app.py

import gradio as gr
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from statsmodels.tsa.seasonal import seasonal_decompose
from sklearn.preprocessing import MinMaxScaler
from keras.preprocessing.sequence import TimeseriesGenerator
from keras.models import Sequential
from keras.layers import Dense, LSTM
import yfinance as yf



# Function to execute the entire code, handling potential errors
def forecast_stock(ticker, period, interval):
    try:
        # Download data
        df = yf.download(ticker, period=period, interval=interval)
        df = df.asfreq('D').fillna(method='ffill')

        # Filter columns
        ts = df[['Close']]

        # Perform seasonal decomposition
        decomposition = seasonal_decompose(ts, model='additive')

        # Train/test split
        ts_train = ts.iloc[:int(ts.size * .8)]
        ts_test = ts.iloc[int(ts.size * .8):]

        # Normalize the data
        scaler = MinMaxScaler()
        scaler.fit(ts_train.values)
        scaled_ts_train_values = scaler.transform(ts_train.values)
        scaled_ts_test_values = scaler.transform(ts_test.values)

        # Create LSTM model
        n_input = 24
        n_features = 1
        generator = TimeseriesGenerator(scaled_ts_train_values, scaled_ts_train_values, length=n_input, batch_size=1)
        model = Sequential()
        model.add(LSTM(100, activation='relu', input_shape=(n_input, n_features)))
        model.add(Dense(1))
        model.compile(optimizer='adam', loss='mse')
        model.fit(generator, epochs=50, verbose=0)

        # Make predictions
        test_predictions = []
        first_eval_batch = scaled_ts_train_values[-n_input:]
        current_batch = first_eval_batch.reshape((1, n_input, n_features))
        for i in range(len(ts_test)):
            current_pred = model.predict(current_batch, verbose=0)[0]
            test_predictions.append(current_pred)
            current_batch = np.append(current_batch[:, 1:, :], [[current_pred]], axis=1)

        true_predictions = scaler.inverse_transform(test_predictions)

        
        fig = plt.figure(figsize=(10, 6))
        plt.plot(ts.index, ts.values, label='Original Data')
        plt.plot(ts_test.index, true_predictions, label='Forecasted Data')
        plt.legend()
        plt.xlabel('Time')
        plt.ylabel('Value')
        plt.title('Stock Price Forecast')

        return fig


    except Exception as e:
        print(f"An error occurred: {e}")
        return gr.Markdown(f"An error occurred: {e}")  # Display error in interface

tickers_info="""
        **Ticker Examples**
        ---
        **Common Stocks**

        - **AAPL:** Apple Inc.
        - **MSFT:** Microsoft Corp.
        - **GOOG:** Alphabet Inc. (Google)
        - **AMZN:** Amazon.com Inc.
        - **TSLA:** Tesla Inc.
        - **FB:** Meta Platforms Inc.

        **Indices**

        - **^GSPC:** S&P 500 Index
        - **^IXIC:** Nasdaq Composite Index
        - **^DJI:** Dow Jones Industrial Average

        **ETFs ️**

        - **SPY:** SPDR S&P 500 ETF Trust
        - **QQQ:** Invesco QQQ Trust
        - **IWM:** iShares Russell 2000 ETF
        """



with gr.Blocks() as demo:
    gr.Interface(
        forecast_stock,  # Function to execute on submit
        [
            gr.Textbox(label="Ticker", placeholder="e.g., AAPL, MSFT, GOOG"),
            gr.Textbox(label="Period", placeholder="e.g., 1mo, 5y, max"),
            gr.Textbox(label="Interval", placeholder="e.g., 1d, 1wk, 1h"),
            
        ],
        "plot",  # Output type
        live=False,  # Disable live updates
        title="Stock Price Forecast",
        description="Enter a stock ticker, desired data period, and interval to generate a forecast.",
        examples = [
        ["AAPL", "2mo", "1d"],  # Apple stock, data for 2 months, daily intervals
        ["GOOG", "1y", "1d"],   # Google stock, data for 1 year, daily intervals
        ["MSFT", "5y", "1wk"],  # Microsoft stock, data for 5 years, weekly intervals
        ["TSLA", "max", "1h"],  # Tesla stock, maximum available data, hourly intervals
        ["AMZN", "1y", "1h"],   # Amazon stock, data for 1 year, hourly intervals
        ["NVDA", "3mo", "1d"],  # NVIDIA stock, data for 3 months, daily intervals
        ["FB", "1y", "1wk"],    # Meta Platforms (Facebook) stock, data for 1 year, weekly intervals
        ["JNJ", "2y", "1d"],    # Johnson & Johnson stock, data for 2 years, daily intervals
        ["BAC", "6mo", "1d"],   # Bank of America stock, data for 6 months, daily intervals
        ["XOM", "1y", "1wk"],   # Exxon Mobil stock, data for 1 year, weekly intervals
        ]
    )
    with gr.Accordion("Open for More info"):
        gr.Markdown(tickers_info)

demo.launch()