import streamlit as st
import numpy as np
import pickle

# Function to load the trained model # Cache the model so it's loaded only once per session
def load_model(model_file):
    with open(model_file, 'rb') as f:
        model = pickle.load(f)
    return model

# Function to predict wine quality
def predict_quality(model, features):
    input_features = np.array(features).reshape(1, -1)
    prediction = model.predict(input_features)
    predicted_quality = prediction[0]
    return predicted_quality

# Load the model
model_file = 'model.pkl'  # Replace with your model file path
model = load_model(model_file)

# Streamlit UI
st.title('Wine Quality Prediction')

# Input fields for each feature
st.header('Input Features')
fixed_acidity = st.slider('Fixed Acidity', min_value=4.0, max_value=16.0, value=8.0, step=0.1)
volatile_acidity = st.slider('Volatile Acidity', min_value=0.1, max_value=2.0, value=0.5, step=0.01)
citric_acid = st.slider('Citric Acid', min_value=0.0, max_value=1.0, value=0.5, step=0.01)
residual_sugar = st.slider('Residual Sugar', min_value=0.0, max_value=20.0, value=10.0, step=0.1)
chlorides = st.slider('Chlorides', min_value=0.0, max_value=1.0, value=0.08, step=0.01)
free_sulfur_dioxide = st.slider('Free Sulfur Dioxide', min_value=1, max_value=100, value=30, step=1)
total_sulfur_dioxide = st.slider('Total Sulfur Dioxide', min_value=1, max_value=300, value=100, step=1)
density = st.slider('Density', min_value=0.8, max_value=1.0, value=0.996, step=0.001)
pH = st.slider('pH', min_value=2.0, max_value=4.0, value=3.0, step=0.01)
sulphates = st.slider('Sulphates', min_value=0.2, max_value=2.0, value=0.5, step=0.01)
alcohol = st.slider('Alcohol', min_value=8.0, max_value=15.0, value=10.0, step=0.1)

# Predict button
if st.button('Predict'):
    features = [fixed_acidity, volatile_acidity, citric_acid, residual_sugar, chlorides,
                free_sulfur_dioxide, total_sulfur_dioxide, density, pH, sulphates, alcohol]
    predicted_quality = predict_quality(model, features)
    st.success(f'Predicted wine quality: {predicted_quality}')