data_preprocess_2021_2023.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import pyreadstat\n",
    "# !pip install pyreadstat\n",
    "import json\n",
    "import os, glob"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### BRFSS Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Read the XPT file using pyreadstat\n",
    "df, meta = pyreadstat.read_xport('BRFSS/BFRSS2021.XPT')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "shape of BRFSS 2021: (438693, 303)\n",
      "number of columns: 303\n"
     ]
    }
   ],
   "source": [
    "# store column name and labels pair \n",
    "column_names=meta.column_names_to_labels\n",
    "with open('BRFSS/BRFSS_column_labels.json', 'w') as json_file:\n",
    "    json.dump(column_names, json_file, indent=1, separators=(',', ': '))\n",
    "print(f\"shape of BRFSS 2021: {df.shape}\")\n",
    "print(f\"number of columns: {len(column_names)}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# common_col = ['_STATE', 'IYEAR', '_AGE80', '_SEX', 'BPHIGH6', 'DIABETE4', 'CHCKDNY2', 'NUMADULT', 'INCOME3', 'CHILDREN', 'EDUCA', 'MARITAL', 'PREGNANT', '_BMI5']\n",
    "common_col =  ['_STATE', 'IYEAR', '_AGE80', '_SEX', 'CHCKDNY2', 'DIABETE4', 'NUMADULT', 'INCOME3', 'CHILDREN', 'EDUCA', 'MARITAL', 'PREGNANT', '_BMI5'] # remove BPHIGH6, DIABETE4\n",
    "add_col = ['_RFHLTH', '_URBSTAT', '_PHYS14D', '_MENT14D', '_TOTINDA', '_SMOKER3']\n",
    "common_col = common_col + add_col"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\frank\\anaconda3\\envs\\deep_learning\\lib\\site-packages\\pandas\\io\\sas\\sas_xport.py:475: PerformanceWarning: DataFrame is highly fragmented.  This is usually the result of calling `frame.insert` many times, which has poor performance.  Consider joining all columns at once using pd.concat(axis=1) instead.  To get a de-fragmented frame, use `newframe = frame.copy()`\n",
      "  df[x] = v\n"
     ]
    }
   ],
   "source": [
    "# get 2022 and 2023 data\n",
    "df_2022 = pd.read_sas('BRFSS/LLCP2022.XPT')    \n",
    "df_2022 = df_2022[common_col]\n",
    "df_2023 = pd.read_sas('BRFSS/LLCP2023.XPT')\n",
    "df_2023 = df_2023[common_col]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train = pd.concat([df[common_col], df_2022, df_2023], axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train['_BMI5'] = df_train['_BMI5'] / 100 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "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>_STATE</th>\n",
       "      <th>IYEAR</th>\n",
       "      <th>_AGE80</th>\n",
       "      <th>_SEX</th>\n",
       "      <th>CHCKDNY2</th>\n",
       "      <th>DIABETE4</th>\n",
       "      <th>NUMADULT</th>\n",
       "      <th>INCOME3</th>\n",
       "      <th>CHILDREN</th>\n",
       "      <th>EDUCA</th>\n",
       "      <th>MARITAL</th>\n",
       "      <th>PREGNANT</th>\n",
       "      <th>_BMI5</th>\n",
       "      <th>_RFHLTH</th>\n",
       "      <th>_URBSTAT</th>\n",
       "      <th>_PHYS14D</th>\n",
       "      <th>_MENT14D</th>\n",
       "      <th>_TOTINDA</th>\n",
       "      <th>_SMOKER3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2021</td>\n",
       "      <td>70.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>88.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>14.54</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2021</td>\n",
       "      <td>67.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>77.0</td>\n",
       "      <td>88.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2021</td>\n",
       "      <td>72.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>88.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>28.29</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2021</td>\n",
       "      <td>62.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>88.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>33.47</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2021</td>\n",
       "      <td>76.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>88.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>28.73</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   _STATE IYEAR  _AGE80  _SEX  CHCKDNY2  DIABETE4  NUMADULT  INCOME3  \\\n",
       "0     1.0  2021    70.0   2.0       2.0       3.0       2.0      5.0   \n",
       "1     1.0  2021    67.0   2.0       1.0       1.0       2.0     77.0   \n",
       "2     1.0  2021    72.0   2.0       2.0       1.0       2.0      3.0   \n",
       "3     1.0  2021    62.0   2.0       2.0       1.0       2.0      7.0   \n",
       "4     1.0  2021    76.0   1.0       2.0       1.0       2.0      4.0   \n",
       "\n",
       "   CHILDREN  EDUCA  MARITAL  PREGNANT  _BMI5  _RFHLTH  _URBSTAT  _PHYS14D  \\\n",
       "0      88.0    4.0      1.0       NaN  14.54      2.0       1.0       3.0   \n",
       "1      88.0    6.0      9.0       NaN    NaN      1.0       1.0       1.0   \n",
       "2      88.0    4.0      3.0       NaN  28.29      1.0       1.0       1.0   \n",
       "3      88.0    4.0      1.0       NaN  33.47      1.0       1.0       1.0   \n",
       "4      88.0    3.0      1.0       NaN  28.73      2.0       1.0       3.0   \n",
       "\n",
       "   _MENT14D  _TOTINDA  _SMOKER3  \n",
       "0       2.0       2.0       3.0  \n",
       "1       1.0       1.0       4.0  \n",
       "2       1.0       2.0       4.0  \n",
       "3       2.0       1.0       4.0  \n",
       "4       1.0       1.0       4.0  "
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train = df_train.dropna()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Numeric: poverty_index, '_DRNKWK1', '_FRUTSU1', 'FRNCHDA_', 'POTADA1_', '_VEGESU1', '_BMI5'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "29.007255345201134"
      ]
     },
     "execution_count": 141,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train['_BMI5'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\frank\\anaconda3\\envs\\deep_learning\\lib\\site-packages\\pandas\\core\\generic.py:6619: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  return self._update_inplace(result)\n"
     ]
    }
   ],
   "source": [
    "df_train['_DRNKWK1'].replace(99900, np.nan, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# get number of people in household\n",
    "df_train['CHILDREN'].replace(88, 0, inplace=True)\n",
    "df_train = df_train[df_train['CHILDREN'] < 99] \n",
    "df_train['NUMPEOPLE'] = df_train['NUMADULT'] + df_train['CHILDREN']\n",
    " # calculate poverty level\n",
    "df_train = df_train.dropna(subset=['NUMPEOPLE', 'INCOME3'])\n",
    "df_train['NUMPEOPLE'].value_counts()\n",
    "# drop 102, 78 and 57 \n",
    "df_train = df_train[df_train['NUMPEOPLE'] < 50]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2021    111132\n",
       "2022      3845\n",
       "Name: IYEAR, dtype: int64"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train['IYEAR'] = df_train['IYEAR'].astype(int)\n",
    "df_train['IYEAR'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "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>_STATE</th>\n",
       "      <th>IYEAR</th>\n",
       "      <th>_AGE80</th>\n",
       "      <th>_SEX</th>\n",
       "      <th>BPHIGH6</th>\n",
       "      <th>DIABETE4</th>\n",
       "      <th>CHCKDNY2</th>\n",
       "      <th>INCOME3</th>\n",
       "      <th>EDUCA</th>\n",
       "      <th>MARITAL</th>\n",
       "      <th>...</th>\n",
       "      <th>_MICHD</th>\n",
       "      <th>_LTASTH1</th>\n",
       "      <th>_LMTACT3</th>\n",
       "      <th>_SMOKER3</th>\n",
       "      <th>_DRNKWK1</th>\n",
       "      <th>_FRUTSU1</th>\n",
       "      <th>FRNCHDA_</th>\n",
       "      <th>POTADA1_</th>\n",
       "      <th>INCOME</th>\n",
       "      <th>FPL_RATIO</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2021</td>\n",
       "      <td>70.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>75000.0</td>\n",
       "      <td>4.305396</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2021</td>\n",
       "      <td>67.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>77.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2021</td>\n",
       "      <td>72.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>30000.0</td>\n",
       "      <td>1.722158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2021</td>\n",
       "      <td>62.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>300.0</td>\n",
       "      <td>114.0</td>\n",
       "      <td>57.0</td>\n",
       "      <td>27.0</td>\n",
       "      <td>200000.0</td>\n",
       "      <td>5.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2021</td>\n",
       "      <td>76.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>42500.0</td>\n",
       "      <td>2.439724</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   _STATE  IYEAR  _AGE80  _SEX  BPHIGH6  DIABETE4  CHCKDNY2  INCOME3  EDUCA  \\\n",
       "0     1.0   2021    70.0   2.0      3.0       3.0       2.0      5.0    4.0   \n",
       "1     1.0   2021    67.0   2.0      1.0       1.0       1.0     77.0    6.0   \n",
       "2     1.0   2021    72.0   2.0      1.0       1.0       2.0      3.0    4.0   \n",
       "3     1.0   2021    62.0   2.0      1.0       1.0       2.0      7.0    4.0   \n",
       "4     1.0   2021    76.0   1.0      4.0       1.0       2.0      4.0    3.0   \n",
       "\n",
       "   MARITAL  ...  _MICHD  _LTASTH1  _LMTACT3  _SMOKER3  _DRNKWK1  _FRUTSU1  \\\n",
       "0      1.0  ...     2.0       2.0       2.0       3.0       0.0     100.0   \n",
       "1      9.0  ...     1.0       1.0       1.0       4.0       0.0     100.0   \n",
       "2      3.0  ...     1.0       1.0       3.0       4.0       0.0     100.0   \n",
       "3      1.0  ...     2.0       1.0       3.0       4.0     300.0     114.0   \n",
       "4      1.0  ...     1.0       1.0       3.0       4.0       0.0     100.0   \n",
       "\n",
       "   FRNCHDA_  POTADA1_    INCOME  FPL_RATIO  \n",
       "0      43.0      14.0   75000.0   4.305396  \n",
       "1       0.0      14.0       NaN        NaN  \n",
       "2      14.0      14.0   30000.0   1.722158  \n",
       "3      57.0      27.0  200000.0   5.000000  \n",
       "4      29.0      29.0   42500.0   2.439724  \n",
       "\n",
       "[5 rows x 31 columns]"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# define 2021 federal poverty level per household size, more than 8 people, add 4540 for each additional person, write up to 16 people\n",
    "fpl_2021 = {1: 12880, 2: 17420, 3: 21960, 4: 26500, 5: 31040, 6: 35580, 7: 40120, 8: 44660}\n",
    "fpl_2022 = {1: 13590, 2: 18310, 3: 23030, 4: 27750, 5: 32470, 6: 37190, 7: 41910, 8: 46630}\n",
    "fpl_2023 = {1: 14580, 2: 19720, 3: 24860, 4: 30000, 5: 35140, 6: 40280, 7: 45420, 8: 50560}\n",
    "fpl_2024 = {1: 15060, 2: 20440, 3: 25820, 4: 31200, 5: 36580, 6: 41960, 7: 47340, 8: 52720}\n",
    "def fpl_level(num_people, year):\n",
    "    if num_people > 8:\n",
    "        if year == 2021:\n",
    "            return 44660 + 4540*(num_people-8)\n",
    "        elif year == 2023:\n",
    "            return 50560 + 5140*(num_people-8)\n",
    "        elif year == 2024:\n",
    "            return 52720 + 5380*(num_people-8)\n",
    "        elif year == 2022:\n",
    "            return 46630 + 4720*(num_people-8)\n",
    "    else:\n",
    "        if year == 2021:\n",
    "            return fpl_2021[num_people]\n",
    "        elif year == 2023:\n",
    "            return fpl_2023[num_people]\n",
    "        elif year == 2024:\n",
    "            return fpl_2024[num_people]\n",
    "        elif year == 2022:\n",
    "            return fpl_2022[num_people]\n",
    "\n",
    "# convert income category to income \n",
    "def get_income(income_cat):\n",
    "    if income_cat == 1:\n",
    "        return 7500 \n",
    "    elif income_cat == 2:\n",
    "        return (15000 + 25000)/2\n",
    "    elif income_cat == 3:\n",
    "        return (25000 + 35000)/2\n",
    "    elif income_cat == 4:\n",
    "        return (35000 + 50000)/2\n",
    "    elif income_cat == 5:\n",
    "        return (50000 + 100000)/2\n",
    "    elif income_cat == 6:\n",
    "        return (100000 + 200000)/2\n",
    "    elif income_cat == 7:\n",
    "        return 200000\n",
    "    else:\n",
    "        return \n",
    "# convert income to poverty level \n",
    "df_train['INCOME'] = df_train['INCOME3'].apply(get_income)\n",
    "# Convert FPL based on the year\n",
    "df_train['FPL'] = df_train.apply(lambda row: fpl_level(row['NUMPEOPLE'], row['IYEAR']), axis=1)\n",
    "df_train['FPL_RATIO'] = df_train['INCOME']/df_train['FPL']\n",
    "df_train['FPL_RATIO'] = df_train['FPL_RATIO'].apply(lambda x: 5 if x > 5 else x)\n",
    "df_train.drop(columns=['FPL', 'NUMPEOPLE', 'CHILDREN', 'NUMADULT'], inplace=True)\n",
    "df_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot histogram of age_G\n",
    "df_train['_AGE80'].hist(bins=20)\n",
    "plt.title('Age Group Distribution')\n",
    "plt.xlabel('Age Group')\n",
    "plt.ylabel('Number of People')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "26.63    1230\n",
       "27.46    1042\n",
       "27.44     995\n",
       "24.41     732\n",
       "28.34     693\n",
       "         ... \n",
       "22.88       1\n",
       "43.51       1\n",
       "59.41       1\n",
       "33.32       1\n",
       "41.30       1\n",
       "Name: _BMI5, Length: 2723, dtype: int64"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train['_BMI5'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0    59928\n",
       "2.0    42388\n",
       "3.0    11702\n",
       "Name: MARITAL, dtype: int64"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# convert marital status to match with NHANES, 1,6->1, 2,3,4_>2, 5->3 \n",
    "def process_marital(x):\n",
    "    if x in [1,6]:\n",
    "        return 1\n",
    "    elif x in [2,3,4]:\n",
    "        return 2\n",
    "    elif x == 5:\n",
    "        return 3\n",
    "    else:\n",
    "        return \n",
    "df_train['MARITAL'] = df_train['MARITAL'].apply(process_marital)\n",
    "df_train['MARITAL'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df_train['_HLTHPLN'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "# standardize health insurance\n",
    "df_train = df_train[df_train['_HLTHPLN']!=9]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# standadize chronic kidney disease, select 1 and 2\n",
    "df_train = df_train[(df_train['CHCKDNY2']==1) | (df_train['CHCKDNY2']==2)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "def process_weight(value):\n",
    "    if value == 7777 or value == 9999 :  # Handle None values\n",
    "        return np.nan\n",
    "    if value < 1000:  # Convert pounds to kilograms\n",
    "        kg = value * 0.453592\n",
    "        return round(kg, 2)  # Round to two decimal places\n",
    "    elif 9000 < value < 9400:  # Remove leading '9'\n",
    "        return float(str(value)[1:])  # Remove the first character\n",
    "    return np.nan  # Return None for all other values\n",
    "# df_train['Weight'] = df_train['WEIGHT2'].apply(process_weight)\n",
    "# df_train = df_train.drop(columns=['WEIGHT2'])\n",
    "def process_height(value):\n",
    "    if value == 7777 or value == 9999 :  # Handle None values\n",
    "        return np.nan\n",
    "    if value < 1000:  # Convert pounds to kilograms\n",
    "        feet = float(str(value)[0])\n",
    "        inch = float(str(value)[1:])\n",
    "        cm = inch * 2.54 + feet * 30.48\n",
    "        return round(cm, 2)  # Round to two decimal places\n",
    "    elif 9000 < value < 9400:  # Remove leading '9'\n",
    "        return float(str(value)[1:])  # Remove the first character\n",
    "    return np.nan  # Return None for all other values\n",
    "# df_train['Height'] = df_train['HEIGHT3'].apply(process_height)\n",
    "# df_train = df_train.drop(columns=['HEIGHT3'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0    93612\n",
       "2.0     7575\n",
       "5.0     4441\n",
       "3.0     2964\n",
       "4.0     1473\n",
       "Name: _RACEPRV, dtype: int64"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def process_race(value):\n",
    "    # convert American Indians 3, Pacific Islanders 5, other races 6, multiracial 7 to 5 \n",
    "    if value == 3 or value == 6 or value ==7:\n",
    "        return 5\n",
    "    # convert Hispanic 8 to 3 \n",
    "    elif value == 8:\n",
    "        return 3 \n",
    "    else:\n",
    "        return value\n",
    "df_train['_RACEPRV'] = df_train['_RACEPRV'].apply(process_race)\n",
    "df_train['_RACEPRV'].value_counts()    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df_train['BMI'] = df_train['Weight'] / ((df_train['Height']/100) ** 2)\n",
    "# df_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# remove unknown BP status, DM status and convert to binary: 3-->2 \n",
    "df_train = df_train[(df_train['BPHIGH6'] == 1) | (df_train['BPHIGH6'] == 3)]\n",
    "df_train['BPHIGH6'] = df_train['BPHIGH6'].replace(3, 2)\n",
    "df_train = df_train[(df_train['DIABETE4'] == 1) | (df_train['DIABETE4'] == 3)] \n",
    "df_train['DIABETE4'] = df_train['DIABETE4'].replace(3, 2)\n",
    "# remove non adult (age<18)\n",
    "# df_train = df_train[df_train['CADULT1'] == 1]\n",
    "# remove pregnant responder and drop the column \n",
    "df_train = df_train[df_train['PREGNANT'] != 1]\n",
    "df_train = df_train.drop(columns=['PREGNANT'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "# change education status to match NHANES\n",
    "value_mapping = {\n",
    "    1: 1,  # Never attended school or only kindergarten\n",
    "    2: 1,  # Grades 1 through 8\n",
    "    3: 2,  # Grades 9 through 11\n",
    "    4: 3,  # high school graduate or GED\n",
    "    5: 4,  # Some college or technical school\n",
    "    6: 5   # College graduate\n",
    "}\n",
    "df_train['EDUCA'] = df_train['EDUCA'].replace(value_mapping)\n",
    "\n",
    "# Remove rows where 'EDUCA' is 9(refuse) or NaN\n",
    "df_train = df_train[df_train['EDUCA'].notna() & (df_train['EDUCA'] != 9)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train = df_train.rename(columns={\n",
    "    'CADULT1': 'Is_Adult',\n",
    "    '_SEX': 'Gender',     # 1: Male, 2: Female  \n",
    "    'BPHIGH6': 'Hypertension',\n",
    "    'DIABETE4': 'Diabetes',\n",
    "    'EDUCA': 'Education',\n",
    "    '_RACEPRV': 'Race',\n",
    "    '_AGE80': 'Age',\n",
    "    'MARITAL': 'Marriage',\n",
    "    '_HLTHPLN': 'Health_Insurance',\n",
    "    'CHCKDNY2': 'CKD',\n",
    "    'FPL_RATIO': 'Poverty_Index',\n",
    "    '_BMI5': 'BMI'\n",
    "})\n",
    "# df_train = df_train.drop(columns=['IYEAR'])\n",
    "# bmi = df_train['BMI']\n",
    "# df_train = df_train.drop(columns=['BMI'])\n",
    "# df_train['BMI'] = bmi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "df_train = df_train.dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(40620, 30)"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0    34439\n",
       "2.0     3131\n",
       "5.0     1651\n",
       "3.0     1041\n",
       "4.0      358\n",
       "Name: Race, dtype: int64"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train['Race'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0    24368\n",
       "2.0    16252\n",
       "Name: Hypertension, dtype: int64"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train['Hypertension'].value_counts()    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "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>_STATE</th>\n",
       "      <th>IYEAR</th>\n",
       "      <th>Age</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Hypertension</th>\n",
       "      <th>Diabetes</th>\n",
       "      <th>CKD</th>\n",
       "      <th>INCOME3</th>\n",
       "      <th>Education</th>\n",
       "      <th>Marriage</th>\n",
       "      <th>...</th>\n",
       "      <th>_MICHD</th>\n",
       "      <th>_LTASTH1</th>\n",
       "      <th>_LMTACT3</th>\n",
       "      <th>_SMOKER3</th>\n",
       "      <th>_DRNKWK1</th>\n",
       "      <th>_FRUTSU1</th>\n",
       "      <th>FRNCHDA_</th>\n",
       "      <th>POTADA1_</th>\n",
       "      <th>INCOME</th>\n",
       "      <th>Poverty_Index</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2021</td>\n",
       "      <td>70.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>75000.0</td>\n",
       "      <td>4.305396</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2021</td>\n",
       "      <td>72.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>100.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>30000.0</td>\n",
       "      <td>1.722158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2021</td>\n",
       "      <td>62.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>300.0</td>\n",
       "      <td>114.0</td>\n",
       "      <td>57.0</td>\n",
       "      <td>27.0</td>\n",
       "      <td>200000.0</td>\n",
       "      <td>5.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2021</td>\n",
       "      <td>80.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>29.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>150000.0</td>\n",
       "      <td>5.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2021</td>\n",
       "      <td>67.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>30000.0</td>\n",
       "      <td>2.329193</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 30 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    _STATE  IYEAR   Age  Gender  Hypertension  Diabetes  CKD  INCOME3  \\\n",
       "0      1.0   2021  70.0     2.0           2.0       2.0  2.0      5.0   \n",
       "2      1.0   2021  72.0     2.0           1.0       1.0  2.0      3.0   \n",
       "3      1.0   2021  62.0     2.0           1.0       1.0  2.0      7.0   \n",
       "5      1.0   2021  80.0     1.0           2.0       2.0  2.0      6.0   \n",
       "10     1.0   2021  67.0     2.0           1.0       2.0  2.0      3.0   \n",
       "\n",
       "    Education  Marriage  ...  _MICHD  _LTASTH1  _LMTACT3  _SMOKER3  _DRNKWK1  \\\n",
       "0         3.0       1.0  ...     2.0       2.0       2.0       3.0       0.0   \n",
       "2         3.0       2.0  ...     1.0       1.0       3.0       4.0       0.0   \n",
       "3         3.0       1.0  ...     2.0       1.0       3.0       4.0     300.0   \n",
       "5         4.0       1.0  ...     2.0       1.0       3.0       3.0       0.0   \n",
       "10        4.0       2.0  ...     2.0       2.0       3.0       2.0       0.0   \n",
       "\n",
       "    _FRUTSU1  FRNCHDA_  POTADA1_    INCOME  Poverty_Index  \n",
       "0      100.0      43.0      14.0   75000.0       4.305396  \n",
       "2      100.0      14.0      14.0   30000.0       1.722158  \n",
       "3      114.0      57.0      27.0  200000.0       5.000000  \n",
       "5       29.0       0.0      14.0  150000.0       5.000000  \n",
       "10      43.0      14.0       0.0   30000.0       2.329193  \n",
       "\n",
       "[5 rows x 30 columns]"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7.0    11776\n",
       "6.0    10145\n",
       "5.0     8680\n",
       "4.0     4115\n",
       "3.0     2768\n",
       "2.0     2000\n",
       "1.0     1136\n",
       "Name: INCOME3, dtype: int64"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train['INCOME3'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df_train.to_csv('BRFSS/BRFSS_2021_2023.csv', index=False)\n",
    "# df_train.to_csv('BRFSS/BRFSS_2021.csv', index=False)\n",
    "# df_train.to_csv('BRFSS/BRFSS_2021_long.csv', index=False)\n",
    "df_train.to_csv('BRFSS/BRFSS_2021_long.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create json file for column names and labels\n",
    "variables_meta = {\n",
    "    'Race': {\n",
    "        1: 'White',\n",
    "        2: 'Black',\n",
    "        3: 'Hispanic',\n",
    "        4: 'Asian',\n",
    "        5: 'Other'\n",
    "    },\n",
    "    'Gender': {\n",
    "        1: 'Male', \n",
    "        2: 'Female'\n",
    "    },\n",
    "    'Education': {\n",
    "        1: 'Less than high school',\n",
    "        2: 'some high school',\n",
    "        3: 'high school graduate',\n",
    "        4: 'some college',\n",
    "        5: 'college graduate'\n",
    "    },\n",
    "    'Hypertension': {\n",
    "        1: 'Yes',\n",
    "        2: 'No'\n",
    "    },\n",
    "    'Diabetes': {\n",
    "        1: 'Yes',\n",
    "        2: 'No'\n",
    "    },\n",
    "    'Marrige': {\n",
    "        1: 'Married or living with partner',\n",
    "        2: 'divorced or separated or widowed',\n",
    "        3: 'never married'\n",
    "    },\n",
    "    'Health_Insurance': {\n",
    "        1: 'Yes',\n",
    "        2: 'No'\n",
    "    },\n",
    "    'CKD': {\n",
    "        1: 'Yes',\n",
    "        2: 'No'\n",
    "    },\n",
    "    'Age': {\n",
    "        1: '18-24',\n",
    "        2: '25-34',\n",
    "        3: '35-44',\n",
    "        4: '45-54',\n",
    "        5: '55-64',\n",
    "        6:  '>='\n",
    "    },\n",
    "}\n",
    "with open('BRFSS/BRFSS_variables_meta.json', 'w') as json_file:\n",
    "    json.dump(variables_meta, json_file, indent=1, separators=(',', ': '))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### NHANES 2021-2023 Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_path = os.path.join(\"NHANES\", \"2021-2023\")\n",
    "file_pattern = os.path.join(data_path, \"*.XPT\")\n",
    "\n",
    "# Use glob to find files that match the pattern\n",
    "files = glob.glob(file_pattern)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(11933, 27)\n",
      "Index(['SEQN', 'SDDSRVYR', 'RIDSTATR', 'RIAGENDR', 'RIDAGEYR', 'RIDAGEMN',\n",
      "       'RIDRETH1', 'RIDRETH3', 'RIDEXMON', 'RIDEXAGM', 'DMQMILIZ', 'DMDBORN4',\n",
      "       'DMDYRUSR', 'DMDEDUC2', 'DMDMARTZ', 'RIDEXPRG', 'DMDHHSIZ', 'DMDHRGND',\n",
      "       'DMDHRAGZ', 'DMDHREDZ', 'DMDHRMAZ', 'DMDHSEDZ', 'WTINT2YR', 'WTMEC2YR',\n",
      "       'SDMVSTRA', 'SDMVPSU', 'INDFMPIR'],\n",
      "      dtype='object')\n"
     ]
    }
   ],
   "source": [
    "demo_df = pd.read_sas(os.path.join(data_path, \"DEMO_L.XPT\"))\n",
    "print(demo_df.shape)\n",
    "print(demo_df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "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>SEQN</th>\n",
       "      <th>RIAGENDR</th>\n",
       "      <th>RIDAGEYR</th>\n",
       "      <th>RIDRETH3</th>\n",
       "      <th>DMDEDUC2</th>\n",
       "      <th>DMDMARTZ</th>\n",
       "      <th>DMDHHSIZ</th>\n",
       "      <th>INDFMPIR</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>130378.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>130379.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>66.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>130380.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>44.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1.41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>130381.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>130382.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.60</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       SEQN  RIAGENDR  RIDAGEYR  RIDRETH3  DMDEDUC2  DMDMARTZ  DMDHHSIZ  \\\n",
       "0  130378.0       1.0      43.0       6.0       5.0       1.0       4.0   \n",
       "1  130379.0       1.0      66.0       3.0       5.0       1.0       2.0   \n",
       "2  130380.0       2.0      44.0       2.0       3.0       1.0       7.0   \n",
       "3  130381.0       2.0       5.0       7.0       NaN       NaN       2.0   \n",
       "4  130382.0       1.0       2.0       3.0       NaN       NaN       4.0   \n",
       "\n",
       "   INDFMPIR  \n",
       "0      5.00  \n",
       "1      5.00  \n",
       "2      1.41  \n",
       "3      1.53  \n",
       "4      3.60  "
      ]
     },
     "execution_count": 144,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "demo_df_select = demo_df[['SEQN', 'RIAGENDR', 'RIDAGEYR', \n",
    "       'RIDRETH3', 'DMDEDUC2', 'DMDMARTZ',  'DMDHHSIZ',  'INDFMPIR', ]]\n",
    "demo_df_select.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(8860, 22)\n",
      "Index(['SEQN', 'BMDSTATS', 'BMXWT', 'BMIWT', 'BMXRECUM', 'BMIRECUM', 'BMXHEAD',\n",
      "       'BMIHEAD', 'BMXHT', 'BMIHT', 'BMXBMI', 'BMDBMIC', 'BMXLEG', 'BMILEG',\n",
      "       'BMXARML', 'BMIARML', 'BMXARMC', 'BMIARMC', 'BMXWAIST', 'BMIWAIST',\n",
      "       'BMXHIP', 'BMIHIP'],\n",
      "      dtype='object')\n"
     ]
    }
   ],
   "source": [
    "bmi_df = pd.read_sas(os.path.join(data_path, \"BodyMeasure_L.XPT\"))\n",
    "print(bmi_df.shape)\n",
    "print(bmi_df.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "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>SEQN</th>\n",
       "      <th>BMXBMI</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>130378.0</td>\n",
       "      <td>27.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>130379.0</td>\n",
       "      <td>33.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>130380.0</td>\n",
       "      <td>29.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>130381.0</td>\n",
       "      <td>23.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>130382.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       SEQN  BMXBMI\n",
       "0  130378.0    27.0\n",
       "1  130379.0    33.5\n",
       "2  130380.0    29.7\n",
       "3  130381.0    23.8\n",
       "4  130382.0     NaN"
      ]
     },
     "execution_count": 146,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bmi_df_valid = bmi_df[['SEQN', 'BMXBMI']]\n",
    "# filter out rows with comments(invalid) or does not have weight or height \n",
    "# bmi_valid = bmi_df_select[bmi_df_select[\"BMIWT\"].isna() &\n",
    "#                           bmi_df_select[\"BMIHT\"].isna() & \n",
    "#                           bmi_df_select[\"BMXWT\"].notna() &\n",
    "#                           bmi_df_select[\"BMXHT\"].notna()]\n",
    "# Drop 'BMIWT' and 'BMIHT' columns\n",
    "# bmi_df_valid = bmi_valid.drop(['BMIWT', 'BMIHT'], axis=1)\n",
    "bmi_df_valid.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "27.24666509266917"
      ]
     },
     "execution_count": 147,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bmi_df_valid['BMXBMI'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "self reported BMI ratio by Interval:\n",
      "18.5-20: 0.0414\n",
      "20-30: 0.5860\n",
      "30-35: 0.1830\n",
      "35-40: 0.0967\n",
      ">40: 0.0929\n",
      "measured Dataset BMI ratio by Interval:\n",
      "18.5-20: 0.0579\n",
      "20-30: 0.5728\n",
      "30-35: 0.1897\n",
      "35-40: 0.0937\n",
      ">40: 0.0830\n"
     ]
    }
   ],
   "source": [
    "df_train = df_train[df_train['_BMI5'] >= 18.5]\n",
    "bmi_df_valid = bmi_df_valid[bmi_df_valid['BMXBMI'] >= 18.5]\n",
    "# Define the BMI intervals\n",
    "intervals = [(18.5, 20), (20, 30), (30, 35), (35, 40), (40, np.inf)]\n",
    "\n",
    "# Initialize dictionaries to store means for each interval in both datasets\n",
    "means_train = {}\n",
    "means_valid = {}\n",
    "\n",
    "for lower, upper in intervals:\n",
    "    # Calculate the mean for each interval in the training dataset\n",
    "    if upper == np.inf:\n",
    "        means_train[f'>{lower}'] = np.mean(df_train['_BMI5'] > lower)\n",
    "        means_valid[f'>{lower}'] = np.mean(bmi_df_valid['BMXBMI'] > lower)\n",
    "    else:\n",
    "        means_train[f'{lower}-{upper}'] = np.mean((df_train['_BMI5'] > lower) & (df_train['_BMI5'] <= upper))\n",
    "        means_valid[f'{lower}-{upper}'] = np.mean((bmi_df_valid['BMXBMI'] > lower) & (bmi_df_valid['BMXBMI'] <= upper))\n",
    "\n",
    "print(\"self reported BMI ratio by Interval:\")\n",
    "for interval, mean in means_train.items():\n",
    "    print(f\"{interval}: {mean:.4f}\")\n",
    "\n",
    "print(\"measured Dataset BMI ratio by Interval:\")\n",
    "for interval, mean in means_valid.items():\n",
    "    print(f\"{interval}: {mean:.4f}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "29.211037093836914"
      ]
     },
     "execution_count": 149,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train['_BMI5'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "29.054934029686727"
      ]
     },
     "execution_count": 150,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bmi_df_valid['BMXBMI'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(bmi_df_valid['BMXBMI'], bins=20)\n",
    "plt.title('measured BMI Distribution')\n",
    "plt.xlabel('BMI')\n",
    "plt.ylabel('Number of People')\n",
    "\n",
    "plt.show()\n",
    "\n",
    "plt.hist(df_train['_BMI5'], bins=20)\n",
    "plt.title('self report BMI Distribution')\n",
    "plt.xlabel('BMI')\n",
    "plt.ylabel('Number of People')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "29.007255345201134\n",
      "27.24666509266917\n"
     ]
    }
   ],
   "source": [
    "print(df_train['_BMI5'].mean())\n",
    "print(bmi_df_valid['BMXBMI'].mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(8860, 9)\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>SEQN</th>\n",
       "      <th>RIAGENDR</th>\n",
       "      <th>RIDAGEYR</th>\n",
       "      <th>RIDRETH3</th>\n",
       "      <th>DMDEDUC2</th>\n",
       "      <th>DMDMARTZ</th>\n",
       "      <th>DMDHHSIZ</th>\n",
       "      <th>INDFMPIR</th>\n",
       "      <th>BMXBMI</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>130378.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.00</td>\n",
       "      <td>27.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>130379.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>66.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.00</td>\n",
       "      <td>33.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>130380.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>44.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1.41</td>\n",
       "      <td>29.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>130381.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.53</td>\n",
       "      <td>23.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>130382.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>3.60</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       SEQN  RIAGENDR  RIDAGEYR  RIDRETH3  DMDEDUC2  DMDMARTZ  DMDHHSIZ  \\\n",
       "0  130378.0       1.0      43.0       6.0       5.0       1.0       4.0   \n",
       "1  130379.0       1.0      66.0       3.0       5.0       1.0       2.0   \n",
       "2  130380.0       2.0      44.0       2.0       3.0       1.0       7.0   \n",
       "3  130381.0       2.0       5.0       7.0       NaN       NaN       2.0   \n",
       "4  130382.0       1.0       2.0       3.0       NaN       NaN       4.0   \n",
       "\n",
       "   INDFMPIR  BMXBMI  \n",
       "0      5.00    27.0  \n",
       "1      5.00    33.5  \n",
       "2      1.41    29.7  \n",
       "3      1.53    23.8  \n",
       "4      3.60     NaN  "
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Merge two DataFrames on the 'SEQN' column\n",
    "merge_df = pd.merge(demo_df_select, bmi_df_valid, on='SEQN')\n",
    "print(merge_df.shape)\n",
    "merge_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Rename columns in the DataFrame\n",
    "merge_df = merge_df.rename(columns={\n",
    "    'RIAGENDR': 'Gender',\n",
    "    'RIDAGEYR': 'Age',\n",
    "    'RIDRETH3': 'Race',\n",
    "    'DMDEDUC2': 'Education',\n",
    "    'DMDMARTZ': 'Marriage',\n",
    "    'INDFMPIR': 'Poverty_Index',\n",
    "    'BMXBMI': 'BMI'\n",
    "})\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "White       4684\n",
       "Hispanic    1841\n",
       "Black       1159\n",
       "Other        647\n",
       "Asian        529\n",
       "Name: Race, dtype: int64"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# convert race to match BRFSS\n",
    "def process_race_NHA(n):\n",
    "    if n == 1 or n == 2:\n",
    "        return 'Hispanic'\n",
    "    elif n == 3:\n",
    "        return 'White' # White\n",
    "    elif n == 4:\n",
    "        return 'Black' # Black\n",
    "    elif n == 6:\n",
    "        return 'Asian' # Asian\n",
    "    elif n == 7:\n",
    "        return 'Other' # Other\n",
    "    else:\n",
    "        return \n",
    "merge_df['Race'] = merge_df['Race'].apply(process_race_NHA)\n",
    "merge_df['Race'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Subgroup analysis of  NHANES data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8860, 9)"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "merge_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "mean bmi national 2021-2023: 27.24666509266917\n",
      "mean bmi per race 2021-2023\n",
      "Race\n",
      "Asian       23.011133\n",
      "Black       28.503755\n",
      "Hispanic    26.708094\n",
      "Other       26.792471\n",
      "White       27.682976\n",
      "Name: BMI, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "# race level analysis\n",
    "mean_bmi_national = merge_df['BMI'].mean()\n",
    "mean_bmi_per_race = merge_df.groupby('Race')['BMI'].mean()\n",
    "print(f'mean bmi national 2021-2023: {mean_bmi_national}')\n",
    "print('mean bmi per race 2021-2023')\n",
    "print(mean_bmi_per_race)\n",
    "# convert to csv\n",
    "merge_df.to_csv('NHANES/NHANES_2021_2023_race.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# histogram of BMI\n",
    "merge_df['BMI'].hist(bins=20)\n",
    "plt.title('BMI Distribution')\n",
    "plt.xlabel('BMI')\n",
    "plt.ylabel('Number of People')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5.0    2143\n",
       "4.0    1846\n",
       "3.0    1297\n",
       "2.0     482\n",
       "1.0     292\n",
       "9.0       4\n",
       "Name: Education, dtype: int64"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "merge_df['Education'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "# remove unknown education status \n",
    "merge_df = merge_df[(merge_df['Education'] != 9) & (merge_df['Education'].notna()) & (merge_df['Education'] != 7)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.0     3277\n",
      "2.0     1533\n",
      "3.0     1245\n",
      "99.0       3\n",
      "77.0       1\n",
      "Name: Marriage, dtype: int64\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "1.0    3277\n",
       "2.0    1533\n",
       "3.0    1245\n",
       "Name: Marriage, dtype: int64"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# standardize marital status to match BRFSS, only select 1,2,3\n",
    "print(merge_df['Marriage'].value_counts())\n",
    "def process_marital_NHA(x):\n",
    "    if x == 1:\n",
    "        return 1\n",
    "    elif x == 2:\n",
    "        return 2\n",
    "    elif x == 3:\n",
    "        return 3\n",
    "    else:\n",
    "        return\n",
    "merge_df['Marriage'] = merge_df['Marriage'].apply(process_marital_NHA)\n",
    "merge_df['Marriage'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "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>SEQN</th>\n",
       "      <th>Gender</th>\n",
       "      <th>Age</th>\n",
       "      <th>Race</th>\n",
       "      <th>Education</th>\n",
       "      <th>Marriage</th>\n",
       "      <th>DMDHHSIZ</th>\n",
       "      <th>Poverty_Index</th>\n",
       "      <th>BMI</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>130378.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>4</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.00</td>\n",
       "      <td>27.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>130379.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>66.0</td>\n",
       "      <td>1</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>5.00</td>\n",
       "      <td>33.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>130380.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>44.0</td>\n",
       "      <td>3</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>1.41</td>\n",
       "      <td>29.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>130386.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>34.0</td>\n",
       "      <td>3</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.33</td>\n",
       "      <td>30.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>130387.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>68.0</td>\n",
       "      <td>1</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.32</td>\n",
       "      <td>42.6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       SEQN  Gender   Age  Race  Education  Marriage  DMDHHSIZ  Poverty_Index  \\\n",
       "0  130378.0     1.0  43.0     4        5.0       1.0       4.0           5.00   \n",
       "1  130379.0     1.0  66.0     1        5.0       1.0       2.0           5.00   \n",
       "2  130380.0     2.0  44.0     3        3.0       1.0       7.0           1.41   \n",
       "5  130386.0     1.0  34.0     3        4.0       1.0       3.0           1.33   \n",
       "6  130387.0     2.0  68.0     1        5.0       3.0       1.0           1.32   \n",
       "\n",
       "    BMI  \n",
       "0  27.0  \n",
       "1  33.5  \n",
       "2  29.7  \n",
       "5  30.2  \n",
       "6  42.6  "
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "merge_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(8501, 6)\n"
     ]
    }
   ],
   "source": [
    "# extract hypertension status \n",
    "variables_meta[\"BPQ020\"] =  {\n",
    "    \"Description\": \"Ever told you had high blood pressure\",\n",
    "    \"Values\": {\n",
    "      \"1\": \"Yes\",\n",
    "      \"2\": \"No\",\n",
    "      \"7\": \"Refused\",\n",
    "      \"9\": \"Don't know\",\n",
    "      \".\": \"Missing\"\n",
    "    }\n",
    "  } \n",
    "\n",
    "df_htn = pd.read_sas(os.path.join(data_path, \"BPchol_L.XPT\"))\n",
    "print(df_htn.shape)\n",
    "df_htn = df_htn[['SEQN', 'BPQ020']]\n",
    "df_htn = df_htn[(df_htn['BPQ020'] == 1) | (df_htn['BPQ020'] == 2)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(11744, 9)\n",
      "(11736, 2)\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>SEQN</th>\n",
       "      <th>DIQ010</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>130378.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>130379.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>130380.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>130381.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>130382.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       SEQN  DIQ010\n",
       "0  130378.0     2.0\n",
       "1  130379.0     2.0\n",
       "2  130380.0     1.0\n",
       "3  130381.0     2.0\n",
       "4  130382.0     2.0"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# extract DM status \n",
    "variables_meta[\"DIQ010\"] = {  \"Description\": \"Doctor told you have diabetes\",\n",
    "    \"Values\": {\n",
    "      \"1\": \"Yes\",\n",
    "      \"2\": \"No\",\n",
    "      \"3\": \"Borderline\",\n",
    "      \"7\": \"Refused\",\n",
    "      \"9\": \"Don't know\",\n",
    "      \".\": \"Missing\"\n",
    "    }\n",
    "  }\n",
    "df_dm = pd.read_sas(os.path.join(data_path, \"DiabetesQ_L.XPT\"))\n",
    "print(df_dm.shape)\n",
    "df_dm = df_dm[['SEQN', 'DIQ010']]\n",
    "df_dm.loc[df_dm['DIQ010'] == 3, 'DIQ010'] = 1\n",
    "df_dm = df_dm[(df_dm['DIQ010'] == 1) | (df_dm['DIQ010'] == 2)]\n",
    "print(df_dm.shape)\n",
    "df_dm.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [],
   "source": [
    "# extract health insurance status\n",
    "df_hiq = pd.read_sas(os.path.join(data_path, \"HIQ_L.XPT\"))\n",
    "df_hiq = df_hiq[['SEQN', 'HIQ011']]\n",
    "df_hiq = df_hiq[(df_hiq['HIQ011'] == 1) | (df_hiq['HIQ011'] == 2)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [],
   "source": [
    "# extract CKD status\n",
    "df_ckd = pd.read_sas(os.path.join(data_path, \"KIQ_U_L.XPT\"))\n",
    "df_ckd = df_ckd[['SEQN', 'KIQ022']]\n",
    "df_ckd = df_ckd[(df_ckd['KIQ022'] == 1) | (df_ckd['KIQ022'] == 2)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [],
   "source": [
    "# merge with the main dataframe\n",
    "merge_df = pd.merge(merge_df, df_htn, on='SEQN')\n",
    "merge_df = pd.merge(merge_df, df_dm, on='SEQN')\n",
    "merge_df = pd.merge(merge_df, df_hiq, on='SEQN')\n",
    "merge_df = pd.merge(merge_df, df_ckd, on='SEQN')\n",
    "merge_df = merge_df.rename(columns={\n",
    "    'BPQ020': 'Hypertension',\n",
    "    'DIQ010': 'Diabetes',\n",
    "    'HIQ011': 'Health_Insurance',\n",
    "    'KIQ022': 'CKD',\n",
    "    'DMDHHSIZ': 'Household_Size',\n",
    "})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [],
   "source": [
    "# drop na values and drop SEQN column\n",
    "merge_df = merge_df.drop(columns=['SEQN', 'Household_Size'])\n",
    "merge_df = merge_df.dropna()\n",
    "bmi = merge_df['BMI']\n",
    "merge_df = merge_df.drop(columns=['BMI'])\n",
    "merge_df['BMI'] = bmi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "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>Gender</th>\n",
       "      <th>Age</th>\n",
       "      <th>Race</th>\n",
       "      <th>Education</th>\n",
       "      <th>Marriage</th>\n",
       "      <th>Poverty_Index</th>\n",
       "      <th>Hypertension</th>\n",
       "      <th>Diabetes</th>\n",
       "      <th>Health_Insurance</th>\n",
       "      <th>CKD</th>\n",
       "      <th>BMI</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>4</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>27.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>66.0</td>\n",
       "      <td>1</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.00</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>33.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.0</td>\n",
       "      <td>44.0</td>\n",
       "      <td>3</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.41</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>29.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>34.0</td>\n",
       "      <td>3</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.33</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>30.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2.0</td>\n",
       "      <td>68.0</td>\n",
       "      <td>1</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.32</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>42.6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Gender   Age  Race  Education  Marriage  Poverty_Index  Hypertension  \\\n",
       "0     1.0  43.0     4        5.0       1.0           5.00           1.0   \n",
       "1     1.0  66.0     1        5.0       1.0           5.00           1.0   \n",
       "2     2.0  44.0     3        3.0       1.0           1.41           2.0   \n",
       "3     1.0  34.0     3        4.0       1.0           1.33           2.0   \n",
       "4     2.0  68.0     1        5.0       3.0           1.32           1.0   \n",
       "\n",
       "   Diabetes  Health_Insurance  CKD   BMI  \n",
       "0       2.0               1.0  2.0  27.0  \n",
       "1       2.0               1.0  2.0  33.5  \n",
       "2       1.0               1.0  2.0  29.7  \n",
       "3       2.0               1.0  2.0  30.2  \n",
       "4       2.0               1.0  2.0  42.6  "
      ]
     },
     "execution_count": 134,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "merge_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5169, 11)"
      ]
     },
     "execution_count": 135,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "merge_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjsAAAHFCAYAAAAUpjivAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8pXeV/AAAACXBIWXMAAA9hAAAPYQGoP6dpAAA/kUlEQVR4nO3de3zP9f//8fvbDm/DtmzYe6uZlVM1JMqpciaRahWKUCrlkDl8sVTUB5MKHyl9yjKSpgMdPiITrY+kWC2HfESNprZWzOYwG/P8/dHPu8/b0N7znve83K6Xy+ty8X6+nu/n6/F6WnbvdbQZY4wAAAAsqpK3CwAAAChPhB0AAGBphB0AAGBphB0AAGBphB0AAGBphB0AAGBphB0AAGBphB0AAGBphB0AAGBphB3gIjB79mzZbDbFxMR4uxTZbDbn4uPjo+rVq6tJkyYaPHiwNmzYUKL/7t27ZbPZlJSU5NZ2Fi9erFmzZrn1ndNta9KkSbLZbPrjjz/cGutsvv/+e02aNEm7d+8usW7gwIGqU6eOx7YFgLADXBRef/11SdK2bdv01Vdfebka6a677tKXX36pdevWKTk5Wf3799eGDRvUqlUrjRgxwqVveHi4vvzyS3Xv3t2tbZQl7JR1W+76/vvv9fTTT5827Dz55JNatmxZuW4fuNj4ersAAOVr06ZN+u6779S9e3ctX75ciYmJatGihVdrCgsLU8uWLZ2fu3btqri4OD388MOaPXu2GjZsqEcffVSSZLfbXfqWh+LiYh0/fvy8bOvvXHHFFV7dPmBFHNkBLC4xMVGSNG3aNLVu3VrJyck6cuRIiX579+7VXXfdpcDAQF1yySXq27evNm7ceNpTSJs2bVLPnj0VEhKiypUrq2nTpnr77bfPqU4fHx/NmTNHNWrU0HPPPedsP92ppd9//10PP/ywIiMjZbfbVbNmTbVp00arV6+WJLVr107Lly/Xnj17XE6b/e9406dP1+TJkxUdHS273a61a9ee9ZRZZmamYmNjFRQUpODgYPXr10+///67Sx+bzaZJkyaV+G6dOnU0cOBASVJSUpLuvvtuSVL79u2dtZ3c5ulOYx09elTx8fGKjo6Wv7+/Lr30Ug0dOlQHDhwosZ0ePXpo5cqVuvbaaxUQEKCGDRs6j+wBFyvCDmBhBQUFeuutt3TdddcpJiZGDzzwgA4ePKh33nnHpd/hw4fVvn17rV27Vs8++6zefvtthYWFqXfv3iXGXLt2rdq0aaMDBw7olVde0QcffKBrrrlGvXv3dvu6mlMFBASoU6dOysjI0N69e8/Y77777tP777+vp556SqtWrdK8efPUqVMn7du3T5L08ssvq02bNnI4HPryyy+dy/+aPXu21qxZo+eff14rVqxQw4YNz1rbHXfcobp16+rdd9/VpEmT9P7776tr1646duyYW/vYvXt3TZ06VZL00ksvOWs706kzY4xuv/12Pf/887rvvvu0fPlyjRo1SgsWLFCHDh1UWFjo0v+7777T6NGjNXLkSH3wwQdq3LixBg0apM8//9ytOgFLMQAsa+HChUaSeeWVV4wxxhw8eNBUq1bN3HjjjS79XnrpJSPJrFixwqV98ODBRpKZP3++s61hw4amadOm5tixYy59e/ToYcLDw01xcfFZa5Jkhg4desb148aNM5LMV199ZYwxJiMjo0QN1apVM3FxcWfdTvfu3U1UVFSJ9pPjXXHFFaaoqOi06/53WxMnTjSSzMiRI136vvnmm0aSWbRokcu+TZw4scQ2o6KizIABA5yf33nnHSPJrF27tkTfAQMGuNS9cuVKI8lMnz7dpd+SJUuMJPPqq6+6bKdy5cpmz549zraCggITEhJiBg8eXGJbwMWCIzuAhSUmJiogIEB9+vSRJFWrVk133323/vOf/2jnzp3OfqmpqQoMDNTNN9/s8v177rnH5fOuXbv03//+V3379pUkHT9+3LnccsstysrK0o4dO86pZmPM3/a5/vrrlZSUpMmTJ2vDhg1uH12RpJ49e8rPz6/U/U/u80m9evWSr6+v1q5d6/a23bFmzRpJcp4GO+nuu+9W1apV9emnn7q0X3PNNapdu7bzc+XKlVW/fn3t2bOnXOsEKjLCDmBRu3bt0ueff67u3bvLGKMDBw7owIEDuuuuuyTJ5TqOffv2KSwsrMQYp7b99ttvkqQxY8bIz8/PZRkyZIgknfMt2id/KUdERJyxz5IlSzRgwADNmzdPrVq1UkhIiPr376/s7OxSbyc8PNytuhwOh8tnX19fhYaGOk+dlZd9+/bJ19dXNWvWdGm32WxyOBwlth8aGlpiDLvdroKCgnKtE6jICDuARb3++usyxujdd99V9erVncvJa0MWLFig4uJiSX/+gjwZZP7XqeGhRo0akqT4+Hht3LjxtMs111xT5poLCgq0evVqXXHFFbrsssvO2K9GjRqaNWuWdu/erT179ighIUFLly4tcfTjbE5esFxap87F8ePHtW/fPpdwYbfbS1xDI+mcAlFoaKiOHz9e4mJoY4yys7OdfycAzoywA1hQcXGxFixYoCuuuEJr164tsYwePVpZWVlasWKFJKlt27Y6ePCg8/NJycnJLp8bNGigevXq6bvvvlPz5s1PuwQGBpa55mHDhmnfvn0aN25cqb9Xu3ZtDRs2TJ07d9Y333zjbPf00Yw333zT5fPbb7+t48ePq127ds62OnXqaPPmzS791qxZo0OHDrm02e12SSpVfR07dpQkLVq0yKX9vffe0+HDh53rAZwZz9kBLGjFihX69ddf9eyzz7r8Mj4pJiZGc+bMUWJionr06KEBAwZo5syZ6tevnyZPnqy6detqxYoV+uSTTyRJlSr99f9F//rXv9StWzd17dpVAwcO1KWXXqr9+/dr+/bt+uabb0rc6XU6v/32mzZs2CBjjA4ePKitW7dq4cKF+u677zRy5Eg99NBDZ/xuXl6e2rdvr3vvvVcNGzZUYGCgNm7cqJUrVyo2NtbZr1GjRlq6dKnmzp2rZs2aqVKlSmrevLkbs+hq6dKl8vX1VefOnbVt2zY9+eSTatKkiXr16uXsc9999+nJJ5/UU089pbZt2+r777/XnDlzFBwc7DLWySdZv/rqqwoMDFTlypUVHR192lNQnTt3VteuXTVu3Djl5+erTZs22rx5syZOnKimTZvqvvvuK/M+ARcNr14eDaBc3H777cbf39/k5OScsU+fPn2Mr6+vyc7ONsYY8/PPP5vY2FhTrVo1ExgYaO68807z8ccfG0nmgw8+cPnud999Z3r16mVq1apl/Pz8jMPhMB06dHDe9XU2kpxLpUqVTFBQkGnUqJF5+OGHzZdfflmi/6l3SB09etQ88sgjpnHjxiYoKMgEBASYBg0amIkTJ5rDhw87v7d//35z1113mUsuucTYbDZz8p+7k+M999xzf7stY/66GystLc3ceuutzvm55557zG+//eby/cLCQjN27FgTGRlpAgICTNu2bU16enqJu7GMMWbWrFkmOjra+Pj4uGzz1LuxjPnzjqpx48aZqKgo4+fnZ8LDw82jjz5qcnNzXfpFRUWZ7t27l9ivtm3bmrZt25ZoBy4WNmNKcesDgIvS1KlT9cQTT+jnn38+6zU0AFCRcRoLgCRpzpw5kqSGDRvq2LFjWrNmjWbPnq1+/foRdABc0Ag7ACRJVapU0cyZM7V7924VFhaqdu3aGjdunJ544glvlwYA54TTWAAAwNK49RwAAFgaYQcAAFgaYQcAAFgaFyhLOnHihH799VcFBga6/Qh5AADgHeb/P5g0IiLC5eGnpyLsSPr1118VGRnp7TIAAEAZZGZmnvURGYQdyfkun8zMTAUFBXm5GgAAUBr5+fmKjIz823fyEXb019uPg4KCCDsAAFxg/u4SFC5QBgAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlubr7QIAAEDFUGf88nIZd/e07uUybmlxZAcAAFgaYQcAAFgaYQcAAFgaYQcAAFgaYQcAAFgaYQcAAFgaYQcAAFgaYQcAAFhahQk7CQkJstlsiouLc7YZYzRp0iRFREQoICBA7dq107Zt21y+V1hYqOHDh6tGjRqqWrWqevbsqb17957n6gEAQEVVIcLOxo0b9eqrr6px48Yu7dOnT9eMGTM0Z84cbdy4UQ6HQ507d9bBgwedfeLi4rRs2TIlJydr3bp1OnTokHr06KHi4uLzvRsAAKAC8nrYOXTokPr27avXXntN1atXd7YbYzRr1ixNmDBBsbGxiomJ0YIFC3TkyBEtXrxYkpSXl6fExES98MIL6tSpk5o2bapFixZpy5YtWr16tbd2CQAAVCBeDztDhw5V9+7d1alTJ5f2jIwMZWdnq0uXLs42u92utm3bav369ZKktLQ0HTt2zKVPRESEYmJinH1Op7CwUPn5+S4LAACwJq++CDQ5OVnffPONNm7cWGJddna2JCksLMylPSwsTHv27HH28ff3dzkidLLPye+fTkJCgp5++ulzLR8AAFwAvHZkJzMzUyNGjNCiRYtUuXLlM/az2Wwun40xJdpO9Xd94uPjlZeX51wyMzPdKx4AAFwwvBZ20tLSlJOTo2bNmsnX11e+vr5KTU3V7Nmz5evr6zyic+oRmpycHOc6h8OhoqIi5ebmnrHP6djtdgUFBbksAADAmrwWdjp27KgtW7YoPT3duTRv3lx9+/ZVenq6Lr/8cjkcDqWkpDi/U1RUpNTUVLVu3VqS1KxZM/n5+bn0ycrK0tatW519AADAxc1r1+wEBgYqJibGpa1q1aoKDQ11tsfFxWnq1KmqV6+e6tWrp6lTp6pKlSq69957JUnBwcEaNGiQRo8erdDQUIWEhGjMmDFq1KhRiQueAQDAxcmrFyj/nbFjx6qgoEBDhgxRbm6uWrRooVWrVikwMNDZZ+bMmfL19VWvXr1UUFCgjh07KikpST4+Pl6sHAAAVBQ2Y4zxdhHelp+fr+DgYOXl5XH9DgCUUZ3xy8tt7N3Tupfb2PhLef0dltffX2l/f3v9OTsAAADlibADAAAsjbADAAAsjbADAAAsjbADAAAsjbADAAAsjbADAAAsjbADAAAsjbADAAAsjbADAAAsjbADAAAsjbADAAAsjbADAAAsjbADAAAsjbADAAAsjbADAAAsjbADAAAszdfbBQAAzq8645d7uwTgvCLsAAAuWuUZ/HZP615uY8M9hB0AQIXH0SicC67ZAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlkbYAQAAlubVsDN37lw1btxYQUFBCgoKUqtWrbRixQrn+oEDB8pms7ksLVu2dBmjsLBQw4cPV40aNVS1alX17NlTe/fuPd+7AgAAKiivhp3LLrtM06ZN06ZNm7Rp0yZ16NBBt912m7Zt2+bsc/PNNysrK8u5fPzxxy5jxMXFadmyZUpOTta6det06NAh9ejRQ8XFxed7dwAAQAXk1bee33rrrS6fp0yZorlz52rDhg26+uqrJUl2u10Oh+O038/Ly1NiYqLeeOMNderUSZK0aNEiRUZGavXq1eratWv57gAAAKjwKsw1O8XFxUpOTtbhw4fVqlUrZ/tnn32mWrVqqX79+nrooYeUk5PjXJeWlqZjx46pS5cuzraIiAjFxMRo/fr157V+AABQMXn1yI4kbdmyRa1atdLRo0dVrVo1LVu2TFdddZUkqVu3brr77rsVFRWljIwMPfnkk+rQoYPS0tJkt9uVnZ0tf39/Va9e3WXMsLAwZWdnn3GbhYWFKiwsdH7Oz88vn50DAABe5/Ww06BBA6Wnp+vAgQN67733NGDAAKWmpuqqq65S7969nf1iYmLUvHlzRUVFafny5YqNjT3jmMYY2Wy2M65PSEjQ008/7dH9AAAAFZPXT2P5+/urbt26at68uRISEtSkSRP985//PG3f8PBwRUVFaefOnZIkh8OhoqIi5ebmuvTLyclRWFjYGbcZHx+vvLw855KZmem5HQIAABWK18POqYwxLqeY/te+ffuUmZmp8PBwSVKzZs3k5+enlJQUZ5+srCxt3bpVrVu3PuM27Ha783b3kwsAALAmr57Gevzxx9WtWzdFRkbq4MGDSk5O1meffaaVK1fq0KFDmjRpku68806Fh4dr9+7devzxx1WjRg3dcccdkqTg4GANGjRIo0ePVmhoqEJCQjRmzBg1atTIeXcWAAC4uHk17Pz222+67777lJWVpeDgYDVu3FgrV65U586dVVBQoC1btmjhwoU6cOCAwsPD1b59ey1ZskSBgYHOMWbOnClfX1/16tVLBQUF6tixo5KSkuTj4+PFPQMAABWFV8NOYmLiGdcFBATok08++dsxKleurBdffFEvvviiJ0sDAAAWUeGu2QEAAPAkwg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0wg4AALA0X28XAACAFdUZv7xcxt09rXu5jGtlHNkBAACWRtgBAACWRtgBAACWRtgBAACW5tWwM3fuXDVu3FhBQUEKCgpSq1attGLFCud6Y4wmTZqkiIgIBQQEqF27dtq2bZvLGIWFhRo+fLhq1KihqlWrqmfPntq7d+/53hUAAFBBeTXsXHbZZZo2bZo2bdqkTZs2qUOHDrrtttucgWb69OmaMWOG5syZo40bN8rhcKhz5846ePCgc4y4uDgtW7ZMycnJWrdunQ4dOqQePXqouLjYW7sFAAAqEK+GnVtvvVW33HKL6tevr/r162vKlCmqVq2aNmzYIGOMZs2apQkTJig2NlYxMTFasGCBjhw5osWLF0uS8vLylJiYqBdeeEGdOnVS06ZNtWjRIm3ZskWrV6/25q4BAIAKosJcs1NcXKzk5GQdPnxYrVq1UkZGhrKzs9WlSxdnH7vdrrZt22r9+vWSpLS0NB07dsylT0REhGJiYpx9TqewsFD5+fkuCwAAsCavh50tW7aoWrVqstvteuSRR7Rs2TJdddVVys7OliSFhYW59A8LC3Ouy87Olr+/v6pXr37GPqeTkJCg4OBg5xIZGenhvQIAABWF18NOgwYNlJ6erg0bNujRRx/VgAED9P333zvX22w2l/7GmBJtp/q7PvHx8crLy3MumZmZ57YTAACgwvJ62PH391fdunXVvHlzJSQkqEmTJvrnP/8ph8MhSSWO0OTk5DiP9jgcDhUVFSk3N/eMfU7Hbrc77wA7uQAAAGvyetg5lTFGhYWFio6OlsPhUEpKinNdUVGRUlNT1bp1a0lSs2bN5Ofn59InKytLW7dudfYBAAAXN6++CPTxxx9Xt27dFBkZqYMHDyo5OVmfffaZVq5cKZvNpri4OE2dOlX16tVTvXr1NHXqVFWpUkX33nuvJCk4OFiDBg3S6NGjFRoaqpCQEI0ZM0aNGjVSp06dvLlrAACggvBq2Pntt9903333KSsrS8HBwWrcuLFWrlypzp07S5LGjh2rgoICDRkyRLm5uWrRooVWrVqlwMBA5xgzZ86Ur6+vevXqpYKCAnXs2FFJSUny8fHx1m4BAIAKxGaMMd4uwtvy8/MVHBysvLw8rt8BYHl1xi/3dgk4B7undS+3scvrZ6O8ai7t7+8Kd80OAACAJ5Up7Bw4cEDz5s1TfHy89u/fL0n65ptv9Msvv3i0OAAAgHPl9jU7mzdvVqdOnRQcHKzdu3froYceUkhIiJYtW6Y9e/Zo4cKF5VEnAABAmbh9ZGfUqFEaOHCgdu7cqcqVKzvbu3Xrps8//9yjxQEAAJwrt8POxo0bNXjw4BLtl1566Vlf0QAAAOANboedypUrn/bFmTt27FDNmjU9UhQAAICnuB12brvtNj3zzDM6duyYpD/fXfXzzz9r/PjxuvPOOz1eIAAAwLlwO+w8//zz+v3331WrVi0VFBSobdu2qlu3rgIDAzVlypTyqBEAAKDM3L4bKygoSOvWrdOaNWv0zTff6MSJE7r22mt5PQMAAKiQyvy6iA4dOqhDhw6erAUAAMDjShV2Zs+eXeoBH3vssTIXAwAA4GmlCjszZ84s1WA2m42wAwAAKpRShZ2MjIzyrgMAAKBcnNOLQI0x4qXpAACgIitT2ElMTFRMTIwqV66sypUrKyYmRvPmzfN0bQAAAOfM7buxnnzySc2cOVPDhw9Xq1atJElffvmlRo4cqd27d2vy5MkeLxIAAKCs3A47c+fO1WuvvaZ77rnH2dazZ081btxYw4cPJ+wAAIAKxe3TWMXFxWrevHmJ9mbNmun48eMeKQoAAMBT3A47/fr109y5c0u0v/rqq+rbt69HigIAAPCUMj1BOTExUatWrVLLli0lSRs2bFBmZqb69++vUaNGOfvNmDHDM1UCwEWmzvjl3i4BsAy3w87WrVt17bXXSpJ+/PFHSVLNmjVVs2ZNbd261dnPZrN5qEQAAICyczvsrF27tjzqAAAAKBfn9FDBvXv36pdffvFULQAAAB7ndtg5ceKEnnnmGQUHBysqKkq1a9fWJZdcon/84x86ceJEedQIAABQZm6fxpowYYISExM1bdo0tWnTRsYYffHFF5o0aZKOHj2qKVOmlEedAAAAZeJ22FmwYIHmzZunnj17OtuaNGmiSy+9VEOGDCHsAACACsXtsLN//341bNiwRHvDhg21f/9+jxQFlEV53aq7e1r3chkXAHB+uH3NTpMmTTRnzpwS7XPmzFGTJk08UhQAAICnuH1kZ/r06erevbtWr16tVq1ayWazaf369crMzNTHH39cHjUCAACUmdtHdtq2basffvhBd9xxhw4cOKD9+/crNjZWO3bs0I033lgeNQIAAJRZmV4XERERwYXIAADgglCmhwr+5z//Ub9+/dS6dWvnQwXfeOMNrVu3zq1xEhISdN111ykwMFC1atXS7bffrh07drj0GThwoGw2m8ty8p1cJxUWFmr48OGqUaOGqlatqp49e2rv3r1l2TUAAGAxboed9957T127dlVAQIC++eYbFRYWSpIOHjyoqVOnujVWamqqhg4dqg0bNiglJUXHjx9Xly5ddPjwYZd+N998s7KyspzLqdcGxcXFadmyZUpOTta6det06NAh9ejRQ8XFxe7uHgAAsBi3T2NNnjxZr7zyivr376/k5GRne+vWrfXMM8+4NdbKlStdPs+fP1+1atVSWlqabrrpJme73W6Xw+E47Rh5eXlKTEzUG2+8oU6dOkmSFi1apMjISK1evVpdu3Z1qyYAAGAtbh/Z2bFjh0sQOSkoKEgHDhw4p2Ly8vIkSSEhIS7tn332mWrVqqX69evroYceUk5OjnNdWlqajh07pi5dujjbIiIiFBMTo/Xr1592O4WFhcrPz3dZAACANbkddsLDw7Vr164S7evWrdPll19e5kKMMRo1apRuuOEGxcTEONu7deumN998U2vWrNELL7ygjRs3qkOHDs7TZ9nZ2fL391f16tVdxgsLC1N2dvZpt5WQkKDg4GDnEhkZWea6AQBAxeb2aazBgwdrxIgRev3112Wz2fTrr7/qyy+/1JgxY/TUU0+VuZBhw4Zp8+bNJS5y7t27t/PPMTExat68uaKiorR8+XLFxsaecTxjjGw222nXxcfHa9SoUc7P+fn5BB4AACzK7bAzduxY5eXlqX379jp69Khuuukm2e12jRkzRsOGDStTEcOHD9eHH36ozz//XJdddtlZ+4aHhysqKko7d+6UJDkcDhUVFSk3N9fl6E5OTo5at2592jHsdrvsdnuZagUAABeWMt16PmXKFP3xxx/6+uuvtWHDBv3+++/6xz/+4fY4xhgNGzZMS5cu1Zo1axQdHf2339m3b58yMzMVHh4uSWrWrJn8/PyUkpLi7JOVlaWtW7eeMewAAICLR6nDzpEjRzR06FBdeumlqlWrlh588EHVqVNH119/vapVq1amjQ8dOlSLFi3S4sWLFRgYqOzsbGVnZ6ugoECSdOjQIY0ZM0Zffvmldu/erc8++0y33nqratSooTvuuEOSFBwcrEGDBmn06NH69NNP9e2336pfv35q1KiR8+4sAABw8Sr1aayJEycqKSlJffv2VeXKlfXWW2/p0Ucf1TvvvFPmjc+dO1eS1K5dO5f2+fPna+DAgfLx8dGWLVu0cOFCHThwQOHh4Wrfvr2WLFmiwMBAZ/+ZM2fK19dXvXr1UkFBgTp27KikpCT5+PiUuTYAAGANpQ47S5cuVWJiovr06SNJ6tevn9q0aaPi4uIyhwpjzFnXBwQE6JNPPvnbcSpXrqwXX3xRL774YpnqAAAA1lXq01iZmZkuL/q8/vrr5evrq19//bVcCgMAAPCEUoed4uJi+fv7u7T5+vrq+PHjHi8KAADAU0p9GssYo4EDB7rcsn306FE98sgjqlq1qrNt6dKlnq0QAADgHJQ67AwYMKBEW79+/TxaDAAAgKeVOuzMnz+/POsAAAAoF2V6qCAAAMCFgrADAAAsjbADAAAsze0XgQIA/lRn/HJvlwCgFEp1ZOfaa69Vbm6uJOmZZ57RkSNHyrUoAAAATylV2Nm+fbsOHz4sSXr66ad16NChci0KAADAU0p1Guuaa67R/fffrxtuuEHGGD3//PNnfNP5U0895dECAbivPE+v7J7WvdzGBoDyUKqwk5SUpIkTJ+rf//63bDabVqxYIV/fkl+12WyEHQAAUKGUKuw0aNBAycnJkqRKlSrp008/Va1atcq1MFgTF3QCAM43t+/GOnHiRHnUAQAAUC7KdOv5jz/+qFmzZmn79u2y2Wy68sorNWLECF1xxRWerg8AAOCcuP1QwU8++URXXXWVvv76azVu3FgxMTH66quvdPXVVyslJaU8agQAACgzt4/sjB8/XiNHjtS0adNKtI8bN06dO3f2WHEAAADnyu0jO9u3b9egQYNKtD/wwAP6/vvvPVIUAACAp7gddmrWrKn09PQS7enp6dyhBQAAKhy3T2M99NBDevjhh/XTTz+pdevWstlsWrdunZ599lmNHj26PGoELItb8QG4i3833Od22HnyyScVGBioF154QfHx8ZKkiIgITZo0SY899pjHCwQAADgXbocdm82mkSNHauTIkTp48KAkKTAw0OOFAQAAeEKZnrNzEiEHAABUdOcUdoCLAefHL3z8HQIXN7fvxgIAALiQEHYAAICluRV2jh07pvbt2+uHH34or3oAAAA8yq2w4+fnp61bt8pms5VXPQAAAB7l9mms/v37KzExsTxqAQAA8Di378YqKirSvHnzlJKSoubNm6tq1aou62fMmOGx4gAAAM6V20d2tm7dqmuvvVZBQUH64Ycf9O233zqX070z62wSEhJ03XXXKTAwULVq1dLtt9+uHTt2uPQxxmjSpEmKiIhQQECA2rVrp23btrn0KSws1PDhw1WjRg1VrVpVPXv21N69e93dNQAAYEFuH9lZu3atxzaempqqoUOH6rrrrtPx48c1YcIEdenSRd9//73ziNH06dM1Y8YMJSUlqX79+po8ebI6d+6sHTt2OB9qGBcXp48++kjJyckKDQ3V6NGj1aNHD6WlpcnHx8dj9QIAgAtPmR8quGvXLv3444+66aabFBAQIGOM2xcur1y50uXz/PnzVatWLaWlpemmm26SMUazZs3ShAkTFBsbK0lasGCBwsLCtHjxYg0ePFh5eXlKTEzUG2+8oU6dOkmSFi1apMjISK1evVpdu3Yt6y4CAAALcPs01r59+9SxY0fVr19ft9xyi7KysiRJDz744Dm/9TwvL0+SFBISIknKyMhQdna2unTp4uxjt9vVtm1brV+/XpKUlpamY8eOufSJiIhQTEyMs8+pCgsLlZ+f77IAAABrcvvIzsiRI+Xn56eff/5ZV155pbO9d+/eGjlypF544YUyFWKM0ahRo3TDDTcoJiZGkpSdnS1JCgsLc+kbFhamPXv2OPv4+/urevXqJfqc/P6pEhIS9PTTT5epTuBiV16vXtg9rXu5jAsAbh/ZWbVqlZ599llddtllLu316tVzBpCyGDZsmDZv3qy33nqrxLpTT4+V5pTZ2frEx8crLy/PuWRmZpa5bgAAULG5HXYOHz6sKlWqlGj/448/ZLfby1TE8OHD9eGHH2rt2rUuIcrhcEhSiSM0OTk5zqM9DodDRUVFys3NPWOfU9ntdgUFBbksAADAmtwOOzfddJMWLlzo/Gyz2XTixAk999xzat++vVtjGWM0bNgwLV26VGvWrFF0dLTL+ujoaDkcDqWkpDjbioqKlJqaqtatW0uSmjVrJj8/P5c+WVlZ2rp1q7MPAAC4eLl9zc5zzz2ndu3aadOmTSoqKtLYsWO1bds27d+/X1988YVbYw0dOlSLFy/WBx98oMDAQOcRnODgYAUEBMhmsykuLk5Tp05VvXr1VK9ePU2dOlVVqlTRvffe6+w7aNAgjR49WqGhoQoJCdGYMWPUqFEj591ZAADg4uV22Lnqqqu0efNmzZ07Vz4+Pjp8+LBiY2M1dOhQhYeHuzXW3LlzJUnt2rVzaZ8/f74GDhwoSRo7dqwKCgo0ZMgQ5ebmqkWLFlq1apXzGTuSNHPmTPn6+qpXr14qKChQx44dlZSUxDN2AACAbMYY4+0ivC0/P1/BwcHKy8vj+p1yVl538uDCV553Y/FzB3hXef33Xdrf32V6qGBubq4SExO1fft22Ww2XXnllbr//vudz8cBAACoKNy+QDk1NVXR0dGaPXu2cnNztX//fs2ePVvR0dFKTU0tjxoBAADKzO0jO0OHDlWvXr2c1+xIUnFxsYYMGaKhQ4dq69atHi8SAACgrNw+svPjjz9q9OjRLhf/+vj4aNSoUfrxxx89WhwAAMC5cjvsXHvttdq+fXuJ9u3bt+uaa67xRE0AAAAeU6rTWJs3b3b++bHHHtOIESO0a9cutWzZUpK0YcMGvfTSS5o2bVr5VAkAAFBGpQo711xzjWw2m/73LvWxY8eW6Hfvvfeqd+/enqsOAADgHJUq7GRkZJR3HQAAAOWiVGEnKiqqvOsAAAAoF2V6qOAvv/yiL774Qjk5OTpx4oTLuscee8wjhQEAAHiC22Fn/vz5euSRR+Tv76/Q0FDZbDbnOpvNRtgBAAAVitth56mnntJTTz2l+Ph4Vark9p3rAHBavL8KQHlxO60cOXJEffr0IegAAIALgtuJZdCgQXrnnXfKoxYAAACPc/s0VkJCgnr06KGVK1eqUaNG8vPzc1k/Y8YMjxUHAABwrtwOO1OnTtUnn3yiBg0aSFKJC5RhDVw/AQCwCrfDzowZM/T6669r4MCB5VAOAACAZ7l9zY7dblebNm3KoxYAAACPczvsjBgxQi+++GJ51AIAAOBxbp/G+vrrr7VmzRr9+9//1tVXX13iAuWlS5d6rDgAAIBz5XbYueSSSxQbG1setQAAAHhcmV4XAQAAcKHgMcgAAMDS3D6yEx0dfdbn6fz000/nVBAAAIAnuR124uLiXD4fO3ZM3377rVauXKn/+7//81RdAAAAHuF22BkxYsRp21966SVt2rTpnAsCAADwJJsxxnhioJ9++knXXHON8vPzPTHceZWfn6/g4GDl5eUpKCjI2+WUGq90AABcCHZP614u45b297fHLlB+9913FRIS4qnhAAAAPMLt01hNmzZ1uUDZGKPs7Gz9/vvvevnllz1aHAAAwLlyO+zcfvvtLp8rVaqkmjVrql27dmrYsKGn6gIAAPAIt8POxIkTy6MOAACAcsFDBQEAgKWVOuxUqlRJPj4+Z118fd07UPT555/r1ltvVUREhGw2m95//32X9QMHDpTNZnNZWrZs6dKnsLBQw4cPV40aNVS1alX17NlTe/fudasOAABgXaVOJ8uWLTvjuvXr1+vFF1+Uu3exHz58WE2aNNH999+vO++887R9br75Zpf3cfn7+7usj4uL00cffaTk5GSFhoZq9OjR6tGjh9LS0uTj4+NWPQAAwHpKHXZuu+22Em3//e9/FR8fr48++kh9+/bVP/7xD7c23q1bN3Xr1u2sfex2uxwOx2nX5eXlKTExUW+88YY6deokSVq0aJEiIyO1evVqde3a1a16AACA9ZTpmp1ff/1VDz30kBo3bqzjx48rPT1dCxYsUO3atT1dnz777DPVqlVL9evX10MPPaScnBznurS0NB07dkxdunRxtkVERCgmJkbr168/45iFhYXKz893WQAAgDW5FXby8vI0btw41a1bV9u2bdOnn36qjz76SDExMeVSXLdu3fTmm29qzZo1euGFF7Rx40Z16NBBhYWFkqTs7Gz5+/urevXqLt8LCwtTdnb2GcdNSEhQcHCwc4mMjCyX+gEAgPeV+jTW9OnT9eyzz8rhcOitt9467WktT+vdu7fzzzExMWrevLmioqK0fPlyxcbGnvF7xpizvpk9Pj5eo0aNcn7Oz88n8AAAYFGlDjvjx49XQECA6tatqwULFmjBggWn7bd06VKPFXeq8PBwRUVFaefOnZIkh8OhoqIi5ebmuhzdycnJUevWrc84jt1ul91uL7c6AQBAxVHqsNO/f/+zHi05H/bt26fMzEyFh4dLkpo1ayY/Pz+lpKSoV69ekqSsrCxt3bpV06dP92apAACggih12ElKSvL4xg8dOqRdu3Y5P2dkZCg9PV0hISEKCQnRpEmTdOeddyo8PFy7d+/W448/rho1auiOO+6QJAUHB2vQoEEaPXq0QkNDFRISojFjxqhRo0bOu7MAAMDFze3XRXjSpk2b1L59e+fnk9fRDBgwQHPnztWWLVu0cOFCHThwQOHh4Wrfvr2WLFmiwMBA53dmzpwpX19f9erVSwUFBerYsaOSkpJ4xg4AAJAk2Yy7TwK0oPz8fAUHBysvL09BQUHeLqfU6oxf7u0SAAD4W7undS+XcUv7+9urR3YuBgQSAAC8ixeBAgAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAASyPsAAAAS/Nq2Pn888916623KiIiQjabTe+//77LemOMJk2apIiICAUEBKhdu3batm2bS5/CwkINHz5cNWrUUNWqVdWzZ0/t3bv3PO4FAACoyLwadg4fPqwmTZpozpw5p10/ffp0zZgxQ3PmzNHGjRvlcDjUuXNnHTx40NknLi5Oy5YtU3JystatW6dDhw6pR48eKi4uPl+7AQAAKjBfb268W7du6tat22nXGWM0a9YsTZgwQbGxsZKkBQsWKCwsTIsXL9bgwYOVl5enxMREvfHGG+rUqZMkadGiRYqMjNTq1avVtWvX87YvAACgYqqw1+xkZGQoOztbXbp0cbbZ7Xa1bdtW69evlySlpaXp2LFjLn0iIiIUExPj7HM6hYWFys/Pd1kAAIA1Vdiwk52dLUkKCwtzaQ8LC3Ouy87Olr+/v6pXr37GPqeTkJCg4OBg5xIZGenh6gEAQEVRYcPOSTabzeWzMaZE26n+rk98fLzy8vKcS2ZmpkdqBQAAFU+FDTsOh0OSShyhycnJcR7tcTgcKioqUm5u7hn7nI7dbldQUJDLAgAArKnChp3o6Gg5HA6lpKQ424qKipSamqrWrVtLkpo1ayY/Pz+XPllZWdq6dauzDwAAuLh59W6sQ4cOadeuXc7PGRkZSk9PV0hIiGrXrq24uDhNnTpV9erVU7169TR16lRVqVJF9957ryQpODhYgwYN0ujRoxUaGqqQkBCNGTNGjRo1ct6dBQAALm5eDTubNm1S+/btnZ9HjRolSRowYICSkpI0duxYFRQUaMiQIcrNzVWLFi20atUqBQYGOr8zc+ZM+fr6qlevXiooKFDHjh2VlJQkHx+f874/AACg4rEZY4y3i/C2/Px8BQcHKy8vz+PX79QZv9yj4wEAcKHZPa17uYxb2t/fFfaaHQAAAE8g7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEsj7AAAAEur0GFn0qRJstlsLovD4XCuN8Zo0qRJioiIUEBAgNq1a6dt27Z5sWIAAFDRVOiwI0lXX321srKynMuWLVuc66ZPn64ZM2Zozpw52rhxoxwOhzp37qyDBw96sWIAAFCRVPiw4+vrK4fD4Vxq1qwp6c+jOrNmzdKECRMUGxurmJgYLViwQEeOHNHixYu9XDUAAKgoKnzY2blzpyIiIhQdHa0+ffrop59+kiRlZGQoOztbXbp0cfa12+1q27at1q9ff9YxCwsLlZ+f77IAAABrqtBhp0WLFlq4cKE++eQTvfbaa8rOzlbr1q21b98+ZWdnS5LCwsJcvhMWFuZcdyYJCQkKDg52LpGRkeW2DwAAwLsqdNjp1q2b7rzzTjVq1EidOnXS8uXLJUkLFixw9rHZbC7fMcaUaDtVfHy88vLynEtmZqbniwcAABVChQ47p6pataoaNWqknTt3Ou/KOvUoTk5OTomjPaey2+0KCgpyWQAAgDVdUGGnsLBQ27dvV3h4uKKjo+VwOJSSkuJcX1RUpNTUVLVu3dqLVQIAgIrE19sFnM2YMWN06623qnbt2srJydHkyZOVn5+vAQMGyGazKS4uTlOnTlW9evVUr149TZ06VVWqVNG9997r7dIBAEAFUaHDzt69e3XPPffojz/+UM2aNdWyZUtt2LBBUVFRkqSxY8eqoKBAQ4YMUW5urlq0aKFVq1YpMDDQy5UDAICKwmaMMd4uwtvy8/MVHBysvLw8j1+/U2f8co+OBwDAhWb3tO7lMm5pf39fUNfsAAAAuIuwAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALI2wAwAALM0yYefll19WdHS0KleurGbNmuk///mPt0sCAAAVgCXCzpIlSxQXF6cJEybo22+/1Y033qhu3brp559/9nZpAADAyywRdmbMmKFBgwbpwQcf1JVXXqlZs2YpMjJSc+fO9XZpAADAyy74sFNUVKS0tDR16dLFpb1Lly5av369l6oCAAAVha+3CzhXf/zxh4qLixUWFubSHhYWpuzs7NN+p7CwUIWFhc7PeXl5kqT8/HyP13ei8IjHxwQA4EJSHr9f/3dcY8xZ+13wYeckm83m8tkYU6LtpISEBD399NMl2iMjI8ulNgAALmbBs8p3/IMHDyo4OPiM6y/4sFOjRg35+PiUOIqTk5NT4mjPSfHx8Ro1apTz84kTJ7R//36FhoaeMSCVRX5+viIjI5WZmamgoCCPjWtVzFfpMVelx1yVHnNVesxV6ZXnXBljdPDgQUVERJy13wUfdvz9/dWsWTOlpKTojjvucLanpKTotttuO+137Ha77Ha7S9sll1xSbjUGBQXxH4MbmK/SY65Kj7kqPeaq9Jir0iuvuTrbEZ2TLviwI0mjRo3Sfffdp+bNm6tVq1Z69dVX9fPPP+uRRx7xdmkAAMDLLBF2evfurX379umZZ55RVlaWYmJi9PHHHysqKsrbpQEAAC+zRNiRpCFDhmjIkCHeLsOF3W7XxIkTS5wyw+kxX6XHXJUec1V6zFXpMVelVxHmymb+7n4tAACAC9gF/1BBAACAsyHsAAAASyPsAAAASyPsAAAASyPseEBCQoKuu+46BQYGqlatWrr99tu1Y8cOlz7GGE2aNEkREREKCAhQu3bttG3bNi9V7D1z585V48aNnQ+XatWqlVasWOFczzydWUJCgmw2m+Li4pxtzNefJk2aJJvN5rI4HA7neubJ1S+//KJ+/fopNDRUVapU0TXXXKO0tDTneubrT3Xq1Cnxc2Wz2TR06FBJzNP/On78uJ544glFR0crICBAl19+uZ555hmdOHHC2cer82Vwzrp27Wrmz59vtm7datLT00337t1N7dq1zaFDh5x9pk2bZgIDA817771ntmzZYnr37m3Cw8NNfn6+Fys//z788EOzfPlys2PHDrNjxw7z+OOPGz8/P7N161ZjDPN0Jl9//bWpU6eOady4sRkxYoSznfn608SJE83VV19tsrKynEtOTo5zPfP0l/3795uoqCgzcOBA89VXX5mMjAyzevVqs2vXLmcf5utPOTk5Lj9TKSkpRpJZu3atMYZ5+l+TJ082oaGh5t///rfJyMgw77zzjqlWrZqZNWuWs48354uwUw5ycnKMJJOammqMMebEiRPG4XCYadOmOfscPXrUBAcHm1deecVbZVYY1atXN/PmzWOezuDgwYOmXr16JiUlxbRt29YZdpivv0ycONE0adLktOuYJ1fjxo0zN9xwwxnXM19nNmLECHPFFVeYEydOME+n6N69u3nggQdc2mJjY02/fv2MMd7/ueI0VjnIy8uTJIWEhEiSMjIylJ2drS5dujj72O12tW3bVuvXr/dKjRVBcXGxkpOTdfjwYbVq1Yp5OoOhQ4eqe/fu6tSpk0s78+Vq586dioiIUHR0tPr06aOffvpJEvN0qg8//FDNmzfX3XffrVq1aqlp06Z67bXXnOuZr9MrKirSokWL9MADD8hmszFPp7jhhhv06aef6ocffpAkfffdd1q3bp1uueUWSd7/ubLME5QrCmOMRo0apRtuuEExMTGS5Hwj+6lvYQ8LC9OePXvOe43etmXLFrVq1UpHjx5VtWrVtGzZMl111VXOH3jm6S/Jycn65ptvtHHjxhLr+Ln6S4sWLbRw4ULVr19fv/32myZPnqzWrVtr27ZtzNMpfvrpJ82dO1ejRo3S448/rq+//lqPPfaY7Ha7+vfvz3ydwfvvv68DBw5o4MCBkvjv71Tjxo1TXl6eGjZsKB8fHxUXF2vKlCm65557JHl/vgg7HjZs2DBt3rxZ69atK7HOZrO5fDbGlGi7GDRo0EDp6ek6cOCA3nvvPQ0YMECpqanO9czTnzIzMzVixAitWrVKlStXPmM/5kvq1q2b88+NGjVSq1atdMUVV2jBggVq2bKlJObppBMnTqh58+aaOnWqJKlp06batm2b5s6dq/79+zv7MV+uEhMT1a1bN0VERLi0M09/WrJkiRYtWqTFixfr6quvVnp6uuLi4hQREaEBAwY4+3lrvjiN5UHDhw/Xhx9+qLVr1+qyyy5ztp+8K+Rksj0pJyenRMq9GPj7+6tu3bpq3ry5EhIS1KRJE/3zn/9knk6RlpamnJwcNWvWTL6+vvL19VVqaqpmz54tX19f55wwXyVVrVpVjRo10s6dO/m5OkV4eLiuuuoql7Yrr7xSP//8syT+vTqdPXv2aPXq1XrwwQedbcyTq//7v//T+PHj1adPHzVq1Ej33XefRo4cqYSEBEneny/CjgcYYzRs2DAtXbpUa9asUXR0tMv66OhoORwOpaSkONuKioqUmpqq1q1bn+9yKxxjjAoLC5mnU3Ts2FFbtmxRenq6c2nevLn69u2r9PR0XX755czXGRQWFmr79u0KDw/n5+oUbdq0KfFojB9++EFRUVGS+PfqdObPn69atWqpe/fuzjbmydWRI0dUqZJrpPDx8XHeeu71+Sr3S6AvAo8++qgJDg42n332mcttikeOHHH2mTZtmgkODjZLly41W7ZsMffcc89FeYtifHy8+fzzz01GRobZvHmzefzxx02lSpXMqlWrjDHM09/537uxjGG+Tho9erT57LPPzE8//WQ2bNhgevToYQIDA83u3buNMczT//r666+Nr6+vmTJlitm5c6d58803TZUqVcyiRYucfZivvxQXF5vatWubcePGlVjHPP1lwIAB5tJLL3Xeer506VJTo0YNM3bsWGcfb84XYccDJJ12mT9/vrPPiRMnzMSJE43D4TB2u93cdNNNZsuWLd4r2kseeOABExUVZfz9/U3NmjVNx44dnUHHGObp75wadpivP518Xoefn5+JiIgwsbGxZtu2bc71zJOrjz76yMTExBi73W4aNmxoXn31VZf1zNdfPvnkEyPJ7Nixo8Q65ukv+fn5ZsSIEaZ27dqmcuXK5vLLLzcTJkwwhYWFzj7enC+bMcaU//EjAAAA7+CaHQAAYGmEHQAAYGmEHQAAYGmEHQAAYGmEHQAAYGmEHQAAYGmEHQAAYGmEHQAAYGmEHQAXrPXr18vHx0c333yzt0sBUIHxBGUAF6wHH3xQ1apV07x58/T999+rdu3a3i4JQAXEkR0AF6TDhw/r7bff1qOPPqoePXooKSnJZf2HH36oevXqKSAgQO3bt9eCBQtks9l04MABZ5/169frpptuUkBAgCIjI/XYY4/p8OHD53dHAJQ7wg6AC9KSJUvUoEEDNWjQQP369dP8+fN18kD17t27ddddd+n2229Xenq6Bg8erAkTJrh8f8uWLeratatiY2O1efNmLVmyROvWrdOwYcO8sTsAyhGnsQBckNq0aaNevXppxIgROn78uMLDw/XWW2+pU6dOGj9+vJYvX64tW7Y4+z/xxBOaMmWKcnNzdckll6h///4KCAjQv/71L2efdevWqW3btjp8+LAqV67sjd0CUA44sgPggrNjxw59/fXX6tOnjyTJ19dXvXv31uuvv+5cf91117l85/rrr3f5nJaWpqSkJFWrVs25dO3aVSdOnFBGRsb52REA54WvtwsAAHclJibq+PHjuvTSS51txhj5+fkpNzdXxhjZbDaX75x6EPvEiRMaPHiwHnvssRLjc6EzYC2EHQAXlOPHj2vhwoV64YUX1KVLF5d1d955p9588001bNhQH3/8scu6TZs2uXy+9tprtW3bNtWtW7fcawbgXVyzA+CC8v7776t3797KyclRcHCwy7oJEybo448/1tKlS9WgQQONHDlSgwYNUnp6ukaPHq29e/fqwIEDCg4O1ubNm9WyZUvdf//9euihh1S1alVt375dKSkpevHFF720dwDKA9fsALigJCYmqlOnTiWCjvTnkZ309HTl5ubq3Xff1dKlS9W4cWPNnTvXeTeW3W6XJDVu3FipqanauXOnbrzxRjVt2lRPPvmkwsPDz+v+ACh/HNkBcFGYMmWKXnnlFWVmZnq7FADnGdfsALCkl19+Wdddd51CQ0P1xRdf6LnnnuMZOsBFirADwJJ27typyZMna//+/apdu7ZGjx6t+Ph4b5cFwAs4jQUAACyNC5QBAIClEXYAAIClEXYAAIClEXYAAIClEXYAAIClEXYAAIClEXYAAIClEXYAAIClEXYAAICl/T91kOXU0KxFUgAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# age \n",
    "plt.hist(merge_df['Age'], bins=20)\n",
    "plt.title('Age Distribution')\n",
    "plt.xlabel('Age')\n",
    "plt.ylabel('Number of People')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [],
   "source": [
    "# write to csv file\n",
    "merge_df.to_csv('NHANES/NHANES_2021_2023.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Write Json file to store selected meta data from NHANES 2021-2023"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 925,
   "metadata": {},
   "outputs": [],
   "source": [
    "# File path for the JSON file\n",
    "json_file_path = os.path.join(data_path, 'NHANES_metadata.json')\n",
    "\n",
    "# Write the dictionary to a JSON file\n",
    "with open(json_file_path, 'w') as json_file:\n",
    "    json.dump(variables_meta, json_file, indent=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "deep_learning",
   "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": "3.8.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}