order=['Male', 'Female'], kind=\"box\", data=df_cxr)\n",
    "sns.stripplot(x='sex', y='age', color='k', order=['Male', 'Female'], alpha=0.01, data=df_cxr, ax=fig.ax)\n",
    "plt.xlabel('Sex', fontsize=fontsize)\n",
    "plt.ylabel('Age (years)', fontsize=fontsize)\n",
    "plt.ylim([0, 100])\n",
    "plt.show()\n",
    "# fig.savefig(\"MIMIC-age-sex.png\", bbox_inches='tight', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = sns.catplot(x='race', y='age', order=['White', 'Asian', 'Black'], kind=\"box\", data=df_cxr)\n",
    "sns.stripplot(x='race', y='age', color='k', order=['White', 'Asian', 'Black'], alpha=0.01, data=df_cxr, ax=fig.ax)\n",
    "plt.xlabel('Race', fontsize=fontsize)\n",
    "plt.ylabel('Age (years)', fontsize=fontsize)\n",
    "plt.ylim([0, 100])\n",
    "plt.show()\n",
    "# fig.savefig(\"MIMIC-age-race.png\", bbox_inches='tight', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = sns.catplot(x='race', order=['White', 'Asian', 'Black'], data=df_cxr, kind='count')\n",
    "plt.xlabel('Race', fontsize=fontsize)\n",
    "plt.ylabel('Count', fontsize=fontsize)\n",
    "plt.show()\n",
    "# fig.savefig(\"MIMIC-race.png\", bbox_inches='tight', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = sns.catplot(x='sex', order=['Male', 'Female'], data=df_cxr, kind='count')\n",
    "plt.xlabel('Sex', fontsize=fontsize)\n",
    "plt.ylabel('Count', fontsize=fontsize)\n",
    "plt.show()\n",
    "# fig.savefig(\"MIMIC-sex.png\", bbox_inches='tight', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = sns.catplot(x='disease', order=['Other', 'Pleural Effusion', 'No Finding'], data=df_cxr, kind='count')\n",
    "plt.xlabel('Disease', fontsize=fontsize)\n",
    "plt.ylabel('Count', fontsize=fontsize)\n",
    "plt.show()\n",
    "# fig.savefig(\"MIMIC-pathology.png\", bbox_inches='tight', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432.125x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = sns.catplot(x='race', hue='sex', order=['White', 'Asian', 'Black'], hue_order=['Male', 'Female'], data=df_cxr, kind='count')\n",
    "fig._legend.remove()\n",
    "plt.xlabel('Race', fontsize=fontsize)\n",
    "plt.ylabel('Count', fontsize=fontsize)\n",
    "plt.legend(fontsize=14)\n",
    "plt.show()\n",
    "# fig.savefig(\"MIMIC-sex-race.png\", bbox_inches='tight', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 425.125x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = sns.catplot(x='sex', hue='race', order=['Male', 'Female'], hue_order=['White', 'Asian', 'Black'], data=df_cxr, kind='count')\n",
    "fig._legend.remove()\n",
    "plt.xlabel('Sex', fontsize=fontsize)\n",
    "plt.ylabel('Count', fontsize=fontsize)\n",
    "plt.legend(fontsize=14)\n",
    "plt.show()\n",
    "# fig.savefig(\"MIMIC-race-sex.png\", bbox_inches='tight', dpi=300)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}