Code Editor for app.py

import streamlit as st
import pandas as pd
import numpy as np
import pickle
from sklearn.datasets import make_classification
from sklearn.model_selection import train_test_split
from sklearn.ensemble import RandomForestClassifier

# Define a function to create synthetic data
def create_data(samples=100):
    X, y = make_classification(n_samples=samples, n_features=20, n_informative=2, n_redundant=10)
    return pd.DataFrame(X), pd.Series(y)

# Training section
st.title('Train a Model')

if st.button('Generate Sample Data'):
    X, y = create_data(samples=100)
    st.session_state.X_train = X
    st.session_state.y_train = y
    st.write('Sample data generated!')
    st.dataframe(X.head())

if 'X_train' in st.session_state and 'y_train' in st.session_state:
    if st.button('Train Model'):
        X_train, X_test, y_train, y_test = train_test_split(st.session_state.X_train, st.session_state.y_train, test_size=0.2, random_state=42)
        model = RandomForestClassifier()
        model.fit(X_train, y_train)
        st.session_state.model = model
        st.write('Model trained!')

    if st.button('Save Model'):
        if 'model' in st.session_state:
            model_filename = 'trained_model.pkl'
            with open(model_filename, 'wb') as file:
                pickle.dump(st.session_state.model, file)
            st.write(f'Model saved as {model_filename}!')

# Prediction section
st.title('Make Predictions')

if st.button('Load Model'):
    model_filename = 'trained_model.pkl'
    with open(model_filename, 'rb') as file:
        loaded_model = pickle.load(file)
    st.session_state.loaded_model = loaded_model
    st.write('Model loaded!')

if 'loaded_model' in st.session_state:
    if st.button('Generate New Data and Predict'):
        X_new, _ = create_data(samples=10)
        print("X_new", X_new)
        predictions = st.session_state.loaded_model.predict(X_new)
        st.write('Predictions:', predictions)
        st.dataframe(X_new)