Update app.py

app.py

@@ -1,711 +1,237 @@
-# ╔══════════════════════════════════════════════════════════╗
-# ║          CRYPTO SIGNAL PRO  —  OKX Edition              ║
-# ║   OKX REST API · XGBoost AI · Canlı Otomatik Analiz     ║
-# ╚══════════════════════════════════════════════════════════╝
-
-import subprocess, sys, warnings
-warnings.filterwarnings("ignore")
-
-def _pip(pkg):
-    subprocess.check_call([sys.executable,"-m","pip","install",pkg,"-q"],
-                          stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
-
-for _p in ["requests","pandas","numpy","plotly","xgboost","lightgbm","scikit-learn"]:
-    try: _pip(_p)
-    except: pass
 
 import gradio as gr
-import requests, time, traceback
 import pandas as pd
 import numpy as np
-import plotly.graph_objects as go
-from plotly.subplots import make_subplots
-from datetime import datetime, timezone
-
-try:
-    from xgboost import XGBClassifier
-    XGB_OK = True
-except ImportError:
-    XGB_OK = False
-
-try:
-    from lightgbm import LGBMClassifier
-    LGB_OK = True
-except ImportError:
-    LGB_OK = False
-
-from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier
+from datetime import datetime, timedelta
+import yfinance as yf
+from sklearn.ensemble import RandomForestRegressor, GradientBoostingRegressor, AdaBoostRegressor
+from sklearn.linear_model import Ridge
+from sklearn.svm import SVR
 from sklearn.preprocessing import StandardScaler
+import warnings
+warnings.filterwarnings('ignore')
 
-# ──────────────────────────────────────────────────────────
-# OKX  REST  API
-# ──────────────────────────────────────────────────────────
-OKX_BASE = "https://www.okx.com"
-
-# OKX instId formatı:  BTC-USDT  (tire ile)
-# Kullanıcıya gösterilecek etiket: BTCUSDT  (tirensiz)
-SYMBOLS_LABEL = [
-    "BTCUSDT","ETHUSDT","SOLUSDT","BNBUSDT","XRPUSDT",
-    "ADAUSDT","DOGEUSDT","AVAXUSDT","DOTUSDT","LINKUSDT",
-    "LTCUSDT","UNIUSDT","ATOMUSDT","NEARUSDT","OPUSDT",
-    "ARBUSDT","INJUSDT","TIAUSDT","SUIUSDT","APTUSDT",
-]
-def _to_instid(label: str) -> str:
-    """BTCUSDT  →  BTC-USDT"""
-    base = label.replace("USDT","")
-    return f"{base}-USDT"
-
-# OKX bar formatları
-INTERVALS = ["5m","15m","1H","4H","1D"]
-_OKX_BAR = {"5m":"5m","15m":"15m","1H":"1H","4H":"4H","1D":"1D"}
-
-def fetch_okx(label: str, interval: str, limit: int = 300) -> pd.DataFrame:
-    """OKX /api/v5/market/candles  →  OHLCV DataFrame"""
-    inst = _to_instid(label)
-    bar  = _OKX_BAR.get(interval, interval)
-    url  = f"{OKX_BASE}/api/v5/market/candles"
-    params = {"instId": inst, "bar": bar, "limit": min(limit, 300)}
-    r = requests.get(url, params=params, timeout=12)
-    r.raise_for_status()
-    data = r.json()
-    if data.get("code") != "0":
-        raise RuntimeError(f"OKX hata: {data.get('msg','bilinmeyen')}")
-    raw = data["data"]           # liste: [[ts,o,h,l,c,vol,volCcy,volCcyQuote,confirm],...]
-    if not raw:
-        raise RuntimeError("OKX boş veri döndürdü.")
-    df = pd.DataFrame(raw, columns=["ts","open","high","low","close","vol","volCcy","volCcyQ","confirm"])
-    for c in ["open","high","low","close","vol"]:
-        df[c] = df[c].astype(float)
-    df["ts"] = pd.to_datetime(df["ts"].astype(np.int64), unit="ms", utc=True)
-    df = df.sort_values("ts").set_index("ts")
-    df.index.name = "open_time"
-    return df[["open","high","low","close"]].rename(columns={"vol":"volume"}).assign(
-        volume=df["vol"])
-
-def fetch_ticker(label: str) -> dict:
-    """Anlık ticker: bid/ask/last/24h değişim"""
-    inst = _to_instid(label)
-    r = requests.get(f"{OKX_BASE}/api/v5/market/ticker",
-                     params={"instId": inst}, timeout=8)
-    r.raise_for_status()
-    d = r.json()
-    if d.get("code") != "0" or not d["data"]:
-        return {}
-    t = d["data"][0]
-    return {
-        "last":   float(t.get("last",0)),
-        "bid":    float(t.get("bidPx",0)),
-        "ask":    float(t.get("askPx",0)),
-        "chg24":  float(t.get("chgUtc8",0) or t.get("sodUtc8",0) or 0),
-        "vol24":  float(t.get("vol24h",0)),
-        "high24": float(t.get("high24h",0)),
-        "low24":  float(t.get("low24h",0)),
-    }
-
-# ──────────────────────────────────────────────────────────
-# TEKNİK GÖSTERGELer
-# ──────────────────────────────────────────────────────────
-def ema(s,n):   return s.ewm(span=n,adjust=False).mean()
-def rsi(s,n=14):
-    d=s.diff(); g=d.clip(lower=0).rolling(n).mean()
-    l=(-d.clip(upper=0)).rolling(n).mean()
-    return 100-(100/(1+g/l.replace(0,np.nan)))
-def macd(s):
-    ml=s.ewm(span=12,adjust=False).mean()-s.ewm(span=26,adjust=False).mean()
-    sig=ml.ewm(span=9,adjust=False).mean(); return ml,sig,ml-sig
-def atr(df,n=14):
-    tr=pd.concat([df.high-df.low,
-                  (df.high-df.close.shift()).abs(),
-                  (df.low -df.close.shift()).abs()],axis=1).max(axis=1)
-    return tr.rolling(n).mean()
-def vwap(df):
-    tp=(df.high+df.low+df.close)/3
-    return (tp*df.volume).cumsum()/df.volume.cumsum()
-
-def add_indicators(df):
-    df=df.copy()
-    df["ema20"] =ema(df.close,20)
-    df["ema50"] =ema(df.close,50)
-    df["ema200"]=ema(df.close,200)
-    df["rsi"]   =rsi(df.close)
-    df["macd"],df["macd_sig"],df["macd_hist"]=macd(df.close)
-    df["atr"]   =atr(df)
-    df["vwap"]  =vwap(df)
-    df["vol_ma"]=df.volume.rolling(20).mean()
-    df["vol_ratio"]=df.volume/df.vol_ma.replace(0,np.nan)
-    return df.dropna()
-
-# ──────────────────────────────────────────────────────────
-# SİNYAL MANTIĞI
-# ──────────────────────────────────────────────────────────
-def generate_signal(df):
-    l,p=df.iloc[-1],df.iloc[-2]
-    bc=[l.ema20>l.ema50, l.ema50>l.ema200, l.rsi>50,
-        bool((l.macd>l.macd_sig) and (p.macd<=p.macd_sig)),
-        l.close>l.vwap]
-    sc=[l.ema20<l.ema50, l.ema50<l.ema200, l.rsi<50,
-        bool((l.macd<l.macd_sig) and (p.macd>=p.macd_sig)),
-        l.close<l.vwap]
-    bs,ss=sum(bc),sum(sc)
-    px,at=l.close,l.atr
-    if bs>=3:
-        sig="AL 📈"; conf=int(bs/5*100)
-        sl=round(px-1.5*at,6); tp=round(px+3.0*at,6)
-    elif ss>=3:
-        sig="SAT 📉"; conf=int(ss/5*100)
-        sl=round(px+1.5*at,6); tp=round(px-3.0*at,6)
-    else:
-        sig="BEKLE ⏳"; conf=30
-        sl=round(px-1.5*at,6); tp=round(px+1.5*at,6)
-    rr=round(abs(tp-px)/abs(sl-px),2) if abs(sl-px)>0 else 0
-    cond_txt=(
-        f"EMA20>50:{'✅' if bc[0] else '❌'}  "
-        f"EMA50>200:{'✅' if bc[1] else '❌'}  "
-        f"RSI>50:{'✅' if bc[2] else '❌'}  "
-        f"MACD✗:{'✅' if bc[3] else '❌'}  "
-        f"Fiyat>VWAP:{'✅' if bc[4] else '❌'}"
-    )
-    return sig,conf,sl,tp,rr,cond_txt
-
-# ──────────────────────────────────────────────────────────
-# AI  — PREMIUM ENSEMBLE  (XGBoost + LightGBM + RF + GBM)
-# ──────────────────────────────────────────────────────────
-FCOLS = ["r1","r3","r5","r10",
-         "e2050","e50200",
-         "rsi_n","rsi_slope",
-         "macd_n","mhist_n","macd_slope",
-         "vwap_d","atr_n","volr",
-         "hl_ratio","body_ratio"]
-
-def featurize(df):
-    d = df.copy()
-    # Fiyat değişimleri
-    d["r1"]  = d.close.pct_change(1)
-    d["r3"]  = d.close.pct_change(3)
-    d["r5"]  = d.close.pct_change(5)
-    d["r10"] = d.close.pct_change(10)
-    # EMA mesafeleri
-    d["e2050"]  = (d.ema20  - d.ema50)  / d.close
-    d["e50200"] = (d.ema50  - d.ema200) / d.close
-    # RSI
-    d["rsi_n"]    = d.rsi / 100
-    d["rsi_slope"]= d.rsi.diff(3) / 100
-    # MACD
-    d["macd_n"]   = d.macd       / d.close
-    d["mhist_n"]  = d.macd_hist  / d.close
-    d["macd_slope"]= d.macd.diff(3) / d.close
-    # Fiyat konumu
-    d["vwap_d"]   = (d.close - d.vwap)  / d.close
-    d["atr_n"]    = d.atr / d.close
-    d["volr"]     = d.vol_ratio.clip(0, 5)
-    # Mum şekli
-    d["hl_ratio"] = (d.high - d.low)   / d.close
-    d["body_ratio"]= (d.close - d.open).abs() / (d.high - d.low + 1e-9)
-    return d.dropna()
-
-# Her modelin bilgisi: isim, ağırlık, renk
-ENSEMBLE_META = [
-    {"name": "XGBoost",   "w": 0.25, "color": "#64b5f6"},
-    {"name": "LightGBM",  "w": 0.25, "color": "#a5d6a7"},
-    {"name": "RndForest", "w": 0.25, "color": "#ffcc02"},
-    {"name": "GradBoost", "w": 0.25, "color": "#ce93d8"},
-]
-
-def _build_models():
-    """Tüm ensemble modellerini döndür; eksik kütüphane varsa atlar."""
-    models = []
-    if XGB_OK:
-        models.append((
-            ENSEMBLE_META[0],
-            XGBClassifier(n_estimators=80, max_depth=3, learning_rate=0.08,
-                          subsample=0.8, colsample_bytree=0.8,
-                          eval_metric="logloss", verbosity=0, n_jobs=1)
-        ))
-    if LGB_OK:
-        models.append((
-            ENSEMBLE_META[1],
-            LGBMClassifier(n_estimators=80, max_depth=4, learning_rate=0.08,
-                           subsample=0.8, colsample_bytree=0.8,
-                           verbose=-1, n_jobs=1)
-        ))
-    # scikit-learn modelleri her zaman mevcut
-    models.append((
-        ENSEMBLE_META[2],
-        RandomForestClassifier(n_estimators=120, max_depth=5,
-                               min_samples_leaf=3, n_jobs=1, random_state=42)
-    ))
-    models.append((
-        ENSEMBLE_META[3],
-        GradientBoostingClassifier(n_estimators=80, max_depth=3,
-                                   learning_rate=0.08, subsample=0.8,
-                                   random_state=42)
-    ))
-    return models
-
-def train_ensemble(df):
-    """
-    Premium Ensemble eğitimi.
-    Returns:
-        ai_conf   : int  0-100  (ağırlıklı ortalama prob)
-        model_probs: list[dict]  her modelin detayı
-    """
-    fd = featurize(df)
-    if len(fd) < 80:
-        return 50, []
-
-    fd = fd.copy()
-    fd["tgt"] = (fd.close.shift(-3) > fd.close).astype(int)
-    fd = fd.dropna()
-    if len(fd) < 50:
-        return 50, []
-
-    X = fd[FCOLS].values
-    y = fd["tgt"].values
-    sc = StandardScaler()
-    Xs = sc.fit_transform(X)
-    sp = int(len(Xs) * 0.80)
-
-    last_feat = featurize(df.tail(80)).tail(1)[FCOLS].values
-    last_scaled = sc.transform(last_feat)
-
-    total_w  = 0.0
-    weighted = 0.0
-    model_probs = []
-
-    for meta, clf in _build_models():
-        try:
-            clf.fit(Xs[:sp], y[:sp])
-            prob = clf.predict_proba(last_scaled)[0][1]
-            model_probs.append({
-                "name":  meta["name"],
-                "prob":  prob,
-                "color": meta["color"],
-                "w":     meta["w"],
-            })
-            weighted += meta["w"] * prob
-            total_w  += meta["w"]
-        except Exception:
-            pass
-
-    if total_w == 0:
-        return 50, []
-
-    ai_conf = int((weighted / total_w) * 100)
-    return ai_conf, model_probs
-
-# Geriye dönük uyumluluk — eski çağrı noktası
-def train_ai(df):
-    conf, _ = train_ensemble(df)
-    return conf
-
-# ──────────────────────────────────────────────────────────
-# GRAFİK
-# ──────────────────────────────────────────────────────────
-def build_chart(df, label, interval, signal, sl, tp, ticker):
-    p=df.tail(150).copy()
-    fig=make_subplots(
-        rows=4,cols=1,shared_xaxes=True,
-        row_heights=[0.52,0.16,0.16,0.16],
-        vertical_spacing=0.02,
-        subplot_titles=["","RSI","MACD","Volume"],
-    )
-
-    # ── Mum grafiği
-    fig.add_trace(go.Candlestick(
-        x=p.index,open=p.open,high=p.high,low=p.low,close=p.close,
-        name="Fiyat",
-        increasing_line_color="#00e676", decreasing_line_color="#ff5252",
-        increasing_fillcolor="rgba(0,200,83,0.53)",decreasing_fillcolor="rgba(255,82,82,0.53)",
-        line_width=1,
-    ),row=1,col=1)
-
-    # ── EMAs
-    for n,c,w,dash in [
-        (20,"#64b5f6",1.5,"solid"),
-        (50,"#ffb74d",1.5,"solid"),
-        (200,"#ce93d8",2.0,"dot"),
-    ]:
-        fig.add_trace(go.Scatter(x=p.index,y=p[f"ema{n}"],name=f"EMA{n}",
-            line=dict(color=c,width=w,dash=dash),opacity=0.85),row=1,col=1)
-
-    # ── VWAP
-    fig.add_trace(go.Scatter(x=p.index,y=p.vwap,name="VWAP",
-        line=dict(color="#f06292",width=1.4,dash="dashdot"),opacity=0.8),row=1,col=1)
-
-    # ── SL / TP çizgileri
-    fig.add_hline(y=sl,line_dash="dash",line_color="rgba(255,82,82,0.6)",line_width=1.2,
-                  annotation_text=f"SL {sl:,.4f}",annotation_font_color="#ff5252",
-                  annotation_position="right",row=1,col=1)
-    fig.add_hline(y=tp,line_dash="dash",line_color="rgba(0,230,118,0.6)",line_width=1.2,
-                  annotation_text=f"TP {tp:,.4f}",annotation_font_color="#00e676",
-                  annotation_position="right",row=1,col=1)
-
-    # ── Sinyal oku (son barda)
-    li,lc=p.index[-1],p.close.iloc[-1]
-    if "AL" in signal:
-        fig.add_trace(go.Scatter(x=[li],y=[lc*0.992],mode="markers+text",
-            marker=dict(symbol="triangle-up",size=18,color="#00e676"),
-            text=["AL"],textposition="bottom center",
-            textfont=dict(color="#00e676",size=11),showlegend=False),row=1,col=1)
-    elif "SAT" in signal:
-        fig.add_trace(go.Scatter(x=[li],y=[lc*1.008],mode="markers+text",
-            marker=dict(symbol="triangle-down",size=18,color="#ff5252"),
-            text=["SAT"],textposition="top center",
-            textfont=dict(color="#ff5252",size=11),showlegend=False),row=1,col=1)
-
-    # ── RSI
-    fig.add_trace(go.Scatter(x=p.index,y=p.rsi,name="RSI",
-        line=dict(color="#80cbc4",width=1.4),fill="tozeroy",
-        fillcolor="rgba(128,203,196,0.06)"),row=2,col=1)
-    for yv,c in [(70,"rgba(255,82,82,0.4)"),(50,"rgba(85,85,85,0.4)"),(30,"rgba(0,230,118,0.4)")]:
-        fig.add_hline(y=yv,line_dash="dot",line_color=c,line_width=0.8,row=2,col=1)
-
-    # ── MACD histogram + çizgiler
-    clrs=["rgba(0,200,83,0.67)" if v>=0 else "rgba(255,23,68,0.67)" for v in p.macd_hist]
-    fig.add_trace(go.Bar(x=p.index,y=p.macd_hist,name="Hist",
-        marker_color=clrs,showlegend=False),row=3,col=1)
-    fig.add_trace(go.Scatter(x=p.index,y=p.macd,name="MACD",
-        line=dict(color="#64b5f6",width=1.2)),row=3,col=1)
-    fig.add_trace(go.Scatter(x=p.index,y=p.macd_sig,name="Signal",
-        line=dict(color="#ffb74d",width=1.2)),row=3,col=1)
-
-    # ── Volume
-    vcols=["rgba(0,200,83,0.4)" if c>=o else "rgba(255,82,82,0.4)"
-           for c,o in zip(p.close,p.open)]
-    fig.add_trace(go.Bar(x=p.index,y=p.volume,name="Vol",
-        marker_color=vcols,showlegend=False),row=4,col=1)
-    fig.add_trace(go.Scatter(x=p.index,y=p.vol_ma,name="Vol MA",
-        line=dict(color="rgba(255,204,2,0.6)",width=1.2)),row=4,col=1)
-
-    # ── Son fiyat annotation
-    price_str=f"${lc:,.4f}"
-    fig.add_annotation(x=1,y=lc,xref="paper",yref="y",
-        text=f" {price_str} ",showarrow=False,
-        font=dict(color="#fff",size=11,family="Courier New"),
-        bgcolor="#1565c0",bordercolor="#42a5f5",borderwidth=1,
-        xanchor="left",yanchor="middle")
-
-    # ── Başlık
-    chg = ticker.get("chg24",0)
-    chg_col="#00e676" if chg>=0 else "#ff5252"
-    chg_sym="▲" if chg>=0 else "▼"
-    title_txt=(f"<b>{label}</b>  <span style='font-size:13px;color:{chg_col}'>"
-               f"{chg_sym} {abs(chg):.2f}%</span>  "
-               f"<span style='font-size:11px;color:#666'>{interval} · OKX</span>")
-
-    fig.update_layout(
-        title=dict(text=title_txt,font=dict(color="#e0e0e0",size=15,
-                   family="'Courier New',monospace"),x=0.01,xanchor="left"),
-        paper_bgcolor="#0b0b10",plot_bgcolor="#0f0f16",
-        font=dict(color="#9e9e9e",family="'Courier New',monospace",size=10),
-        xaxis_rangeslider_visible=False,
-        height=800,
-        legend=dict(bgcolor="rgba(15,15,22,0.85)",bordercolor="#222",
-                    borderwidth=1,font=dict(size=9),x=0.01,y=0.99,
-                    xanchor="left",yanchor="top"),
-        margin=dict(l=55,r=80,t=55,b=15),
-        hoverlabel=dict(bgcolor="#1a1a2e",bordercolor="#333",
-                        font_color="#e0e0e0",font_family="Courier New"),
-    )
-    for i in range(1,5):
-        fig.update_xaxes(gridcolor="#191924",zeroline=False,
-                         showspikes=True,spikecolor="#444",
-                         spikethickness=1,row=i,col=1)
-        fig.update_yaxes(gridcolor="#191924",zeroline=False,
-                         showspikes=True,spikecolor="#444",row=i,col=1)
-    return fig
-
-# ──────────────────────────────────────────────────────────
-# PANEL HTML
-# ──────────────────────────────────────────────────────────
-def build_panel(label,interval,signal,tech_conf,ai_conf,sl,tp,rr,
-                cond_txt,ticker,df,model_probs=None):
-    final=int(0.5*tech_conf+0.5*ai_conf)
-    price=df.close.iloc[-1]
-    rsi_v=df.rsi.iloc[-1]
-    atr_v=df.atr.iloc[-1]
-    ema20=df.ema20.iloc[-1]; ema50=df.ema50.iloc[-1]; ema200=df.ema200.iloc[-1]
-    vwap_v=df.vwap.iloc[-1]
-    now=datetime.now(timezone.utc).strftime("%H:%M:%S UTC")
-
-    if "AL" in signal:   dir_txt,dir_col,bg="🟢 YUKARI","#00e676","rgba(0,200,83,0.08)"
-    elif "SAT" in signal: dir_txt,dir_col,bg="🔴 AŞAĞI","#ff5252","rgba(255,82,82,0.08)"
-    else:                dir_txt,dir_col,bg="🟡 YATAY","#ffb74d","rgba(255,183,77,0.08)"
-
-    bar_col=("#238636" if final>=65 else "#f0883e" if final>=40 else "#da3633")
-    chg=ticker.get("chg24",0); chg_col="#00e676" if chg>=0 else "#ff5252"
-    h24=ticker.get("high24",0); l24=ticker.get("low24",0)
-    vol24=ticker.get("vol24",0)
-
-    def row(label,val,extra="",bg_tr=""):
-        bg_style=f"background:{bg_tr};" if bg_tr else ""
-        return (f'<tr style="{bg_style}">'
-                f'<td style="padding:7px 10px;color:#6e7681;font-size:.85em">{label}</td>'
-                f'<td style="padding:7px 10px;font-size:.9em">{val}</td>'
-                f'{"<td style=padding:7px_10px;font-size:.8em;color:#555>"+extra+"</td>" if extra else ""}'
-                f'</tr>')
-
-    # ── Premium Ensemble Breakdown HTML
-    if model_probs:
-        rows_html = ""
-        for mp in model_probs:
-            p_pct  = int(mp["prob"] * 100)
-            w_pct  = int(mp["w"] * 100)
-            col    = mp["color"]
-            bull   = "AL ▲" if mp["prob"] >= 0.5 else "SAT ▼"
-            b_col  = "#00e676" if mp["prob"] >= 0.5 else "#ff5252"
-            rows_html += f"""
-    <tr>
-      <td style="padding:5px 8px;color:#8b949e;font-size:.75em;white-space:nowrap">
-        <span style="display:inline-block;width:8px;height:8px;border-radius:50%;
-              background:{col};margin-right:5px;box-shadow:0 0 5px {col}88"></span>
-        {mp["name"]}
-      </td>
-      <td style="padding:5px 8px;font-size:.75em">
-        <div style="background:#0d1117;border-radius:4px;overflow:hidden;height:7px;width:80px;display:inline-block;vertical-align:middle;margin-right:6px">
-          <div style="width:{p_pct}%;height:100%;background:linear-gradient(90deg,{col}99,{col})"></div>
-        </div>
-        <span style="color:{col}">{p_pct}%</span>
-      </td>
-      <td style="padding:5px 8px;font-size:.72em;color:{b_col}">{bull}</td>
-      <td style="padding:5px 8px;font-size:.68em;color:#484f58">ağırlık {w_pct}%</td>
-    </tr>"""
-        ensemble_html = f"""
-  <!-- Ensemble Breakdown -->
-  <div style="margin-bottom:14px;padding:10px;background:#0a0f1a;
-       border:1px solid #1c2a3a;border-radius:10px;">
-    <div style="font-size:.68em;color:#3d5afe;letter-spacing:2px;
-         margin-bottom:8px;display:flex;align-items:center;gap:6px">
-      <span style="background:linear-gradient(90deg,#3d5afe,#00e5ff);
-            -webkit-background-clip:text;-webkit-text-fill-color:transparent;
-            font-weight:bold">⚡ PREMIUM AI ENSEMBLE</span>
-      <span style="color:#1c2a4a">4 Model · Eşit Ağırlıklı Oylama</span>
-    </div>
-    <table style="width:100%;border-collapse:collapse">{rows_html}</table>
-  </div>"""
-    else:
-        ensemble_html = ""
-
-    html=f"""
-<div style="background:linear-gradient(160deg,#0d1117 0%,#161b22 100%);
-     border:1px solid #21262d;border-radius:14px;padding:18px 16px;
-     font-family:'Courier New',monospace;color:#e6edf3;min-width:260px;">
-
-  <!-- Header -->
-  <div style="display:flex;justify-content:space-between;align-items:center;margin-bottom:14px;">
-    <div>
-      <span style="font-size:1.1em;font-weight:bold;color:#58a6ff">{label}</span>
-      <span style="margin-left:8px;font-size:.75em;color:#444;letter-spacing:1px">{interval} · OKX</span>
-    </div>
-    <div style="font-size:.7em;color:#30363d">{now}</div>
-  </div>
-
-  <!-- Büyük sinyal kutusu -->
-  <div style="background:{bg};border:1px solid {dir_col}33;border-radius:10px;
-       padding:14px;text-align:center;margin-bottom:14px;">
-    <div style="font-size:1.6em;font-weight:bold;letter-spacing:2px">{signal}</div>
-    <div style="font-size:1.1em;color:{dir_col};margin-top:4px">{dir_txt}</div>
-  </div>
-
-  <!-- Güven çubuğu -->
-  <div style="margin-bottom:14px">
-    <div style="display:flex;justify-content:space-between;font-size:.75em;
-         color:#8b949e;margin-bottom:4px">
-      <span>Güven Skoru</span><span style="color:#e6edf3;font-weight:bold">{final}%</span>
-    </div>
-    <div style="background:#21262d;border-radius:6px;overflow:hidden;height:10px">
-      <div style="background:linear-gradient(90deg,{bar_col},{bar_col}bb);
-           width:{final}%;height:100%;border-radius:6px;
-           box-shadow:0 0 8px {bar_col}66"></div>
-    </div>
-    <div style="display:flex;justify-content:space-between;font-size:.7em;
-         color:#484f58;margin-top:3px">
-      <span>🔬 Teknik: {tech_conf}%</span><span>🤖 AI Ensemble: {ai_conf}%</span>
-    </div>
-  </div>
-
-  {ensemble_html}
-
-  <!-- Tablo -->
-  <table style="width:100%;border-collapse:collapse;font-size:.88em">
-    {row("💰 Güncel Fiyat", f'<b>${price:,.6g}</b>')}
-    {row("24h Değişim", f'<span style="color:{chg_col}">{"▲" if chg>=0 else "▼"} {abs(chg):.2f}%</span>',
-         bg_tr="rgba(255,255,255,.025)")}
-    {row("24h Yüksek / Düşük", f'<span style="color:#00e676">{h24:,.6g}</span> / <span style="color:#ff5252">{l24:,.6g}</span>')}
-    {row("🛑 Stop Loss",   f'<b style="color:#ff5252">${sl:,.6g}</b>',
-         bg_tr="rgba(255,82,82,.04)")}
-    {row("✅ Take Profit", f'<b style="color:#00e676">${tp:,.6g}</b>',
-         bg_tr="rgba(0,200,83,.04)")}
-    {row("⚖️ Risk/Ödül",  f'<b>{rr}x</b>')}
-    {row("📉 RSI (14)",   f'{rsi_v:.1f}', bg_tr="rgba(255,255,255,.025)")}
-    {row("📏 ATR (14)",   f'${atr_v:,.6g}')}
-    {row("EMA 20/50/200", f'{ema20:,.6g} / {ema50:,.6g} / {ema200:,.6g}',
-         bg_tr="rgba(255,255,255,.025)")}
-    {row("VWAP",          f'${vwap_v:,.6g}')}
-    {row("24h Hacim",     f'{vol24:,.0f}')}
-  </table>
-
-  <!-- Koşullar -->
-  <div style="margin-top:12px;padding:10px;background:#0d1117;border-radius:8px;
-       border:1px solid #1c2128;font-size:.72em;color:#8b949e;line-height:1.8">
-    <div style="color:#444;margin-bottom:4px;letter-spacing:1px">SİNYAL KOŞULLARI</div>
-    {cond_txt.replace("  ","&nbsp;&nbsp;")}
-  </div>
-
-  <p style="margin:12px 0 0;font-size:.65em;color:#30363d;text-align:center;letter-spacing:.5px">
-    ⚠️ Eğitim amaçlıdır — finansal tavsiye değildir. DYOR.
-  </p>
-</div>"""
-    return html
-
-# ──────────────────────────────────────────────────────────
-# ANA ANALİZ FONKSİYONU
-# ──────────────────────────────────────────────────────────
-def analyze(label, interval):
-    # 1. Veri çek
+def get_forex_data(symbol="EURUSD=X", period="60d"):
+    """Forex verilerini indir"""
     try:
-        df_raw=fetch_okx(label, interval, limit=300)
+        data = yf.download(symbol, period=period, interval="1d", progress=False)
+        return data
     except Exception as e:
-        err=f"<div style='color:#ff5252;font-family:monospace;padding:20px'>❌ OKX veri hatası:<br>{e}</div>"
-        return None, err
-
-    # 2. Göstergeler
-    df=add_indicators(df_raw)
-    if len(df)<50:
-        return None,"<div style='color:#ff5252;font-family:monospace;padding:20px'>❌ Yeterli veri yok.</div>"
-
-    # 3. Sinyal
-    sig,tech_conf,sl,tp,rr,cond_txt=generate_signal(df)
-
-    # 4. AI — Premium Ensemble
-    ai_conf, model_probs = train_ensemble(df)
-
-    # 5. Ticker
+        return None
+
+def create_features(data):
+    """Teknik göstergeler ve özellikler oluştur"""
+    df = data.copy()
+    
+    # Temel özellikler
+    df['Returns'] = df['Close'].pct_change()
+    df['High_Low'] = df['High'] - df['Low']
+    df['Close_Open'] = df['Close'] - df['Open']
+    
+    # Hareketli ortalamalar
+    df['SMA_5'] = df['Close'].rolling(window=5).mean()
+    df['SMA_10'] = df['Close'].rolling(window=10).mean()
+    df['SMA_20'] = df['Close'].rolling(window=20).mean()
+    
+    # RSI
+    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))
+    
+    # MACD
+    exp1 = df['Close'].ewm(span=12, adjust=False).mean()
+    exp2 = df['Close'].ewm(span=26, adjust=False).mean()
+    df['MACD'] = exp1 - exp2
+    
+    # Bollinger Bands
+    df['BB_middle'] = df['Close'].rolling(window=20).mean()
+    df['BB_std'] = df['Close'].rolling(window=20).std()
+    df['BB_upper'] = df['BB_middle'] + (df['BB_std'] * 2)
+    df['BB_lower'] = df['BB_middle'] - (df['BB_std'] * 2)
+    
+    # Lag features
+    for i in range(1, 6):
+        df[f'Lag_{i}'] = df['Close'].shift(i)
+    
+    df = df.dropna()
+    return df
+
+def train_ensemble_models(df):
+    """5 farklı model eğit"""
+    feature_cols = ['Returns', 'High_Low', 'Close_Open', 'SMA_5', 'SMA_10', 
+                    'SMA_20', 'RSI', 'MACD', 'BB_middle', 'BB_upper', 'BB_lower',
+                    'Lag_1', 'Lag_2', 'Lag_3', 'Lag_4', 'Lag_5']
+    
+    X = df[feature_cols].values
+    y = df['Close'].values
+    
+    # Train/test split (son 5 gün test için)
+    split_idx = -5
+    X_train, X_test = X[:split_idx], X[split_idx:]
+    y_train, y_test = y[:split_idx], y[split_idx:]
+    
+    # Normalizasyon
+    scaler = StandardScaler()
+    X_train_scaled = scaler.fit_transform(X_train)
+    X_test_scaled = scaler.transform(X_test)
+    
+    # 5 Ensemble Model
+    models = {
+        'Random Forest': RandomForestRegressor(n_estimators=100, random_state=42),
+        'Gradient Boosting': GradientBoostingRegressor(n_estimators=100, random_state=42),
+        'AdaBoost': AdaBoostRegressor(n_estimators=100, random_state=42),
+        'Ridge': Ridge(alpha=1.0),
+        'SVR': SVR(kernel='rbf', C=100, gamma=0.1)
+    }
+    
+    predictions = {}
+    for name, model in models.items():
+        model.fit(X_train_scaled, y_train)
+        pred = model.predict(X_test_scaled)
+        predictions[name] = pred
+    
+    return models, scaler, predictions, y_test, feature_cols
+
+def predict_next_day(models, scaler, df, feature_cols):
+    """Ertesi gün tahmini yap"""
+    latest_features = df[feature_cols].iloc[-1].values.reshape(1, -1)
+    latest_features_scaled = scaler.transform(latest_features)
+    
+    predictions = {}
+    for name, model in models.items():
+        pred = model.predict(latest_features_scaled)[0]
+        predictions[name] = pred
+    
+    # Ensemble ortalaması
+    ensemble_pred = np.mean(list(predictions.values()))
+    
+    return predictions, ensemble_pred
+
+def predict_eurusd():
+    """Ana tahmin fonksiyonu"""
     try:
-        ticker=fetch_ticker(label)
-    except:
-        ticker={}
-
-    # 6. Grafik & Panel
-    fig=build_chart(df,label,interval,sig,sl,tp,ticker)
-    panel=build_panel(label,interval,sig,tech_conf,ai_conf,sl,tp,rr,cond_txt,ticker,df,model_probs)
-    return fig, panel
+        # Veri indir
+        data = get_forex_data()
+        if data is None or len(data) == 0:
+            return "❌ Veri indirilemedi. Lütfen internet bağlantınızı kontrol edin.", None
+        
+        # Özellikler oluştur
+        df = create_features(data)
+        
+        # Modelleri eğit
+        models, scaler, test_predictions, y_test, feature_cols = train_ensemble_models(df)
+        
+        # Ertesi gün tahmini
+        next_day_preds, ensemble_pred = predict_next_day(models, scaler, df, feature_cols)
+        
+        # Mevcut fiyat
+        current_price = df['Close'].iloc[-1]
+        
+        # Sonuçları hazırla
+        result = f"""
+🔄 **GÜNCELLEME ZAMANI:** {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}
+
+���� **MEVCUT EUR/USD FİYATI:** {current_price:.5f}
+
+🎯 **ERTESİ GÜN TAHMİNLERİ:**
 
-# ──────────────────────────────────────────────────────────
-# GRADIO  UI
-# ──────────────────────────────────────────────────────────
-CSS="""
-body,.gradio-container{background:#07070d !important;}
-footer{display:none !important;}
-.gr-button.primary{
-  background:linear-gradient(135deg,#1565c0,#4a148c) !important;
-  border:1px solid #42a5f566 !important;color:#fff !important;
-  font-family:'Courier New',monospace !important;letter-spacing:1px !important;
-  transition:all .2s !important;
-}
-.gr-button.primary:hover{
-  background:linear-gradient(135deg,#1976d2,#6a1b9a) !important;
-  box-shadow:0 0 16px #42a5f544 !important;
-}
-.gr-button{
-  background:#0f0f1a !important;border:1px solid #1e1e3f !important;
-  color:#8b949e !important;font-family:'Courier New',monospace !important;
-}
-select,input{background:#0d1117 !important;border:1px solid #21262d !important;
-  color:#e6edf3 !important;font-family:'Courier New',monospace !important;}
 """
-
-HEADER="""
-<div style="background:linear-gradient(135deg,#07070d,#0d0d1a,#070712);
-     border:1px solid #1c1c3a;border-radius:16px;padding:24px 28px;
-     text-align:center;font-family:'Courier New',monospace;
-     margin-bottom:6px;position:relative;overflow:hidden;">
-  <div style="position:absolute;top:0;left:0;right:0;height:2px;
-       background:linear-gradient(90deg,transparent,#3d5afe,#00e5ff,#7c4dff,transparent)"></div>
-  <div style="position:absolute;bottom:0;left:0;right:0;height:1px;
-       background:linear-gradient(90deg,transparent,#1c2a4a,transparent)"></div>
-  <h1 style="margin:0 0 4px;font-size:1.75em;letter-spacing:5px;
-       background:linear-gradient(90deg,#3d5afe,#00e5ff,#7c4dff);
-       -webkit-background-clip:text;-webkit-text-fill-color:transparent;
-       background-clip:text;text-transform:uppercase;font-weight:900">
-    ⚡ CRYPTO SIGNAL PRO
-  </h1>
-  <p style="margin:0;color:#2a3550;letter-spacing:3px;font-size:.68em;text-transform:uppercase">
-    OKX REST API  ·  XGBoost AI  ·  Canlı Otomatik Analiz
-  </p>
-  <div style="display:flex;justify-content:center;gap:18px;margin-top:10px;
-       font-size:.65em;color:#1c2a4a;letter-spacing:2px">
-    <span>EMA 20/50/200</span><span>·</span><span>RSI · MACD</span>
-    <span>·</span><span>VWAP · ATR</span><span>·</span><span>Volume Profile</span>
-  </div>
-</div>"""
-
-STATUS_TMPL="""
-<div style="display:flex;align-items:center;gap:10px;padding:8px 14px;
-     background:#0d1117;border:1px solid #21262d;border-radius:8px;
-     font-family:'Courier New',monospace;font-size:.78em;color:#8b949e;">
-  <span style="width:8px;height:8px;border-radius:50%;background:{col};
-       box-shadow:0 0 6px {col};display:inline-block"></span>
-  {msg}
-</div>"""
-
-def make_status(msg,ok=True):
-    return STATUS_TMPL.format(col="#00e676" if ok else "#ff5252",msg=msg)
-
-# Refresh count (hafıza için state)
-_refresh_count = {"n": 0}
-
-def run_and_status(label, interval):
-    _refresh_count["n"] += 1
-    n = _refresh_count["n"]
-    fig, panel = analyze(label, interval)
-    now = datetime.now(timezone.utc).strftime("%H:%M:%S UTC")
-    if fig is None:
-        st = make_status(f"#{n} · {now}  ❌ Hata",ok=False)
-    else:
-        st = make_status(f"#{n} · {now}  ✅ {label} / {interval} güncellendi")
-    return fig, panel, st
-
-with gr.Blocks(css=CSS, title="⚡ Crypto Signal Pro · OKX") as demo:
-    gr.HTML(HEADER)
-
-    with gr.Row(equal_height=True):
-        sym = gr.Dropdown(choices=SYMBOLS_LABEL, value="BTCUSDT",
-                          label="📊 Parite", scale=3, interactive=True)
-        ivl = gr.Dropdown(choices=INTERVALS, value="1H",
-                          label="⏱ Zaman Dilimi", scale=2, interactive=True)
-        auto_cb = gr.Checkbox(value=True, label="🔄 Otomatik Yenile", scale=1)
-        btn = gr.Button("🔍 ANALİZ ET", variant="primary", scale=2)
-
-    status_box = gr.HTML(make_status("Hazır — Analiz Et butonuna bas veya otomatik yenilemeyi bekle."))
-
+        
+        for model_name, pred in next_day_preds.items():
+            change = ((pred - current_price) / current_price) * 100
+            arrow = "📈" if change > 0 else "📉"
+            result += f"{arrow} **{model_name}:** {pred:.5f} ({change:+.2f}%)\n"
+        
+        ensemble_change = ((ensemble_pred - current_price) / current_price) * 100
+        ensemble_arrow = "📈" if ensemble_change > 0 else "📉"
+        
+        result += f"\n{ensemble_arrow} **🏆 ENSEMBLE ORTALAMA:** {ensemble_pred:.5f} ({ensemble_change:+.2f}%)\n"
+        
+        # Model performansları
+        result += f"\n\n📈 **MODEL PERFORMANSLARI (Son 5 Gün):**\n"
+        for i, (model_name, preds) in enumerate(test_predictions.items()):
+            mse = np.mean((preds - y_test) ** 2)
+            mae = np.mean(np.abs(preds - y_test))
+            result += f"• {model_name}: MAE={mae:.5f}, MSE={mse:.8f}\n"
+        
+        # Grafik için veri hazırla
+        import matplotlib.pyplot as plt
+        
+        fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(12, 10))
+        
+        # Üst grafik: Fiyat geçmişi
+        ax1.plot(df.index[-30:], df['Close'].iloc[-30:], label='Gerçek Fiyat', linewidth=2)
+        ax1.axhline(y=current_price, color='gray', linestyle='--', alpha=0.5, label='Mevcut Fiyat')
+        ax1.axhline(y=ensemble_pred, color='red', linestyle='--', alpha=0.7, label='Tahmin')
+        ax1.set_title('EUR/USD Son 30 Gün Fiyat Hareketi', fontsize=14, fontweight='bold')
+        ax1.set_xlabel('Tarih')
+        ax1.set_ylabel('Fiyat')
+        ax1.legend()
+        ax1.grid(True, alpha=0.3)
+        
+        # Alt grafik: Model tahminleri karşılaştırması
+        model_names = list(next_day_preds.keys())
+        model_values = list(next_day_preds.values())
+        colors = ['#FF6B6B', '#4ECDC4', '#45B7D1', '#FFA07A', '#98D8C8']
+        
+        bars = ax2.barh(model_names, model_values, color=colors)
+        ax2.axvline(x=current_price, color='black', linestyle='--', linewidth=2, label='Mevcut Fiyat')
+        ax2.set_title('Model Tahminleri Karşılaştırması', fontsize=14, fontweight='bold')
+        ax2.set_xlabel('Tahmin Edilen Fiyat')
+        ax2.legend()
+        ax2.grid(True, alpha=0.3, axis='x')
+        
+        # Değerleri barların üzerine yaz
+        for i, (bar, val) in enumerate(zip(bars, model_values)):
+            change = ((val - current_price) / current_price) * 100
+            ax2.text(val, i, f' {val:.5f} ({change:+.2f}%)', 
+                    va='center', fontsize=9, fontweight='bold')
+        
+        plt.tight_layout()
+        
+        return result, fig
+        
+    except Exception as e:
+        return f"❌ Hata oluştu: {str(e)}", None
+
+# Gradio arayüzü
+with gr.Blocks(title="EUR/USD Fiyat Tahmini", theme=gr.themes.Soft()) as app:
+    gr.Markdown("""
+    # 💹 EUR/USD Fiyat Tahmin Sistemi
+    ## 5 Ensemble Model ile Otomatik Forex Tahmini
+    
+    Bu uygulama **Random Forest, Gradient Boosting, AdaBoost, Ridge Regression ve SVR** modellerini kullanarak EUR/USD paritesi için ertesi gün fiyat tahmini yapar.
+    
+    🔄 **Yenile** butonuna tıklayarak güncel verileri çekip yeni tahmin alabilirsiniz.
+    """)
+    
     with gr.Row():
-        chart_out = gr.Plot(label="Canlı Grafik", scale=3)
-        panel_out = gr.HTML(label="Sinyal Paneli", scale=1)
-
-    # Manuel buton
-    btn.click(fn=run_and_status,
-              inputs=[sym, ivl],
-              outputs=[chart_out, panel_out, status_box])
-
-    # Otomatik yenileme — gr.Timer (Gradio 4.x+)
-    # Her 30 saniyede bir tetiklenir; checkbox kapalıysa atlar
-    def auto_refresh(label, interval, enabled):
-        if not enabled:
-            return gr.update(), gr.update(), gr.update()
-        return run_and_status(label, interval)
-
-    timer = gr.Timer(value=30)   # 30 sn
-    timer.tick(fn=auto_refresh,
-               inputs=[sym, ivl, auto_cb],
-               outputs=[chart_out, panel_out, status_box])
-
-    gr.HTML("""
-<div style="text-align:center;padding:10px;color:#1c2a4a;
-     font-size:.62em;font-family:'Courier New',monospace;letter-spacing:1px">
-  CRYPTO SIGNAL PRO  ·  OKX REST API  ·  XGBoost AI  ·  Plotly  ·  Gradio<br>
-  ⚠️ Yatırım tavsiyesi değildir — DYOR
-</div>""")
+        predict_btn = gr.Button("🚀 Tahmin Yap / Yenile", variant="primary", scale=2)
+    
+    with gr.Row():
+        output_text = gr.Textbox(label="📊 Tahmin Sonuçları", lines=25)
+    
+    with gr.Row():
+        output_plot = gr.Plot(label="📈 Grafikler")
+    
+    predict_btn.click(
+        fn=predict_eurusd,
+        inputs=[],
+        outputs=[output_text, output_plot]
+    )
+    
+    gr.Markdown("""
+    ---
+    ### 📝 Notlar:
+    - Veriler Yahoo Finance'ten otomatik olarak çekilir
+    - Tahminler son 60 günlük veriye dayanır
+    - 5 farklı makine öğrenmesi modeli kullanılır
+    - Ensemble tahmini, tüm modellerin ortalamasıdır
+    - **Finansal tavsiye değildir, sadece eğitim amaçlıdır**
+    """)
 
 if __name__ == "__main__":
-    demo.launch(server_name="0.0.0.0", server_port=7860)
+    app.launch()
+