Refactor trend detection application to enhance CSV and Excel file handling, improve future forecasting logic, and update JSON output to use temporary files. Adjust Gradio UI for better user experience and output display.

app.py

@@ -1,24 +1,29 @@
-# 傾向検出アプリ（Gradio版）
+# 傾向検出アプリ（Gradio版・重要項目ごとの詳細＋JSONダウンロード）修正版
 
 import gradio as gr
 import pandas as pd
 import numpy as np
 from sklearn.linear_model import LinearRegression
 import json
-import io
+import tempfile
+import os
 
 # --- 状態判定＆未来予測関数 ---
 def detect_trends_with_forecast(process_name, datetime_str, window_minutes, forecast_minutes, csv_file, excel_file):
     try:
+        # ★ 修正ポイント: gr.File(type="filepath") は文字列パスが来る
+        csv_path = csv_file
+        excel_path = excel_file
+
         # CSV読み込み（3行ヘッダー）
-        df = pd.read_csv(csv_file.name, header=[0, 1, 2])
+        df = pd.read_csv(csv_path, header=[0, 1, 2])
         timestamp_col = df.iloc[:, 0]
         df = df.drop(df.columns[0], axis=1)
         df.insert(0, "timestamp", timestamp_col)
         df["timestamp"] = pd.to_datetime(df["timestamp"], errors="coerce")
 
         # 閾値テーブル読み込み
-        thresholds_df = pd.read_excel(excel_file.name)
+        thresholds_df = pd.read_excel(excel_path)
         thresholds_df["Important"] = thresholds_df["Important"].astype(str).str.upper().map({"TRUE": True, "FALSE": False})
         for col in ["LL", "L", "H", "HH"]:
             if col in thresholds_df.columns:
@@ -35,7 +40,7 @@ def detect_trends_with_forecast(process_name, datetime_str, window_minutes, fore
 
         # サンプリング間隔
         interval = df_window["timestamp"].diff().median()
-        if pd.isna(interval):
+        if pd.isna(interval) or interval == pd.Timedelta(0):
             return None, "⚠ サンプリング間隔を検出できません", None
         interval_minutes = interval.total_seconds() / 60
 
@@ -64,8 +69,8 @@ def detect_trends_with_forecast(process_name, datetime_str, window_minutes, fore
             last_val = series.iloc[-1]
             n = len(series)
 
-            # 未来予測
-            forecast_steps = int(forecast_minutes / interval_minutes)
+            # 未来予測（最小1ステップ）
+            forecast_steps = max(1, int(round(forecast_minutes / interval_minutes)))
             forecast_index = n + forecast_steps
             forecast_val = model.predict([[forecast_index]])[0][0]
             forecast_time = target_time + pd.Timedelta(minutes=forecast_minutes)
@@ -95,22 +100,33 @@ def detect_trends_with_forecast(process_name, datetime_str, window_minutes, fore
             results.append({
                 "ItemName": thr["ItemName"],
                 "傾向": status,
-                "傾き": round(slope, 4),
-                "最終値": round(float(last_val), 3) if pd.notna(last_val) else None,
-                "予測値": round(float(forecast_val), 3),
+                "傾き": float(round(slope, 4)),
+                "最終値": float(round(float(last_val), 3)) if pd.notna(last_val) else None,
+                "予測値": float(round(float(forecast_val), 3)),
                 "予測時刻": str(forecast_time),
                 "予測傾向": forecast_status,
-                "サンプリング間隔(分)": interval_minutes,
-                "LL": ll, "L": l, "H": h, "HH": hh
+                "サンプリング間隔(分)": float(interval_minutes),
+                "LL": None if pd.isna(ll) else float(ll),
+                "L":  None if pd.isna(l)  else float(l),
+                "H":  None if pd.isna(h)  else float(h),
+                "HH": None if pd.isna(hh) else float(hh)
             })
 
         result_df = pd.DataFrame(results)
+
+        # JSONテキスト
         result_json = json.dumps(results, ensure_ascii=False, indent=2)
 
-        # JSONをファイルとして出力
-        json_bytes = io.BytesIO(result_json.encode("utf-8"))
+        # ★ 修正ポイント: BytesIO ではなく、一時ファイルに保存して「パス」を返す
+        fd, tmp_path = tempfile.mkstemp(suffix=".json", prefix="trend_result_")
+        os.close(fd)
+        with open(tmp_path, "w", encoding="utf-8") as f:
+            f.write(result_json)
+
+        summary = f"✅ {process_name} の傾向検出＋未来予測完了（{start_time}～{end_time}, 予測+{forecast_minutes}分）"
 
-        return result_df, "✅ 傾向検出＋未来予測完了", json_bytes
+        # gr.File 出力には「ファイルパス（文字列）」を返す
+        return result_df, summary, tmp_path
 
     except Exception as e:
         return None, f"❌ エラー: {str(e)}", None
@@ -134,8 +150,9 @@ with gr.Blocks() as demo:
     run_btn = gr.Button("実行")
 
     with gr.Row():
-        result_table = gr.Dataframe(label="傾向＋未来予測結果")
+        result_table = gr.Dataframe(label="傾向＋未来予測結果", wrap=True)
     summary_output = gr.Textbox(label="サマリー")
+    # ★ 修正ポイント: gr.File の出力にはパスを渡す。type指定は不要（出力）
     json_download = gr.File(label="JSON結果ダウンロード")
 
     run_btn.click(
@@ -145,7 +162,6 @@ with gr.Blocks() as demo:
     )
 
 if __name__ == "__main__":
-    import os
     use_mcp = os.getenv("USE_MCP", "0") == "1"
     if use_mcp:
         demo.launch(server_name="0.0.0.0", mcp_server=True)