import streamlit as st
import pandas as pd
import joblib

# -------------------------------
# PAGE CONFIG
# -------------------------------
st.set_page_config(
    page_title="Engine Predictive Maintenance",
    page_icon="🚗",
    layout="wide"
)

st.title("🚗 Engine Predictive Maintenance System")
st.write("Predict engine health using sensor values")

# -------------------------------
# LOAD MODEL (LOCAL FILE)
# -------------------------------
@st.cache_resource
def load_model():
    return joblib.load("best_model.pkl")

model = load_model()

st.success("✅ Model loaded successfully")

# Expected feature order
features = [
    "Engine_RPM",
    "Lub_Oil_Pressure",
    "Fuel_Pressure",
    "Coolant_Pressure",
    "Lub_Oil_Temperature",
    "Coolant_Temperature"
]

# =========================================================
# 🔍 SINGLE PREDICTION
# =========================================================
st.header("🔍 Single Engine Prediction")

col1, col2 = st.columns(2)

with col1:
    rpm = st.number_input("Engine RPM", min_value=0.0)
    oil_p = st.number_input("Lub Oil Pressure")
    fuel_p = st.number_input("Fuel Pressure")

with col2:
    cool_p = st.number_input("Coolant Pressure")
    oil_t = st.number_input("Lub Oil Temperature")
    cool_t = st.number_input("Coolant Temperature")

if st.button("Predict Engine Condition"):

    input_df = pd.DataFrame([[rpm, oil_p, fuel_p, cool_p, oil_t, cool_t]],
                            columns=features)

    prediction = model.predict(input_df)[0]

    try:
        prob = model.predict_proba(input_df)[0][1]
        confidence = round(prob * 100, 2)
    except:
        confidence = None

    if prediction == 1:
        st.error("⚠ Engine Fault Detected")
    else:
        st.success("✅ Engine Operating Normally")

    if confidence is not None:
        st.write(f"**Fault Probability:** {confidence}%")

# =========================================================
# 📂 BATCH PREDICTION
# =========================================================
st.markdown("---")
st.header("📂 Batch Prediction (CSV Upload)")

st.write("Upload a CSV file containing engine sensor values.")

st.code(", ".join(features))

uploaded_file = st.file_uploader("Upload CSV File", type=["csv"])

if uploaded_file is not None:

    try:
        batch_data = pd.read_csv(uploaded_file)

        st.subheader("📊 Uploaded Data Preview")
        st.dataframe(batch_data.head())

        # Check required columns
        if not all(col in batch_data.columns for col in features):
            st.error("❌ CSV must contain these columns:")
            st.write(features)

        else:
            with st.spinner("Running predictions..."):

                predictions = model.predict(batch_data[features])

                batch_data["Prediction"] = predictions
                batch_data["Status"] = batch_data["Prediction"].map({
                    0: "Normal",
                    1: "Faulty"
                })

                # Add probabilities if available
                try:
                    probs = model.predict_proba(batch_data[features])[:, 1]
                    batch_data["Fault_Probability_%"] = (probs * 100).round(2)
                except:
                    pass

            st.success("✅ Batch prediction complete!")

            st.subheader("📈 Prediction Results")
            st.dataframe(batch_data)

            # Download results
            csv = batch_data.to_csv(index=False).encode("utf-8")

            st.download_button(
                label="📥 Download Predictions",
                data=csv,
                file_name="engine_predictions.csv",
                mime="text/csv"
            )

    except Exception as e:
        st.error(f"Error processing file: {e}")

# =========================================================
# 📄 SAMPLE CSV FORMAT
# =========================================================
st.markdown("---")
st.subheader("📄 Sample CSV Format")

sample_df = pd.DataFrame(columns=features)
st.dataframe(sample_df)

csv_sample = sample_df.to_csv(index=False).encode("utf-8")

st.download_button(
    label="Download Sample CSV",
    data=csv_sample,
    file_name="sample_format.csv",
    mime="text/csv"
)

# =========================================================
# FOOTER
# =========================================================
st.markdown("---")
st.caption("Engine Condition: 0 = Normal | 1 = Faulty")
st.caption("Built for predictive maintenance monitoring")