Update app.py

app.py

@@ -9,20 +9,18 @@ from sklearn.neighbors import KNeighborsClassifier
 from sklearn.preprocessing import StandardScaler
 from groq import Groq
 
+# ==========================================
+# 🔌 1. GROQ CLIENT
+# ==========================================
 api_key = os.environ.get("GROQ_API_KEY")
 client = Groq(api_key=api_key) if api_key else None
 
-def load_dataframe(file_obj, url):
+# ==========================================
+# 📊 2. VERİ OKUMA (Tab/CSV, başlıksız)
+# ==========================================
+def parse_data(text):
+    """Tab veya virgülle ayrılmış, başlıksız veriyi okur."""
     try:
-        if url and url.strip():
-            r = requests.get(url.strip(), headers={"User-Agent": "Mozilla/5.0"}, timeout=30)
-            r.raise_for_status()
-            text = r.text
-        elif file_obj is not None:
-            text = open(file_obj.name, "r", encoding="utf-8", errors="ignore").read()
-        else:
-            return None, "Lütfen dosya yükleyin veya link girin."
-
         df = pd.read_csv(io.StringIO(text), sep="\t", header=None)
         if df.shape[1] < 6:
             df = pd.read_csv(io.StringIO(text), sep=",", header=None)
@@ -31,9 +29,11 @@ def load_dataframe(file_obj, url):
 
         df = df.iloc[:, :6]
         df.columns = ["date", "open", "high", "low", "close", "volume"]
+
         for c in ["open", "high", "low", "close", "volume"]:
             df[c] = pd.to_numeric(df[c], errors="coerce")
         df = df.dropna(subset=["close"]).reset_index(drop=True)
+
         if len(df) < 60:
             return None, f"En az 60 satır gerekli, {len(df)} var."
         return df, None
@@ -41,6 +41,22 @@ def load_dataframe(file_obj, url):
         return None, f"Veri okunamadı: {e}"
 
 
+def get_dataframe(raw_text, url):
+    """Önce yapıştırılan metni, yoksa URL'yi dener."""
+    if raw_text and raw_text.strip():
+        return parse_data(raw_text)
+    if url and url.strip():
+        try:
+            r = requests.get(url.strip(), headers={"User-Agent": "Mozilla/5.0"}, timeout=30)
+            r.raise_for_status()
+            return parse_data(r.text)
+        except Exception as e:
+            return None, f"URL okunamadı: {e}"
+    return None, "Lütfen CSV içeriğini yapıştırın veya bir URL girin."
+
+# ==========================================
+# 🧮 3. TEKNİK GÖSTERGELER + MODEL
+# ==========================================
 def compute_rsi(series, period=14):
     delta = series.diff()
     gain = delta.clip(lower=0).rolling(period).mean()
@@ -56,9 +72,10 @@ def train_models(df):
     df["rsi"] = compute_rsi(df["close"], 14)
     df["macd"] = df["close"].ewm(span=12).mean() - df["close"].ewm(span=26).mean()
     df["target"] = (df["close"].shift(-1) > df["close"]).astype(int)
+
     df_model = df.dropna().reset_index(drop=True)
     if len(df_model) < 20:
-        return None, None, None, None, "Yetersiz veri."
+        return None, None, None, None, "Göstergelerden sonra yetersiz veri."
 
     features = ["ma_5", "ma_20", "rsi", "macd", "volume"]
     X = df_model[features].values
@@ -87,9 +104,11 @@ def train_models(df):
     }
     return xgb_model, knn_model, scaler, metrics, None
 
-
-def predict_and_analyze(file_obj, url):
-    df, err = load_dataframe(file_obj, url)
+# ==========================================
+# 🧠 4. ANALİZ FONKSİYONU
+# ==========================================
+def predict_and_analyze(raw_text, url):
+    df, err = get_dataframe(raw_text, url)
     if err:
         return err, {}
 
@@ -97,7 +116,9 @@ def predict_and_analyze(file_obj, url):
     if err:
         return err, {}
 
-    feat = np.array([[metrics["ma5"], metrics["ma20"], metrics["rsi"], metrics["macd"], metrics["volume"]]])
+    feat = np.array([[metrics["ma5"], metrics["ma20"], metrics["rsi"],
+                      metrics["macd"], metrics["volume"]]])
+
     xgb_up = float(xgb_model.predict_proba(feat)[0][1])
     knn_up = float(knn_model.predict_proba(scaler.transform(feat))[0][1])
     ens_up = (xgb_up + knn_up) / 2
@@ -138,30 +159,48 @@ def predict_and_analyze(file_obj, url):
     header = (
         f"**Son: ${metrics['last_close']:,.2f}** | RSI: {metrics['rsi']:.1f} | "
         f"MA5: ${metrics['ma5']:,.0f} | MA20: ${metrics['ma20']:,.0f}\n\n"
-        f"**XGB:** %{xgb_up*100:.1f} | **KNN:** %{knn_up*100:.1f} | **Ens:** %{ens_up*100:.1f}\n\n---\n\n"
+        f"**XGB:** %{xgb_up*100:.1f} | **KNN:** %{knn_up*100:.1f} | "
+        f"**Ens:** %{ens_up*100:.1f}\n\n---\n\n"
     )
     return header + analysis, prob_dict
 
-
+# ==========================================
+# 🎨 5. GRADIO ARAYÜZÜ
+# ==========================================
 with gr.Blocks(theme=gr.themes.Soft(primary_hue="orange", secondary_hue="slate")) as demo:
     gr.Markdown(
         """
         # ₿ BTC/USD XGBoost + KNN & Groq AI Tahmin Radarı
-        **Başlıksız** TSV/CSV yükleyin (Date, Open, High, Low, Close, Volume) **veya** link girin.
+        **Dosya seçici kaldırıldı.** CSV içeriğini doğrudan aşağıya yapıştırın **veya** bir URL girin.
+        Veri **tab veya virgülle ayrılmış, başlıksız** olmalı:
+        `Date, Open, High, Low, Close, Volume`
         """
     )
+
     with gr.Row():
         with gr.Column():
-            file_input = gr.File(label="Dosya Yükle (.csv / .tsv / .txt)", file_types=[".csv", ".tsv", ".txt"])
-            url_input = gr.Textbox(label="veya Link (URL)", placeholder="https://.../BTCUSD30.csv")
+            gr.Markdown("### 📋 CSV İçeriğini Yapıştır")
+            raw_input = gr.Textbox(
+                label="CSV/TSV İçeriği",
+                placeholder="2021-02-04 11:30\t37826.1\t37916.7\t37359.8\t37359.8\t5032\n...",
+                lines=10
+            )
+
+            gr.Markdown("### 🔗 veya URL")
+            url_input = gr.Textbox(
+                label="CSV Linki",
+                placeholder="https://raw.githubusercontent.com/.../BTCUSD30.csv"
+            )
+
             submit_btn = gr.Button("Analiz Et", variant="primary")
+
         with gr.Column():
             prob_output = gr.Label(label="Model Tahminleri")
-            analysis_output = gr.Markdown(value="Veri yükleyip butona basın.")
+            analysis_output = gr.Markdown(value="Veri girip butona basın.")
 
     submit_btn.click(
         fn=predict_and_analyze,
-        inputs=[file_input, url_input],
+        inputs=[raw_input, url_input],
         outputs=[analysis_output, prob_output]
     )