import gradio as gr
import pandas as pd
import requests
import random
import plotly.graph_objects as go
from datetime import datetime

# -----------------------------
# Fetch BTC Data via CoinGecko
# -----------------------------
def fetch_btc_data():
    try:
        url = "https://api.coingecko.com/api/v3/coins/bitcoin/market_chart?vs_currency=usd&days=50"
        res = requests.get(url, timeout=10)
        data = res.json()
        prices = data['prices']  # list of [timestamp, price]
        df = pd.DataFrame(prices, columns=['time', 'close'])
        df['time'] = pd.to_datetime(df['time'], unit='ms')
        return df
    except:
        # fallback if API fails
        return pd.DataFrame({'time': [datetime.now()], 'close': [0]})

# -----------------------------
# Simulated LSTM Prediction
# -----------------------------
def simulate_lstm_prediction(df):
    current_price = df.iloc[-1]['close']
    momentum = (df.iloc[-1]['close'] - df.iloc[-5]['close']) / df.iloc[-5]['close']
    variance = (random.random() - 0.5) * 0.03

    pred_price = current_price * (1 + (momentum * 0.4) + variance)
    trend = "HIGH" if pred_price > current_price else "LOW"
    confidence = random.uniform(0.7, 0.95)

    reasons = [
        "Local LSTM weights detected a hidden bullish divergence in the volume-price vector.",
        "Sequence analysis indicates the vanishing gradient problem is minimized for this 30-day window.",
        "Neural pattern matching suggests a 68% correlation with previous historical breakout cycles.",
        "Recursive hidden states are currently favoring a consolidation phase."
    ]

    return random.choice(reasons), f"${pred_price:,.2f}", f"{trend} ({int(confidence*100)}%)"

# -----------------------------
# Plot BTC Price Chart
# -----------------------------
def create_plot(df):
    fig = go.Figure()
    fig.add_trace(go.Scatter(
        x=df['time'],
        y=df['close'],
        mode='lines',
        name='BTC Price',
        line=dict(color='#3b82f6', width=4),
        fill='tozeroy',
        fillcolor='rgba(59, 130, 246, 0.1)'
    ))
    fig.update_layout(
        template="plotly_dark",
        paper_bgcolor='rgba(0,0,0,0)',
        plot_bgcolor='rgba(0,0,0,0)',
        margin=dict(l=0, r=0, t=0, b=0),
        height=400,
        xaxis=dict(showgrid=False),
        yaxis=dict(showgrid=True, gridcolor='rgba(255,255,255,0.05)')
    )
    return fig

# -----------------------------
# Main Dashboard Logic
# -----------------------------
def run_dashboard():
    df = fetch_btc_data()

    if df.empty or df['close'].iloc[-1] == 0:
        plot = go.Figure()
        plot.update_layout(template="plotly_dark")
        current_price = "Data not available"
        price = "N/A"
        trend = "N/A"
        reasoning = "Failed to fetch BTC data. Try again later."
        return plot, current_price, price, trend, reasoning

    reasoning, price, trend = simulate_lstm_prediction(df)
    plot = create_plot(df)
    current_price = f"${df.iloc[-1]['close']:,.2f}"

    return plot, str(current_price), str(price), str(trend), str(reasoning)

# -----------------------------
# Custom CSS
# -----------------------------
custom_css = """
.gradio-container { background-color: #020617 !important; color: white !important; }
.gr-button-primary { background: linear-gradient(90deg, #2563eb, #4f46e5) !important; border: none !important; }
"""

# -----------------------------
# Gradio UI
# -----------------------------
with gr.Blocks(theme=gr.themes.Soft(primary_hue="blue"), css=custom_css) as demo:
    gr.HTML("""
        <div style="text-align: center; padding: 40px 20px;">
            <div style="display: inline-block; padding: 10px; background: #f59e0b; border-radius: 12px; margin-bottom: 15px; color: #020617;">
                <i class="fa fa-bitcoin" style="font-size: 24px;"></i>
            </div>
            <h1 style="font-weight: 900; font-size: 3rem; margin-bottom: 0; letter-spacing: -2px; color: white;">BTC PREDICT</h1>
            <p style="color: #f59e0b; font-weight: 900; text-transform: uppercase; font-size: 0.7rem; letter-spacing: 3px; margin-top: 5px;">Built By Nadish • LSTM Neural Engine</p>
        </div>
    """)

    with gr.Row():
        with gr.Column(scale=1):
            price_display = gr.Label(label="Current Market Price")
        with gr.Column(scale=1):
            pred_display = gr.Label(label="AI Forecast Target")
        with gr.Column(scale=1):
            trend_display = gr.Label(label="Signal Direction")

    chart = gr.Plot(label="Market Trend Visualizer")

    with gr.Column():
        analysis = gr.Textbox(label="LSTM Inference Logic (Local Computation)", lines=3)
        predict_btn = gr.Button("INITIALIZE NEURAL INFERENCE", variant="primary", size="lg")

    predict_btn.click(
        fn=run_dashboard,
        outputs=[chart, price_display, pred_display, trend_display, analysis]
    )

# -----------------------------
# Launch App
# -----------------------------
if __name__ == "__main__":
    demo.launch(server_name="0.0.0.0", server_port=7860)