Update app.py

app.py

@@ -141,6 +141,9 @@ def fetch_current_prices(item_ids: List[str], cities: List[str], qualities: str
     if not all_data:
         return pd.DataFrame()
     df = pd.DataFrame(all_data)
+    for col in ["sell_price_min", "sell_price_max", "buy_price_min", "buy_price_max"]:
+        if col in df.columns:
+            df[col] = pd.to_numeric(df[col], errors="coerce").astype("float64")
     for col in ["sell_price_min_date", "sell_price_max_date", "buy_price_min_date", "buy_price_max_date"]:
         if col in df.columns:
             df[col] = pd.to_datetime(df[col], errors="coerce")
@@ -176,14 +179,18 @@ def fetch_history(item_ids: List[str], cities: List[str], time_scale: int = 24,
         return pd.DataFrame()
     df = pd.DataFrame(all_records)
     df["timestamp"] = pd.to_datetime(df["timestamp"], errors="coerce")
-    df["avg_price"] = pd.to_numeric(df["avg_price"], errors="coerce")
-    df["item_count"] = pd.to_numeric(df["item_count"], errors="coerce").fillna(0)
+    df["avg_price"] = pd.to_numeric(df["avg_price"], errors="coerce").astype("float64")
+    df["item_count"] = pd.to_numeric(df["item_count"], errors="coerce").fillna(0).astype("float64")
     return df
 
 
 def clean_market_data(df: pd.DataFrame) -> pd.DataFrame:
     if df.empty:
         return df
+    df = df.copy()
+    df["avg_price"] = pd.to_numeric(df["avg_price"], errors="coerce").astype("float64")
+    if "item_count" in df.columns:
+        df["item_count"] = pd.to_numeric(df["item_count"], errors="coerce").fillna(0).astype("float64")
     df = df[df["avg_price"] > 0].copy()
     if "item_id" in df.columns:
         for item in df["item_id"].unique():
@@ -192,7 +199,8 @@ def clean_market_data(df: pd.DataFrame) -> pd.DataFrame:
                 continue
             lower = df.loc[mask, "avg_price"].quantile(0.005)
             upper = df.loc[mask, "avg_price"].quantile(0.995)
-            df.loc[mask, "avg_price"] = df.loc[mask, "avg_price"].clip(lower, upper)
+            if pd.notna(lower) and pd.notna(upper) and upper > lower:
+                df.loc[mask, "avg_price"] = df.loc[mask, "avg_price"].clip(lower, upper)
     df = df.sort_values(["item_id", "city", "timestamp"]).reset_index(drop=True)
     return df
 
@@ -224,6 +232,8 @@ def build_features(df: pd.DataFrame) -> pd.DataFrame:
         return df
 
     df = df.sort_values(["item_id", "city", "timestamp"]).copy()
+    df["avg_price"] = pd.to_numeric(df["avg_price"], errors="coerce").astype("float64")
+    df["item_count"] = pd.to_numeric(df["item_count"], errors="coerce").fillna(0).astype("float64")
 
     df["hour"] = df["timestamp"].dt.hour
     df["dayofweek"] = df["timestamp"].dt.dayofweek
@@ -293,8 +303,8 @@ def prepare_model_data(featured_df: pd.DataFrame) -> Tuple[pd.DataFrame, pd.Seri
         upper = model_df["target_return_24h"].quantile(0.999)
         model_df["target_return_24h"] = model_df["target_return_24h"].clip(lower, upper)
 
-    X = model_df[FEATURE_COLS]
-    y = model_df["target_return_24h"]
+    X = model_df[FEATURE_COLS].astype("float64")
+    y = model_df["target_return_24h"].astype("float64")
     meta = model_df[["item_id", "city", "timestamp", "avg_price", "target_return_24h"]].copy()
 
     return X, y, meta
@@ -348,7 +358,7 @@ def predict_24h_returns(model: lgb.LGBMRegressor, latest_features: pd.DataFrame)
         print(f"Eksik özellikler: {missing}")
         return pd.DataFrame()
 
-    X_pred = latest_features[FEATURE_COLS].dropna()
+    X_pred = latest_features[FEATURE_COLS].astype("float64").dropna()
     if X_pred.empty:
         return pd.DataFrame()
 
@@ -512,8 +522,11 @@ class DataPipeline:
         self.prices_df = prices
         self.status_message = f"{len(self.history_df)} satır tarihsel veri, {len(self.prices_df)} anlık fiyat alındı."
 
-        save_to_csv(self.history_df, "history")
-        save_to_csv(self.prices_df, "prices")
+        try:
+            save_to_csv(self.history_df, "history")
+            save_to_csv(self.prices_df, "prices")
+        except Exception as e:
+            print(f"CSV kaydetme hatası: {e}")
 
         self.featured_df = build_features(self.history_df)
         self._backtest_cache = None
@@ -549,6 +562,8 @@ class DataPipeline:
 
         if not preds.empty and not self.prices_df.empty:
             price_info = self.prices_df[["item_id", "city", "sell_price_min", "buy_price_max"]].copy()
+            price_info["sell_price_min"] = pd.to_numeric(price_info["sell_price_min"], errors="coerce").astype("float64")
+            price_info["buy_price_max"] = pd.to_numeric(price_info["buy_price_max"], errors="coerce").astype("float64")
             preds = preds.merge(price_info, on=["item_id", "city"], how="left")
             preds["spread"] = preds["sell_price_min"] - preds["buy_price_max"]
         else:
@@ -600,10 +615,10 @@ def get_analysis_table():
     latest = pipeline.featured_df.sort_values("timestamp").groupby(["item_id", "city"]).tail(1).copy()
 
     if not pipeline.prices_df.empty:
-        latest = latest.merge(
-            pipeline.prices_df[["item_id", "city", "sell_price_min", "buy_price_max"]],
-            on=["item_id", "city"], how="left"
-        )
+        price_info = pipeline.prices_df[["item_id", "city", "sell_price_min", "buy_price_max"]].copy()
+        price_info["sell_price_min"] = pd.to_numeric(price_info["sell_price_min"], errors="coerce").astype("float64")
+        price_info["buy_price_max"] = pd.to_numeric(price_info["buy_price_max"], errors="coerce").astype("float64")
+        latest = latest.merge(price_info, on=["item_id", "city"], how="left")
     else:
         latest["sell_price_min"] = np.nan
         latest["buy_price_max"] = np.nan
@@ -658,6 +673,8 @@ def get_arbitrage_table():
         return pd.DataFrame({"Bilgi": ["Anlık fiyat verisi yok."]})
 
     prices = pipeline.prices_df.copy()
+    prices["sell_price_min"] = pd.to_numeric(prices["sell_price_min"], errors="coerce").astype("float64")
+    prices["buy_price_max"] = pd.to_numeric(prices["buy_price_max"], errors="coerce").astype("float64")
     prices = prices[prices["sell_price_min"] > 0]
 
     if prices.empty:
@@ -765,37 +782,89 @@ def list_csv_files() -> pd.DataFrame:
 
 
 def refresh_data_ui():
-    pipeline.fetch_data()
-    if not pipeline.featured_df.empty:
-        pipeline.train_model()
+    try:
+        pipeline.fetch_data()
+        if not pipeline.featured_df.empty:
+            pipeline.train_model()
+    except Exception as e:
+        pipeline.status_message = f"Beklenmeyen hata: {e}"
+        print(f"Refresh hatası: {e}")
+        import traceback
+        traceback.print_exc()
+
+    try:
+        bt_table, bt_fig, sim_table = get_backtest_outputs()
+    except Exception as e:
+        print(f"Backtest hatası: {e}")
+        bt_table = pd.DataFrame({"Bilgi": [f"Backtest hatası: {e}"]})
+        bt_fig = go.Figure().update_layout(title="Backtest hatası", template="plotly_dark")
+        sim_table = pd.DataFrame()
+
+    try:
+        dash = get_dashboard_metrics()
+    except Exception as e:
+        dash = pd.DataFrame({"Bilgi": [f"Hata: {e}"]})
+
+    try:
+        ana = get_analysis_table()
+    except Exception as e:
+        ana = pd.DataFrame({"Bilgi": [f"Hata: {e}"]})
+
+    try:
+        pred = get_prediction_table()
+    except Exception as e:
+        pred = pd.DataFrame({"Bilgi": [f"Hata: {e}"]})
+
+    try:
+        arb = get_arbitrage_table()
+    except Exception as e:
+        arb = pd.DataFrame({"Bilgi": [f"Hata: {e}"]})
+
+    try:
+        perf = get_model_performance_plot()
+    except Exception as e:
+        perf = go.Figure().update_layout(title=f"Hata: {e}", template="plotly_dark")
 
-    bt_table, bt_fig, sim_table = get_backtest_outputs()
+    try:
+        csv_df = list_csv_files()
+    except Exception:
+        csv_df = pd.DataFrame(columns=["Dosya", "Boyut (KB)", "Son Değişiklik"])
 
     return (
         pipeline.status_message,
-        get_dashboard_metrics(),
-        get_analysis_table(),
-        get_prediction_table(),
-        get_arbitrage_table(),
-        get_model_performance_plot(),
+        dash,
+        ana,
+        pred,
+        arb,
+        perf,
         bt_table,
         bt_fig,
         sim_table,
-        list_csv_files(),
+        csv_df,
     )
 
 
 def initial_load():
+    try:
+        dash = get_dashboard_metrics()
+    except Exception:
+        dash = pd.DataFrame({"Metrik": ["Durum"], "Değer": ["Hazır"]})
+
+    try:
+        csv_df = list_csv_files()
+    except Exception:
+        csv_df = pd.DataFrame(columns=["Dosya", "Boyut (KB)", "Son Değişiklik"])
+
     return (
-        get_dashboard_metrics(),
-        get_analysis_table(),
-        get_prediction_table(),
-        get_arbitrage_table(),
-        get_model_performance_plot(),
+        dash,
+        pd.DataFrame(),
+        pd.DataFrame(),
         pd.DataFrame(),
         go.Figure().update_layout(title="Verileri yenilemek için butona tıklayın", template="plotly_dark"),
         pd.DataFrame(),
-        list_csv_files(),
+        go.Figure().update_layout(title="Backtest için veri gerekli", template="plotly_dark"),
+        pd.DataFrame(),
+        csv_df,
     )