pyleecan.Classes.OptiBayesAlg module

Method code available at https://github.com/Eomys/pyleecan/tree/master/pyleecan/Methods/Optimization/OptiBayesAlg

class OptiBayesAlg(nb_iter=15, nb_start=300, criterion='PI', kernel=0, problem=- 1, xoutput=- 1, logger_name='Pyleecan.OptiSolver', is_keep_all_output=False, init_dict=None, init_str=None)

Bases: OptiSolver

Bayesian algorithm class

VERSION = 1

save(save_path='', is_folder=False, type_handle_old=2, type_compression=0)

Save the object to the save_path

Parameters:

• self – A pyleecan object

• save_path (str) – path to the folder to save the object

• is_folder (bool) – to split the object in different files: separate simulation machine and materials (json only)

• type_handle_old (int) – How to handle old file in folder mode (0:Nothing, 1:Delete, 2:Move to “Backup” folder)

• type_compression (int) – Available only for json, 0: no compression, 1: gzip

get_logger()

Get the object logger or its parent’s one

Parameters:

obj – A pyleecan object

Returns:

logger – Pyleecan object dedicated logger

Return type:

logging.Logger

compare(other, name='self', ignore_list=None, is_add_value=False)

Compare two objects and return list of differences

as_dict(type_handle_ndarray=0, keep_function=False, **kwargs)

Convert this object in a json serializable dict (can be use in __init__). type_handle_ndarray: int

How to handle ndarray (0: tolist, 1: copy, 2: nothing)

keep_functionbool

True to keep the function object, else return str

Optional keyword input parameter is for internal use only and may prevent json serializability.

copy()

Creates a deepcopy of the object

property nb_iter

Number of iterations

Type:

int

Min:

1

property nb_start

Number of starting points

Type:

int

Min:

1

property criterion

Point selection criteria

Type:

str

property kernel

Type of kernel

Type:

int

Min:

0

Max:

0