Implement initial project structure and setup

app.py

@@ -0,0 +1,125 @@
+import os
+import gradio as gr
+import pandas as pd
+import numpy as np
+import json
+from sklearn.linear_model import LinearRegression
+
+# =========================================================
+# データロード部分（共通）
+# =========================================================
+csv_path = r"sample.csv"  # あなたのCSVパスに置き換え
+excel_path = r"thresholds.xlsx"  # あなたの閾値テーブルExcelパスに置き換え
+lag_excel_path = r"lag_table.xlsx"  # タイムラグ表Excelパスに置き換え
+
+# CSV（3行ヘッダー）
+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
+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:
+        thresholds_df[col] = pd.to_numeric(thresholds_df[col], errors="coerce")
+
+# タイムラグ表
+lag_matrix = pd.read_excel(lag_excel_path, index_col=0)
+
+
+# =========================================================
+# 閾値診断関数
+# =========================================================
+def diagnose(process_name, datetime_str, window_minutes):
+    try:
+        target_time = pd.to_datetime(datetime_str)
+    except Exception:
+        return None, "⚠ 日時が不正です", None
+
+    start_time = target_time - pd.Timedelta(minutes=int(window_minutes))
+    df_window = df[(df["timestamp"] >= start_time) & (df["timestamp"] <= target_time)]
+    if df_window.empty:
+        return None, "⚠ データなし", None
+
+    proc_thresholds = thresholds_df[thresholds_df["ProcessNo_ProcessName"] == process_name]
+    results = []
+    for _, thr in proc_thresholds.iterrows():
+        col_tuple = (thr["ColumnID"], thr["ItemName"], thr["ProcessNo_ProcessName"])
+        if col_tuple not in df.columns:
+            continue
+        series = df_window[col_tuple].dropna()
+        if series.empty:
+            continue
+
+        value = series.iloc[-1]
+        status = "OK"
+        if pd.notna(thr["LL"]) and value < thr["LL"]:
+            status = "LOW-LOW"
+        elif pd.notna(thr["L"]) and value < thr["L"]:
+            status = "LOW"
+        elif pd.notna(thr["HH"]) and value > thr["HH"]:
+            status = "HIGH-HIGH"
+        elif pd.notna(thr["H"]) and value > thr["H"]:
+            status = "HIGH"
+
+        results.append({
+            "ItemName": thr["ItemName"],
+            "値": value,
+            "判定": status,
+            "重要項目": bool(thr.get("Important", False))
+        })
+
+    result_df = pd.DataFrame(results)
+    result_json = json.dumps(results, ensure_ascii=False, indent=2, default=lambda x: x.item() if hasattr(x, "item") else x)
+    summary = f"✅ {process_name} の診断完了（{start_time}～{target_time}）"
+    return result_df, summary, result_json
+
+
+# =========================================================
+# Gradio UI 構築
+# =========================================================
+with gr.Blocks() as demo:
+    gr.Markdown("## 水処理データ解析アプリ")
+
+    with gr.Tab("閾値診断アプリ"):
+        process_in = gr.Textbox(label="プロセス名", value="E018-A012_除害RO")
+        datetime_in = gr.Textbox(label="基準日時 (例: 2025/8/1 0:05)", value="2025/8/1 0:05")
+        window_in = gr.Number(label="時間幅（分）", value=60)
+
+        run_btn = gr.Button("診断実行")
+        summary_out = gr.Textbox(label="サマリー")
+        table_out = gr.Dataframe(label="診断結果", interactive=False)
+        json_out = gr.Textbox(label="JSON出力")
+
+        def run_diagnose(process, dt, win):
+            df_out, summary, js = diagnose(process, dt, win)
+            return summary, df_out, js
+
+        run_btn.click(
+            fn=run_diagnose,
+            inputs=[process_in, datetime_in, window_in],
+            outputs=[summary_out, table_out, json_out]
+        )
+
+
+# =========================================================
+# 実行部分
+# =========================================================
+if __name__ == "__main__":
+    if os.getenv("USE_MCP", "0") == "1":
+        # Hugging Face 用: MCP + UI
+        demo.launch(
+            mcp_server=True,
+            server_name="0.0.0.0",
+            share=False
+        )
+    else:
+        # ローカル用: UI のみ
+        demo.launch(
+            server_name="0.0.0.0",
+            share=False,
+            ssr_mode=False
+        )

requirements.txt

@@ -0,0 +1,13 @@
+# Webアプリ/UI
+gradio>=4.44.0
+
+# データ処理
+pandas>=2.2.0
+numpy>=1.26.0
+openpyxl>=3.1.2
+
+# 機械学習（回帰や予兆解析で利用）
+scikit-learn>=1.5.0
+
+# 可視化（将来グラフ表示を追加する可能性を考慮）
+matplotlib>=3.8.0