import streamlit as st
import pickle
import numpy as np
import os

# Load all trained models
MODEL_FILES = {
    "KNN": "knn_model.pkl",
    "Random Forest": "random_forest_model.pkl",
    "Decision Tree": "decision_tree_model.pkl",
    "Bagging": "bagging_model.pkl",
    "Voting": "voting_model.pkl",
}

# Streamlit UI Config
st.set_page_config(page_title="🍷 Wine Quality Predictor", layout="centered")

# Custom Styling for Background & UI
st.markdown(
    """
    <style>
    .stApp {
        background: linear-gradient(to right, #4B0101, #800020);
        color: white;
    }
    .title {
        font-size: 36px !important;
        font-weight: bold;
        color: #FFD700; /* Gold */
        text-align: center;
    }
    .subtitle {
        font-size: 24px !important;
        font-weight: bold;
        color: #FFA500; /* Orange */
    }
    .stSelectbox label, .stSlider label, .stNumberInput label {
        font-size: 18px !important;
        font-weight: bold;
        color: white;
    }
    .stButton>button {
        background-color: #FFD700; /* Gold */
        color: #4B0101; /* Wine Red */
        font-size: 18px;
        font-weight: bold;
        border-radius: 10px;
    }
    .stButton>button:hover {
        background-color: #FFA500; /* Orange */
        color: white;
    }
    .prediction {
        font-size: 26px;
        font-weight: bold;
        color: #32CD32; /* Bright Green */
        text-align: center;
    }
    </style>
    """,
    unsafe_allow_html=True,
)

# Title and Description
st.markdown('<h1 class="title">🍷 Wine Quality Prediction</h1>', unsafe_allow_html=True)
st.write("Predict the quality of wine based on its chemical properties.")

# Select Model
st.markdown('<h2 class="subtitle">🔍 Select Prediction Model</h2>', unsafe_allow_html=True)
selected_model = st.selectbox("Choose a Model", list(MODEL_FILES.keys()))

# Load Selected Model
model_path = MODEL_FILES[selected_model]
if os.path.exists(model_path):
    with open(model_path, "rb") as f:
        model = pickle.load(f)
    model_loaded = True
else:
    model_loaded = False
    st.error(f"Model file '{model_path}' not found. Please upload the correct model file.")

# User Inputs for Wine Features
st.markdown('<h2 class="subtitle">🍷 Enter Wine Properties</h2>', unsafe_allow_html=True)
fixed_acidity = st.number_input("Fixed Acidity", min_value=3.0, max_value=15.0, value=7.0)
volatile_acidity = st.number_input("Volatile Acidity", min_value=0.0, max_value=2.0, value=0.5)
citric_acid = st.number_input("Citric Acid", min_value=0.0, max_value=1.5, value=0.2)
residual_sugar = st.number_input("Residual Sugar", min_value=0.1, max_value=15.0, value=2.0)
chlorides = st.number_input("Chlorides", min_value=0.01, max_value=0.2, value=0.05)
free_sulfur_dioxide = st.number_input("Free Sulfur Dioxide", min_value=1, max_value=100, value=30)
total_sulfur_dioxide = st.number_input("Total Sulfur Dioxide", min_value=5, max_value=300, value=120)
density = st.number_input("Density", min_value=0.98, max_value=1.1, value=0.995)
pH = st.number_input("pH", min_value=2.5, max_value=4.5, value=3.2)
sulphates = st.number_input("Sulphates", min_value=0.3, max_value=2.0, value=0.8)
alcohol = st.number_input("Alcohol Content", min_value=5.0, max_value=20.0, value=10.0)

# Prepare input for model
input_data = np.array([[fixed_acidity, volatile_acidity, citric_acid, residual_sugar,
                        chlorides, free_sulfur_dioxide, total_sulfur_dioxide, density,
                        pH, sulphates, alcohol]])

# Prediction Button
if st.button("Predict Quality"):
    if model_loaded:
        prediction = model.predict(input_data)
        st.markdown(f'<p class="prediction">Predicted Wine Quality: {int(prediction[0])}/10</p>', unsafe_allow_html=True)
    else:
        st.error(f"Model file '{model_path}' not found. Please upload the correct model file.")

st.write("*Powered by Machine Learning & AI* 🚀")