import gradio as gr
import pandas as pd
import pickle

# ------------------ Load Model ------------------
with open("logistic_regression_model.pkl", "rb") as f:
    model = pickle.load(f)

with open("feature_columns.pkl", "rb") as f:
    feature_columns = pickle.load(f)

# ------------------ Prediction Function ------------------
def predict_status(air_temp, proc_temp, speed, torque, tool_wear):

    sample_input = {
        "Air temperature [K]": air_temp,
        "Process temperature [K]": proc_temp,
        "Rotational speed [rpm]": speed,
        "Torque [Nm]": torque,
        "Tool wear [min]": tool_wear
    }

    df = pd.DataFrame([sample_input])

    for col in feature_columns:
        if col not in df.columns:
            df[col] = 0

    df = df[feature_columns]
    prediction = model.predict(df)[0]

    reasons = []

    if air_temp < 150 or air_temp > 500:
        reasons.append("Air temperature outside optimal range (150–500 K).")
    if proc_temp < 150 or proc_temp > 500:
        reasons.append("Process temperature not within safe limits.")
    if speed < 500 or speed > 1500:
        reasons.append("Rotational speed is outside recommended range.")
    if torque < 100 or torque > 800:
        reasons.append("Torque level may damage machine components.")
    if tool_wear > 300:
        reasons.append("Tool wear is high — maintenance required.")

    if prediction == 0:
        message = "❌ **Machine Status: NOT OPERATIONAL**\n\n"
        message += "\n".join([f"- {r}" for r in reasons]) if reasons else "- No clear fault detected."
    else:
        message = "✅ **Machine Status: OPERATING NORMALLY**\n\nAll parameters are within acceptable ranges."

    return message


# ------------------ Professional CSS ------------------
custom_css = """
body {
    background-color: #0f172a;
}

.gradio-container {
    max-width: 1100px;
    margin: auto;
    padding: 32px;
    background: #020617;
    border-radius: 16px;
    box-shadow: 0 20px 40px rgba(0,0,0,0.4);
}

h1 {
    text-align: center;
    font-size: 32px;
    color: #e5e7eb;
}

.subtitle {
    text-align: center;
    color: #94a3b8;
    margin-bottom: 32px;
}

.panel {
    background: #020617;
    border: 1px solid #1e293b;
    border-radius: 14px;
    padding: 24px;
}

label {
    color: #cbd5f5 !important;
    font-weight: 600;
}

.output-box {
    min-height: 240px;
    font-size: 16px;
}

footer {
    text-align: center;
    color: #64748b;
    font-size: 13px;
    margin-top: 24px;
}
"""

# ------------------ Gradio Layout ------------------
with gr.Blocks(css=custom_css) as app:

    gr.Markdown(
        "<h1>Smart Machine Health Predictor</h1>"
        "<div class='subtitle'>Industrial Machine Failure Prediction System</div>"
    )

    with gr.Row():
        with gr.Column(scale=1):
            with gr.Group(elem_classes="panel"):
                air = gr.Slider(0, 1000, value=300, label="Air Temperature (K)")
                proc = gr.Slider(0, 1000, value=300, label="Process Temperature (K)")
                speed = gr.Slider(0, 2000, value=1000, label="Rotational Speed (rpm)")
                torque = gr.Slider(0, 1000, value=500, label="Torque (Nm)")
                wear = gr.Slider(0, 1000, value=100, label="Tool Wear (min)")

                predict_btn = gr.Button("Run Prediction", variant="primary")

        with gr.Column(scale=1):
            output = gr.Markdown(elem_classes="panel output-box")

    predict_btn.click(
        predict_status,
        inputs=[air, proc, speed, torque, wear],
        outputs=output
    )

    gr.Markdown("<footer>Built with Gradio • Predictive Maintenance Dashboard</footer>")

# ------------------ Launch ------------------
if __name__ == "__main__":
    app.launch()