CS_1_Model_Pipeline.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "naPWqtUVigTd"
   },
   "source": [
    "---\n",
    "---\n",
    "# **`CASE STUDY - 1` :: Healthcare Provider Fraudulent Detection**\n",
    "\n",
    "<img src=\"Display_Pic.png\" alt=\"drawing\" width=\"950\"/>\n",
    "\n",
    "---\n",
    "---\n",
    "\n",
    "\n",
    "**This notebook contains the feature engineering :-**\n",
    "   - **Final Model Pipelining**\n",
    "\n",
    "\n",
    "**Above steps are performed on the publicly available dataset at [Kaggle](https://www.kaggle.com/datasets/rohitrox/healthcare-provider-fraud-detection-analysis).**\n",
    "\n",
    "\n",
    "**Kindly checkout below link for gaining BUSINESS related insights related to this problem ::**\n",
    "\n",
    "- [Deck : Detailed Explanation](https://docs.google.com/presentation/d/1Thuw_eZskafkl9W3xYuEVsjgTJKFeEkzEfAmczH96Uw/)\n",
    "\n",
    "\n",
    "**Kindly checkout below link for TECHNICAL design document ::**\n",
    "\n",
    "- [Technical Document](https://docs.google.com/document/d/10z9xbn4dZWkforlAszCDaa1M0roAW-MOf2pjJHSnEXk/)\n",
    "\n",
    "\n",
    "**Kindly checkout below link for In-depth Description and Reasoning of all the Features ::**\n",
    "\n",
    "- [Features Description](https://docs.google.com/spreadsheets/d/1ktwjad3U-hGT_7yccGyZC4AzGhWAMs3RHrIMk8Gi8xQ/)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "PHkkU2ExpgPR"
   },
   "source": [
    "# **`Notebook Contents`**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "toc",
    "id": "ZsqH5e9kpIzv"
   },
   "source": [
    ">[CASE STUDY - 1 :: Healthcare Provider Fraudulent Detection](#scrollTo=naPWqtUVigTd)\n",
    "\n",
    ">[Notebook Contents](#scrollTo=PHkkU2ExpgPR)\n",
    "\n",
    ">>[Importing_Libraries](#scrollTo=ThPD8j4PnuVH)\n",
    "\n",
    ">>[Unseen_Data_Files](#scrollTo=c16_htZKnucm)\n",
    "\n",
    ">>[Final_Pipeline](#scrollTo=eAXKdt1pyngk)\n",
    "\n",
    ">>>[SUMMARY](#scrollTo=r-VhXYclyng3)\n",
    "\n",
    ">>[Performance_on_TRAIN_and_VALIDATION_Sets](#scrollTo=0IC2VKpXyng4)\n",
    "\n",
    ">>>[SUMMARY](#scrollTo=ZfOButQ2ynhB)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ThPD8j4PnuVH"
   },
   "source": [
    "## **Importing_Libraries**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T11:54:58.198503Z",
     "iopub.status.busy": "2022-09-02T11:54:58.196502Z",
     "iopub.status.idle": "2022-09-02T11:55:06.753538Z",
     "shell.execute_reply": "2022-09-02T11:55:06.750537Z",
     "shell.execute_reply.started": "2022-09-02T11:54:58.198503Z"
    },
    "id": "l6qT6r78nuZC"
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import sys\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import joblib\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score, roc_auc_score, confusion_matrix, roc_curve, auc\n",
    "from sklearn.calibration import CalibratedClassifierCV\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T11:55:06.760541Z",
     "iopub.status.busy": "2022-09-02T11:55:06.758537Z",
     "iopub.status.idle": "2022-09-02T11:55:06.801537Z",
     "shell.execute_reply": "2022-09-02T11:55:06.799537Z",
     "shell.execute_reply.started": "2022-09-02T11:55:06.760541Z"
    },
    "id": "ggC6AoW9u3VV"
   },
   "outputs": [],
   "source": [
    "pd.set_option('display.max_columns',80)\n",
    "label_font_dict = {'family':'sans-serif','size':13.5,'color':'brown','style':'italic'}\n",
    "title_font_dict = {'family':'sans-serif','size':16.5,'color':'Blue','style':'italic'}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "c16_htZKnucm"
   },
   "source": [
    "## **Unseen_Data_Files**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T11:55:06.807536Z",
     "iopub.status.busy": "2022-09-02T11:55:06.806536Z",
     "iopub.status.idle": "2022-09-02T11:55:06.895533Z",
     "shell.execute_reply": "2022-09-02T11:55:06.892537Z",
     "shell.execute_reply.started": "2022-09-02T11:55:06.807536Z"
    },
    "id": "Ge75LVO0yngh",
    "outputId": "4e80ec65-ccb1-4258-da89-6a5fdbc5f1c2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "### Location where we have all the unseen data files --> E:\\STUDY\\PROJECTS\\AAIC_Case_Study_1\\Dataset\\TEST\n"
     ]
    }
   ],
   "source": [
    "# Unseen dataset files\n",
    "bene_file = \"Test_Beneficiarydata-1542969243754.csv\"\n",
    "ip_file = \"Test_Inpatientdata-1542969243754.csv\"\n",
    "op_file = \"Test_Outpatientdata-1542969243754.csv\"\n",
    "tgt_lbl_file = \"Test-1542969243754.csv\"\n",
    "\n",
    "# Folder hierarchies \n",
    "folder_1 = \"Dataset\"\n",
    "folder_2 = \"TEST\"\n",
    "data_files_loc = os.path.join(os.path.join(os.getcwd(), folder_1), folder_2)\n",
    "print(\"### Location where we have all the unseen data files --> {}\".format(data_files_loc))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "eAXKdt1pyngk"
   },
   "source": [
    "## **Final_Pipeline**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T11:55:06.903538Z",
     "iopub.status.busy": "2022-09-02T11:55:06.901539Z",
     "iopub.status.idle": "2022-09-02T11:55:06.975538Z",
     "shell.execute_reply": "2022-09-02T11:55:06.972539Z",
     "shell.execute_reply.started": "2022-09-02T11:55:06.903538Z"
    },
    "id": "jAJxDhqhyngl"
   },
   "outputs": [],
   "source": [
    "def unseen_data_files_names(files_loc, bene_data, ip_data, op_data, tgt_lbl_data):\n",
    "    \"\"\"\n",
    "    Description : This function is created for reading the files names of BENEFICIARIES, IN-PATIENT, OUT-PATIENT and TGT_LABELS.\n",
    "    \"\"\"\n",
    "    # Generating the unseen data files names\n",
    "    bene_file_name = os.path.join(files_loc, bene_data)\n",
    "    ip_file_name = os.path.join(files_loc, ip_data)\n",
    "    op_file_name = os.path.join(files_loc, op_data)\n",
    "    tgt_label_name = os.path.join(files_loc, tgt_lbl_data)\n",
    "    return bene_file_name, ip_file_name, op_file_name, tgt_label_name"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T11:55:06.983536Z",
     "iopub.status.busy": "2022-09-02T11:55:06.979537Z",
     "iopub.status.idle": "2022-09-02T11:55:07.038545Z",
     "shell.execute_reply": "2022-09-02T11:55:07.036542Z",
     "shell.execute_reply.started": "2022-09-02T11:55:06.983536Z"
    },
    "id": "TwcBKVzcyngm"
   },
   "outputs": [],
   "source": [
    "def read_unseen_data(unseen_data_loc, b_file, i_file, o_file, tgt_file):\n",
    "    \"\"\"\n",
    "    Description : This function is created for reading the data based on the files names of BENEFICIARIES, IN-PATIENT, OUT-PATIENT\n",
    "    and TGT_LABELS.\n",
    "    \"\"\"\n",
    "    # Reading the data files names\n",
    "    bene_unseen_data, ip_unseen_data, op_unseen_data, tgt_lbl_unseen_data = unseen_data_files_names(files_loc=unseen_data_loc,\n",
    "                                                                                                    bene_data=b_file,\n",
    "                                                                                                    ip_data=i_file,\n",
    "                                                                                                    op_data=o_file,\n",
    "                                                                                                    tgt_lbl_data=tgt_file)\n",
    "    # Reading the data files\n",
    "    unseen_bene_df = pd.read_csv(bene_unseen_data)\n",
    "    unseen_ip_df = pd.read_csv(ip_unseen_data)\n",
    "    unseen_op_df = pd.read_csv(op_unseen_data)\n",
    "    unseen_tgt_lbls_df = pd.read_csv(tgt_lbl_unseen_data)\n",
    "    \n",
    "    # Creating a dictionary which will contain all the unseen data files\n",
    "    entire_unseen_data = {\"Unseen_Provider\":unseen_tgt_lbls_df,\n",
    "                          \"Unseen_Beneficiary\":unseen_bene_df,\n",
    "                          \"Unseen_Inpatient\":unseen_ip_df,\n",
    "                          \"Unseen_Outpatient\":unseen_op_df}\n",
    "    return entire_unseen_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T11:55:07.043551Z",
     "iopub.status.busy": "2022-09-02T11:55:07.042548Z",
     "iopub.status.idle": "2022-09-02T11:55:09.018934Z",
     "shell.execute_reply": "2022-09-02T11:55:09.015941Z",
     "shell.execute_reply.started": "2022-09-02T11:55:07.043551Z"
    },
    "id": "4PUQeRzayngo"
   },
   "outputs": [],
   "source": [
    "# Generating the unseen data dictionary\n",
    "unseen_data_dict = read_unseen_data(unseen_data_loc=data_files_loc,\n",
    "                                    b_file=bene_file,\n",
    "                                    i_file=ip_file,\n",
    "                                    o_file=op_file,\n",
    "                                    tgt_file=tgt_lbl_file)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T11:55:09.034931Z",
     "iopub.status.busy": "2022-09-02T11:55:09.023931Z",
     "iopub.status.idle": "2022-09-02T11:55:09.700564Z",
     "shell.execute_reply": "2022-09-02T11:55:09.698705Z",
     "shell.execute_reply.started": "2022-09-02T11:55:09.034931Z"
    },
    "id": "oxJMIX9Cyngp"
   },
   "outputs": [],
   "source": [
    "def pre_process(X_input):\n",
    "    \"\"\"\n",
    "    Description : This function is created for pre-processing the data files based on the feature engineering performed while training \n",
    "    the best model.\n",
    "    \"\"\" \n",
    "    # Load the unseen data files\n",
    "    test_tgt_lbls_df = X_input['Unseen_Provider']\n",
    "    test_bene_df = X_input['Unseen_Beneficiary']\n",
    "    test_ip_df = X_input['Unseen_Inpatient']\n",
    "    test_op_df = X_input['Unseen_Outpatient']\n",
    "       \n",
    "    # Adding Feature - 1\n",
    "    test_ip_df[\"Admitted?\"] = 1\n",
    "    test_op_df[\"Admitted?\"] = 0\n",
    "\n",
    "    # Commom columns must be 28\n",
    "    common_cols = [col for col in test_ip_df.columns if col in test_op_df.columns]\n",
    "    \n",
    "    # Merging the IP and OP dataset on the basis of common columns\n",
    "    test_ip_op_df = pd.merge(left=test_ip_df, right=test_op_df, left_on=common_cols, right_on=common_cols, how=\"outer\")\n",
    "    \n",
    "    # Joining the IP_OP dataset with the BENE data\n",
    "    test_ip_op_bene_df = pd.merge(left=test_ip_op_df, right=test_bene_df, left_on='BeneID', right_on='BeneID',how='inner')\n",
    "    \n",
    "    # Joining the IP_OP_BENE dataset with the Tgt Label Provider Data\n",
    "    test_iobp_df = pd.merge(left=test_ip_op_bene_df, right=test_tgt_lbls_df, left_on='Provider', right_on='Provider',how='inner')\n",
    "    \n",
    "    # Joining with the PRV Tgt Labels\n",
    "    prvs_claims_df = pd.DataFrame(test_iobp_df.groupby(['Provider'])['ClaimID'].count()).reset_index()\n",
    "    prvs_claims_tgt_lbls_df = pd.merge(left=prvs_claims_df, right=test_tgt_lbls_df, on='Provider', how='inner')\n",
    "    \n",
    "    # Adding Feature - 2\n",
    "    test_iobp_df['DOB'] = pd.to_datetime(test_iobp_df['DOB'], format=\"%Y-%m-%d\")\n",
    "    test_iobp_df['DOD'] = pd.to_datetime(test_iobp_df['DOD'], format=\"%Y-%m-%d\")\n",
    "    test_iobp_df['Is_Alive?'] = test_iobp_df['DOD'].apply(lambda val: 'No' if val != val else 'Yes')\n",
    "\n",
    "    # Adding Feature - 3\n",
    "    test_iobp_df['ClaimStartDt'] = pd.to_datetime(test_iobp_df['ClaimStartDt'], format=\"%Y-%m-%d\")\n",
    "    test_iobp_df['ClaimEndDt'] = pd.to_datetime(test_iobp_df['ClaimEndDt'], format=\"%Y-%m-%d\")\n",
    "    test_iobp_df['Claim_Duration'] = (test_iobp_df['ClaimEndDt'] - test_iobp_df['ClaimStartDt']).dt.days\n",
    "\n",
    "    # Adding Feature - 4\n",
    "    test_iobp_df['AdmissionDt'] = pd.to_datetime(test_iobp_df['AdmissionDt'], format=\"%Y-%m-%d\")\n",
    "    test_iobp_df['DischargeDt'] = pd.to_datetime(test_iobp_df['DischargeDt'], format=\"%Y-%m-%d\")\n",
    "    test_iobp_df['Admitted_Duration'] = (test_iobp_df['DischargeDt'] - test_iobp_df['AdmissionDt']).dt.days\n",
    "\n",
    "    # Adding Feature - 5\n",
    "    # Filling the Null values as MAX Date of Death in the Dataset\n",
    "    test_iobp_df['DOD'].fillna(value=pd.to_datetime('2009-12-01',format ='%Y-%m-%d'), inplace=True)\n",
    "    test_iobp_df['Bene_Age'] = round(((test_iobp_df['DOD'] - test_iobp_df['DOB']).dt.days)/365,1)\n",
    "\n",
    "    # Adding Feature - 6\n",
    "    test_iobp_df['Att_Phy_tot_claims'] = test_iobp_df.groupby(['AttendingPhysician'])['ClaimID'].transform('count')\n",
    "    test_iobp_df['Opr_Phy_tot_claims'] = test_iobp_df.groupby(['OperatingPhysician'])['ClaimID'].transform('count')\n",
    "    test_iobp_df['Oth_Phy_tot_claims'] = test_iobp_df.groupby(['OtherPhysician'])['ClaimID'].transform('count')\n",
    "    test_iobp_df['Att_Phy_tot_claims'].fillna(value=0, inplace=True)\n",
    "    test_iobp_df['Opr_Phy_tot_claims'].fillna(value=0, inplace=True)\n",
    "    test_iobp_df['Oth_Phy_tot_claims'].fillna(value=0, inplace=True)\n",
    "    test_iobp_df['Att_Opr_Oth_Phy_Tot_Claims'] = test_iobp_df['Att_Phy_tot_claims'] + test_iobp_df['Opr_Phy_tot_claims'] + test_iobp_df['Oth_Phy_tot_claims']\n",
    "    test_iobp_df.drop(['Att_Phy_tot_claims', 'Opr_Phy_tot_claims', 'Oth_Phy_tot_claims'], axis=1, inplace=True)\n",
    "\n",
    "    # Adding Feature - 7\n",
    "    test_iobp_df[\"Prv_Tot_Att_Phy\"] = test_iobp_df.groupby(['Provider'])['AttendingPhysician'].transform('count')\n",
    "    test_iobp_df[\"Prv_Tot_Opr_Phy\"] = test_iobp_df.groupby(['Provider'])['OperatingPhysician'].transform('count')\n",
    "    test_iobp_df[\"Prv_Tot_Oth_Phy\"] = test_iobp_df.groupby(['Provider'])['OtherPhysician'].transform('count')\n",
    "    test_iobp_df['Prv_Tot_Att_Phy'].fillna(value=0, inplace=True)\n",
    "    test_iobp_df['Prv_Tot_Opr_Phy'].fillna(value=0, inplace=True)\n",
    "    test_iobp_df['Prv_Tot_Oth_Phy'].fillna(value=0, inplace=True)\n",
    "    test_iobp_df['Prv_Tot_Att_Opr_Oth_Phys'] = test_iobp_df['Prv_Tot_Att_Phy'] + test_iobp_df['Prv_Tot_Opr_Phy'] + test_iobp_df['Prv_Tot_Oth_Phy']\n",
    "    test_iobp_df.drop(['Prv_Tot_Att_Phy', 'Prv_Tot_Opr_Phy', 'Prv_Tot_Oth_Phy'], axis=1, inplace=True)\n",
    "\n",
    "    # Adding Feature - 8\n",
    "    test_iobp_df['PRV_Tot_Admit_DCodes'] = test_iobp_df.groupby(['Provider'])['ClmAdmitDiagnosisCode'].transform('nunique')\n",
    "\n",
    "    # Adding Feature - 9\n",
    "    test_iobp_df['PRV_Tot_DGrpCodes'] = test_iobp_df.groupby(['Provider'])['DiagnosisGroupCode'].transform('nunique')\n",
    "\n",
    "    # Adding Feature - 10\n",
    "    test_iobp_df['DOB_Year'] = test_iobp_df['DOB'].dt.year\n",
    "    test_iobp_df['PRV_Tot_Unq_DOB_Years'] = test_iobp_df.groupby(['Provider'])['DOB_Year'].transform('nunique')\n",
    "    test_iobp_df.drop(['DOB_Year'], axis=1, inplace=True)\n",
    "\n",
    "    # Adding Feature - 11\n",
    "    test_iobp_df['PRV_Bene_Age_Sum'] = test_iobp_df.groupby(['Provider'])['Bene_Age'].transform('sum')\n",
    "\n",
    "    # Adding Feature - 12\n",
    "    test_iobp_df['PRV_Insc_Clm_ReImb_Amt'] = test_iobp_df.groupby(['Provider'])['InscClaimAmtReimbursed'].transform('sum')\n",
    "\n",
    "    # Adding Feature - 13\n",
    "    test_iobp_df['RenalDiseaseIndicator'] = test_iobp_df['RenalDiseaseIndicator'].apply(lambda val: 1 if val == \"Y\" else 0)\n",
    "    test_iobp_df['PRV_Tot_RKD_Patients'] = test_iobp_df.groupby(['Provider'])['RenalDiseaseIndicator'].transform('sum')\n",
    "    test_iobp_df.drop(['NoOfMonths_PartACov', 'NoOfMonths_PartBCov'], axis=1, inplace=True)\n",
    "    test_iobp_df['Admitted_Duration'].fillna(value=0,inplace=True)\n",
    "\n",
    "    # PRV Aggregate features\n",
    "    test_iobp_df[\"PRV_CoPayment\"] = test_iobp_df.groupby('Provider')['DeductibleAmtPaid'].transform('sum')\n",
    "    test_iobp_df[\"PRV_IP_Annual_ReImb_Amt\"] = test_iobp_df.groupby('Provider')['IPAnnualReimbursementAmt'].transform('sum')\n",
    "    test_iobp_df[\"PRV_IP_Annual_Ded_Amt\"] = test_iobp_df.groupby('Provider')['IPAnnualDeductibleAmt'].transform('sum')\n",
    "    test_iobp_df[\"PRV_OP_Annual_ReImb_Amt\"] = test_iobp_df.groupby('Provider')['OPAnnualReimbursementAmt'].transform('sum')\n",
    "    test_iobp_df[\"PRV_OP_Annual_Ded_Amt\"] = test_iobp_df.groupby('Provider')['OPAnnualDeductibleAmt'].transform('sum')\n",
    "    test_iobp_df[\"PRV_Admit_Duration\"] = test_iobp_df.groupby('Provider')['Admitted_Duration'].transform('sum')\n",
    "    test_iobp_df[\"PRV_Claim_Duration\"] = test_iobp_df.groupby('Provider')['Claim_Duration'].transform('sum')\n",
    "\n",
    "    def create_agg_feats(grp_col, feat_name, operation='sum'):\n",
    "        \"\"\"\n",
    "        Description :: This function is created for adding the aggregated features in the dataset for every level like:\n",
    "            - Beneficiary\n",
    "            - Attending Physician\n",
    "            - Operating Physician\n",
    "            - Other Physician and etc..\n",
    "\n",
    "        Input Parameters :: It accepts below inputs:\n",
    "            - grp_col : `str`\n",
    "                - It represents the feature or level at which you want to perform the aggregation.\n",
    "\n",
    "            - feat_name : `str`\n",
    "                - It represents the feature whose aggregated aspect you want to capture.\n",
    "\n",
    "            - operation : `str`\n",
    "                - It represents the aggregation operation you want to perform.(By default it is SUM)\n",
    "        \"\"\"\n",
    "        feat_1 = feat_name + \"_Insc_ReImb_Amt\"\n",
    "        test_iobp_df[feat_1] = test_iobp_df.groupby(grp_col)['InscClaimAmtReimbursed'].transform(operation)\n",
    "\n",
    "        feat_2 = feat_name + \"_CoPayment\"\n",
    "        test_iobp_df[feat_2] = test_iobp_df.groupby(grp_col)['DeductibleAmtPaid'].transform(operation)\n",
    "\n",
    "        feat_3 = feat_name + \"_IP_Annual_ReImb_Amt\"\n",
    "        test_iobp_df[feat_3] = test_iobp_df.groupby(grp_col)['IPAnnualReimbursementAmt'].transform(operation)\n",
    "\n",
    "        feat_4 = feat_name + \"_IP_Annual_Ded_Amt\"\n",
    "        test_iobp_df[feat_4] = test_iobp_df.groupby(grp_col)['IPAnnualDeductibleAmt'].transform(operation)\n",
    "\n",
    "        feat_5 = feat_name + \"_OP_Annual_ReImb_Amt\"\n",
    "        test_iobp_df[feat_5] = test_iobp_df.groupby(grp_col)['OPAnnualReimbursementAmt'].transform(operation)\n",
    "\n",
    "        feat_6 = feat_name + \"_OP_Annual_Ded_Amt\"\n",
    "        test_iobp_df[feat_6] = test_iobp_df.groupby(grp_col)['OPAnnualDeductibleAmt'].transform(operation)\n",
    "\n",
    "        feat_7 = feat_name + \"_Admit_Duration\"\n",
    "        test_iobp_df[feat_7] = test_iobp_df.groupby(grp_col)['Admitted_Duration'].transform(operation)\n",
    "\n",
    "        feat_8 = feat_name + \"_Claim_Duration\"\n",
    "        test_iobp_df[feat_8] = test_iobp_df.groupby(grp_col)['Claim_Duration'].transform(operation)\n",
    "      \n",
    "    # BENE, PHYs, Diagnosis Admit and Group Codes columns\n",
    "    create_agg_feats(grp_col='BeneID', feat_name=\"BENE\")\n",
    "    create_agg_feats(grp_col='AttendingPhysician', feat_name=\"ATT_PHY\")\n",
    "    create_agg_feats(grp_col='OperatingPhysician', feat_name=\"OPT_PHY\")\n",
    "    create_agg_feats(grp_col='OtherPhysician', feat_name=\"OTH_PHY\")\n",
    "    create_agg_feats(grp_col='ClmAdmitDiagnosisCode', feat_name=\"Claim_Admit_Diag_Code\")\n",
    "    create_agg_feats(grp_col='DiagnosisGroupCode', feat_name=\"Diag_GCode\")\n",
    "\n",
    "    # Dropping these 3 columns as there 99% of values are same\n",
    "    test_iobp_df.drop(['ClmProcedureCode_4', 'ClmProcedureCode_5', 'ClmProcedureCode_6'], axis=1, inplace=True)\n",
    "\n",
    "    # Diagnosis Codes columns\n",
    "    create_agg_feats(grp_col='ClmDiagnosisCode_1', feat_name=\"Claim_DiagCode1\")\n",
    "    create_agg_feats(grp_col='ClmDiagnosisCode_2', feat_name=\"Claim_DiagCode2\")\n",
    "    create_agg_feats(grp_col='ClmDiagnosisCode_3', feat_name=\"Claim_DiagCode3\")\n",
    "    create_agg_feats(grp_col='ClmDiagnosisCode_4', feat_name=\"Claim_DiagCode4\")\n",
    "    create_agg_feats(grp_col='ClmDiagnosisCode_5', feat_name=\"Claim_DiagCode5\")\n",
    "    create_agg_feats(grp_col='ClmDiagnosisCode_6', feat_name=\"Claim_DiagCode6\")\n",
    "    create_agg_feats(grp_col='ClmDiagnosisCode_7', feat_name=\"Claim_DiagCode7\")\n",
    "    create_agg_feats(grp_col='ClmDiagnosisCode_8', feat_name=\"Claim_DiagCode8\")\n",
    "    create_agg_feats(grp_col='ClmDiagnosisCode_9', feat_name=\"Claim_DiagCode9\")\n",
    "    create_agg_feats(grp_col='ClmDiagnosisCode_10', feat_name=\"Claim_DiagCode10\")\n",
    "\n",
    "    # Medical Procedure Codes columns\n",
    "    create_agg_feats(grp_col='ClmProcedureCode_1', feat_name=\"Claim_ProcCode1\")\n",
    "    create_agg_feats(grp_col='ClmProcedureCode_2', feat_name=\"Claim_ProcCode2\")\n",
    "    create_agg_feats(grp_col='ClmProcedureCode_3', feat_name=\"Claim_ProcCode3\")\n",
    "\n",
    "    # PROVIDER <--> other features :: To get claim counts\n",
    "    test_iobp_df[\"ClmCount_Provider\"]=test_iobp_df.groupby(['Provider'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID\"]=test_iobp_df.groupby(['Provider','BeneID'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_AttendingPhysician\"]=test_iobp_df.groupby(['Provider','AttendingPhysician'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_OtherPhysician\"]=test_iobp_df.groupby(['Provider','OtherPhysician'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_OperatingPhysician\"]=test_iobp_df.groupby(['Provider','OperatingPhysician'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_ClmAdmitDiagnosisCode\"]=test_iobp_df.groupby(['Provider','ClmAdmitDiagnosisCode'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_ClmProcedureCode_1\"]=test_iobp_df.groupby(['Provider','ClmProcedureCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_ClmProcedureCode_2\"]=test_iobp_df.groupby(['Provider','ClmProcedureCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_ClmProcedureCode_3\"]=test_iobp_df.groupby(['Provider','ClmProcedureCode_3'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_ClmDiagnosisCode_1\"]=test_iobp_df.groupby(['Provider','ClmDiagnosisCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_ClmDiagnosisCode_2\"]=test_iobp_df.groupby(['Provider','ClmDiagnosisCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_ClmDiagnosisCode_3\"]=test_iobp_df.groupby(['Provider','ClmDiagnosisCode_3'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_ClmDiagnosisCode_4\"]=test_iobp_df.groupby(['Provider','ClmDiagnosisCode_4'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_ClmDiagnosisCode_5\"]=test_iobp_df.groupby(['Provider','ClmDiagnosisCode_5'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_ClmDiagnosisCode_6\"]=test_iobp_df.groupby(['Provider','ClmDiagnosisCode_6'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_ClmDiagnosisCode_7\"]=test_iobp_df.groupby(['Provider','ClmDiagnosisCode_7'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_ClmDiagnosisCode_8\"]=test_iobp_df.groupby(['Provider','ClmDiagnosisCode_8'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_ClmDiagnosisCode_9\"]=test_iobp_df.groupby(['Provider','ClmDiagnosisCode_9'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_ClmDiagnosisCode_10\"]=test_iobp_df.groupby(['Provider','ClmDiagnosisCode_10'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_DiagnosisGroupCode\"]=test_iobp_df.groupby(['Provider','DiagnosisGroupCode'])['ClaimID'].transform('count')\n",
    "\n",
    "    # PROVIDER <--> BENE <--> PHYSICIANS :: To get claim counts\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_AttendingPhysician\"]=test_iobp_df.groupby(['Provider','BeneID','AttendingPhysician'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OtherPhysician\"]=test_iobp_df.groupby(['Provider','BeneID','OtherPhysician'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OperatingPhysician\"]=test_iobp_df.groupby(['Provider','BeneID','OperatingPhysician'])['ClaimID'].transform('count')\n",
    "\n",
    "    # PROVIDER <--> BENE <--> ATTENDING PHYSICIAN <--> PROCEDURE CODES :: To get claim counts\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_AttendingPhysician_ClmProcedureCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','AttendingPhysician','ClmProcedureCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_AttendingPhysician_ClmProcedureCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','AttendingPhysician','ClmProcedureCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_AttendingPhysician_ClmProcedureCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','AttendingPhysician','ClmProcedureCode_3'])['ClaimID'].transform('count')\n",
    "\n",
    "    # PROVIDER <--> BENE <--> OPERATING PHYSICIAN <--> PROCEDURE CODES :: To get claim counts\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OperatingPhysician_ClmProcedureCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','OperatingPhysician','ClmProcedureCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OperatingPhysician_ClmProcedureCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','OperatingPhysician','ClmProcedureCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OperatingPhysician_ClmProcedureCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','OperatingPhysician','ClmProcedureCode_3'])['ClaimID'].transform('count')\n",
    "\n",
    "    # PROVIDER <--> BENE <--> OTHER PHYSICIAN <--> PROCEDURE CODES :: To get claim counts\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OtherPhysician_ClmProcedureCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','OtherPhysician','ClmProcedureCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OtherPhysician_ClmProcedureCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','OtherPhysician','ClmProcedureCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OtherPhysician_ClmProcedureCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','OtherPhysician','ClmProcedureCode_3'])['ClaimID'].transform('count')\n",
    "\n",
    "    # PROVIDER <--> BENE <--> ATTENDING PHYSICIAN <--> DIAGNOSIS CODES :: To get claim counts\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_AttendingPhysician_ClmDiagnosisCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','AttendingPhysician','ClmDiagnosisCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_AttendingPhysician_ClmDiagnosisCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','AttendingPhysician','ClmDiagnosisCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_AttendingPhysician_ClmDiagnosisCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','AttendingPhysician','ClmDiagnosisCode_3'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_AttendingPhysician_ClmDiagnosisCode_4\"]=test_iobp_df.groupby(['Provider','BeneID','AttendingPhysician','ClmDiagnosisCode_4'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_AttendingPhysician_ClmDiagnosisCode_5\"]=test_iobp_df.groupby(['Provider','BeneID','AttendingPhysician','ClmDiagnosisCode_5'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_AttendingPhysician_ClmDiagnosisCode_6\"]=test_iobp_df.groupby(['Provider','BeneID','AttendingPhysician','ClmDiagnosisCode_6'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_AttendingPhysician_ClmDiagnosisCode_7\"]=test_iobp_df.groupby(['Provider','BeneID','AttendingPhysician','ClmDiagnosisCode_7'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_AttendingPhysician_ClmDiagnosisCode_8\"]=test_iobp_df.groupby(['Provider','BeneID','AttendingPhysician','ClmDiagnosisCode_8'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_AttendingPhysician_ClmDiagnosisCode_9\"]=test_iobp_df.groupby(['Provider','BeneID','AttendingPhysician','ClmDiagnosisCode_9'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_AttendingPhysician_ClmDiagnosisCode_10\"]=test_iobp_df.groupby(['Provider','BeneID','AttendingPhysician','ClmDiagnosisCode_10'])['ClaimID'].transform('count')\n",
    "\n",
    "    # PROVIDER <--> BENE <--> OPERATING PHYSICIAN <--> DIAGNOSIS CODES :: To get claim counts\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OperatingPhysician_ClmDiagnosisCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','OperatingPhysician','ClmDiagnosisCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OperatingPhysician_ClmDiagnosisCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','OperatingPhysician','ClmDiagnosisCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OperatingPhysician_ClmDiagnosisCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','OperatingPhysician','ClmDiagnosisCode_3'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OperatingPhysician_ClmDiagnosisCode_4\"]=test_iobp_df.groupby(['Provider','BeneID','OperatingPhysician','ClmDiagnosisCode_4'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OperatingPhysician_ClmDiagnosisCode_5\"]=test_iobp_df.groupby(['Provider','BeneID','OperatingPhysician','ClmDiagnosisCode_5'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OperatingPhysician_ClmDiagnosisCode_6\"]=test_iobp_df.groupby(['Provider','BeneID','OperatingPhysician','ClmDiagnosisCode_6'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OperatingPhysician_ClmDiagnosisCode_7\"]=test_iobp_df.groupby(['Provider','BeneID','OperatingPhysician','ClmDiagnosisCode_7'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OperatingPhysician_ClmDiagnosisCode_8\"]=test_iobp_df.groupby(['Provider','BeneID','OperatingPhysician','ClmDiagnosisCode_8'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OperatingPhysician_ClmDiagnosisCode_9\"]=test_iobp_df.groupby(['Provider','BeneID','OperatingPhysician','ClmDiagnosisCode_9'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OperatingPhysician_ClmDiagnosisCode_10\"]=test_iobp_df.groupby(['Provider','BeneID','OperatingPhysician','ClmDiagnosisCode_10'])['ClaimID'].transform('count')\n",
    "\n",
    "    # PROVIDER <--> BENE <--> OTHER PHYSICIAN <--> DIAGNOSIS CODES :: To get claim counts\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OtherPhysician_ClmDiagnosisCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','OtherPhysician','ClmDiagnosisCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OtherPhysician_ClmDiagnosisCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','OtherPhysician','ClmDiagnosisCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OtherPhysician_ClmDiagnosisCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','OtherPhysician','ClmDiagnosisCode_3'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OtherPhysician_ClmDiagnosisCode_4\"]=test_iobp_df.groupby(['Provider','BeneID','OtherPhysician','ClmDiagnosisCode_4'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OtherPhysician_ClmDiagnosisCode_5\"]=test_iobp_df.groupby(['Provider','BeneID','OtherPhysician','ClmDiagnosisCode_5'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OtherPhysician_ClmDiagnosisCode_6\"]=test_iobp_df.groupby(['Provider','BeneID','OtherPhysician','ClmDiagnosisCode_6'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OtherPhysician_ClmDiagnosisCode_7\"]=test_iobp_df.groupby(['Provider','BeneID','OtherPhysician','ClmDiagnosisCode_7'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OtherPhysician_ClmDiagnosisCode_8\"]=test_iobp_df.groupby(['Provider','BeneID','OtherPhysician','ClmDiagnosisCode_8'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OtherPhysician_ClmDiagnosisCode_9\"]=test_iobp_df.groupby(['Provider','BeneID','OtherPhysician','ClmDiagnosisCode_9'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_OtherPhysician_ClmDiagnosisCode_10\"]=test_iobp_df.groupby(['Provider','BeneID','OtherPhysician','ClmDiagnosisCode_10'])['ClaimID'].transform('count')\n",
    "\n",
    "    # PROVIDER <--> BENE <--> PROCEDURE CODES :: To get claim counts\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmProcedureCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','ClmProcedureCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmProcedureCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','ClmProcedureCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmProcedureCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','ClmProcedureCode_3'])['ClaimID'].transform('count')\n",
    "\n",
    "    # PROVIDER <--> BENE <--> DIAGNOSIS CODES :: To get claim counts\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_3'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_4\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_4'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_5\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_5'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_6\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_6'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_7\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_7'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_8\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_8'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_9\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_9'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_10\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_10'])['ClaimID'].transform('count')\n",
    "\n",
    "    # PROVIDER <--> BENE <--> DIAGNOSIS CODES <--> PROCEDURE CODES :: To get claim counts\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_1_ClmProcedureCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_1','ClmProcedureCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_1_ClmProcedureCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_1','ClmProcedureCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_1_ClmProcedureCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_1','ClmProcedureCode_3'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_2_ClmProcedureCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_2','ClmProcedureCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_2_ClmProcedureCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_2','ClmProcedureCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_2_ClmProcedureCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_2','ClmProcedureCode_3'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_3_ClmProcedureCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_3','ClmProcedureCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_3_ClmProcedureCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_3','ClmProcedureCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_3_ClmProcedureCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_3','ClmProcedureCode_3'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_4_ClmProcedureCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_4','ClmProcedureCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_4_ClmProcedureCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_4','ClmProcedureCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_4_ClmProcedureCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_4','ClmProcedureCode_3'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_5_ClmProcedureCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_5','ClmProcedureCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_5_ClmProcedureCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_5','ClmProcedureCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_5_ClmProcedureCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_5','ClmProcedureCode_3'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_6_ClmProcedureCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_6','ClmProcedureCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_6_ClmProcedureCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_6','ClmProcedureCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_6_ClmProcedureCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_6','ClmProcedureCode_3'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_7_ClmProcedureCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_7','ClmProcedureCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_7_ClmProcedureCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_7','ClmProcedureCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_7_ClmProcedureCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_7','ClmProcedureCode_3'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_8_ClmProcedureCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_8','ClmProcedureCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_8_ClmProcedureCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_8','ClmProcedureCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_8_ClmProcedureCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_8','ClmProcedureCode_3'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_9_ClmProcedureCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_9','ClmProcedureCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_9_ClmProcedureCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_9','ClmProcedureCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_9_ClmProcedureCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_9','ClmProcedureCode_3'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_10_ClmProcedureCode_1\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_10','ClmProcedureCode_1'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_10_ClmProcedureCode_2\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_10','ClmProcedureCode_2'])['ClaimID'].transform('count')\n",
    "    test_iobp_df[\"ClmCount_Provider_BeneID_ClmDiagnosisCode_10_ClmProcedureCode_3\"]=test_iobp_df.groupby(['Provider','BeneID','ClmDiagnosisCode_10','ClmProcedureCode_3'])['ClaimID'].transform('count')\n",
    "\n",
    "    # Removing unwanted columns\n",
    "    remove_unwanted_columns=['BeneID', 'ClaimID', 'ClaimStartDt','ClaimEndDt','AttendingPhysician','OperatingPhysician', 'OtherPhysician',\n",
    "                          'AdmissionDt', 'ClmAdmitDiagnosisCode', 'DischargeDt', 'DiagnosisGroupCode',\n",
    "                          'ClmDiagnosisCode_1', 'ClmDiagnosisCode_2', 'ClmDiagnosisCode_3', 'ClmDiagnosisCode_4', 'ClmDiagnosisCode_5', \n",
    "                          'ClmDiagnosisCode_6', 'ClmDiagnosisCode_7', 'ClmDiagnosisCode_8', 'ClmDiagnosisCode_9', 'ClmDiagnosisCode_10',\n",
    "                          'ClmProcedureCode_1', 'ClmProcedureCode_2', 'ClmProcedureCode_3', 'DOB', 'DOD', 'State', 'County']\n",
    "\n",
    "    test_iobp_df.drop(columns=remove_unwanted_columns, axis=1, inplace=True)\n",
    "\n",
    "    # Filling Nulls in Deductible Amt Paid by Patient\n",
    "    test_iobp_df['DeductibleAmtPaid'].fillna(value=0, inplace=True)\n",
    "\n",
    "    # Binary encoding the categorical features --> 0 means No and 1 means Yes\n",
    "    test_iobp_df['Gender'] = test_iobp_df['Gender'].apply(lambda val: 0 if val == 2 else val)\n",
    "    test_iobp_df['Is_Alive?'] = test_iobp_df['Is_Alive?'].apply(lambda val: 0 if val == \"No\" else 1)\n",
    "\n",
    "    test_iobp_df['ChronicCond_Alzheimer'] = test_iobp_df['ChronicCond_Alzheimer'].apply(lambda val: 0 if val == 2 else val)\n",
    "    test_iobp_df['ChronicCond_Heartfailure'] = test_iobp_df['ChronicCond_Heartfailure'].apply(lambda val: 0 if val == 2 else val)\n",
    "    test_iobp_df['ChronicCond_KidneyDisease'] = test_iobp_df['ChronicCond_KidneyDisease'].apply(lambda val: 0 if val == 2 else val)\n",
    "    test_iobp_df['ChronicCond_Cancer'] = test_iobp_df['ChronicCond_Cancer'].apply(lambda val: 0 if val == 2 else val)\n",
    "    test_iobp_df['ChronicCond_ObstrPulmonary'] = test_iobp_df['ChronicCond_ObstrPulmonary'].apply(lambda val: 0 if val == 2 else val)\n",
    "    test_iobp_df['ChronicCond_Depression'] = test_iobp_df['ChronicCond_Depression'].apply(lambda val: 0 if val == 2 else val)\n",
    "    test_iobp_df['ChronicCond_Diabetes'] = test_iobp_df['ChronicCond_Diabetes'].apply(lambda val: 0 if val == 2 else val)\n",
    "    test_iobp_df['ChronicCond_IschemicHeart'] = test_iobp_df['ChronicCond_IschemicHeart'].apply(lambda val: 0 if val == 2 else val)\n",
    "    test_iobp_df['ChronicCond_Osteoporasis'] = test_iobp_df['ChronicCond_Osteoporasis'].apply(lambda val: 0 if val == 2 else val)\n",
    "    test_iobp_df['ChronicCond_rheumatoidarthritis'] = test_iobp_df['ChronicCond_rheumatoidarthritis'].apply(lambda val: 0 if val == 2 else val)\n",
    "    test_iobp_df['ChronicCond_stroke'] = test_iobp_df['ChronicCond_stroke'].apply(lambda val: 0 if val == 2 else val)\n",
    "\n",
    "    # Encoding the Categorical features\n",
    "    test_iobp_df = pd.get_dummies(test_iobp_df,columns=['Gender', 'Race', 'Admitted?', 'Is_Alive?'], drop_first=True)\n",
    "\n",
    "    # Filling Nulls in the aggregated features\n",
    "    test_iobp_df.fillna(value=0, inplace=True)\n",
    "\n",
    "    # Grouping the records\n",
    "    test_iobp_df = test_iobp_df.groupby(['Provider'],as_index=False).agg('sum')\n",
    "\n",
    "    # Segregating the sets\n",
    "    X_unseen = test_iobp_df.drop(axis=1, columns=['Provider'])\n",
    "    y_unseen_prvs = test_iobp_df['Provider']\n",
    "\n",
    "    return X_unseen, y_unseen_prvs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T11:55:09.708565Z",
     "iopub.status.busy": "2022-09-02T11:55:09.707565Z",
     "iopub.status.idle": "2022-09-02T11:55:57.813033Z",
     "shell.execute_reply": "2022-09-02T11:55:57.810032Z",
     "shell.execute_reply.started": "2022-09-02T11:55:09.708565Z"
    },
    "id": "7W7Aiknqyngx"
   },
   "outputs": [],
   "source": [
    "X_unseen_pp, y_unseen_pp = pre_process(unseen_data_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T11:55:57.817034Z",
     "iopub.status.busy": "2022-09-02T11:55:57.816035Z",
     "iopub.status.idle": "2022-09-02T11:55:57.843035Z",
     "shell.execute_reply": "2022-09-02T11:55:57.841039Z",
     "shell.execute_reply.started": "2022-09-02T11:55:57.817034Z"
    },
    "id": "ac_7YDkcyngy",
    "outputId": "eb83fe9e-2362-4437-884d-e0f808a8e189"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((1353, 299), (1353,))"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Checking shape of unseen data after pre-processing it should have 299 features\n",
    "X_unseen_pp.shape, y_unseen_pp.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T11:56:06.938983Z",
     "iopub.status.busy": "2022-09-02T11:56:06.937985Z",
     "iopub.status.idle": "2022-09-02T11:56:07.893151Z",
     "shell.execute_reply": "2022-09-02T11:56:07.891155Z",
     "shell.execute_reply.started": "2022-09-02T11:56:06.938983Z"
    }
   },
   "outputs": [],
   "source": [
    "X_unseen_pp.to_csv(\"X_unseen_pp.csv\")\n",
    "y_unseen_pp.to_csv(\"y_unseen_pp.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T11:56:54.482273Z",
     "iopub.status.busy": "2022-09-02T11:56:54.481278Z",
     "iopub.status.idle": "2022-09-02T11:56:54.522274Z",
     "shell.execute_reply": "2022-09-02T11:56:54.520275Z",
     "shell.execute_reply.started": "2022-09-02T11:56:54.482273Z"
    }
   },
   "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>Provider</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PRV51002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>PRV51006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PRV51009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>PRV51010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>PRV51018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1348</th>\n",
       "      <td>PRV57713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1349</th>\n",
       "      <td>PRV57726</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1350</th>\n",
       "      <td>PRV57745</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1351</th>\n",
       "      <td>PRV57749</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1352</th>\n",
       "      <td>PRV57750</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1353 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "      Provider\n",
       "0     PRV51002\n",
       "1     PRV51006\n",
       "2     PRV51009\n",
       "3     PRV51010\n",
       "4     PRV51018\n",
       "...        ...\n",
       "1348  PRV57713\n",
       "1349  PRV57726\n",
       "1350  PRV57745\n",
       "1351  PRV57749\n",
       "1352  PRV57750\n",
       "\n",
       "[1353 rows x 1 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.read_csv(\"y_unseen_pp.csv\", index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T12:33:05.241664Z",
     "iopub.status.busy": "2022-09-02T12:33:05.239668Z",
     "iopub.status.idle": "2022-09-02T12:33:05.386664Z",
     "shell.execute_reply": "2022-09-02T12:33:05.384662Z",
     "shell.execute_reply.started": "2022-09-02T12:33:05.241664Z"
    },
    "id": "WdWTib4Fyngz"
   },
   "outputs": [],
   "source": [
    "# Loading the best model\n",
    "rfc_3 = joblib.load('best_model.pkl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T12:33:05.394666Z",
     "iopub.status.busy": "2022-09-02T12:33:05.390666Z",
     "iopub.status.idle": "2022-09-02T12:33:05.448691Z",
     "shell.execute_reply": "2022-09-02T12:33:05.446662Z",
     "shell.execute_reply.started": "2022-09-02T12:33:05.393665Z"
    },
    "id": "uUDPeMDGyng1"
   },
   "outputs": [],
   "source": [
    "# Making the predictions on the unseen data\n",
    "unseen_preds = rfc_3.predict(X_unseen_pp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T12:35:20.455051Z",
     "iopub.status.busy": "2022-09-02T12:35:20.454047Z",
     "iopub.status.idle": "2022-09-02T12:35:20.519046Z",
     "shell.execute_reply": "2022-09-02T12:35:20.516055Z",
     "shell.execute_reply.started": "2022-09-02T12:35:20.455051Z"
    }
   },
   "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>Class_0</th>\n",
       "      <th>Class_1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.088072</td>\n",
       "      <td>0.911928</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.273612</td>\n",
       "      <td>0.726388</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.107149</td>\n",
       "      <td>0.892851</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.077191</td>\n",
       "      <td>0.922809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.084086</td>\n",
       "      <td>0.915914</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1348</th>\n",
       "      <td>0.803848</td>\n",
       "      <td>0.196152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1349</th>\n",
       "      <td>0.708234</td>\n",
       "      <td>0.291766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1350</th>\n",
       "      <td>0.828519</td>\n",
       "      <td>0.171481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1351</th>\n",
       "      <td>0.222062</td>\n",
       "      <td>0.777938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1352</th>\n",
       "      <td>0.204629</td>\n",
       "      <td>0.795371</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1353 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Class_0   Class_1\n",
       "0     0.088072  0.911928\n",
       "1     0.273612  0.726388\n",
       "2     0.107149  0.892851\n",
       "3     0.077191  0.922809\n",
       "4     0.084086  0.915914\n",
       "...        ...       ...\n",
       "1348  0.803848  0.196152\n",
       "1349  0.708234  0.291766\n",
       "1350  0.828519  0.171481\n",
       "1351  0.222062  0.777938\n",
       "1352  0.204629  0.795371\n",
       "\n",
       "[1353 rows x 2 columns]"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Generating the predictions probabilities on the unseen data\n",
    "unseen_preds_probs = rfc_3.predict_proba(X_unseen_pp)\n",
    "unseen_preds_probs_df = pd.DataFrame(unseen_preds_probs)\n",
    "unseen_preds_probs_df.columns=['Class_0', 'Class_1']\n",
    "unseen_preds_probs_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T12:36:47.060983Z",
     "iopub.status.busy": "2022-09-02T12:36:47.058977Z",
     "iopub.status.idle": "2022-09-02T12:36:47.122988Z",
     "shell.execute_reply": "2022-09-02T12:36:47.119989Z",
     "shell.execute_reply.started": "2022-09-02T12:36:47.059983Z"
    },
    "id": "IE71xD-Tyng2",
    "outputId": "f5fdc542-682b-430a-e1d1-52ee1635ad72"
   },
   "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>Providers</th>\n",
       "      <th>Potential_Fraud?</th>\n",
       "      <th>Non_Fraud_Prob_Score</th>\n",
       "      <th>Fraud_Prob_Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PRV51002</td>\n",
       "      <td>1</td>\n",
       "      <td>0.088072</td>\n",
       "      <td>0.911928</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>PRV51006</td>\n",
       "      <td>1</td>\n",
       "      <td>0.273612</td>\n",
       "      <td>0.726388</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PRV51009</td>\n",
       "      <td>1</td>\n",
       "      <td>0.107149</td>\n",
       "      <td>0.892851</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>PRV51010</td>\n",
       "      <td>1</td>\n",
       "      <td>0.077191</td>\n",
       "      <td>0.922809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>PRV51018</td>\n",
       "      <td>1</td>\n",
       "      <td>0.084086</td>\n",
       "      <td>0.915914</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1348</th>\n",
       "      <td>PRV57713</td>\n",
       "      <td>0</td>\n",
       "      <td>0.803848</td>\n",
       "      <td>0.196152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1349</th>\n",
       "      <td>PRV57726</td>\n",
       "      <td>0</td>\n",
       "      <td>0.708234</td>\n",
       "      <td>0.291766</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1350</th>\n",
       "      <td>PRV57745</td>\n",
       "      <td>0</td>\n",
       "      <td>0.828519</td>\n",
       "      <td>0.171481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1351</th>\n",
       "      <td>PRV57749</td>\n",
       "      <td>1</td>\n",
       "      <td>0.222062</td>\n",
       "      <td>0.777938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1352</th>\n",
       "      <td>PRV57750</td>\n",
       "      <td>1</td>\n",
       "      <td>0.204629</td>\n",
       "      <td>0.795371</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1353 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Providers  Potential_Fraud?  Non_Fraud_Prob_Score  Fraud_Prob_Score\n",
       "0     PRV51002                 1              0.088072          0.911928\n",
       "1     PRV51006                 1              0.273612          0.726388\n",
       "2     PRV51009                 1              0.107149          0.892851\n",
       "3     PRV51010                 1              0.077191          0.922809\n",
       "4     PRV51018                 1              0.084086          0.915914\n",
       "...        ...               ...                   ...               ...\n",
       "1348  PRV57713                 0              0.803848          0.196152\n",
       "1349  PRV57726                 0              0.708234          0.291766\n",
       "1350  PRV57745                 0              0.828519          0.171481\n",
       "1351  PRV57749                 1              0.222062          0.777938\n",
       "1352  PRV57750                 1              0.204629          0.795371\n",
       "\n",
       "[1353 rows x 4 columns]"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Predicted Results of all the providers\n",
    "unseen_results = pd.concat([pd.DataFrame(y_unseen_pp), pd.DataFrame(unseen_preds), unseen_preds_probs_df],axis=1)\n",
    "unseen_results.reset_index(drop=True, inplace=True)\n",
    "unseen_results.columns = ['Providers','Potential_Fraud?', 'Non_Fraud_Prob_Score', 'Fraud_Prob_Score']\n",
    "unseen_results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T13:45:41.458200Z",
     "iopub.status.busy": "2022-09-02T13:45:41.457195Z",
     "iopub.status.idle": "2022-09-02T13:45:41.497191Z",
     "shell.execute_reply": "2022-09-02T13:45:41.495191Z",
     "shell.execute_reply.started": "2022-09-02T13:45:41.458200Z"
    }
   },
   "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>Providers</th>\n",
       "      <th>Potential_Fraud?</th>\n",
       "      <th>Non_Fraud_Prob_Score</th>\n",
       "      <th>Fraud_Prob_Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PRV51009</td>\n",
       "      <td>1</td>\n",
       "      <td>0.107149</td>\n",
       "      <td>0.892851</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Providers  Potential_Fraud?  Non_Fraud_Prob_Score  Fraud_Prob_Score\n",
       "2  PRV51009                 1              0.107149          0.892851"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check whether a specific PROVIDER is frauster or not?\n",
    "unseen_results[unseen_results['Providers'] == 'PRV51009']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "r-VhXYclyng3"
   },
   "source": [
    "### **`SUMMARY`**\n",
    "- Here, we have generated the predictions from the best-trained model on the unseen dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "0IC2VKpXyng4"
   },
   "source": [
    "## **Performance_on_TRAIN_and_VALIDATION_Sets**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T12:03:25.956200Z",
     "iopub.status.busy": "2022-09-02T12:03:25.955199Z",
     "iopub.status.idle": "2022-09-02T12:03:26.044197Z",
     "shell.execute_reply": "2022-09-02T12:03:26.042195Z",
     "shell.execute_reply.started": "2022-09-02T12:03:25.956200Z"
    },
    "id": "Lb5Hc96Jci_h"
   },
   "outputs": [],
   "source": [
    "def pred_prob(clf, data): \n",
    "    \"\"\"\n",
    "    Description :: This function is created for storing the predicted probabability using the trained model.\n",
    "    \n",
    "    Input :: It accepts below input parameters :\n",
    "      - clf : Trained model classifier\n",
    "      - data : Dataset for which we want to generate the predictions\n",
    "    \"\"\"\n",
    "    y_pred = clf.predict_proba(data)[:,1]\n",
    "    return y_pred\n",
    "\n",
    "def draw_roc(train_fpr, train_tpr, test_fpr, test_tpr):\n",
    "    \"\"\"\n",
    "    Description :: This function is created for calculating the AUC score on train and test data. And, plotting the ROC curve.\n",
    "    \n",
    "    Input :: It accepts below input parameters :\n",
    "      - train_fpr : Train False +ve rate\n",
    "      - train_tpr : Train True +ve rate\n",
    "      - test_fpr : Test False +ve rate\n",
    "      - test_tpr : Test True +ve rate\n",
    "    \"\"\"\n",
    "    # calculate auc for train and test\n",
    "    train_auc = auc(train_fpr, train_tpr)\n",
    "    test_auc = auc(test_fpr, test_tpr)\n",
    "    with plt.style.context('seaborn-poster'):\n",
    "      plt.plot(train_fpr, train_tpr, label=\"Train AUC =\"+\"{:.4f}\".format(train_auc), color='blue')\n",
    "      plt.plot(test_fpr, test_tpr, label=\"Test AUC =\"+\"{:.4f}\".format(test_auc), color='red')\n",
    "      plt.legend()\n",
    "      plt.xlabel(\"False Positive Rate(FPR)\", fontdict=label_font_dict)\n",
    "      plt.ylabel(\"True Positive Rate(TPR)\", fontdict=label_font_dict)\n",
    "      plt.title(\"Area Under Curve\", fontdict=title_font_dict)\n",
    "      plt.grid(b=True, which='major', color='lightgrey', linestyle='--')\n",
    "      plt.minorticks_on()\n",
    "      plt.show()\n",
    "    \n",
    "def find_best_threshold(threshold, fpr, tpr):\n",
    "    \"\"\"\n",
    "    Description :: This function is created for finding the best threshold value.\n",
    "    \"\"\"\n",
    "    t = threshold[np.argmax(tpr * (1-fpr))]\n",
    "    return t\n",
    "\n",
    "def predict_with_best_t(proba, threshold):\n",
    "    \"\"\"\n",
    "    Description :: This function is created for generating the predictions based on the best threshold value.\n",
    "    \"\"\"\n",
    "    predictions = []\n",
    "    for i in proba:\n",
    "        if i>=threshold:\n",
    "            predictions.append(1)\n",
    "        else:\n",
    "            predictions.append(0)\n",
    "    return predictions\n",
    "\n",
    "def draw_confusion_matrix(best_t, x_train, x_test, y_train, y_test, y_train_pred, y_test_pred):\n",
    "    \"\"\"\n",
    "    Description :: This function is created for plotting the confusion matrix of TRAIN and TEST sets.\n",
    "    \"\"\"\n",
    "    fig, ax = plt.subplots(1,2, figsize=(20,6))\n",
    "\n",
    "    train_prediction = predict_with_best_t(y_train_pred, best_t)\n",
    "    cm = confusion_matrix(y_train, train_prediction)\n",
    "    with plt.style.context('seaborn'):\n",
    "        sns.heatmap(cm, annot=True, fmt='d', ax=ax[0], cmap='viridis')\n",
    "        ax[0].set_title('Train Dataset Confusion Matrix', fontdict=title_font_dict)\n",
    "        ax[0].set_xlabel(\"Predicted Label\", fontdict=label_font_dict)\n",
    "        ax[0].set_ylabel(\"Actual Label\", fontdict=label_font_dict)\n",
    "\n",
    "    test_prediction = predict_with_best_t(y_test_pred, best_t)\n",
    "    cm = confusion_matrix(y_test, test_prediction)\n",
    "    with plt.style.context('seaborn'):\n",
    "        sns.heatmap(cm, annot=True, fmt='d', ax=ax[1], cmap='summer')\n",
    "        ax[1].set_title('Test Dataset Confusion Matrix', fontdict=title_font_dict)\n",
    "        ax[1].set_xlabel(\"Predicted Label\", fontdict=label_font_dict)\n",
    "        ax[1].set_ylabel(\"Actual Label\", fontdict=label_font_dict)\n",
    "    \n",
    "    plt.show()\n",
    "    \n",
    "    return train_prediction, test_prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T12:03:26.047198Z",
     "iopub.status.busy": "2022-09-02T12:03:26.046200Z",
     "iopub.status.idle": "2022-09-02T12:03:26.094198Z",
     "shell.execute_reply": "2022-09-02T12:03:26.089201Z",
     "shell.execute_reply.started": "2022-09-02T12:03:26.047198Z"
    },
    "id": "_r3xu3XXeS9v"
   },
   "outputs": [],
   "source": [
    "def validate_model(clf, x_train, x_test, y_train, y_test):\n",
    "    \"\"\"\n",
    "    Description :: This function is created for performing the evaluation of the trained model.\n",
    "    \"\"\"\n",
    "    # predict the probability of train data\n",
    "    y_train_pred = pred_prob(clf, x_train)\n",
    "    \n",
    "    # predict the probability of test data\n",
    "    y_test_pred = pred_prob(clf, x_test)\n",
    "    \n",
    "    # calculate tpr, fpr using roc_curve\n",
    "    train_fpr, train_tpr, tr_thresholds = roc_curve(y_train, y_train_pred)\n",
    "    test_fpr, test_tpr, te_thresholds = roc_curve(y_test, y_test_pred)\n",
    "    \n",
    "    # calculate auc for train and test\n",
    "    train_auc = auc(train_fpr, train_tpr)\n",
    "    print(\"### Train AUC = {}\".format(train_auc))\n",
    "    test_auc = auc(test_fpr, test_tpr)\n",
    "    print(\"### Test AUC = {}\".format(test_auc))\n",
    "    \n",
    "    # plotting the ROC curve\n",
    "    draw_roc(train_fpr, train_tpr, test_fpr, test_tpr)\n",
    "    \n",
    "    # Best threshold value\n",
    "    best_t = find_best_threshold(tr_thresholds, train_fpr, train_tpr)\n",
    "    \n",
    "    # Plotting the confusion matrices\n",
    "    train_prediction, test_prediction = draw_confusion_matrix(best_t, x_train, x_test, y_train, y_test, y_train_pred, y_test_pred)\n",
    "    \n",
    "    # Generating the F1-scores\n",
    "    train_f1_score = f1_score(y_train, train_prediction)\n",
    "    test_f1_score = f1_score(y_test, test_prediction)\n",
    "    \n",
    "    return test_auc, train_f1_score, test_f1_score, best_t"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T12:03:26.099197Z",
     "iopub.status.busy": "2022-09-02T12:03:26.097199Z",
     "iopub.status.idle": "2022-09-02T12:03:27.781399Z",
     "shell.execute_reply": "2022-09-02T12:03:27.777398Z",
     "shell.execute_reply.started": "2022-09-02T12:03:26.099197Z"
    },
    "id": "zK3dEaKAnugE"
   },
   "outputs": [],
   "source": [
    "# Reading the TRAIN and VALIDATION sets for checking the model performance\n",
    "X_train_std = pd.read_csv(\"X_train_std.csv\", index_col=0)\n",
    "X_test_std = pd.read_csv(\"X_test_std.csv\", index_col=0)\n",
    "\n",
    "y_train = pd.read_csv(\"y_train.csv\", index_col=0)\n",
    "y_test = pd.read_csv(\"y_test.csv\", index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T12:03:27.786397Z",
     "iopub.status.busy": "2022-09-02T12:03:27.785400Z",
     "iopub.status.idle": "2022-09-02T12:03:27.814401Z",
     "shell.execute_reply": "2022-09-02T12:03:27.810404Z",
     "shell.execute_reply.started": "2022-09-02T12:03:27.786397Z"
    },
    "id": "J37T7AWqVl3F",
    "outputId": "d52b24a2-e29f-4ff1-8c94-714ffae7e1ff"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0, 0, 0, ..., 0, 0, 1], dtype=int64)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.ravel(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T12:03:27.819400Z",
     "iopub.status.busy": "2022-09-02T12:03:27.818399Z",
     "iopub.status.idle": "2022-09-02T12:03:31.413777Z",
     "shell.execute_reply": "2022-09-02T12:03:31.409777Z",
     "shell.execute_reply.started": "2022-09-02T12:03:27.819400Z"
    },
    "id": "Yfji-CNSVl3s",
    "outputId": "7ef2422f-f66a-4523-8684-7a06a6abd775"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "### Train AUC = 0.9464100603279804\n",
      "### Test AUC = 0.9509593161150978\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 921.6x633.6 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x432 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "### Best Threshold = 0.5567\n",
      "### Model AUC is : 0.9510\n",
      "### Model Train F1 Score is : 0.5918\n",
      "### Model Test F1 Score is : 0.5828\n"
     ]
    }
   ],
   "source": [
    "# Validate model\n",
    "test_auc, train_f1_score, test_f1_score, best_t = validate_model(rfc_3, X_train_std, X_test_std, np.ravel(y_train), np.ravel(y_test))\n",
    "\n",
    "print(\"\\n\")\n",
    "print(\"### Best Threshold = {:.4f}\".format(best_t))\n",
    "print(\"### Model AUC is : {:.4f}\".format(test_auc))\n",
    "print(\"### Model Train F1 Score is : {:.4f}\".format(train_f1_score))\n",
    "print(\"### Model Test F1 Score is : {:.4f}\".format(test_f1_score))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T12:03:31.424777Z",
     "iopub.status.busy": "2022-09-02T12:03:31.423778Z",
     "iopub.status.idle": "2022-09-02T12:03:31.490781Z",
     "shell.execute_reply": "2022-09-02T12:03:31.488923Z",
     "shell.execute_reply.started": "2022-09-02T12:03:31.424777Z"
    },
    "id": "01uoJOm2Vl3t",
    "outputId": "56d9e582-f19f-4ec1-869a-b48145fba06c"
   },
   "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>Features</th>\n",
       "      <th>Importance_Model_1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>InscClaimAmtReimbursed</td>\n",
       "      <td>0.065302</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>DeductibleAmtPaid</td>\n",
       "      <td>0.054782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>RenalDiseaseIndicator</td>\n",
       "      <td>0.000297</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ChronicCond_ObstrPulmonary</td>\n",
       "      <td>0.000880</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ChronicCond_Osteoporasis</td>\n",
       "      <td>0.000377</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     Features  Importance_Model_1\n",
       "0      InscClaimAmtReimbursed            0.065302\n",
       "1           DeductibleAmtPaid            0.054782\n",
       "2       RenalDiseaseIndicator            0.000297\n",
       "3  ChronicCond_ObstrPulmonary            0.000880\n",
       "4    ChronicCond_Osteoporasis            0.000377"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "feats_imps_4 = pd.DataFrame({'Features': X_train_std.columns, 'Importance_Model_1': rfc_3.feature_importances_})\n",
    "feats_imps_4 = feats_imps_4[feats_imps_4['Importance_Model_1'] != 0]\n",
    "feats_imps_4.reset_index(drop=True, inplace=True)\n",
    "feats_imps_4.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T12:03:31.495775Z",
     "iopub.status.busy": "2022-09-02T12:03:31.494782Z",
     "iopub.status.idle": "2022-09-02T12:03:31.554764Z",
     "shell.execute_reply": "2022-09-02T12:03:31.551762Z",
     "shell.execute_reply.started": "2022-09-02T12:03:31.495775Z"
    },
    "id": "XU0dckbeVl3u"
   },
   "outputs": [],
   "source": [
    "top_20_pos_feats_4 = feats_imps_4.sort_values(by='Importance_Model_1',axis=0,ascending=False)['Features'].iloc[0:20]\n",
    "top_20_pos_feats_scores_4 = feats_imps_4.sort_values(by='Importance_Model_1',axis=0,ascending=False)['Importance_Model_1'].iloc[0:20]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T12:03:31.558767Z",
     "iopub.status.busy": "2022-09-02T12:03:31.557766Z",
     "iopub.status.idle": "2022-09-02T12:03:33.573379Z",
     "shell.execute_reply": "2022-09-02T12:03:33.571382Z",
     "shell.execute_reply.started": "2022-09-02T12:03:31.558767Z"
    },
    "id": "A0779i_KVl3v",
    "outputId": "b1dc2ebb-4311-4db9-cec4-3f40439ae3cf"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 921.6x633.6 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "with plt.style.context('seaborn-poster'):\n",
    "    sns.barplot(y=top_20_pos_feats_4, x=top_20_pos_feats_scores_4, orient='h', palette='coolwarm')\n",
    "    plt.xlabel(\"\\nFeatures Importance\", fontdict=label_font_dict)\n",
    "    plt.ylabel(\"Features\\n\", fontdict=label_font_dict)\n",
    "    plt.title(\"Top 20 Importance Positive Features\\n\", fontdict=title_font_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-02T12:03:33.581394Z",
     "iopub.status.busy": "2022-09-02T12:03:33.579379Z",
     "iopub.status.idle": "2022-09-02T12:03:33.622379Z",
     "shell.execute_reply": "2022-09-02T12:03:33.619397Z",
     "shell.execute_reply.started": "2022-09-02T12:03:33.581394Z"
    },
    "id": "GM5nuXaXVl3w",
    "outputId": "478ab880-4c19-4f3e-d323-fb79f07fca16"
   },
   "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>Feature</th>\n",
       "      <th>Imp_Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>PRV_CoPayment</td>\n",
       "      <td>0.081920</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>OPT_PHY_CoPayment</td>\n",
       "      <td>0.070996</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>InscClaimAmtReimbursed</td>\n",
       "      <td>0.065302</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>145</th>\n",
       "      <td>ClmCount_Provider_BeneID_OperatingPhysician_Cl...</td>\n",
       "      <td>0.056550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>DeductibleAmtPaid</td>\n",
       "      <td>0.054782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>144</th>\n",
       "      <td>ClmCount_Provider_BeneID_OperatingPhysician_Cl...</td>\n",
       "      <td>0.052589</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>142</th>\n",
       "      <td>ClmCount_Provider_BeneID_AttendingPhysician_Cl...</td>\n",
       "      <td>0.051732</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153</th>\n",
       "      <td>ClmCount_Provider_BeneID_ClmDiagnosisCode_7</td>\n",
       "      <td>0.051206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Admitted_Duration</td>\n",
       "      <td>0.035899</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>Claim_Admit_Diag_Code_Admit_Duration</td>\n",
       "      <td>0.030555</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>154</th>\n",
       "      <td>ClmCount_Provider_BeneID_ClmDiagnosisCode_8</td>\n",
       "      <td>0.028236</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>146</th>\n",
       "      <td>ClmCount_Provider_BeneID_OperatingPhysician_Cl...</td>\n",
       "      <td>0.028039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131</th>\n",
       "      <td>ClmCount_Provider_ClmDiagnosisCode_9</td>\n",
       "      <td>0.026697</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140</th>\n",
       "      <td>ClmCount_Provider_BeneID_AttendingPhysician_Cl...</td>\n",
       "      <td>0.026592</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>Claim_DiagCode9_OP_Annual_Ded_Amt</td>\n",
       "      <td>0.026252</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>113</th>\n",
       "      <td>Claim_ProcCode1_Insc_ReImb_Amt</td>\n",
       "      <td>0.025006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>Claim_DiagCode1_Admit_Duration</td>\n",
       "      <td>0.024637</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>Claim_DiagCode8_IP_Annual_ReImb_Amt</td>\n",
       "      <td>0.023743</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>129</th>\n",
       "      <td>ClmCount_Provider_ClmDiagnosisCode_7</td>\n",
       "      <td>0.023474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>Claim_DiagCode8_Insc_ReImb_Amt</td>\n",
       "      <td>0.023008</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                               Feature  Imp_Score\n",
       "17                                       PRV_CoPayment   0.081920\n",
       "35                                   OPT_PHY_CoPayment   0.070996\n",
       "0                               InscClaimAmtReimbursed   0.065302\n",
       "145  ClmCount_Provider_BeneID_OperatingPhysician_Cl...   0.056550\n",
       "1                                    DeductibleAmtPaid   0.054782\n",
       "144  ClmCount_Provider_BeneID_OperatingPhysician_Cl...   0.052589\n",
       "142  ClmCount_Provider_BeneID_AttendingPhysician_Cl...   0.051732\n",
       "153        ClmCount_Provider_BeneID_ClmDiagnosisCode_7   0.051206\n",
       "10                                   Admitted_Duration   0.035899\n",
       "50                Claim_Admit_Diag_Code_Admit_Duration   0.030555\n",
       "154        ClmCount_Provider_BeneID_ClmDiagnosisCode_8   0.028236\n",
       "146  ClmCount_Provider_BeneID_OperatingPhysician_Cl...   0.028039\n",
       "131               ClmCount_Provider_ClmDiagnosisCode_9   0.026697\n",
       "140  ClmCount_Provider_BeneID_AttendingPhysician_Cl...   0.026592\n",
       "105                  Claim_DiagCode9_OP_Annual_Ded_Amt   0.026252\n",
       "113                     Claim_ProcCode1_Insc_ReImb_Amt   0.025006\n",
       "63                      Claim_DiagCode1_Admit_Duration   0.024637\n",
       "94                 Claim_DiagCode8_IP_Annual_ReImb_Amt   0.023743\n",
       "129               ClmCount_Provider_ClmDiagnosisCode_7   0.023474\n",
       "92                      Claim_DiagCode8_Insc_ReImb_Amt   0.023008"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result_feats_scrs_4 = pd.DataFrame({'Feature': top_20_pos_feats_4, 'Imp_Score': top_20_pos_feats_scores_4})\n",
    "result_feats_scrs_4"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ZfOButQ2ynhB"
   },
   "source": [
    "### **`SUMMARY`**\n",
    "- Here, we have the most important features on the basis of which model is giving the predictions."
   ]
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "CS_1_Model_Pipeline.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.6.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}