app.py

import numpy as np
import pandas as pd
import tensorflow as tf
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense, LSTM
from sklearn.ensemble import RandomForestRegressor
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import MinMaxScaler

# Function to preprocess data
def preprocess_data(data):
    # Convert date column to datetime
    data['date'] = pd.to_datetime(data['date'])
    
    # Create a sequence feature
    data['sequence'] = np.arange(len(data))
    
    # Scale the demand values
    scaler = MinMaxScaler()
    data['demand_scaled'] = scaler.fit_transform(data['demand'].values.reshape(-1, 1))
    
    return data, scaler

# Function for demand forecasting
def forecast_demand(data, model_type='LSTM'):
    data, scaler = preprocess_data(data)
    
    X = data[['sequence']]
    y = data['demand_scaled']
    
    # Split the data into train and test sets
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
    
    if model_type == 'LSTM':
        # Reshape the input data for LSTM
        X_train = X_train.values.reshape(-1, 1, 1)
        X_test = X_test.values.reshape(-1, 1, 1)
        
        model = Sequential()
        model.add(LSTM(64, input_shape=(1, 1)))
        model.add(Dense(1))
        model.compile(optimizer='adam', loss='mean_squared_error')
        model.fit(X_train, y_train, epochs=50, batch_size=32, verbose=0)
        forecast = model.predict(X_test)
        
        # Inverse scale the forecasted values
        forecast = scaler.inverse_transform(forecast)
    elif model_type == 'RandomForest':
        model = RandomForestRegressor(n_estimators=100, random_state=42)
        model.fit(X_train, y_train)
        forecast = model.predict(X_test)
        
        # Inverse scale the forecasted values
        forecast = scaler.inverse_transform(forecast.reshape(-1, 1)).squeeze()
    
    # Inverse scale the actual test values
    y_test = scaler.inverse_transform(y_test.values.reshape(-1, 1)).squeeze()
    
    return forecast, y_test

# Gradio interface
import gradio as gr

def run_app():
    with gr.Blocks() as demo:
        gr.Markdown("# Intelligent Inventory and Supply Chain Management")

        with gr.Tab("Demand Forecasting"):
            with gr.Row():
                model_type = gr.Radio(["LSTM", "RandomForest"], label="Model Type")
            with gr.Row():
                data_upload = gr.File(label="Upload Data")
                forecast_button = gr.Button("Forecast Demand")
            with gr.Row():
                forecast_plot = gr.Plot()
                forecast_output = gr.Dataframe(label="Forecasted Demand")

        forecast_button.click(forecast_demand_wrapper, inputs=[data_upload, model_type], outputs=[forecast_plot, forecast_output])

    demo.launch()

if __name__ == "__main__":
    run_app()