python - Trying to do project which suppouse to fetch Recent news on BTC and historical data on it to predict future price - Sta

admin管理员组

文章数量:1305987

Here's error i receive, what can it be. Any ideas?

INFO: root:Fetched Bitcoin news successfully! INFO:root:Average sentiment of the latest news: 0.38 INFO:root:Raw Response from CoinMarketCap: {'status': {'timestamp': '2025-02-03T16:46:55.598Z', 'error_code': 0, 'error_message': None, 'elapsed': 40, 'credit_count': 1, 'notice': None}, 'data': {'id': 1, 'name': 'Bitcoin', 'symbol': 'BTC', 'quotes': [{'time_open': '2025-01-05T00:00:00.000Z', 'time_close': '2025-01-05T23:59:59.999Z', 'time_high': '2025-01-05T23:03:00.000Z', 'time_low': '2025-01-05T14:21:00.000Z', 'quote': {'USD': {'open': 98233.90577658577, 'high': 98813.30855441149, 'low': 97291.76460634715, 'close': 98314.95944363723, 'volume': 20525254824.64, 'market_cap': 1947189546143.58, 'timestamp': '2025-01-05T23:59:59.999Z'}}}, ERROR:root:Invalid response structure or no historical data found. ERROR:root:Failed to fetch historical Bitcoin data. ERROR:root:Failed to predict Bitcoin price.

import os
    import requests
    import numpy as np
    import matplotlib.pyplot as plt
    import logging
    from datetime import datetime, timedelta, timezone
    from nltk.sentiment.vader import SentimentIntensityAnalyzer
    import nltk
    from sklearn.linear_model import LinearRegression
    from sklearn.preprocessing import PolynomialFeatures
    from sklearn.pipeline import make_pipeline
    from dotenv import load_dotenv, dotenv_values
    
    # Download the vader_lexicon
    nltk.download('vader_lexicon')
    
    # Load environment variables
    load_dotenv(dotenv_path="MainKeys.env")
    env_vars = dotenv_values(".env")  # Loads environment variables as a dictionary
    print(env_vars)  # For debugging: should show your API keys
    
    # Retrieve API keys
    NEWS_API_KEY = os.getenv('NEWS_API_KEY')
    COINMARKETCAP_API_KEY = os.getenv('COINMARKETCAP_API_KEY')

# Set up logging
logging.basicConfig(level=logging.INFO)

def get_bitcoin_news(api_key, page_size=5):
    url = ''
    params = {'q': 'bitcoin', 'apiKey': api_key, 'language': 'en', 'sortBy': 'publishedAt', 'pageSize': page_size}
    try:
        response = requests.get(url, params=params)
        response.raise_for_status()
        return response.json()
    except requests.exceptions.RequestException as e:
        logging.error(f"Error fetching Bitcoin news: {e}")
        return None

def get_historical_bitcoin_data(api_key, start_date, end_date):
    url = ''
    params = {'id': 1, 'convert': 'USD', 'time_start': start_date, 'time_end': end_date, 'interval': 'daily'}
    headers = {'X-CMC_PRO_API_KEY': api_key, 'Accept': 'application/json'}
    try:
        response = requests.get(url, headers=headers, params=params)
        response.raise_for_status()

        # Debugging: print the raw API response to understand the structure
        logging.info(f"Raw Response from CoinMarketCap: {response.json()}")

        data = response.json()

        # Check for the 'data' key and validate structure
        if 'data' not in data or 'BTC' not in data['data'] or 'quotes' not in data['data']['BTC']:
            logging.error("Invalid response structure or no historical data found.")
            return None

        return data['data']['BTC']['quotes']
    except requests.exceptions.RequestException as e:
        logging.error(f"Error fetching historical Bitcoin data: {e}")
        return None

def analyze_sentiment(news_data):
    sid = SentimentIntensityAnalyzer()
    sentiments = []
    for article in news_data.get('articles', []):
        title = article.get('title', '')
        description = article.get('description', '')
        text = f"{title} {description}".strip()
        sentiment_score = sid.polarity_scores(text)
        sentiments.append(sentiment_score['compound'])  # Only storing compound score for simplicity
    return sentiments

def predict_price(api_key):
    # Get the current date and calculate the start date for one month ago
    end_date = datetime.now(timezone.utc)
    start_date = end_date - timedelta(days=30)
    
    # Format dates as YYYY-MM-DD strings
    start_date_str = start_date.strftime('%Y-%m-%d')
    end_date_str = end_date.strftime('%Y-%m-%d')
    
    # Fetch the historical Bitcoin data for the last month
    historical_data = get_historical_bitcoin_data(api_key, start_date_str, end_date_str)
    if not historical_data:
        logging.error("Failed to fetch historical Bitcoin data.")
        return None

    # Extract prices and dates
    prices = [item.get('close', 0) for item in historical_data]
    dates = [item.get('time_open', '').split('T')[0] for item in historical_data]  # Extract only the date part
    if len(prices) == 0 or len(dates) == 0:
        logging.error("No historical price data available.")
        return None

    years = np.array([int(date.split('-')[0]) for date in dates]).reshape(-1, 1)

    # Using Polynomial Regression
    model = make_pipeline(PolynomialFeatures(2), LinearRegression())
    model.fit(years, prices)

    # Predict future Bitcoin price for 2025
    future_year = np.array([[2025]])
    predicted_price = model.predict(future_year)
    plot_bitcoin_prices(years, prices, future_year, predicted_price)
    return predicted_price[0]

def plot_bitcoin_prices(years, prices, future_year, predicted_price):
    plt.figure(figsize=(8, 6))
    plt.plot(years, prices, color='blue', label='Historical Prices')
    plt.plot(future_year, predicted_price, 'ro', label=f'Predicted Price (2025): ${predicted_price[0]:.2f}')
    plt.xlabel("Year")
    plt.ylabel("Price (in USD)")
    plt.title("Bitcoin Price Prediction")
    plt.legend(loc="upper left")
    plt.grid(True)
    plt.show()

def main():
    if not NEWS_API_KEY or not COINMARKETCAP_API_KEY:
        logging.error("Missing API keys. Set them in the .env file.")
        return
    
    # Fetch Bitcoin news and analyze sentiment
    news = get_bitcoin_news(NEWS_API_KEY)
    if news and news.get('status') == 'ok':
        logging.info("Fetched Bitcoin news successfully!")
        sentiments = analyze_sentiment(news)
        if sentiments:
            avg_sentiment = np.mean(sentiments)
            logging.info(f"Average sentiment of the latest news: {avg_sentiment:.2f}")
        else:
            logging.warning("No valid sentiment data available.")
    else:
        logging.error("Failed to fetch Bitcoin news")

    # Predict Bitcoin price in 2025
    predicted_price = predict_price(COINMARKETCAP_API_KEY)
    if predicted_price is not None:
        logging.info(f"Predicted Bitcoin Price in 2025: ${predicted_price:.2f}")
    else:
        logging.error("Failed to predict Bitcoin price.")

if __name__ == '__main__':
    main()

本文标签：