pyfemtet 0.8.4__py3-none-any.whl → 0.8.7__py3-none-any.whl

pyfemtet/opt/_femopt_core.py

@@ -1,4 +1,5 @@
 # typing
+import json
 from typing import List, TYPE_CHECKING
 
 # built-in
@@ -524,7 +525,6 @@ class History:
         obj_names (List[str], optional): The names of objectives. Defaults to None.
         cns_names (List[str], optional): The names of constraints. Defaults to None.
         client (dask.distributed.Client): Dask client.
-        additional_metadata (str, optional): metadata of optimization process.
         hv_reference (str or list[float or np.ndarray, optional):
             The method to calculate hypervolume or
             the reference point itself.
@@ -550,7 +550,6 @@ class History:
             obj_names=None,
             cns_names=None,
             client=None,
-            additional_metadata=None,
             hv_reference=None,
     ):
         # hypervolume 計算メソッド
@@ -561,7 +560,8 @@ class History:
         self.prm_names = prm_names
         self.obj_names = obj_names
         self.cns_names = cns_names
-        self.additional_metadata = additional_metadata or ''
+        self.extra_data = dict()
+        self.meta_columns = None
         self.__scheduler_address = client.scheduler.address if client is not None else None
 
         # 最適化実行中かどうか
@@ -575,15 +575,15 @@ class History:
 
             # 続きからなら df を読み込んで df にコピー
             if self.is_restart:
-                self.load()
+                self.load()  # 中で meta_columns を読む
 
             # そうでなければ df を初期化
             else:
-                columns, metadata = self.create_df_columns()
+                columns, meta_columns = self.create_df_columns()
                 df = pd.DataFrame()
                 for c in columns:
                     df[c] = None
-                self.metadata = metadata
+                self.meta_columns = meta_columns
                 self.set_df(df)
 
             # 一時ファイルに書き込みを試み、UnicodeEncodeError が出ないかチェック
@@ -609,16 +609,16 @@ class History:
         # df を読み込む
         df = pd.read_csv(self.path, encoding=self.ENCODING, header=self.HEADER_ROW)
 
-        # metadata を読み込む
+        # meta_columns を読み込む
         with open(self.path, mode='r', encoding=self.ENCODING, newline='\n') as f:
             reader = csv.reader(f, delimiter=',')
-            self.metadata = reader.__next__()
+            self.meta_columns = reader.__next__()
 
         # 最適化問題を読み込む
         columns = df.columns
-        prm_names = [column for i, column in enumerate(columns) if self.metadata[i] == 'prm']
-        obj_names = [column for i, column in enumerate(columns) if self.metadata[i] == 'obj']
-        cns_names = [column for i, column in enumerate(columns) if self.metadata[i] == 'cns']
+        prm_names = [column for i, column in enumerate(columns) if self.meta_columns[i] == 'prm']
+        obj_names = [column for i, column in enumerate(columns) if self.meta_columns[i] == 'obj']
+        cns_names = [column for i, column in enumerate(columns) if self.meta_columns[i] == 'cns']
 
         # is_restart の場合、読み込んだ names と引数の names が一致するか確認しておく
         if self.is_restart:
@@ -662,7 +662,7 @@ class History:
                         logger.debug('Access df of History before it is initialized.')
                         return pd.DataFrame()
             except OSError:
-                logger.error('Scheduler is already dead. Most frequent reasen to show this message is that the pyfemtet monitor UI is not refreshed even if the main optimization process is terminated.')
+                logger.error('Scheduler is already dead. Most frequent reason to show this message is that the pyfemtet monitor UI is not refreshed even if the main optimization process is terminated.')
                 return pd.DataFrame()
 
     def set_df(self, df: pd.DataFrame):
@@ -687,46 +687,46 @@ class History:
         columns = list()
 
         # columns のメタデータを作成
-        metadata = list()
+        meta_columns = list()
 
         # trial
         columns.append('trial')  # index
-        metadata.append(self.additional_metadata)
+        meta_columns.append('')  # extra_data. save 時に中身を記入する。
 
         # parameter
         for prm_name in self.prm_names:
             columns.extend([prm_name, prm_name + '_lower_bound', prm_name + '_upper_bound'])
-            metadata.extend(['prm', 'prm_lb', 'prm_ub'])
+            meta_columns.extend(['prm', 'prm_lb', 'prm_ub'])
 
         # objective relative
         for name in self.obj_names:
             columns.append(name)
-            metadata.append('obj')
+            meta_columns.append('obj')
             columns.append(name + '_direction')
-            metadata.append('obj_direction')
+            meta_columns.append('obj_direction')
         columns.append('non_domi')
-        metadata.append('')
+        meta_columns.append('')
 
         # constraint relative
         for name in self.cns_names:
             columns.append(name)
-            metadata.append('cns')
+            meta_columns.append('cns')
             columns.append(name + '_lower_bound')
-            metadata.append('cns_lb')
+            meta_columns.append('cns_lb')
             columns.append(name + '_upper_bound')
-            metadata.append('cns_ub')
+            meta_columns.append('cns_ub')
         columns.append('feasible')
-        metadata.append('')
+        meta_columns.append('')
 
         # the others
         columns.append('hypervolume')
-        metadata.append('')
+        meta_columns.append('')
         columns.append('message')
-        metadata.append('')
+        meta_columns.append('')
         columns.append('time')
-        metadata.append('')
+        meta_columns.append('')
 
-        return columns, metadata
+        return columns, meta_columns
 
     def record(
             self,
@@ -973,13 +973,16 @@ class History:
 
         df = self.get_df()
 
+        # extra_data の更新
+        self.meta_columns[0] = json.dumps(self.extra_data)
+
         if _f is None:
             # save df with columns with prefix
             with open(self.path, 'w', encoding=self.ENCODING) as f:
                 writer = csv.writer(f, delimiter=',', lineterminator="\n")
-                writer.writerow(self.metadata)
+                writer.writerow(self.meta_columns)
                 for i in range(self.HEADER_ROW-1):
-                    writer.writerow([''] * len(self.metadata))
+                    writer.writerow([''] * len(self.meta_columns))
                 df.to_csv(f, index=None, encoding=self.ENCODING, lineterminator='\n')
         else:  # test
             df.to_csv(_f, index=None, encoding=self.ENCODING, lineterminator='\n')
@@ -1088,3 +1091,35 @@ class OptimizationStatus:
     def get_text(self) -> str:
         """Get optimization status message."""
         return self._actor.status
+
+
+class _MonitorHostRecordActor:
+    host = None
+    port = None
+
+    def set(self, host, port):
+        self.host = host
+        self.port = port
+
+
+class MonitorHostRecord:
+
+    def __init__(self, client, worker_name):
+        self._future = client.submit(
+            _MonitorHostRecordActor,
+            actor=True,
+            workers=(worker_name,),
+            allow_other_workers=False,
+        )
+        self._actor = self._future.result()
+
+    def set(self, host, port):
+        self._actor.set(host, port).result()
+
+    def get(self):
+        host = self._actor.host
+        port = self._actor.port
+        if host is None and port is None:
+            return dict()
+        else:
+            return dict(host=host, port=port)

pyfemtet/opt/_test_utils/record_history.py

@@ -10,7 +10,7 @@ from pyfemtet.opt import FEMOpt
 from pyfemtet._message import encoding as ENCODING
 
 
-def remove_femprj_metadata_from_csv(csv_path, encoding=ENCODING):
+def remove_extra_data_from_csv(csv_path, encoding=ENCODING):
 
     with open(csv_path, mode="r", encoding=encoding, newline="\n") as f:
         reader = csv.reader(f, delimiter=",")
@@ -59,32 +59,39 @@ def _get_obj_from_csv(csv_path, encoding=ENCODING):
         reader = csv.reader(f, delimiter=",")
         meta = reader.__next__()
     obj_indices = np.where(np.array(meta) == "obj")[0]
-    out = df.iloc[:, obj_indices]
+    out: pd.DataFrame = df.iloc[:, obj_indices]
+    out = out.dropna(axis=0)
     return out, columns
 
 
-def is_equal_result(ref_path, dif_path, log_path):
+def is_equal_result(ref_path, dif_path, log_path=None, threashold=0.05):
     """Check the equality of two result csv files."""
     ref_df, ref_columns = _get_obj_from_csv(ref_path)
     dif_df, dif_columns = _get_obj_from_csv(dif_path)
 
-    with open(log_path, "a", newline="\n", encoding=ENCODING) as f:
-        f.write("\n\n===== 結果の分析 =====\n\n")
-        f.write(f"        \tref\tdif\n")
-        f.write(f"---------------------\n")
-        f.write(f"len(col)\t{len(ref_columns)}\t{len(dif_columns)}\n")
-        f.write(f"len(df) \t{len(ref_df)}\t{len(dif_df)}\n")
-        try:
-            difference = (
+    if log_path is not None:
+        with open(log_path, "a", newline="\n", encoding=ENCODING) as f:
+            f.write("\n\n===== 結果の分析 =====\n\n")
+            f.write(f"        \tref\tdif\n")
+            f.write(f"---------------------\n")
+            f.write(f"len(col)\t{len(ref_columns)}\t{len(dif_columns)}\n")
+            f.write(f"len(df) \t{len(ref_df)}\t{len(dif_df)}\n")
+            try:
+                difference = (
+                    np.abs(ref_df.values - dif_df.values) / np.abs(dif_df.values)
+                ).mean()
+                f.write(f"diff    \t{int(difference*100)}%\n")
+            except Exception:
+                f.write(f"diff    \tcannot calc\n")
+
+    else:
+        difference = (
                 np.abs(ref_df.values - dif_df.values) / np.abs(dif_df.values)
-            ).mean()
-            f.write(f"diff    \t{int(difference*100)}%\n")
-        except Exception:
-            f.write(f"diff    \tcannot calc\n")
+        ).mean()
 
     assert len(ref_columns) == len(dif_columns), "結果 csv の column 数が異なります。"
     assert len(ref_df) == len(dif_df), "結果 csv の row 数が異なります。"
-    assert difference <= 0.05, "前回の結果との平均差異が 5% を超えています。"
+    assert difference <= threashold*100, f"前回の結果との平均差異が {int(difference)}% で {int(threashold*100)}% を超えています。"
 
 
 def _get_simplified_df_values(csv_path, exclude_columns=None):

pyfemtet/opt/interface/__init__.py

@@ -21,7 +21,9 @@ else:
     FemtetWithNXInterface = type('FemtetWithNXInterface', (FemtetInterface,), {})
     ExcelInterface = type('FemtetInterface', (NotAvailableForWindows,), {})
 
-from pyfemtet.opt.interface._surrogate import PoFBoTorchInterface
+from pyfemtet.opt.interface._surrogate._base import SurrogateModelInterfaceBase
+from pyfemtet.opt.interface._surrogate._singletaskgp import PoFBoTorchInterface
+
 
 __all__ =[
     'FEMInterface',
@@ -30,5 +32,6 @@ __all__ =[
     'FemtetWithSolidworksInterface',
     'FemtetWithNXInterface',
     'ExcelInterface',
+    'SurrogateModelInterfaceBase',
     'PoFBoTorchInterface',
 ]

pyfemtet/opt/interface/_femtet.py

@@ -1,3 +1,4 @@
+import warnings
 from typing import Optional, List, Final
 
 import os
@@ -90,6 +91,19 @@ class FemtetInterface(FEMInterface):
             it will be None and no parametric outputs are used
             as objectives.
 
+
+            Note:
+                Indexes start at 0, but the parametric analysis
+                output settings in the Femtet dialog box indicate
+                setting numbers starting at 1.
+
+
+            Warning:
+                **Setting this argument deletes the parametric
+                analysis swept table set in the femprj file.**
+                If you do not want to delete the swept table,
+                make a copy of the original file.
+
         **kwargs: Additional arguments from inherited classes.
 
     Warning:
@@ -117,6 +131,9 @@ class FemtetInterface(FEMInterface):
             parametric_output_indexes_use_as_objective: dict[int, str or float] = None,
             **kwargs  # 継承されたクラスからの引数
     ):
+        # warning
+        if parametric_output_indexes_use_as_objective is not None:
+            warnings.warn('解析モデルに設定された既存のスイープテーブルは削除されます。')
 
         # win32com の初期化
         CoInitialize()

pyfemtet/opt/interface/_femtet_parametric.py

@@ -1,12 +1,147 @@
+import os
+import ctypes
+# from ctypes import wintypes
 import logging
+import warnings
+from time import sleep, time
+from packaging.version import Version
+
+import numpy as np
+import pandas as pd
+from femtetutils import util
+from femtetutils import logger as util_logger
 
-from femtetutils import util, logger
 from pyfemtet.dispatch_extensions import _get_pid
+from pyfemtet.core import SolveError
+from pyfemtet._message.messages import encoding, Message
+from pyfemtet.logger import get_module_logger
 
-import ctypes
+logger = get_module_logger('opt.fem.ParametricIF', __name__)
+
+util_logger.setLevel(logging.ERROR)
+
+# singleton pattern
+_P_CSV: 'ParametricResultCSVProcessor' = None
+
+
+def get_csv_processor(Femtet):
+    global _P_CSV
+    if _P_CSV is None:
+        _P_CSV = ParametricResultCSVProcessor(Femtet)
+    return _P_CSV
+
+
+class ParametricResultCSVProcessor:
+
+    def __init__(self, Femtet):
+        self.Femtet = Femtet
+
+    def refresh_csv(self):
+        # 存在するならば削除する
+        csv_paths = self.get_csv_paths()
+        for path in csv_paths:
+            if os.path.exists(path):
+                os.remove(path)
+
+    def get_csv_paths(self):
+        # 結果フォルダを取得
+        path: str = self.Femtet.Project
+        res_dir_path = path.removesuffix('.femprj') + '.Results'
+
+        # csv を取得
+        model_name = self.Femtet.AnalysisModelName
+        csv_path = os.path.join(res_dir_path, f'{model_name}.csv')
+        table_csv_path = os.path.join(res_dir_path, f'{model_name}_table.csv')
+
+        return csv_path, table_csv_path
+
+    def check_csv_after_succeeded_PrmCalcExecute(self):
+        """Parametric Solve の後に呼ぶこと。"""
+
+        csv_path, table_csv_path = self.get_csv_paths()
+
+        # csv が生成されているか
+        start = time()
+        while not os.path.exists(csv_path):
+            # solve が succeeded であるにもかかわらず
+            # 数秒経過しても csv が存在しないのはおかしい
+            if time() - start > 3.:
+                return False
+            sleep(0.25)
+
+        # csv は存在するが、Femtet が古いと
+        # table は生成されない
+        if not os.path.exists(table_csv_path):
+            warnings.warn('テーブル形式 csv が生成されていないため、'
+                          '結果出力エラーチェックが行われません。'
+                          'そのため、結果出力にエラーがある場合は'
+                          '目的関数が 0 と記録される場合があります。'
+                          '結果出力エラーチェック機能を利用するためには、'
+                          'Femtet を最新バージョンにアップデートして'
+                          'ください。')
+
+        return True
+
+    def is_succeeded(self, parametric_output_index):
+
+        # まず csv 保存が成功しているかどうか。通常あるはず。
+        if not self.check_csv_after_succeeded_PrmCalcExecute():
+            return False, 'Reason: output csv not found.'
+
+        # 成功しているならば table があるかどうか
+        csv_path, table_csv_path = self.get_csv_paths()
+
+        # なければエラーチェックできないので
+        # エラーなしとみなす (warning の記載通り)
+        if not os.path.exists(table_csv_path):
+            return True, None
+
+        # table があれば読み込む
+        df = pd.read_csv(table_csv_path, encoding=encoding)
 
+        # 結果出力用に行番号を付記する
+        df['row_num'] = range(2, len(df) + 2)  # row=0 は header, excel は 1 始まり
 
-logger.setLevel(logging.ERROR)
+        # 「結果出力設定番号」カラムが存在するか
+        if '結果出力設定番号' in df.columns:  # TODO: 英語版対応
+
+            # 与えられた output_number に関連する行だけ抜き出し
+            # エラーがあるかどうかチェックする
+            pdf = df['結果出力設定番号'] == parametric_output_index + 1
+
+        # 結果出力設定番号 カラムが存在しない
+        else:
+            # output_number に関係なくエラーがあればエラーにする
+            pdf = df
+
+        # エラーの有無を確認
+        if 'エラー' in pdf.columns:  # TODO: 英語版対応
+            is_no_error = np.all(pdf['エラー'].isna().values)
+
+            if not is_no_error:
+                error_message_row_numbers = pdf['row_num'][~pdf['エラー'].isna()].values.astype(str)
+                error_messages = pdf['エラー'][~pdf['エラー'].isna()].values.astype(str)
+
+                def add_st_or_nd_or_th(n_: int):
+                    if n_ == 1:
+                        return f'{n_}st'
+                    elif n_ == 2:
+                        return f'{n_}nd'
+                    elif n_ == 3:
+                        return f'{n_}rd'
+                    else:
+                        return f'{n_}th'
+
+                error_msg = f'Error message(s) from {os.path.basename(table_csv_path)}: ' + ', '.join(
+                    [f'({add_st_or_nd_or_th(row)} row) {message}' for row, message in zip(error_message_row_numbers, error_messages)])
+            else:
+                error_msg = None
+
+        else:
+            raise RuntimeError('Internal Error! Parametric Analysis '
+                               'output csv has no error column.')
+
+        return is_no_error, error_msg
 
 
 def _get_dll():
@@ -46,7 +181,6 @@ def add_parametric_results_as_objectives(femopt, indexes, directions) -> bool:
 
     # get objective names
     dll.GetPrmnResult.restype = ctypes.c_int
-    n = dll.GetPrmnResult()
     for i, direction in zip(indexes, directions):
         # objective name
         dll.GetPrmResultName.restype = ctypes.c_char_p
@@ -58,6 +192,15 @@ def add_parametric_results_as_objectives(femopt, indexes, directions) -> bool:
 
 
 def _parametric_objective(Femtet, parametric_result_index):
+    # csv から結果取得エラーの有無を確認する
+    # (解析自体は成功していないと objective は呼ばれないはず)
+    csv_processor = get_csv_processor(Femtet)
+    succeeded, error_msg = csv_processor.is_succeeded(parametric_result_index)
+    if not succeeded:
+        logger.error(Message.ERR_PARAMETRIC_CSV_CONTAINS_ERROR)
+        logger.error(error_msg)
+        raise SolveError
+
     # load dll and set target femtet
     dll = _get_dll_with_set_femtet(Femtet)
     dll.GetPrmResult.restype = ctypes.c_double  # 複素数の場合は実部しか取らない
@@ -65,9 +208,65 @@ def _parametric_objective(Femtet, parametric_result_index):
 
 
 def solve_via_parametric_dll(Femtet) -> bool:
+    csv_processor = get_csv_processor(Femtet)
+
+    # remove previous csv if exists
+    #   消さなくても解析はできるが
+    #   エラーハンドリングのため
+    csv_processor.refresh_csv()
+
     # load dll and set target femtet
     dll = _get_dll_with_set_femtet(Femtet)
+
+    # reset existing sweep table
+    dll.ClearPrmSweepTable.restype = ctypes.c_bool
+    succeed = dll.ClearPrmSweepTable()
+    if not succeed:
+        logger.error('Failed to remove existing sweep table!')  # 通常ありえないので error
+        return False
+
     # solve
     dll.PrmCalcExecute.restype = ctypes.c_bool
     succeed = dll.PrmCalcExecute()
+    if not succeed:
+        logger.warning('Failed to solve!')  # 通常起こりえるので warn
+        return False
+
+    # Check post-processing error
+    #   現時点では table csv に index の情報がないので、
+    #   エラーがどの番号のものかわからない。
+    #   ただし、エラーがそのまま出力されるよりマシなので
+    #   安全目に引っ掛けることにする
+    succeed = csv_processor.check_csv_after_succeeded_PrmCalcExecute()
+    if not succeed:
+        logger.error('Failed to save parametric result csv!')
+        return False  # 通常ありえないので error
+
     return succeed  # 成功した場合はTRUE、失敗した場合はFALSEを返す
+
+
+if __name__ == '__main__':
+    from win32com.client import Dispatch
+
+    g_Femtet = Dispatch('FemtetMacro.Femtet')
+    g_dll = _get_dll_with_set_femtet(g_Femtet)
+
+    # solve
+    g_succeeded = solve_via_parametric_dll(g_Femtet)
+    if not g_succeeded:
+        g_dll.GetLastErrorMsg.restype = ctypes.c_char_p  # or wintypes.LPCSTR
+        g_error_msg: bytes = g_dll.GetLastErrorMsg()
+        g_error_msg: str = g_error_msg.decode(encoding='932')
+
+    # 結果取得：内部的にはエラーになっているはず
+    g_parametric_result_index = 1
+    g_dll = _get_dll_with_set_femtet(g_Femtet)
+    g_dll.GetPrmResult.restype = ctypes.c_double  # 複素数やベクトルの場合は実部や第一成分しか取らない PIF の仕様
+    g_output = g_dll.GetPrmResult(g_parametric_result_index)
+
+    # ... だが、下記のコードでそれは出てこない。
+    # 値が実際に 0 である場合と切り分けられないので、
+    # csv を見てエラーがあるかどうか判断せざるを得ない。
+    g_dll.GetLastErrorMsg.restype = ctypes.c_char_p  # or wintypes.LPCSTR
+    g_error_msg: bytes = g_dll.GetLastErrorMsg()
+    g_error_msg: str = g_error_msg.decode(encoding='932')

pyfemtet/opt/interface/_surrogate/_base.py

@@ -17,6 +17,7 @@ class SurrogateModelInterfaceBase(FEMInterface, ABC):
             self,
             history_path: str = None,
             history: History = None,
+            override_objective: bool = True,
     ):
 
         self.history: History
@@ -25,6 +26,7 @@ class SurrogateModelInterfaceBase(FEMInterface, ABC):
         self.obj: dict[str, float] = dict()
         self.df_prm: pd.DataFrame
         self.df_obj: pd.DataFrame
+        self.override_objective: bool = override_objective
 
         # history_path が与えられた場合、history をコンストラクトする
         if history_path is not None:
@@ -57,27 +59,25 @@ class SurrogateModelInterfaceBase(FEMInterface, ABC):
         FEMInterface.__init__(
             self,
             history=history,  # コンストラクト済み history を渡せば並列計算時も何もしなくてよい
+            override_objective=self.override_objective
         )
 
     def filter_feasible(self, x: np.ndarray, y: np.ndarray, return_feasibility=False):
         feasible_idx = np.where(~np.isnan(y.sum(axis=1)))
         if return_feasibility:
             # calculated or not
-            y = np.zeros_like(y)
-            y[feasible_idx] = 1.
-            # satisfy weak feasibility or not
-            infeasible_idx = np.where(~self.history.get_df()['feasible'].values)
-            y[infeasible_idx] = .0
-            return x, y.reshape((-1, 1))
+            feas = np.zeros((len(y), 1), dtype=float)
+            feas[feasible_idx] = 1.
+            return x, feas
         else:
             return x[feasible_idx], y[feasible_idx]
 
     def _setup_after_parallel(self, *args, **kwargs):
-
-        opt: AbstractOptimizer = kwargs['opt']
-        obj: Objective
-        for obj_name, obj in opt.objectives.items():
-            obj.fun = lambda: self.obj[obj_name]
+        if self.override_objective:
+            opt: AbstractOptimizer = kwargs['opt']
+            obj: Objective
+            for obj_name, obj in opt.objectives.items():
+                obj.fun = lambda obj_name_=obj_name: self.obj[obj_name_]
 
     def update_parameter(self, parameters: pd.DataFrame, with_warning=False) -> Optional[List[str]]:
         for i, row in parameters.iterrows():

pyfemtet/opt/interface/_surrogate/_singletaskgp.py

@@ -27,8 +27,8 @@ class PoFBoTorchInterface(SurrogateModelInterfaceBase):
     def train_f(self):
         # df そのまま用いて training する
         x, y = self.filter_feasible(self.df_prm.values, self.df_obj.values, return_feasibility=True)
-        if y.min() == 1:
-            self.model_f.predict = lambda *args, **kwargs: (1., 0.001)
+        if y.min() == 1:  # feasible values only
+            self.model_f.predict = lambda *args, **kwargs: (1., 0.001)  # mean, std
         self.model_f.fit(x, y)
 
     def _setup_after_parallel(self, *args, **kwargs):
@@ -64,7 +64,8 @@ class PoFBoTorchInterface(SurrogateModelInterfaceBase):
             raise SolveError(Msg.INFO_POF_IS_LESS_THAN_THRESHOLD)
 
         # 実際の計算(mean は history.obj_names 順)
-        mean, _ = self.model.predict(np.array([x]))
+        _mean, _std = self.model.predict(np.array([x]))
+        mean = _mean[0]
 
         # 目的関数の更新
         self.obj = {obj_name: value for obj_name, value in zip(self.history.obj_names, mean)}

pyfemtet/opt/prediction/single_task_gp.py

@@ -42,16 +42,16 @@ class SingleTaskGPModel(PredictionModelBase):
     def set_bounds_from_history(self, history, df=None):
         from pyfemtet.opt._femopt_core import History
         history: History
-        metadata: str
+        meta_column: str
 
         if df is None:
             df = history.get_df()
 
         columns = df.columns
-        metadata_columns = history.metadata
+
         target_columns = [
-            col for col, metadata in zip(columns, metadata_columns)
-            if metadata == 'prm_lb' or metadata == 'prm_ub'
+            col for col, meta_column in zip(columns, history.meta_columns)
+            if meta_column == 'prm_lb' or meta_column == 'prm_ub'
         ]
 
         bounds_buff = df.iloc[0][target_columns].values  # 2*len(prm_names) array
@@ -85,6 +85,8 @@ class SingleTaskGPModel(PredictionModelBase):
         fit_gpytorch_mll(mll)
 
     def predict(self, x: np.ndarray) -> tuple[np.ndarray, np.ndarray]:
+        assert len(x.shape) >= 2
+
         X = tensor(x)
 
         post = self.gp.posterior(X)