app.py

import streamlit as st
import pickle
import pandas as pd

# Load trained pipeline
with open("uber_fare_amount.pkl", "rb") as model_file:
    model = pickle.load(model_file)

# Streamlit UI
st.title("🚖 Uber Fare Prediction")
st.markdown("Enter trip details to predict the fare.")

# User inputs
pickup_lat = st.number_input("Pickup Latitude")
pickup_long = st.number_input("Pickup Longitude")
dropoff_lat = st.number_input("Dropoff Latitude")
dropoff_long = st.number_input("Dropoff Longitude")
passenger_count = st.number_input("Passenger Count", min_value=1, max_value=6, step=1)
# pickup_datetime = st.text_input("Pickup Date & Time (YYYY-MM-DD HH:MM:SS)")

# Prediction
if st.button("Predict Fare"):
    # Prepare input as a DataFrame with correct column names
    input_data = pd.DataFrame([{
        'pickup_latitude': pickup_lat,
        'pickup_longitude': pickup_long,
        'dropoff_latitude': dropoff_lat,
        'dropoff_longitude': dropoff_long,
        'passenger_count': passenger_count,
    }])
    # if st.button("Predict Fare"):
    # # Prepare input as a DataFrame with correct column names
    # input_data = pd.DataFrame([{
    #     'pickup_latitude': pickup_lat,
    #     'pickup_longitude': pickup_long,
    #     'dropoff_latitude': dropoff_lat,
    #     'dropoff_longitude': dropoff_long,
    #     'passenger_count': passenger_count,
    #     'pickup_datetime': pickup_datetime
    # }])
    
    predicted_fare = model.predict(input_data)[0]
    st.success(f"Estimated Fare: ${predicted_fare:.2f}")