V8055/2ndproject

1# app.py2import streamlit as st3import numpy as np4import pandas as pd5import matplotlib.pyplot as plt6from sklearn.model_selection import train_test_split7from sklearn.linear_model import LinearRegression8from sklearn.metrics import mean_squared_error, r2_score9import io10import base6411 12def generate_sample_data():13    np.random.seed(42)14    X = np.random.rand(100, 1) * 1015    y = 2 * X + 1 + np.random.randn(100, 1) * 216    return pd.DataFrame({'X': X.flatten(), 'y': y.flatten()})17 18def train_model(df):19    X = df[['X']]20    y = df['y']21    22    # Split the data23    X_train, X_test, y_train, y_test = train_test_split(24        X, y, test_size=0.2, random_state=4225    )26    27    # Create and train the model28    model = LinearRegression()29    model.fit(X_train, y_train)30    31    # Make predictions32    y_train_pred = model.predict(X_train)33    y_test_pred = model.predict(X_test)34    35    return {36        'model': model,37        'X_train': X_train, 'X_test': X_test,38        'y_train': y_train, 'y_test': y_test,39        'y_train_pred': y_train_pred, 'y_test_pred': y_test_pred40    }41 42def plot_regression(results):43    fig, ax = plt.subplots(figsize=(10, 6))44    45    # Plot training data46    ax.scatter(results['X_train'], results['y_train'], 47              color='blue', alpha=0.5, label='Training Data')48    # Plot test data49    ax.scatter(results['X_test'], results['y_test'], 50              color='green', alpha=0.5, label='Test Data')51    52    # Plot regression line53    X_line = np.linspace(0, 10, 100).reshape(-1, 1)54    y_line = results['model'].predict(X_line)55    ax.plot(X_line, y_line, color='red', label='Regression Line')56    57    ax.set_xlabel('X')58    ax.set_ylabel('y')59    ax.set_title('Linear Regression: Training and Test Data with Regression Line')60    ax.legend()61    ax.grid(True, alpha=0.3)62    63    return fig64 65def main():66    st.title("Linear Regression Demo")67    st.write("""68    This app demonstrates simple Linear Regression using scikit-learn.69    You can either use the sample dataset or upload your own CSV file.70    """)71    72    # Data selection73    data_option = st.radio(74        "Choose data source:",75        ("Use sample data", "Upload CSV file")76    )77    78    if data_option == "Use sample data":79        df = generate_sample_data()80    else:81        uploaded_file = st.file_uploader("Choose a CSV file", type="csv")82        if uploaded_file is not None:83            try:84                df = pd.read_csv(uploaded_file)85                if len(df.columns) != 2:86                    st.error("Please upload a CSV file with exactly 2 columns (X and y)")87                    return88                df.columns = ['X', 'y']89            except Exception as e:90                st.error(f"Error reading file: {str(e)}")91                return92        else:93            st.info("Please upload a CSV file")94            return95    96    # Display sample of the data97    st.subheader("Data Preview")98    st.write(df.head())99    100    # Train model and display results101    results = train_model(df)102    model = results['model']103    104    # Model metrics105    train_mse = mean_squared_error(results['y_train'], results['y_train_pred'])106    test_mse = mean_squared_error(results['y_test'], results['y_test_pred'])107    train_r2 = r2_score(results['y_train'], results['y_train_pred'])108    test_r2 = r2_score(results['y_test'], results['y_test_pred'])109    110    st.subheader("Model Performance Metrics")111    col1, col2 = st.columns(2)112    with col1:113        st.metric("Training MSE", f"{train_mse:.4f}")114        st.metric("Training R²", f"{train_r2:.4f}")115    with col2:116        st.metric("Test MSE", f"{test_mse:.4f}")117        st.metric("Test R²", f"{test_r2:.4f}")118    119    st.write(f"Model Equation: y = {model.coef_[0]:.4f}x + {model.intercept_:.4f}")120    121    # Plot122    st.subheader("Regression Plot")123    fig = plot_regression(results)124    st.pyplot(fig)125    126    # Prediction interface127    st.subheader("Make Predictions")128    x_input = st.number_input("Enter a value for X:", value=5.0)129    prediction = model.predict([[x_input]])[0]130    st.write(f"Predicted y: {prediction:.4f}")131 132if __name__ == "__main__":133    main()134 135 136 137