import streamlit as st
import joblib
import numpy as np
import pandas as pd

# Set page configuration
st.set_page_config(
    page_title="Engine Predictive Maintenance",
    page_icon="⚙️",
    layout="wide"
)

# Load the trained model
@st.cache_resource
def load_model():
    model = joblib.load('best_xgboost_model.pkl')
    return model

model = load_model()

# Title and description
st.title("⚙️ Engine Predictive Maintenance System")
st.markdown("""
This application predicts whether an engine is **Normal** or **Faulty** based on sensor readings.
Enter the sensor values below to get a prediction.
""")

# Create two columns for better layout
col1, col2 = st.columns(2)

with col1:
    st.subheader("📊 Input Sensor Readings")

    # Input fields for the 6 features
    engine_rpm = st.number_input(
        "Engine RPM",
        min_value=0.0,
        max_value=10000.0,
        value=2000.0,
        step=100.0,
        help="Engine Revolutions Per Minute"
    )

    lub_oil_pressure = st.number_input(
        "Lub Oil Pressure (psi)",
        min_value=0.0,
        max_value=200.0,
        value=50.0,
        step=1.0,
        help="Lubricating Oil Pressure"
    )

    fuel_pressure = st.number_input(
        "Fuel Pressure (psi)",
        min_value=0.0,
        max_value=200.0,
        value=50.0,
        step=1.0,
        help="Fuel Pressure"
    )

with col2:
    st.subheader("🌡️ Temperature & Pressure")

    coolant_pressure = st.number_input(
        "Coolant Pressure (psi)",
        min_value=0.0,
        max_value=200.0,
        value=50.0,
        step=1.0,
        help="Coolant Pressure"
    )

    lub_oil_temp = st.number_input(
        "Lub Oil Temperature (°C)",
        min_value=0.0,
        max_value=200.0,
        value=80.0,
        step=1.0,
        help="Lubricating Oil Temperature"
    )

    coolant_temp = st.number_input(
        "Coolant Temperature (°C)",
        min_value=0.0,
        max_value=150.0,
        value=70.0,
        step=1.0,
        help="Coolant Temperature"
    )

# Predict button
if st.button("🔍 Predict Engine Condition", type="primary"):
    # Create input array with the correct feature order
    input_data = np.array([[
        engine_rpm,
        lub_oil_pressure,
        fuel_pressure,
        coolant_pressure,
        lub_oil_temp,
        coolant_temp
    ]])

    # Make prediction
    prediction = model.predict(input_data)[0]
    prediction_proba = model.predict_proba(input_data)[0]

    # Display results
    st.markdown("---")
    st.subheader("Prediction Result")

    if prediction == 0:
        st.success("**Engine Status: NORMAL**")
        st.metric("Confidence", f"{prediction_proba[0]*100:.2f}%")
        st.info("The engine is operating within normal parameters. Continue regular maintenance schedule.")
    else:
        st.error("**Engine Status: FAULTY**")
        st.metric("Confidence", f"{prediction_proba[1]*100:.2f}%")
        st.warning("The engine shows signs of potential failure. Immediate inspection recommended!")

    # Show probability distribution
    st.markdown("### Prediction Probabilities")
    prob_df = pd.DataFrame({
        'Condition': ['Normal', 'Faulty'],
        'Probability': [prediction_proba[0], prediction_proba[1]]
    })
    st.bar_chart(prob_df.set_index('Condition'))

# Add footer with information
st.markdown("---")
st.markdown("""
**Model Information:**
- Algorithm: XGBoost Classifier
- F1-Score: 0.7630
- Recall: 87.01%
- Training Dataset: 19,535 engine records

**Features Used:**
1. Engine RPM
2. Lubricating Oil Pressure
3. Fuel Pressure
4. Coolant Pressure
5. Lubricating Oil Temperature
6. Coolant Temperature
""")