import yfinance as yf
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from sklearn.preprocessing import MinMaxScaler
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense, LSTM

# Fetch stock data using yfinance
def get_stock_data(ticker, start_date, end_date):
    stock_data = yf.download(ticker, start=start_date, end=end_date)
    return stock_data

# Example usage:
ticker = 'AAPL'
start_date = '2022-01-01'
end_date = '2023-01-01'
stock_data = get_stock_data(ticker, start_date, end_date)
print(stock_data.head())
def preprocess_data(data):
    scaler = MinMaxScaler(feature_range=(0, 1))
    scaled_data = scaler.fit_transform(data[['Close']].values)
    return scaled_data, scaler

# Preprocess stock data
scaled_data, scaler = preprocess_data(stock_data)
def create_lstm_model(input_shape):
    model = Sequential()
    model.add(LSTM(units=50, return_sequences=True, input_shape=input_shape))
    model.add(LSTM(units=50))
    model.add(Dense(units=1))  # Predict the stock price
    model.compile(optimizer='adam', loss='mean_squared_error')
    return model

# Train the LSTM model
def train_model(stock_data, window_size=60):
    X_train, y_train = [], []
    for i in range(window_size, len(stock_data)):
        X_train.append(stock_data[i-window_size:i, 0])
        y_train.append(stock_data[i, 0])
    X_train, y_train = np.array(X_train), np.array(y_train)
    X_train = np.reshape(X_train, (X_train.shape[0], X_train.shape[1], 1))
    
    model = create_lstm_model((X_train.shape[1], 1))
    model.fit(X_train, y_train, epochs=5, batch_size=32)
    
    return model

# Train and prepare the model
model = train_model(scaled_data)
def make_prediction(model, scaler, stock_data, last_data, prediction_days=90):
    # Prepare last data for prediction
    input_data = last_data[-60:].reshape(1, 60, 1)
    predicted_prices = []
    
    for _ in range(prediction_days):
        predicted_price = model.predict(input_data)
        predicted_prices.append(predicted_price[0, 0])
        input_data = np.append(input_data[:, 1:], [[predicted_price]], axis=1)
    
    # Rescale to original prices
    predicted_prices = scaler.inverse_transform(np.array(predicted_prices).reshape(-1, 1))
    return predicted_prices

# Predict future prices
last_data = scaled_data[-60:]  # Last 60 days data
future_prices = make_prediction(model, scaler, scaled_data, last_data)
import gradio as gr

def predict_stock(ticker, start_date, end_date):
    stock_data = get_stock_data(ticker, start_date, end_date)
    scaled_data, scaler = preprocess_data(stock_data)
    model = train_model(scaled_data)
    future_prices = make_prediction(model, scaler, scaled_data, scaled_data[-60:])
    
    # Generate a plot of historical and future prices
    plt.figure(figsize=(14,5))
    plt.plot(stock_data['Close'], label='Historical Prices')
    future_dates = pd.date_range(end=end_date, periods=90, freq='B')
    plt.plot(future_dates, future_prices, label='Predicted Prices', linestyle='--')
    plt.title(f"Stock Prices for {ticker}")
    plt.xlabel('Date')
    plt.ylabel('Price')
    plt.legend()
    plt.savefig('stock_plot.png')
    return future_prices, 'stock_plot.png'

# Define UI components
tickers = ['AAPL', 'MSFT', 'GOOGL', 'TSLA', 'AMZN', 'FB', 'NFLX', 'NVDA', 'INTC', 'IBM']
with gr.Blocks() as app:
    gr.Markdown("## Stock Buy/Sell Prediction App")
    
    ticker = gr.Dropdown(tickers, label="Stock Ticker")
    start_date = gr.Date(label="Start Date")
    end_date = gr.Date(label="End Date")
    
    predict_btn = gr.Button("Predict")
    
    predicted_values = gr.Textbox(label="Predicted Stock Values")
    stock_plot = gr.Image(label="Stock Performance Graph")
    
    predict_btn.click(fn=predict_stock, inputs=[ticker, start_date, end_date], outputs=[predicted_values, stock_plot])

app.launch()