added bigquery

app.py

@@ -1,7 +1,23 @@
 import streamlit as st
 import joblib
 import pandas as pd
+from google.cloud import bigquery
+import datetime
 
+# Retrieve the base64-encoded credentials from environment variables
+encoded_credentials = os.getenv('BIGQUERY_KEY')
+
+# Decode the base64-encoded credentials
+decoded_credentials = base64.b64decode(encoded_credentials).decode('utf-8')
+
+# Set Google Application Credentials
+os.environ['GOOGLE_APPLICATION_CREDENTIALS'] = decoded_credentials
+keypath = r'C:\Users\aruni\Project\Loan_Approval\youtube-analytics-keys.json'
+dataset = 'home_loan'
+table = 'home_loan_approval'
+client = bigquery.Client()
+
+table_ref = client.dataset(dataset).table(table)
 st.title("Welcome to ABC Bank")
 
 model = joblib.load('model_final.joblib')
@@ -11,9 +27,9 @@ with st.form('Loan Form',clear_on_submit=True):
     col1,col2 = st.columns(2)
     with col1:
         Gender = st.selectbox('Gender',('Male','Female'))
-        Applicant_Income = st.number_input('Applicant Income',min_value=150)
+        Applicant_Income = st.number_input('Applicant Income',min_value=15000)
         Coapplicant_Income = st.number_input('Co-applicant Income',min_value=0)
-        Loan_amount = st.number_input('Loan Amount (In thousands)',min_value=5)
+        Loan_amount = st.number_input('Loan Amount (In Lakhs)',min_value=2)
         Loan_Amount_Term = st.number_input('Loan Amount Term (Months)',min_value=12)
     with col2:
         Property_Area = st.selectbox('Property Area',('Urban','Rural','Semiurban'))
@@ -28,8 +44,8 @@ with st.form('Loan Form',clear_on_submit=True):
         'Dependents': [Dependents],
         'Education': [Education],
         'Self_Employed': [Self_Employed],
-        'Applicant_Income': [Applicant_Income],
-        'Coapplicant_Income': [Coapplicant_Income],
+        'Applicant_Income': [Applicant_Income/100],
+        'Coapplicant_Income': [Coapplicant_Income/100],
         'Loan_Amount': [Loan_amount],
         'Loan_Amount_Term': [Loan_Amount_Term],
         'Credit_History': [Credit_History],
@@ -41,14 +57,23 @@ with st.form('Loan Form',clear_on_submit=True):
         emi = ((principle*1000) + interest)/term
         return emi
     df['EMI'] = df.apply(lambda row: emi_calculator(row['Loan_Amount'],row['Loan_Amount_Term']),axis =1)
-    df['Balance_Income'] = df['Applicant_Income'] + df['Coapplicant_Income'] - df['EMI']
+    df['EMI'] = round(df['EMI'],2)
+    df['Balance_Income'] = (df['Applicant_Income'] + df['Coapplicant_Income']) - df['EMI']
+    df['Balance_Income'] = round(df['Balance_Income'],2)
     df.drop(columns = ['Applicant_Income','Coapplicant_Income','Loan_Amount','Loan_Amount_Term'])
 
     submit = st.form_submit_button('Predict')
     if submit:
         prediction = model.predict(df)
+        df['_created_at'] = datetime.datetime.now()
+        df['_created_by'] = 'user'
+        df['Approval_Id'] = df['_created_by']+'-'+df['_created_at'].astype('str')
+        df['Prediction'] = prediction
+        job = client.load_table_from_dataframe(df, table_ref)
+        result = job.result()
         if prediction:
-            st.success('Congratulations, Your Home Loan is Approved!!')
+    
+            st.success(f'Congratulations, Your Home Loan is Approved!!')
         else:
             st.error('We are extremely sorry to inform you that you Home Loan is not approved. Please reach out to nearest branch for further clarification')
 

requirements.txt

@@ -1,4 +1,6 @@
 streamlit==1.12.0
 pandas==2.0.3
 scikit-learn==1.2.2
-joblib==1.3.2
+joblib==1.3.2
+xgboost==2.0.3
+google-cloud-bigquery==3.25.0