Bank_Customer_churn/Model.py at main · Mfundo-debug/Bank_Customer_churn · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
import streamlit as st
import pandas as pd
import joblib
import eli5
from eli5 import explain_weights, explain_prediction

#load xgb model
xgb_model = joblib.load('xgb_model.pkl')
st.title('Bank Customer Churn Prediction')
age = st.number_input('Age', min_value=18, max_value=100, value=18)
credit_score = st.number_input('Credit Score', min_value=300, max_value=850, value=300)
balance = st.number_input('Balance', min_value=0, max_value=250000, value=0)
num_of_products = st.number_input('Number of Products', min_value=1, max_value=4, value=1)
tenure = st.number_input('Tenure', min_value=0, max_value=10, value=0)
has_cr_card = st.selectbox('Has Credit Card', ['Yes','No'])
is_active_member = st.selectbox('Is Active Member', ['Yes','No'])
gender = st.selectbox('Gender',['Male','Female'])
estimated_salary = st.number_input('Estimated Salary', min_value=0, max_value=250000, value=0)
geography = st.selectbox('Geography', ['France', 'Germany', 'Spain'])
#Preprocess user input
if has_cr_card == 'Yes':
    has_cr_card = 1
else:
    has_cr_card = 0
if is_active_member == 'Yes':
    is_active_member = 1
else:
    is_active_member = 0
if gender =='Male':
    gender = 0
else:
    gender = 1
if geography == 'France':
    geography = 0
elif geography == 'Germany':
    geography = 1
else:
    geography = 2
#Create features from user input
features = pd.DataFrame({'CreditScore': credit_score, 'Age': age, 'Tenure': tenure, 'Balance': balance, 'NumOfProducts': num_of_products, 'HasCrCard': has_cr_card, 'IsActiveMember': is_active_member,'Gender':gender, 'EstimatedSalary': estimated_salary, 'Geography': geography}, index=[0])
#Predict
if st.button('Predict'):
    prediction = xgb_model.predict(features)
    pred_prob = xgb_model.predict_proba(features)[0][1]
    if prediction == 0:
        st.write('Customer will not churn')
        st.write('Reason: Prediction Probability is {}'.format(pred_prob))
    else:
        st.write('Customer will churn')
        st.write('Reason: Prediction Probability is {}'.format(pred_prob))

        # Generate features weights
        # Generate feature importance
        eli5_weights = eli5.explain_weights(xgb_model)
        feature_importances = eli5_weights.feature_importances.importances
        # map feature indices to feature names
        feature_names = list(features.columns)
        feature_importances = [(feature_names[int(feature.feature[1:])], feature.weight) for feature in feature_importances]
        #create a dataframe with feature names and weights
        importance_df = pd.DataFrame(feature_importances, columns=['Feature', 'Weight'])
        importance_df = importance_df.sort_values(by='Weight', ascending=False)
        # Add colors to the data frame
        importance_df['Weight'] = importance_df['Weight'].apply(
        lambda x: f"<span style='color: {'red' if x < 0.05 else 'green'}'>{x:.4f}</span>")

        # Display the data frame
        st.write('Feature Importance')
        st.write(importance_df.to_html(escape=False), unsafe_allow_html=True)


        # Generate prediction explanation using eli5
        eli5_prediction = eli5.explain_prediction(xgb_model, features.values[0])
        prediction_text = eli5.format_as_dataframe(eli5_prediction)
        st.write('Prediction Explanation')
        st.write(prediction_text)