app.py

import streamlit as st
import requests
import yfinance as yf
import pandas as pd
import datetime
from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch
from plotly import graph_objs as go

# Load the Hugging Face model and tokenizer
@st.cache_resource
def load_model():
    model_name = "rahilv/financial-sentiment-model-roberta-3"
    model = AutoModelForSequenceClassification.from_pretrained(model_name)
    tokenizer = AutoTokenizer.from_pretrained(model_name)
    return model, tokenizer

sentiment_model, sentiment_tokenizer = load_model()

LABEL_MAP = {0: "bullish", 1: "bearish", 2: "neutral"}

# Function to fetch stock news
def get_stock_news(ticker):
    api_key = "d651109fae5346cbbb6812912c801e73"
    url = f'https://newsapi.org/v2/everything?q={ticker}&apiKey={api_key}'

    response = requests.get(url)
    if response.status_code == 200:
        articles = response.json().get('articles', [])
        return articles
    else:
        st.error(f"Error fetching news: {response.status_code}")
        return []

# Function to analyze sentiment
def classify_sentiment(news_title):
    inputs = sentiment_tokenizer(news_title, return_tensors="pt", truncation=True, padding="max_length", max_length=128)
    with torch.no_grad():
        outputs = sentiment_model(**inputs)
        predictions = torch.argmax(outputs.logits, dim=-1).item()
    return LABEL_MAP[predictions]

# Function to fetch stock data from Yahoo Finance
def fetch_stock_data(ticker):
    stock = yf.Ticker(ticker)
    end_date = datetime.date.today()
    start_date = end_date - datetime.timedelta(days=365)  # 1 year of data
    data = stock.history(start=start_date, end=end_date)
    return data

# Streamlit UI
st.title("Stock Analysis and Sentiment App")

# User input for stock ticker symbol
ticker_symbol = st.text_input("Enter a Stock Ticker Symbol (e.g., AAPL, TSLA, GOOGL):")

if ticker_symbol:
    st.subheader(f"Analysis for {ticker_symbol.upper()}")

    # Fetch news
    articles = get_stock_news(ticker_symbol)

    if articles:
        # Create a DataFrame for chart points
        news_points = []
        
        st.write("## Stock Price Chart")
        # Fetch and plot stock data
        stock_data = fetch_stock_data(ticker_symbol)
        
        if not stock_data.empty:
            fig = go.Figure()
            fig.add_trace(go.Scatter(x=stock_data.index, y=stock_data['Close'], mode='lines', name='Close Price'))

            for article in articles:
                date = article['publishedAt'][:10]  # Extract date
                title = article['title']
                sentiment = classify_sentiment(title)
                news_points.append({'date': date, 'title': title, 'sentiment': sentiment})
                
                color = 'green' if sentiment == 'bullish' else 'red' if sentiment == 'bearish' else 'gray'
                if date in stock_data.index:
                    fig.add_trace(go.Scatter(x=[date], y=[stock_data['Close'][date]],
                                             mode='markers', marker=dict(color=color, size=10)))

            fig.update_layout(title=f"{ticker_symbol.upper()} Stock Price & News Sentiment", xaxis_title="Date", yaxis_title="Price")
            st.plotly_chart(fig)

        else:
            st.write("No stock data available.")

        st.write("## News Analysis")
        for point in news_points:
            color = 'green' if point['sentiment'] == 'bullish' else 'red' if point['sentiment'] == 'bearish' else 'gray'
            st.markdown(f"<div style='border-left: 5px solid {color}; padding: 10px; margin: 10px 0;'>"
                        f"<b>{point['title']}</b><br>{point['date']}</div>", unsafe_allow_html=True)

        # Recommendation based on sentiment
        sentiments = [p['sentiment'] for p in news_points]
        recommendation = "hold"

        if sentiments.count('bullish') > sentiments.count('bearish'):
            recommendation = "buy"
        elif sentiments.count('bearish') > sentiments.count('bullish'):
            recommendation = "sell"

        color_map = {"buy": "green", "sell": "red", "hold": "gray"}
        st.markdown(f"### Recommendation: <span style='color: {color_map[recommendation]}; font-size: 1.5em;'>{recommendation.upper()}</span>", unsafe_allow_html=True)

    else:
        st.write("No news articles found.")