optimisation.design_space.design_space module

Define an object to hold variables and constraints.

class DesignSpace(variables: list[Variable], constraints: list[Constraint])

Bases: object

Hold variables and constraints of an optimisation problem.

__init__(variables: list[Variable], constraints: list[Constraint]) → None
__post_init__() → None

Print out some info.

_check_dimensions(parameters: list[Variable] | list[Constraint]) → int

Ensure that all elements have the same number of var or const.

_parameters_to_single_file_line(element_name: str, parameters: list[DesignSpaceParameter]) → dict[str, float | None | tuple[float, float]]

Prepare a dict containing all info of a single element.

Parameters:

  • element_name (str) – Name of the element, which will be inserted in the output dict.

  • parameters (list[DesignSpaceParameter]) – Parameters concerning the element, which limits (x_0 if appliable) will be inserted in the dict.

Returns:

Contains all Variable or Constraint information of the element.

Return type:

dict[str, float | None | tuple[float, float]]

_str_constraints() → list[str]

Generate information on the constraints that were created.

_str_variables() → list[str]

Generate information on the variables that were created.

_to_file(parameters: list[DesignSpaceParameter], filepath: Path, delimiter: str = ',', **to_csv_kw: Any) → None

Save all the design space parameters in a compact file.

Parameters:

  • parameters (list[DesignSpaceParameter]) – All the defined parameters.

  • filepath (Path) – Where file will be stored.

  • delimiter (str) – Delimiter between two columns. The default is ‘,’.

  • to_csv_kw (dict[str, Any]) – Keyword arguments given to the pandas to_csv method.

compute_constraints(simulation_output: SimulationOutput) → ndarray

Compute constraint violation for simulation_output.

constraints: list[Constraint]
classmethod from_files(elements_names: Sequence[str], filepath_variables: Path, variables_names: Sequence[str], filepath_constraints: Path | None = None, constraints_names: Sequence[str] | None = None, delimiter: str = ',') → Self

Generate design space from files.

Parameters:

  • elements_names (Sequence[str]) – Name of the elements with variables and constraints.

  • filepath_variables (Path) – Path to the variables.csv file.

  • variables_names (Sequence[str]) – Name of the variables to create.

  • filepath_constraints (Path | None, optional) – Path to the constraints.csv file. The default is None, in which case no constraint is created.

  • constraints_names (Sequence[str] | None, optional) – Name of the constraints to create. The default is None, in which case no constraint is created.

  • delimiter (str, optional) – Delimiter in the files. The default is a commma.

Return type:

cls

to_files(basepath: Path, variables_filename: Path = PosixPath('variables'), constraints_filename: Path = PosixPath('constraints'), overwrite: bool = False, **to_csv_kw: Any) → None

Save variables and constraints in files.

Parameters:

  • basepath (Path) – Folder where the files will be stored.

  • variables_filename (Path) – Name of the output files without extension.

  • constraints_filename (Path) – Name of the output files without extension.

  • overwrite (bool, optional) – To overwrite an existing file with the same name or not. The default is False.

  • to_csv_kw (dict[str, Any]) – Keyword arguments given to the pandas to_csv method.

to_pandas_dataframe() → DataFrame

Convert list of variables to a pandas dataframe.

variables: list[Variable]
_from_file(parameter_class: type[Variable], filepath: Path, elements_names: Sequence[str], parameters_names: Sequence[str], delimiter: str = ',') → list[Variable]
_from_file(parameter_class: type[Constraint], filepath: Path, elements_names: Sequence[str], parameters_names: Sequence[str], delimiter: str = ',') → list[Constraint]

Generate list of variables or constraints from a given .csv.

Todo

Add support for when all element do not have the same variables/constraints.

Parameters:

  • parameter_class ({type[Variable], type[Constraint]}) – Object which from_pd_series method will be called.

  • filepath (Path) – Path to the .csv.

  • elements_names (Sequence[str]) – Name of the elements.

  • parameters_names (Sequence[str]) – Name of the parameters.

  • delimiter (str) – Delimiter in the .csv.

Returns:

List of variables or constraints.

Return type:

list[DesignSpaceParameter]

_gather_dicts_by_key(parameters: list[DesignSpaceParameter], key: str) → dict[str, list[DesignSpaceParameter]]

Gather parameters with the same key attribute value in lists.

Parameters:

  • parameters (list[DesignSpaceParameter]) – Objects to study.

  • key (str) – Name of the attribute against which parameters should be gathered.

Returns:

Keys are all existing values of attribute key from parameters. Values are lists of DesignSpaceParameter with key attribute equaling the dict key.

Return type:

dict[Any, list[DesignSpaceParameter]]

_merge(dicts: list[dict]) → dict

Merge a list of dicts in a single dict.

_parameters_to_dict(parameters: list[DesignSpaceParameter], to_get: Sequence[str]) → list[dict]

Convert several design space parameters to dict.

We use the prepend_parameter_name argument to prepend the name of each parameter.name to the name of the values to_get. This way, we avoid dictionaries sharing the same keys in the output list.

Parameters:

  • parameters (list[DesignSpaceParameter]) – Where to_get will be looked for.

  • to_get (Sequence[str]) – Values to get.

Returns:

Contains to_get values in dictionaries for every parameter.

Return type:

list[dict]