import streamlit as st
import yfinance as yf
import pandas as pd
import numpy as np
import plotly.graph_objects as go
from transformers import pipeline
import torch
from sklearn.preprocessing import MinMaxScaler

class AdvancedTradingPredictionApp:
    def __init__(self):
        # Initialize key configurations
        self.trading_pairs = {
            'Bitcoin': 'BTC-USD',
            'Ethereum': 'ETH-USD', 
            'Binance Coin': 'BNB-USD',
            'Cardano': 'ADA-USD',
            'Solana': 'SOL-USD'
        }
        
        # Load pre-trained sentiment and forecasting models
        try:
            self.sentiment_model = pipeline("sentiment-analysis")
            self.forecasting_model = self.load_forecasting_model()
        except Exception as e:
            st.error(f"Model loading error: {e}")
    
    def load_forecasting_model(self):
        # Simulate a more advanced forecasting approach
        class SimpleForecaster:
            def __init__(self):
                self.scaler = MinMaxScaler()
            
            def predict(self, data, days_ahead=7):
                # Normalize data
                scaled_data = self.scaler.fit_transform(data.reshape(-1, 1))
                
                # Simple prediction logic
                trend = np.polyfit(range(len(scaled_data)), scaled_data.flatten(), 1)[0]
                
                # Generate forecast
                last_value = scaled_data[-1]
                forecast = [last_value]
                for _ in range(days_ahead):
                    next_val = forecast[-1] + trend
                    forecast.append(next_val)
                
                # Inverse transform to get actual prices
                forecast = self.scaler.inverse_transform(np.array(forecast[1:]).reshape(-1, 1))
                return forecast.flatten()
        
        return SimpleForecaster()
    
    def fetch_historical_data(self, symbol, period='1y'):
        """Fetch historical price data for a given symbol"""
        try:
            df = yf.download(symbol, period=period)
            return df
        except Exception as e:
            st.error(f"Data fetching error for {symbol}: {e}")
            return None
    
    def calculate_technical_indicators(self, df):
        """Calculate advanced technical indicators"""
        # RSI Calculation
        delta = df['Close'].diff()
        gain = (delta.where(delta > 0, 0)).rolling(window=14).mean()
        loss = (-delta.where(delta < 0, 0)).rolling(window=14).mean()
        rs = gain / loss
        df['RSI'] = 100 - (100 / (1 + rs))
        
        # Moving Averages
        df['SMA_50'] = df['Close'].rolling(window=50).mean()
        df['SMA_200'] = df['Close'].rolling(window=200).mean()
        
        # MACD
        exp1 = df['Close'].ewm(span=12, adjust=False).mean()
        exp2 = df['Close'].ewm(span=26, adjust=False).mean()
        df['MACD'] = exp1 - exp2
        
        return df
    
    def sentiment_analysis(self, text):
        """Perform sentiment analysis on trading-related text"""
        try:
            return self.sentiment_model(text)[0]
        except Exception as e:
            st.error(f"Sentiment analysis error: {e}")
            return None
    
    def advanced_prediction(self, df):
        """Advanced price prediction with multiple techniques"""
        # Prepare data for prediction
        close_prices = df['Close'].values
        
        # Use forecasting model to predict next 7 days
        try:
            predictions = self.forecasting_model.predict(close_prices)
            return predictions
        except Exception as e:
            st.error(f"Prediction error: {e}")
            return None

def main():
    st.set_page_config(page_title="🚀 Advanced Crypto Trading Predictor", layout="wide")
    st.title("🚀 Advanced Cryptocurrency Trading Prediction Platform")
    
    # Initialize the trading prediction app
    app = AdvancedTradingPredictionApp()
    
    # Sidebar for asset and configuration
    st.sidebar.header("Trading Configuration")
    selected_crypto = st.sidebar.selectbox(
        "Select Cryptocurrency",
        list(app.trading_pairs.keys()),
        index=0
    )
    
    # Fetch historical data
    symbol = app.trading_pairs[selected_crypto]
    historical_data = app.fetch_historical_data(symbol)
    
    if historical_data is not None:
        # Calculate technical indicators
        data_with_indicators = app.calculate_technical_indicators(historical_data)
        
        # Prediction Section
        st.subheader(f"🔮 {selected_crypto} Price Prediction")
        
        # Make predictions
        predictions = app.advanced_prediction(data_with_indicators)
        
        if predictions is not None:
            # Visualization
            col1, col2 = st.columns(2)
            
            with col1:
                # Price Chart
                fig_price = go.Figure()
                fig_price.add_trace(go.Scatter(
                    x=historical_data.index, 
                    y=historical_data['Close'], 
                    mode='lines', 
                    name='Historical Price'
                ))
                fig_price.add_trace(go.Scatter(
                    x=pd.date_range(
                        start=historical_data.index[-1], 
                        periods=len(predictions)+1
                    )[1:], 
                    y=predictions, 
                    mode='lines', 
                    name='Price Forecast'
                ))
                fig_price.update_layout(title=f'{selected_crypto} Price Forecast')
                st.plotly_chart(fig_price)
            
            with col2:
                # Technical Indicators
                st.subheader("Technical Indicators")
                st.write(f"Current RSI: {data_with_indicators['RSI'].iloc[-1]:.2f}")
                st.write(f"50-Day SMA: {data_with_indicators['SMA_50'].iloc[-1]:.2f}")
                st.write(f"200-Day SMA: {data_with_indicators['SMA_200'].iloc[-1]:.2f}")
                st.write(f"MACD: {data_with_indicators['MACD'].iloc[-1]:.2f}")
        
        # Sentiment Analysis Section
        st.subheader("🧠 Market Sentiment Analysis")
        sentiment_text = st.text_area("Enter recent news or market commentary")
        if st.button("Analyze Sentiment"):
            if sentiment_text:
                sentiment = app.sentiment_analysis(sentiment_text)
                if sentiment:
                    st.write(f"Sentiment: {sentiment['label']}")
                    st.write(f"Confidence: {sentiment['score']:.2%}")

if __name__ == "__main__":
    main()