app.py

import gradio as gr
import pandas as pd
import joblib

# Load the trained XGBoost model
best_model = joblib.load("xgb_model.pkl")

# Define the prediction function
def predict_fuel_rate(loaded_drv, empty_drv, eng_speed, empty_stop, loading_stop, loaded_stop):
    input_data = {
        'loaded_drv_time_percycle': loaded_drv,
        'empty_drv_time_percycle': empty_drv,
        'Eng_Speed_Ave': eng_speed,
        'empty_stop_time_percycle': empty_stop,
        'loadingstoptime_percycle': loading_stop,
        'loaded_stop_time_percycle': loaded_stop
    }
    input_df = pd.DataFrame([input_data])
    prediction = best_model.predict(input_df)[0]
    return round(prediction, 2)

# Gradio Interface
interface = gr.Interface(
    fn=predict_fuel_rate,
    inputs=[
        gr.Slider(3, 60, value=18, label="Loaded Drive Time per Cycle"),
        gr.Slider(2, 51, value=16, label="Empty Drive Time per Cycle"),
        gr.Slider(1051, 1596, value=1416, label="Engine Speed Average"),
        gr.Slider(0.2, 24.6, value=4.2, label="Empty Stop Time per Cycle"),
        gr.Slider(2, 18, value=2, label="Loading Stop Time per Cycle"),
        gr.Slider(0.4, 9, value=0.4, label="Loaded Stop Time per Cycle"),
    ],
    outputs=gr.Number(label="Predicted Fuel Rate per Cycle (L)"),
    title="🚛 Fuel Rate What-If Simulator",
    description="Adjust the sliders to simulate different operating conditions and estimate fuel consumption."
)

# Launch the app
interface.launch()