Spaces:
Sleeping
Sleeping
File size: 2,917 Bytes
c4bf16e edda823 c4bf16e 3880f80 c4bf16e |
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 78 79 80 81 82 83 84 85 86 87 88 89 |
import streamlit as st
import pandas as pd
import numpy as np
import joblib
from sklearn.ensemble import RandomForestClassifier
from sklearn.preprocessing import StandardScaler,MinMaxScaler, OneHotEncoder
import shap
from streamlit_shap import st_shap
# Page configuration
st.set_page_config(
page_title="Kiva loan amount predictor",
page_icon="💰")
st.title('Predict Kiva loan amounts')
# Load model and preprocessing objects
@st.cache_resource
def load_model_objects():
model_rf = joblib.load('best_reg.joblib')
scaler = joblib.load('scaler.joblib')
ohe = joblib.load('ohe.joblib')
return model_rf, scaler, ohe
model_rf, scaler, ohe = load_model_objects()
# Create SHAP explainer
explainer = shap.TreeExplainer(model_rf)
# App description
with st.expander("What's this app?"):
st.markdown("""
This app helps you determine how much you will be succesfully funded with on Kiva
""")
st.subheader('Describe what you want to loan to')
# User inputs
col1, col2 = st.columns(2)
with col1:
Sector = st.selectbox('sector', options=ohe.categories_[0])
Country = st.selectbox('country', options=ohe.categories_[1])
Gender = st.selectbox('borrower_genders', options=ohe.categories_[2])
with col2:
term_in_months = st.number_input('Lenght of loan in months', min_value=0, value=1)
lender_count = st.number_input('Number of Lenders', min_value=1,value=1)
# Prediction button
if st.button('Predict loan amount 🚀'):
# Prepare categorical features
cat_features = pd.DataFrame({'sector': [Sector], 'country': [Country],'borrower_genders': [Gender]})
cat_encoded = pd.DataFrame(ohe.transform(cat_features).todense(),
columns=ohe.get_feature_names_out(['sector', 'country', 'borrower_genders']))
# Prepare numerical features
num_features = pd.DataFrame({
'term_in_months': [term_in_months],
'lender_count': [lender_count],
})
num_scaled = pd.DataFrame(scaler.transform(num_features), columns=num_features.columns)
# Combine features
features = pd.concat([num_scaled, cat_encoded], axis=1)
# Make prediction
predicted_price = model_rf.predict(features)[0]
# Display prediction
st.metric(label="Predicted loan amount", value=f'{round(predicted_price)} USD')
# SHAP explanation
st.subheader('Price Factors Explained 🤖')
shap_values = explainer.shap_values(features)
st_shap(shap.force_plot(explainer.expected_value, shap_values, features), height=400, width=600)
st.markdown("""
This plot shows how each feature contributes to the predicted price:
- Blue bars push the loan amount lower
- Red bars push the loan amount higher
- The length of each bar indicates the strength of the feature's impact
""")
# Footer
st.markdown("---")
st.markdown("Developed with ❤️ using Streamlit") |