4_LTC_Logistic_regression-female.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "fc28175d",
   "metadata": {},
   "source": [
    "# In this file LR is used for both balanced and imbalanced dataset.\n",
    "\n",
    "* resource :https://www.geeksforgeeks.org/logistic-regression-using-statsmodels/\n",
    "\n",
    "\n",
    "### The sensitivity with balanced class is improved to a great extent"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "f9c3dd0b",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "#Import the liberaries \n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "# visualization \n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "sns.set_style(\"whitegrid\")\n",
    "from sklearn.model_selection import train_test_split\n",
    "import statsmodels.api as sm \n",
    "from sklearn.metrics import confusion_matrix , accuracy_score, classification_report , roc_auc_score, f1_score,precision_score, recall_score\n",
    "from plot_metric.functions import BinaryClassification\n",
    "from imblearn.over_sampling import RandomOverSampler\n",
    "from sklearn.preprocessing import PowerTransformer"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "39ae8cc9",
   "metadata": {},
   "source": [
    "# Read the data after cleaning and pre-processing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "3e1e1ab0",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_final=pd.read_hdf(\"data_final.h5\", key='data_final')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "fcb9880f",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_final=data_final[data_final['patient_gender']=='F']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "2c263967",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['num_elixhauser_2_years_full', 'elix_index', 'num_admits_1_year',\n",
       "       'num_scu_admits_1_year', 'num_procs_dad_1_year',\n",
       "       'num_procs_nacrs_1_year', 'num_procs_tot_1_year', 'spec_pat_num_age',\n",
       "       'specimen_year_month_collected_con', 'collected_in_hospital',\n",
       "       'collected_in_ed', 'collected_in_ltc', 'collected_in_dsl',\n",
       "       'ltc_resident_during_collection', 'dsl_resident_during_collection',\n",
       "       'spec_pat_agecat', 'patient_gender', 'num_elixhauser_2_years_full_cat',\n",
       "       'symptomatic_during_collection', 'interp_result', 'y_m_',\n",
       "       'died_within_60_days', 'spec_pat_agecat_18-29', 'spec_pat_agecat_30-39',\n",
       "       'spec_pat_agecat_40-49', 'spec_pat_agecat_50-59',\n",
       "       'spec_pat_agecat_60-69', 'spec_pat_agecat_70-79', 'spec_pat_agecat_80+',\n",
       "       'patient_gender_F', 'patient_gender_M',\n",
       "       'num_elixhauser_2_years_full_cat_0',\n",
       "       'num_elixhauser_2_years_full_cat_1+', 'symptomatic_during_collection_N',\n",
       "       'symptomatic_during_collection_U', 'symptomatic_during_collection_Y',\n",
       "       'interp_result_Negative', 'interp_result_Positive', 'y_m__2020-10',\n",
       "       'y_m__2020-11', 'y_m__2020-12', 'y_m__2020-3', 'y_m__2020-4',\n",
       "       'y_m__2020-5', 'y_m__2020-6', 'y_m__2020-7', 'y_m__2020-8',\n",
       "       'y_m__2020-9', 'y_m__2021-1', 'y_m__2021-2', 'y_m__2021-3',\n",
       "       'chf_2_years_full_1', 'htn_unc_2_years_full_1',\n",
       "       'cancer_mets_2_years_full_1', 'fluid_elec_dis_2_years_full_1',\n",
       "       'card_arrh_2_years_full_1', 'valv_dis_2_years_full_1',\n",
       "       'pcd_2_years_full_1', 'pvd_2_years_full_1', 'htn_c_2_years_full_1',\n",
       "       'paral_2_years_full_1', 'oth_neur_dis_2_years_full_1',\n",
       "       'cpd_2_years_full_1', 'diab_2_years_full_1',\n",
       "       'hypothyroidism_2_years_full_1', 'ren_fail_2_years_full_1',\n",
       "       'liver_dis_2_years_full_1', 'pep_ulc_exc_bld_2_years_full_1',\n",
       "       'aids_hiv_2_years_full_1', 'lymph_2_years_full_1',\n",
       "       'tumour_no_mets_2_years_full_1', 'rheum_col_vasc_dis_2_years_full_1',\n",
       "       'coag_2_years_full_1', 'obes_2_years_full_1', 'wt_loss_2_years_full_1',\n",
       "       'anemia_2_years_full_1', 'alc_abuse_2_years_full_1',\n",
       "       'drug_abuse_2_years_full_1', 'psych_2_years_full_1',\n",
       "       'depress_2_years_full_1'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_final.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "6544b77f",
   "metadata": {},
   "outputs": [],
   "source": [
    "data_final = data_final.drop(['patient_gender', 'patient_gender_M', 'patient_gender_F'], axis=1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "a62263e7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['num_elixhauser_2_years_full', 'elix_index', 'num_admits_1_year',\n",
       "       'num_scu_admits_1_year', 'num_procs_dad_1_year',\n",
       "       'num_procs_nacrs_1_year', 'num_procs_tot_1_year', 'spec_pat_num_age',\n",
       "       'specimen_year_month_collected_con', 'collected_in_hospital',\n",
       "       'collected_in_ed', 'collected_in_ltc', 'collected_in_dsl',\n",
       "       'ltc_resident_during_collection', 'dsl_resident_during_collection',\n",
       "       'spec_pat_agecat', 'num_elixhauser_2_years_full_cat',\n",
       "       'symptomatic_during_collection', 'interp_result', 'y_m_',\n",
       "       'died_within_60_days', 'spec_pat_agecat_18-29', 'spec_pat_agecat_30-39',\n",
       "       'spec_pat_agecat_40-49', 'spec_pat_agecat_50-59',\n",
       "       'spec_pat_agecat_60-69', 'spec_pat_agecat_70-79', 'spec_pat_agecat_80+',\n",
       "       'num_elixhauser_2_years_full_cat_0',\n",
       "       'num_elixhauser_2_years_full_cat_1+', 'symptomatic_during_collection_N',\n",
       "       'symptomatic_during_collection_U', 'symptomatic_during_collection_Y',\n",
       "       'interp_result_Negative', 'interp_result_Positive', 'y_m__2020-10',\n",
       "       'y_m__2020-11', 'y_m__2020-12', 'y_m__2020-3', 'y_m__2020-4',\n",
       "       'y_m__2020-5', 'y_m__2020-6', 'y_m__2020-7', 'y_m__2020-8',\n",
       "       'y_m__2020-9', 'y_m__2021-1', 'y_m__2021-2', 'y_m__2021-3',\n",
       "       'chf_2_years_full_1', 'htn_unc_2_years_full_1',\n",
       "       'cancer_mets_2_years_full_1', 'fluid_elec_dis_2_years_full_1',\n",
       "       'card_arrh_2_years_full_1', 'valv_dis_2_years_full_1',\n",
       "       'pcd_2_years_full_1', 'pvd_2_years_full_1', 'htn_c_2_years_full_1',\n",
       "       'paral_2_years_full_1', 'oth_neur_dis_2_years_full_1',\n",
       "       'cpd_2_years_full_1', 'diab_2_years_full_1',\n",
       "       'hypothyroidism_2_years_full_1', 'ren_fail_2_years_full_1',\n",
       "       'liver_dis_2_years_full_1', 'pep_ulc_exc_bld_2_years_full_1',\n",
       "       'aids_hiv_2_years_full_1', 'lymph_2_years_full_1',\n",
       "       'tumour_no_mets_2_years_full_1', 'rheum_col_vasc_dis_2_years_full_1',\n",
       "       'coag_2_years_full_1', 'obes_2_years_full_1', 'wt_loss_2_years_full_1',\n",
       "       'anemia_2_years_full_1', 'alc_abuse_2_years_full_1',\n",
       "       'drug_abuse_2_years_full_1', 'psych_2_years_full_1',\n",
       "       'depress_2_years_full_1'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_final.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "c24802a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# these are the features that we have finilized to give the model \n",
    "\n",
    "feature_list_for_training=[\n",
    "'spec_pat_num_age',\n",
    "# Gender-male\n",
    "# 'patient_gender_M',\n",
    "# Specimen Collection location\n",
    "'collected_in_ed', 'collected_in_hospital', 'collected_in_dsl',\n",
    "# Resident During Collection\n",
    "'dsl_resident_during_collection',\n",
    "# Symptomatic During Collection\n",
    "'symptomatic_during_collection_U', 'symptomatic_during_collection_Y',\n",
    "# Result of the covid test\n",
    "'interp_result_Positive',\n",
    "# Specimen year-month collection\n",
    "'y_m__2020-10',\n",
    "'y_m__2020-11', 'y_m__2020-12', 'y_m__2020-3', 'y_m__2020-4',\n",
    "'y_m__2020-5', 'y_m__2021-3', 'y_m__2020-7', 'y_m__2020-8',\n",
    "'y_m__2020-9', 'y_m__2021-1', 'y_m__2021-2', \n",
    "# Num of Elixhauser \n",
    "'chf_2_years_full_1', 'htn_unc_2_years_full_1',\n",
    "'cancer_mets_2_years_full_1', 'fluid_elec_dis_2_years_full_1',\n",
    "'card_arrh_2_years_full_1', 'valv_dis_2_years_full_1',\n",
    "'pcd_2_years_full_1', 'pvd_2_years_full_1', 'htn_c_2_years_full_1',\n",
    "'paral_2_years_full_1', 'oth_neur_dis_2_years_full_1',\n",
    "'cpd_2_years_full_1', 'diab_2_years_full_1',\n",
    "'hypothyroidism_2_years_full_1', 'ren_fail_2_years_full_1',\n",
    "'liver_dis_2_years_full_1', 'pep_ulc_exc_bld_2_years_full_1',\n",
    "'aids_hiv_2_years_full_1', 'lymph_2_years_full_1',\n",
    "'tumour_no_mets_2_years_full_1', 'rheum_col_vasc_dis_2_years_full_1',\n",
    "'coag_2_years_full_1', 'obes_2_years_full_1', 'wt_loss_2_years_full_1',\n",
    "'anemia_2_years_full_1', 'alc_abuse_2_years_full_1',\n",
    "'drug_abuse_2_years_full_1', 'psych_2_years_full_1',\n",
    "'depress_2_years_full_1',\n",
    "#Num of Admits for 1 year\n",
    "'num_admits_1_year', 'num_scu_admits_1_year',\n",
    "#Num Procedures for 1 year\n",
    "'num_procs_dad_1_year', 'num_procs_nacrs_1_year',\n",
    "]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "496dd648",
   "metadata": {},
   "source": [
    "# trian the model by Imbalanced data "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "cfc4e69d",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/fuzzysha/software/miniconda3/envs/hproj/lib/python3.8/site-packages/statsmodels/tsa/tsatools.py:142: FutureWarning: In a future version of pandas all arguments of concat except for the argument 'objs' will be keyword-only.\n",
      "  x = pd.concat(x[::order], 1)\n",
      "/Users/fuzzysha/software/miniconda3/envs/hproj/lib/python3.8/site-packages/statsmodels/base/model.py:566: ConvergenceWarning: Maximum Likelihood optimization failed to converge. Check mle_retvals\n",
      "  warnings.warn(\"Maximum Likelihood optimization failed to \"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "### Classes percentage in the training as there are imbalanceness:\n",
      "0    90.99815\n",
      "1     9.00185\n",
      "Name: died_within_60_days, dtype: float64\n",
      "Warning: Maximum number of iterations has been exceeded.\n",
      "         Current function value: 0.255725\n",
      "         Iterations: 35\n",
      "Confusion Matrix : \n",
      " [[1482    0]\n",
      " [ 136    4]]\n",
      "Test accuracy =  0.9161528976572133\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.92      1.00      0.96      1482\n",
      "           1       1.00      0.03      0.06       140\n",
      "\n",
      "    accuracy                           0.92      1622\n",
      "   macro avg       0.96      0.51      0.51      1622\n",
      "weighted avg       0.92      0.92      0.88      1622\n",
      "\n",
      "F1 Score: 0.05555555555555556\n",
      "AUC: 0.51%\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 500x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# separate the features for training \n",
    "X=data_final[feature_list_for_training]\n",
    "# separate the labels for the training \n",
    "y=data_final['died_within_60_days']\n",
    "\n",
    "# add the constant to the x features\n",
    "X=sm.add_constant(X)\n",
    "# split the dataset into train and tes set \n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=0)\n",
    "\n",
    "######################################\n",
    "\n",
    "print(\"### Classes percentage in the training as there are imbalanceness:\") \n",
    "print(y_train.value_counts()*100/len(y_train))\n",
    "\n",
    "# fit the model\n",
    "log_reg=sm.Logit(y_train,X_train ).fit()\n",
    "\n",
    "# after fitting the model, we need to test the model on the x_test (y_hat are the predicted probabilities)\n",
    "yhat = log_reg.predict(X_test)\n",
    "# recieve the labels from the y_hat for the classification\n",
    "prediction = list(map(round, yhat))\n",
    "\n",
    "# If I understand this – you rounded the yhat to create a binary variable for comparison? \n",
    "# Then you created the confusion matrix for these predictions?\n",
    "# Can you try calculating the auc for y_test using yhat as a continuous measure?\n",
    "\n",
    "\n",
    "# Usually, we’d choose the cutoff so that the outcome is more balanced, using either the youden index \n",
    "# or a cutoff that ensures that the sum 1’s based on the predicted scores = the sum of observed scores=1. \n",
    "# I think this is leading to the poor calibration we see in the confusion matrix.\n",
    "\n",
    "\n",
    "\n",
    "# confusion matrix\n",
    "cm = confusion_matrix(y_test, prediction)\n",
    "print (\"Confusion Matrix : \\n\", cm)\n",
    "\n",
    "# accuracy score of the model\n",
    "print('Test accuracy = ', accuracy_score(y_test, prediction))\n",
    "\n",
    "print(classification_report(y_test, prediction))\n",
    "\n",
    "print(\"F1 Score: {}\".format(f1_score(y_test, prediction)))\n",
    "\n",
    "roc_auc = roc_auc_score(y_test, prediction)\n",
    "print('AUC: %.2f' % roc_auc + \"%\")\n",
    "\n",
    "\n",
    "\n",
    "# Visualisation with plot_metric\n",
    "bc = BinaryClassification(y_test, prediction, labels = [\"Class 1\", \"Class 2\"])\n",
    "\n",
    "# Figures\n",
    "plt.figure(figsize = (5,5))\n",
    "bc.plot_roc_curve()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1257d795",
   "metadata": {},
   "source": [
    "# summary of the model trained by imbalanced classes "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "51da1fb0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            Logit Regression Results                           \n",
      "===============================================================================\n",
      "Dep. Variable:     died_within_60_days   No. Observations:                14597\n",
      "Model:                           Logit   Df Residuals:                    14543\n",
      "Method:                            MLE   Df Model:                           53\n",
      "Date:                 Sat, 29 Jul 2023   Pseudo R-squ.:                  0.1549\n",
      "Time:                         19:05:59   Log-Likelihood:                -3732.8\n",
      "converged:                       False   LL-Null:                       -4416.8\n",
      "Covariance Type:             nonrobust   LLR p-value:                2.373e-251\n",
      "=====================================================================================================\n",
      "                                        coef    std err          z      P>|z|      [0.025      0.975]\n",
      "-----------------------------------------------------------------------------------------------------\n",
      "const                                -7.1966      0.346    -20.786      0.000      -7.875      -6.518\n",
      "spec_pat_num_age                      0.0454      0.004     12.228      0.000       0.038       0.053\n",
      "collected_in_ed                       1.4774      0.169      8.762      0.000       1.147       1.808\n",
      "collected_in_hospital                 1.3302      0.184      7.247      0.000       0.970       1.690\n",
      "collected_in_dsl                     -0.0797      0.200     -0.399      0.690      -0.471       0.311\n",
      "dsl_resident_during_collection       -0.6360      0.186     -3.426      0.001      -1.000      -0.272\n",
      "symptomatic_during_collection_U       0.3826      0.087      4.397      0.000       0.212       0.553\n",
      "symptomatic_during_collection_Y       0.7435      0.090      8.276      0.000       0.567       0.920\n",
      "interp_result_Positive                1.1823      0.093     12.764      0.000       1.001       1.364\n",
      "y_m__2020-10                          0.8286      0.147      5.651      0.000       0.541       1.116\n",
      "y_m__2020-11                          0.7452      0.140      5.322      0.000       0.471       1.020\n",
      "y_m__2020-12                          0.7079      0.139      5.104      0.000       0.436       0.980\n",
      "y_m__2020-3                           1.5030      0.148     10.151      0.000       1.213       1.793\n",
      "y_m__2020-4                           1.0299      0.107      9.601      0.000       0.820       1.240\n",
      "y_m__2020-5                           0.6517      0.131      4.965      0.000       0.394       0.909\n",
      "y_m__2021-3                           0.6365      0.441      1.444      0.149      -0.228       1.500\n",
      "y_m__2020-7                           0.9160      0.157      5.842      0.000       0.609       1.223\n",
      "y_m__2020-8                           0.5565      0.227      2.454      0.014       0.112       1.001\n",
      "y_m__2020-9                           0.7834      0.200      3.909      0.000       0.391       1.176\n",
      "y_m__2021-1                           0.8434      0.157      5.366      0.000       0.535       1.151\n",
      "y_m__2021-2                           0.8325      0.282      2.952      0.003       0.280       1.385\n",
      "chf_2_years_full_1                    0.3276      0.080      4.084      0.000       0.170       0.485\n",
      "htn_unc_2_years_full_1               -0.2958      0.070     -4.241      0.000      -0.432      -0.159\n",
      "cancer_mets_2_years_full_1            0.3394      0.239      1.423      0.155      -0.128       0.807\n",
      "fluid_elec_dis_2_years_full_1         0.1926      0.076      2.531      0.011       0.043       0.342\n",
      "card_arrh_2_years_full_1              0.0732      0.078      0.933      0.351      -0.081       0.227\n",
      "valv_dis_2_years_full_1              -0.1042      0.182     -0.572      0.568      -0.461       0.253\n",
      "pcd_2_years_full_1                    0.1285      0.151      0.852      0.394      -0.167       0.424\n",
      "pvd_2_years_full_1                    0.1610      0.105      1.528      0.126      -0.045       0.368\n",
      "htn_c_2_years_full_1                 -0.6814      0.351     -1.943      0.052      -1.369       0.006\n",
      "paral_2_years_full_1                 -0.3389      0.198     -1.709      0.087      -0.728       0.050\n",
      "oth_neur_dis_2_years_full_1           0.0086      0.065      0.132      0.895      -0.119       0.136\n",
      "cpd_2_years_full_1                    0.1433      0.078      1.830      0.067      -0.010       0.297\n",
      "diab_2_years_full_1                  -0.0081      0.076     -0.107      0.915      -0.157       0.141\n",
      "hypothyroidism_2_years_full_1         0.0246      0.089      0.278      0.781      -0.149       0.198\n",
      "ren_fail_2_years_full_1               0.1319      0.091      1.443      0.149      -0.047       0.311\n",
      "liver_dis_2_years_full_1              0.2708      0.198      1.365      0.172      -0.118       0.660\n",
      "pep_ulc_exc_bld_2_years_full_1       -0.0567      0.261     -0.217      0.828      -0.568       0.455\n",
      "aids_hiv_2_years_full_1             -16.3824   7671.259     -0.002      0.998   -1.51e+04     1.5e+04\n",
      "lymph_2_years_full_1                  0.1275      0.316      0.403      0.687      -0.492       0.747\n",
      "tumour_no_mets_2_years_full_1         0.2985      0.113      2.642      0.008       0.077       0.520\n",
      "rheum_col_vasc_dis_2_years_full_1    -0.2313      0.073     -3.175      0.001      -0.374      -0.089\n",
      "coag_2_years_full_1                   0.0266      0.171      0.156      0.876      -0.308       0.362\n",
      "obes_2_years_full_1                  -0.1758      0.183     -0.961      0.336      -0.534       0.183\n",
      "wt_loss_2_years_full_1                0.0797      0.131      0.608      0.543      -0.177       0.337\n",
      "anemia_2_years_full_1                 0.2087      0.096      2.171      0.030       0.020       0.397\n",
      "alc_abuse_2_years_full_1             -0.0601      0.179     -0.336      0.737      -0.411       0.291\n",
      "drug_abuse_2_years_full_1            -0.0057      0.198     -0.029      0.977      -0.394       0.382\n",
      "psych_2_years_full_1                 -0.0063      0.076     -0.084      0.933      -0.155       0.142\n",
      "depress_2_years_full_1               -0.0741      0.067     -1.113      0.266      -0.204       0.056\n",
      "num_admits_1_year                     0.0165      0.038      0.431      0.666      -0.059       0.092\n",
      "num_scu_admits_1_year                 0.1317      0.221      0.595      0.552      -0.302       0.566\n",
      "num_procs_dad_1_year                  0.0195      0.034      0.578      0.563      -0.047       0.086\n",
      "num_procs_nacrs_1_year                0.0008      0.002      0.537      0.591      -0.002       0.004\n",
      "=====================================================================================================\n"
     ]
    }
   ],
   "source": [
    "# print summary of the model trained with Imbalanced data\n",
    "print(log_reg.summary())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ecf9f1b1",
   "metadata": {},
   "source": [
    "# Adjusted Odds Ratios for female"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "f3851fa9",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "\n",
    "# Assuming you have already fitted your logistic regression model and stored it in log_reg_model\n",
    "# For example: log_reg_model = sm.Logit(y, X).fit()\n",
    "\n",
    "# Get the summary of the logistic regression model\n",
    "log_reg_summary = log_reg.summary()\n",
    "\n",
    "# Convert the summary to HTML\n",
    "html_summary = log_reg_summary.tables[1].as_html()\n",
    "\n",
    "# Read the HTML table into a DataFrame\n",
    "df_summary = pd.read_html(html_summary, header=0, index_col=0)[0]\n",
    "\n",
    "# Calculate the odds ratio and add it to the DataFrame\n",
    "df_summary['Odds Ratio'] = df_summary['coef'].apply(lambda x: round(np.exp(x), 2))\n",
    "df_summary['[0.025'] = df_summary['[0.025'].apply(lambda x: round(np.exp(x), 2))\n",
    "df_summary['0.975]'] = df_summary['0.975]'].apply(lambda x: round(np.exp(x), 2))\n",
    "df_summary=df_summary[['Odds Ratio','[0.025', '0.975]','P>|z|']]\n",
    "# Display the DataFrame with the odds ratio\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "a34b561a",
   "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>Odds Ratio</th>\n",
       "      <th>[0.025</th>\n",
       "      <th>0.975]</th>\n",
       "      <th>P&gt;|z|</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>y_m__2020-3</th>\n",
       "      <td>4.50</td>\n",
       "      <td>3.36</td>\n",
       "      <td>6.01</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>collected_in_ed</th>\n",
       "      <td>4.38</td>\n",
       "      <td>3.15</td>\n",
       "      <td>6.10</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>collected_in_hospital</th>\n",
       "      <td>3.78</td>\n",
       "      <td>2.64</td>\n",
       "      <td>5.42</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>interp_result_Positive</th>\n",
       "      <td>3.26</td>\n",
       "      <td>2.72</td>\n",
       "      <td>3.91</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>y_m__2020-4</th>\n",
       "      <td>2.80</td>\n",
       "      <td>2.27</td>\n",
       "      <td>3.46</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>y_m__2020-7</th>\n",
       "      <td>2.50</td>\n",
       "      <td>1.84</td>\n",
       "      <td>3.40</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>y_m__2021-1</th>\n",
       "      <td>2.32</td>\n",
       "      <td>1.71</td>\n",
       "      <td>3.16</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>y_m__2021-2</th>\n",
       "      <td>2.30</td>\n",
       "      <td>1.32</td>\n",
       "      <td>3.99</td>\n",
       "      <td>0.003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>y_m__2020-10</th>\n",
       "      <td>2.29</td>\n",
       "      <td>1.72</td>\n",
       "      <td>3.05</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>y_m__2020-9</th>\n",
       "      <td>2.19</td>\n",
       "      <td>1.48</td>\n",
       "      <td>3.24</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>y_m__2020-11</th>\n",
       "      <td>2.11</td>\n",
       "      <td>1.60</td>\n",
       "      <td>2.77</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>symptomatic_during_collection_Y</th>\n",
       "      <td>2.10</td>\n",
       "      <td>1.76</td>\n",
       "      <td>2.51</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>y_m__2020-12</th>\n",
       "      <td>2.03</td>\n",
       "      <td>1.55</td>\n",
       "      <td>2.66</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>y_m__2020-5</th>\n",
       "      <td>1.92</td>\n",
       "      <td>1.48</td>\n",
       "      <td>2.48</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>y_m__2021-3</th>\n",
       "      <td>1.89</td>\n",
       "      <td>0.80</td>\n",
       "      <td>4.48</td>\n",
       "      <td>0.149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>y_m__2020-8</th>\n",
       "      <td>1.74</td>\n",
       "      <td>1.12</td>\n",
       "      <td>2.72</td>\n",
       "      <td>0.014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>symptomatic_during_collection_U</th>\n",
       "      <td>1.47</td>\n",
       "      <td>1.24</td>\n",
       "      <td>1.74</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cancer_mets_2_years_full_1</th>\n",
       "      <td>1.40</td>\n",
       "      <td>0.88</td>\n",
       "      <td>2.24</td>\n",
       "      <td>0.155</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>chf_2_years_full_1</th>\n",
       "      <td>1.39</td>\n",
       "      <td>1.19</td>\n",
       "      <td>1.62</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>tumour_no_mets_2_years_full_1</th>\n",
       "      <td>1.35</td>\n",
       "      <td>1.08</td>\n",
       "      <td>1.68</td>\n",
       "      <td>0.008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>liver_dis_2_years_full_1</th>\n",
       "      <td>1.31</td>\n",
       "      <td>0.89</td>\n",
       "      <td>1.93</td>\n",
       "      <td>0.172</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>anemia_2_years_full_1</th>\n",
       "      <td>1.23</td>\n",
       "      <td>1.02</td>\n",
       "      <td>1.49</td>\n",
       "      <td>0.030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fluid_elec_dis_2_years_full_1</th>\n",
       "      <td>1.21</td>\n",
       "      <td>1.04</td>\n",
       "      <td>1.41</td>\n",
       "      <td>0.011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pvd_2_years_full_1</th>\n",
       "      <td>1.17</td>\n",
       "      <td>0.96</td>\n",
       "      <td>1.44</td>\n",
       "      <td>0.126</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cpd_2_years_full_1</th>\n",
       "      <td>1.15</td>\n",
       "      <td>0.99</td>\n",
       "      <td>1.35</td>\n",
       "      <td>0.067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lymph_2_years_full_1</th>\n",
       "      <td>1.14</td>\n",
       "      <td>0.61</td>\n",
       "      <td>2.11</td>\n",
       "      <td>0.687</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>num_scu_admits_1_year</th>\n",
       "      <td>1.14</td>\n",
       "      <td>0.74</td>\n",
       "      <td>1.76</td>\n",
       "      <td>0.552</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ren_fail_2_years_full_1</th>\n",
       "      <td>1.14</td>\n",
       "      <td>0.95</td>\n",
       "      <td>1.36</td>\n",
       "      <td>0.149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pcd_2_years_full_1</th>\n",
       "      <td>1.14</td>\n",
       "      <td>0.85</td>\n",
       "      <td>1.53</td>\n",
       "      <td>0.394</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>card_arrh_2_years_full_1</th>\n",
       "      <td>1.08</td>\n",
       "      <td>0.92</td>\n",
       "      <td>1.25</td>\n",
       "      <td>0.351</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>wt_loss_2_years_full_1</th>\n",
       "      <td>1.08</td>\n",
       "      <td>0.84</td>\n",
       "      <td>1.40</td>\n",
       "      <td>0.543</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>spec_pat_num_age</th>\n",
       "      <td>1.05</td>\n",
       "      <td>1.04</td>\n",
       "      <td>1.05</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>coag_2_years_full_1</th>\n",
       "      <td>1.03</td>\n",
       "      <td>0.73</td>\n",
       "      <td>1.44</td>\n",
       "      <td>0.876</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>num_admits_1_year</th>\n",
       "      <td>1.02</td>\n",
       "      <td>0.94</td>\n",
       "      <td>1.10</td>\n",
       "      <td>0.666</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>num_procs_dad_1_year</th>\n",
       "      <td>1.02</td>\n",
       "      <td>0.95</td>\n",
       "      <td>1.09</td>\n",
       "      <td>0.563</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hypothyroidism_2_years_full_1</th>\n",
       "      <td>1.02</td>\n",
       "      <td>0.86</td>\n",
       "      <td>1.22</td>\n",
       "      <td>0.781</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>oth_neur_dis_2_years_full_1</th>\n",
       "      <td>1.01</td>\n",
       "      <td>0.89</td>\n",
       "      <td>1.15</td>\n",
       "      <td>0.895</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>num_procs_nacrs_1_year</th>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.591</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>diab_2_years_full_1</th>\n",
       "      <td>0.99</td>\n",
       "      <td>0.85</td>\n",
       "      <td>1.15</td>\n",
       "      <td>0.915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>drug_abuse_2_years_full_1</th>\n",
       "      <td>0.99</td>\n",
       "      <td>0.67</td>\n",
       "      <td>1.47</td>\n",
       "      <td>0.977</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>psych_2_years_full_1</th>\n",
       "      <td>0.99</td>\n",
       "      <td>0.86</td>\n",
       "      <td>1.15</td>\n",
       "      <td>0.933</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pep_ulc_exc_bld_2_years_full_1</th>\n",
       "      <td>0.94</td>\n",
       "      <td>0.57</td>\n",
       "      <td>1.58</td>\n",
       "      <td>0.828</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>alc_abuse_2_years_full_1</th>\n",
       "      <td>0.94</td>\n",
       "      <td>0.66</td>\n",
       "      <td>1.34</td>\n",
       "      <td>0.737</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>depress_2_years_full_1</th>\n",
       "      <td>0.93</td>\n",
       "      <td>0.82</td>\n",
       "      <td>1.06</td>\n",
       "      <td>0.266</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>collected_in_dsl</th>\n",
       "      <td>0.92</td>\n",
       "      <td>0.62</td>\n",
       "      <td>1.36</td>\n",
       "      <td>0.690</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>valv_dis_2_years_full_1</th>\n",
       "      <td>0.90</td>\n",
       "      <td>0.63</td>\n",
       "      <td>1.29</td>\n",
       "      <td>0.568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>obes_2_years_full_1</th>\n",
       "      <td>0.84</td>\n",
       "      <td>0.59</td>\n",
       "      <td>1.20</td>\n",
       "      <td>0.336</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>rheum_col_vasc_dis_2_years_full_1</th>\n",
       "      <td>0.79</td>\n",
       "      <td>0.69</td>\n",
       "      <td>0.91</td>\n",
       "      <td>0.001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>htn_unc_2_years_full_1</th>\n",
       "      <td>0.74</td>\n",
       "      <td>0.65</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>paral_2_years_full_1</th>\n",
       "      <td>0.71</td>\n",
       "      <td>0.48</td>\n",
       "      <td>1.05</td>\n",
       "      <td>0.087</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dsl_resident_during_collection</th>\n",
       "      <td>0.53</td>\n",
       "      <td>0.37</td>\n",
       "      <td>0.76</td>\n",
       "      <td>0.001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>htn_c_2_years_full_1</th>\n",
       "      <td>0.51</td>\n",
       "      <td>0.25</td>\n",
       "      <td>1.01</td>\n",
       "      <td>0.052</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>aids_hiv_2_years_full_1</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>inf</td>\n",
       "      <td>0.998</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>const</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   Odds Ratio  [0.025  0.975]  P>|z|\n",
       "y_m__2020-3                              4.50    3.36    6.01  0.000\n",
       "collected_in_ed                          4.38    3.15    6.10  0.000\n",
       "collected_in_hospital                    3.78    2.64    5.42  0.000\n",
       "interp_result_Positive                   3.26    2.72    3.91  0.000\n",
       "y_m__2020-4                              2.80    2.27    3.46  0.000\n",
       "y_m__2020-7                              2.50    1.84    3.40  0.000\n",
       "y_m__2021-1                              2.32    1.71    3.16  0.000\n",
       "y_m__2021-2                              2.30    1.32    3.99  0.003\n",
       "y_m__2020-10                             2.29    1.72    3.05  0.000\n",
       "y_m__2020-9                              2.19    1.48    3.24  0.000\n",
       "y_m__2020-11                             2.11    1.60    2.77  0.000\n",
       "symptomatic_during_collection_Y          2.10    1.76    2.51  0.000\n",
       "y_m__2020-12                             2.03    1.55    2.66  0.000\n",
       "y_m__2020-5                              1.92    1.48    2.48  0.000\n",
       "y_m__2021-3                              1.89    0.80    4.48  0.149\n",
       "y_m__2020-8                              1.74    1.12    2.72  0.014\n",
       "symptomatic_during_collection_U          1.47    1.24    1.74  0.000\n",
       "cancer_mets_2_years_full_1               1.40    0.88    2.24  0.155\n",
       "chf_2_years_full_1                       1.39    1.19    1.62  0.000\n",
       "tumour_no_mets_2_years_full_1            1.35    1.08    1.68  0.008\n",
       "liver_dis_2_years_full_1                 1.31    0.89    1.93  0.172\n",
       "anemia_2_years_full_1                    1.23    1.02    1.49  0.030\n",
       "fluid_elec_dis_2_years_full_1            1.21    1.04    1.41  0.011\n",
       "pvd_2_years_full_1                       1.17    0.96    1.44  0.126\n",
       "cpd_2_years_full_1                       1.15    0.99    1.35  0.067\n",
       "lymph_2_years_full_1                     1.14    0.61    2.11  0.687\n",
       "num_scu_admits_1_year                    1.14    0.74    1.76  0.552\n",
       "ren_fail_2_years_full_1                  1.14    0.95    1.36  0.149\n",
       "pcd_2_years_full_1                       1.14    0.85    1.53  0.394\n",
       "card_arrh_2_years_full_1                 1.08    0.92    1.25  0.351\n",
       "wt_loss_2_years_full_1                   1.08    0.84    1.40  0.543\n",
       "spec_pat_num_age                         1.05    1.04    1.05  0.000\n",
       "coag_2_years_full_1                      1.03    0.73    1.44  0.876\n",
       "num_admits_1_year                        1.02    0.94    1.10  0.666\n",
       "num_procs_dad_1_year                     1.02    0.95    1.09  0.563\n",
       "hypothyroidism_2_years_full_1            1.02    0.86    1.22  0.781\n",
       "oth_neur_dis_2_years_full_1              1.01    0.89    1.15  0.895\n",
       "num_procs_nacrs_1_year                   1.00    1.00    1.00  0.591\n",
       "diab_2_years_full_1                      0.99    0.85    1.15  0.915\n",
       "drug_abuse_2_years_full_1                0.99    0.67    1.47  0.977\n",
       "psych_2_years_full_1                     0.99    0.86    1.15  0.933\n",
       "pep_ulc_exc_bld_2_years_full_1           0.94    0.57    1.58  0.828\n",
       "alc_abuse_2_years_full_1                 0.94    0.66    1.34  0.737\n",
       "depress_2_years_full_1                   0.93    0.82    1.06  0.266\n",
       "collected_in_dsl                         0.92    0.62    1.36  0.690\n",
       "valv_dis_2_years_full_1                  0.90    0.63    1.29  0.568\n",
       "obes_2_years_full_1                      0.84    0.59    1.20  0.336\n",
       "rheum_col_vasc_dis_2_years_full_1        0.79    0.69    0.91  0.001\n",
       "htn_unc_2_years_full_1                   0.74    0.65    0.85  0.000\n",
       "paral_2_years_full_1                     0.71    0.48    1.05  0.087\n",
       "dsl_resident_during_collection           0.53    0.37    0.76  0.001\n",
       "htn_c_2_years_full_1                     0.51    0.25    1.01  0.052\n",
       "aids_hiv_2_years_full_1                  0.00    0.00     inf  0.998\n",
       "const                                    0.00    0.00    0.00  0.000"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_summary.sort_values('Odds Ratio', ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "94aac714",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "75946f55",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}