import yfinance as yfimport pandas as pdimport scheduleimport timeimport ollamafrom datetime import datetime, timedelta

stock = yf.Ticker("AAPL")dow_jones = yf.Ticker("^DJI")data = stock.history(period="1d", interval="1m")dow_data = dow_jones.history(period="1d", interval="1m")data.head()

rolling_window = pd.DataFrame()dow_rolling_window = pd.DataFrame()# Variables to track daily contextdaily_high = float('-inf')daily_low = float('inf')buying_momentum = 0selling_momentum = 0

def get_market_open_duration(window):    # Extract current time from the last element of the window    current_time = window.index[-1].time()  # Returns a datetime.time object        # Get the previous trading day's date    previous_trading_day = datetime.today() - timedelta(days=1)        # Combine the previous trading day with the current time    current_datetime = datetime.combine(previous_trading_day, current_time)        # Define the market opening time as 09:30:00 on the previous trading day    market_start_time = datetime.combine(previous_trading_day, datetime.strptime("09:30:00", "%H:%M:%S").time())        # Calculate the duration the market has been open in minutes    market_open_duration = (current_datetime - market_start_time).total_seconds() / 60  # in minutes        return market_open_duration

def get_natural_language_insights(    rolling_avg, ema, rsi, bollinger_upper, bollinger_lower,    price_change, volume_change, dow_rolling_avg, market_open_duration, dow_price_change, dow_volume_change, daily_high, daily_low, buying_momentum, selling_momentum):    prompt = f"""    You are a professional stock broker. Apple's stock has a 5-minute rolling average of {rolling_avg:.2f}.    The Exponential Moving Average (EMA) is {ema:.2f}, and the Relative Strength Index (RSI) is {rsi:.2f}.    The Bollinger Bands are set with an upper band of {bollinger_upper:.2f} and a lower band of {bollinger_lower:.2f}.    The price has changed by {price_change:.2f}, and the volume has shifted by {volume_change}.    The DOW price has changed by {dow_price_change:.2f}, and the volume has shifted by {dow_volume_change}.    Meanwhile, the Dow Jones index has a 5-minute rolling average of {dow_rolling_avg:.2f}.    The market has been open for {market_open_duration:.2f} minutes.    Today's high was {daily_high:.2f} and low was {daily_low:.2f}.    The buying momentum is {buying_momentum:.2f} and selling momentum is {selling_momentum:.2f}.    Based on this data, provide insights into the current stock trend and the general market sentiment.    The insights should not be longer than 100 words and should not have an introduction.    """    response = ollama.chat(            model="llama3",            messages=[{"role": "user", "content": prompt}]        )    response_text = response['message']['content'].strip()        # Print the natural language insight    print("Natural Language Insight:", response_text)

def calculate_insights(window, dow_window):    if len(window) >= 5:        # Calculate 5-minute rolling average of the 'Close' prices        rolling_avg = window['Close'].rolling(window=5).mean().iloc[-1]        # Calculate price change and volume change        price_change = window['Close'].iloc[-1] - window['Close'].iloc[-2] if len(window) >= 2 else 0        volume_change = window['Volume'].iloc[-1] - window['Volume'].iloc[-2] if len(window) >= 2 else 0        # Calculate DOW price change and volume change        dow_price_change = dow_window['Close'].iloc[-1] - dow_window['Close'].iloc[-2] if len(dow_window) >= 2 else 0        dow_volume_change = dow_window['Volume'].iloc[-1] - dow_window['Volume'].iloc[-2] if len(dow_window) >= 2 else 0            # Calculate Exponential Moving Average (EMA) and Bollinger Bands (with a 5-period window)        ema = window['Close'].ewm(span=5, adjust=False).mean().iloc[-1]        std = window['Close'].rolling(window=5).std().iloc[-1]        bollinger_upper = rolling_avg + (2 * std)        bollinger_lower = rolling_avg - (2 * std)        # Calculate Relative Strength Index (RSI) if there are enough periods (14 is typical)        delta = window['Close'].diff()        gain = delta.where(delta > 0, 0)        loss = -delta.where(delta < 0, 0)        avg_gain = gain.rolling(window=14, min_periods=1).mean().iloc[-1]        avg_loss = loss.rolling(window=14, min_periods=1).mean().iloc[-1]        rs = avg_gain / avg_loss if avg_loss != 0 else float('nan')        rsi = 100 - (100 / (1 + rs))        # Calculate Dow Jones index rolling average        dow_rolling_avg = dow_window['Close'].rolling(window=5).mean().iloc[-1]                market_open_duration = get_market_open_duration(window)        # Print the calculated insights        print(f"5-minute Rolling Average: {rolling_avg:.2f}")        print(f"EMA: {ema:.2f}")        print(f"RSI: {rsi:.2f}")        print(f"Bollinger Upper Band: {bollinger_upper:.2f}, Lower Band: {bollinger_lower:.2f}")        print(f"Price Change: {price_change:.2f}")        print(f"Volume Change: {volume_change}")        print(f"DOW Price Change: {dow_price_change:.2f}")        print(f"DOW Volume Change: {dow_volume_change}")        print(f"Dow Jones 5-minute Rolling Average: {dow_rolling_avg:.2f}")        print(f"Daily High: {daily_high:.2f}, Daily Low: {daily_low:.2f}")        print(f"Buying Momentum: {buying_momentum:.2f}, Selling Momentum: {selling_momentum:.2f}")        print(f"Market has been open for {market_open_duration:.2f} minutes")                if int(market_open_duration) % 5 == 0:  # Trigger LLM every 5 minutes            get_natural_language_insights(                rolling_avg, ema, rsi, bollinger_upper, bollinger_lower,                price_change, volume_change, dow_rolling_avg, market_open_duration, dow_price_change, dow_volume_change, daily_high, daily_low, buying_momentum, selling_momentum            )

def process_stock_update():    global rolling_window, data, dow_rolling_window, dow_data    global daily_high, daily_low, buying_momentum, selling_momentum    if not data.empty and not dow_data.empty:        # Simulate receiving a new data point for AAPL and Dow Jones        update = data.iloc[0].to_frame().T        time_str = update.index[0].time()        print(time_str)  # Output: ['09:30:00']        dow_update = dow_data.iloc[0].to_frame().T        data = data.iloc[1:]  # Safely remove the first row without causing index issues        dow_data = dow_data.iloc[1:]        # Append the new data points to the rolling windows        rolling_window = pd.concat([rolling_window, update], ignore_index=False)        dow_rolling_window = pd.concat([dow_rolling_window, dow_update], ignore_index=False)        # Update daily high and low        daily_high = max(daily_high, update['Close'].values[0])        daily_low = min(daily_low, update['Close'].values[0])        # Calculate momentum based on price changes        if len(rolling_window) >= 2:            price_change = update['Close'].values[0] - rolling_window['Close'].iloc[-2]            if price_change > 0:                buying_momentum += price_change            else:                selling_momentum += abs(price_change)                        # Limit the rolling window to 5 minutes for moving average        if len(rolling_window) > 5:            rolling_window = rolling_window.iloc[1:]        if len(dow_rolling_window) > 5:            dow_rolling_window = dow_rolling_window.iloc[1:]        # Calculate insights (moving averages, Bollinger Bands, RSI, etc.)        calculate_insights(rolling_window, dow_rolling_window)

schedule.every(10).seconds.do(process_stock_update)# Run the scheduled jobsprint("Starting real-time simulation for AAPL stock updates...")while True:    schedule.run_pending()    time.sleep(1)

# Streamlit UIst.title("AI Stock Advisor")logtxtbox = st.empty()logtxt = '09:30:00'logtxtbox.caption(logtxt)message = st.chat_message("assistant")message.write("Starting real-time simulation for AAPL stock updates. First update will be processed in 5 minutes...")