app.py

import streamlit as st
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import pickle
import os

from sklearn.model_selection import train_test_split
from sklearn.linear_model import LogisticRegression
from sklearn.preprocessing import LabelEncoder, StandardScaler
from sklearn.metrics import confusion_matrix, classification_report

# Load the dataset
df = pd.read_csv('WA_Fn-UseC_-Telco-Customer-Churn.csv')

# Display the dataset
st.write("## Telco Customer Churn Dataset")
st.write(df)

# Preprocess the data
df = df.drop(columns=['customerID'])
df['TotalCharges'] = pd.to_numeric(df['TotalCharges'], errors='coerce')
df = df.dropna()

# Encode categorical features
label_encoders = {}
for column in df.select_dtypes(include=['object']).columns:
    le = LabelEncoder()
    df[column] = le.fit_transform(df[column])
    label_encoders[column] = le

# Split the data into features and target
X = df.drop(columns=['Churn'])
y = df['Churn']

# Scale the features
scaler = StandardScaler()
X_scaled = scaler.fit_transform(X)

# Split the data into training and testing sets
X_train, X_test, y_train, y_test = train_test_split(X_scaled, y, test_size=0.2, random_state=42)

# Train the logistic regression model with increased iterations and a different solver
model = LogisticRegression(max_iter=5000, solver='saga')  # 'saga' works well for large datasets
model.fit(X_train, y_train)

# Save the trained model to a pickle file
with open('customer_churn_logres_model.pkl', 'wb') as f:
    pickle.dump(model, f)

# Plot feature importance
importance = model.coef_[0]
feature_importance = pd.Series(importance, index=X.columns).sort_values(ascending=False)

st.write("## Feature Importance")
fig, ax = plt.subplots()
feature_importance.plot(kind='bar', ax=ax)
st.pyplot(fig)

# Model evaluation
y_pred = model.predict(X_test)

# Confusion matrix
st.write("## Confusion Matrix")
cm = confusion_matrix(y_test, y_pred)
fig, ax = plt.subplots()
sns.heatmap(cm, annot=True, fmt='d', cmap='Blues', xticklabels=['No Churn', 'Churn'], yticklabels=['No Churn', 'Churn'])
ax.set_xlabel('Predicted Label')
ax.set_ylabel('True Label')
ax.set_title('Confusion Matrix')
st.pyplot(fig)

# Classification report
st.write("## Classification Report")
st.text(classification_report(y_test, y_pred))

# Upload the trained model to Hugging Face when the button is clicked
if st.button('Upload Model to Hugging Face'):
    hf_token = os.getenv("HF_TOKEN")
    if hf_token:
        from huggingface_hub import HfApi
        
        api = HfApi()
        api.upload_file(
            path_or_fileobj='customer_churn_logres_model.pkl',
            path_in_repo='customer_churn_logres_model.pkl',
            repo_id='wvsu-dti-aidev-team/customer_churn_logres_model',
            token=hf_token,
        )
        st.success("Model uploaded successfully!")
    else:
        st.error("HF_TOKEN environment variable not set.")