pyfemtet.opt.history package

Module contents

class pyfemtet.opt.history.ColumnOrderMode(value)

Bases: StrEnum

The order rule of the history csv columns.

important_first = 'important_first'

per_category = 'per_category'

Sort per each object.

class pyfemtet.opt.history.History

Bases: object

最適化の試行の履歴を管理します。

additional_data: dict

property all_output_names: list[str]

cns_names: list[str]

get_df(equality_filters=None)

Returns the optimization history.

Parameters:

equality_filters (dict, optional) – The {column: value} というフォーマットの matching filter.

Return type:

DataFrame

Returns: The optimization history.

static get_trial_name(trial=None, fidelity=None, sub_sampling=None, row=None)

Parameters:

row (Series | None) –

is_restart: bool

obj_names: list[str]

other_output_names: list[str]

path: str

The existing or destination CSV path.

If not specified, the CSV file is saved in the format “pyfemtet.opt_%Y%m%d_%H%M%S.csv” when the optimization process starts.

prm_names: list[str]

save()

Export the optimization history.

The destination path is History.path.

sub_fidelity_names: list[str]

class pyfemtet.opt.history.TrialState(value)

Bases: StrEnum

An enumeration.

hard_constraint_violation = 'Hard constraint violation'

mesh_error = 'Mesh error'

model_error = 'Model error'

post_error = 'Post-processing error'

skipped = 'Skip'

soft_constraint_violation = 'Soft constraint violation'

solve_error = 'Solve error'

succeeded = 'Success'

undefined = 'undefined'

unknown_error = 'Unknown error'

pyfemtet.opt.history.get_trial_name(trial=None, fidelity=None, sub_sampling=None, row=None)

Parameters:

row (Series | None) –