import streamlit as st
import numpy as np
import pandas as pd
from tensorflow.keras.models import load_model
from tensorflow.keras.losses import MeanSquaredError
import joblib

@st.cache_resource
def load_resources():
    # Load model and scalers
    custom_objects = {
        'mse': MeanSquaredError(),
        'mean_squared_error': MeanSquaredError(),
    }
    
    try:
        # Load model resources
        model = load_model('bike_model.h5', custom_objects=custom_objects)
        scaler_x = joblib.load('scaler_x.pkl')
        scaler_y = joblib.load('scaler_y.pkl')
        features = np.load('feature_names.npy', allow_pickle=True)
        
        # Load original data to get min-max values
        df = pd.read_csv('train.csv')
        feature_ranges = {
            'season': (1, 4, 'Season (1:Spring, 2:Summer, 3:Fall, 4:Winter)'),
            'holiday': (0, 1, 'Holiday (0:No, 1:Yes)'),
            'workingday': (0, 1, 'Working Day (0:No, 1:Yes)'),
            'weather': (1, 4, 'Weather (1:Clear, 2:Mist, 3:Light Rain/Snow, 4:Heavy Rain/Snow)'),
            'temp': (df['temp'].min(), df['temp'].max(), 'Temperature (Celsius)'),
            'atemp': (df['atemp'].min(), df['atemp'].max(), 'Feels Like Temperature (Celsius)'),
            'humidity': (df['humidity'].min(), df['humidity'].max(), 'Humidity (%)'),
            'windspeed': (df['windspeed'].min(), df['windspeed'].max(), 'Wind Speed')
        }
        
        return model, scaler_x, scaler_y, features, feature_ranges
    except Exception as e:
        st.error(f"Error loading model: {str(e)}")
        raise e

st.title('🚲 Bike Sharing Demand Predictor')
st.write('Predict hourly bike rental demand based on weather conditions and time features.')

try:
    model, scaler_x, scaler_y, feature_names, feature_ranges = load_resources()

    col1, col2 = st.columns(2)
    inputs = {}
    
    for i, feature in enumerate(feature_names):
        with col1 if i % 2 == 0 else col2:
            min_val, max_val, description = feature_ranges[feature]
            
            if feature in ['season', 'holiday', 'workingday', 'weather']:
                # Integer input for categorical variables
                inputs[feature] = st.number_input(
                    f'{description}',
                    min_value=int(min_val),
                    max_value=int(max_val),
                    value=int(min_val),
                    step=1
                )
            else:
                # Float input for continuous variables
                inputs[feature] = st.number_input(
                    f'{description}',
                    min_value=float(min_val),
                    max_value=float(max_val),
                    value=float(min_val),
                    step=0.1
                )

    if st.button('Predict Demand', use_container_width=True):
        x = scaler_x.transform(np.array(list(inputs.values())).reshape(1, -1))
        x = x.reshape(1, 1, x.shape[1])
        
        pred_scaled = model.predict(x)
        pred = scaler_y.inverse_transform(pred_scaled)[0][0]
        
        st.success(f'Predicted demand: {max(0, int(pred))} bikes')

except Exception as e:
    st.error(f'Error: {str(e)}')

st.markdown('---')
st.markdown('Made with ❤️ using Streamlit and TensorFlow')