app.py

import gradio as gr
import pandas as pd
import numpy as np
from datetime import datetime
from transformers import pipeline
import pandas_ta as ta
import requests

# 1. تحميل نموذجك المدرب (أو تدريبه هنا)
def load_model():
    try:
        # مثال: تحميل نموذج من Hugging Face Hub
        return pipeline("text-classification", model="finiteautomata/bertweet-base-sentiment-analysis")
    except:
        return None

model = load_model()

# 2. جلب بيانات العملات
def fetch_crypto_data(coin_id="bitcoin", days=30):
    url = f"https://api.coingecko.com/api/v3/coins/{coin_id}/market_chart?vs_currency=usd&days={days}"
    data = requests.get(url).json()
    df = pd.DataFrame(data['prices'], columns=['timestamp', 'price'])
    df['timestamp'] = pd.to_datetime(df['timestamp'], unit='ms')
    return df

# 3. تحليل فني + تنبؤ
def analyze(coin):
    df = fetch_crypto_data(coin)
    
    # حساب المؤشرات الفنية
    df['RSI'] = ta.rsi(df['price'])
    df['MACD'] = ta.macd(df['price'])['MACD_12_26_9']
    
    # تنبؤ مبسط (استبدل بنموذجك الفعلي)
    last_price = df['price'].iloc[-1]
    prediction = last_price * (1 + np.random.uniform(-0.1, 0.1))
    
    # تحليل المشاعر
    sentiment = model("Cryptocurrency market is booming")[0]['label'] if model else "Neutral"
    
    return {
        "price": last_price,
        "prediction": prediction,
        "rsi": df['RSI'].iloc[-1],
        "sentiment": sentiment
    }

# 4. واجهة Gradio
with gr.Blocks() as demo:
    gr.Markdown("## 🚀 محلل العملات المشفرة بالذكاء الاصطناعي")
    
    with gr.Row():
        coin = gr.Dropdown(["bitcoin", "ethereum"], label="اختر العملة")
        btn = gr.Button("حلل الآن")
    
    with gr.Row():
        price = gr.Textbox(label="السعر الحالي")
        prediction = gr.Textbox(label="التنبؤ")
    
    with gr.Row():
        rsi = gr.Textbox(label="مؤشر RSI")
        sentiment = gr.Textbox(label="مشاعر السوق")
    
    btn.click(
        fn=lambda c: analyze(c),
        inputs=coin,
        outputs=[price, prediction, rsi, sentiment]
    )

demo.launch()