Results and metrics

class OMADS._post._postprocess.PostMADS(x_incumbent=[], xmin=None, coords=[], poll_dirs=[], iter=[], bb_eval=[], psize=[], step_name=[], nd_points=[], counter=0)

Bases: object

Results postprocessor

Parameters:

• x_incumbent (List[CandidatePoint])

• xmin (CandidatePoint)

• coords (List[List[CandidatePoint]])

• poll_dirs (List[CandidatePoint])

• iter (List[int])

• bb_eval (List[int])

• psize (List[float])

• step_name (List[str] | None)

• nd_points (List[CandidatePoint])

• counter (int)

output_results(out, all_res=True)

Create a results file from the saved cache

Parameters:

• out (Output)

• all_res (bool)

output_nd_results(out)

Create a results file from the saved cache

Parameters:

out (Output)

output_coordinates(out)

Save spinners in a json file

Parameters:

out (Output)

class OMADS._post._postprocess.Output(file_path=None, vnames=None, fnames=None, file_writer=None, field_names=[], pname='MADS0', runfolder='undefined', replace=True, step_name='Poll', suffix='all', h_max=inf)

Bases: object

Results output file decorator

Parameters:

• file_path (str)

• vnames (List[str])

• fnames (List[str])

• file_writer (Any)

• field_names (List[str])

• pname (str)

• runfolder (str)

• replace (bool)

• step_name (str)

• suffix (str)

• h_max (float)

class OMADS._metrics._metrics.Metrics(nd_solutions=None, nobj=2, ref_point=None)

Bases: object

Parameters:

• nd_solutions (List[CandidatePoint] | None)

• nobj (int)

• ref_point (Point | None)

normalize_data(pareto_front, reference_point)

Normalize Pareto points and the reference point.

Parameters:

• pareto_front – List of Pareto points where each point is a tuple (x, y).

• reference_point – The reference point (rx, ry).

Returns:

Normalized Pareto points and reference point.

hypervolume()

Compute the hypervolume indicator of a Pareto front.

Parameters: - self._ref_point: A list or array representing the reference point. - front: A list of tuples, where each tuple represents a solution’s objective values.

Returns: - The hypervolume indicator value.

calculate_hypervolume(pf, rp)

Calculate the hypervolume of a bi-objective Pareto front.

Parameters:

• pareto_front – A list of Pareto points where each point is a tuple (x, y).

• reference_point – The reference point (rx, ry) to compute the hypervolume against.

Returns:

Hypervolume of the Pareto front.

generational_distance(true_pareto_front, approximate_pareto_front)

Compute the generational distance (GD) metric between two Pareto fronts.

Parameters: - true_pareto_front: A list of tuples representing the true Pareto front. - approximate_pareto_front: A list of tuples representing the approximate Pareto front.

Returns: - The generational distance metric value.

inverted_generational_distance(true_pareto_front, approximate_pareto_front)

Compute the inverted generational distance (IGD) metric between two Pareto fronts.

Parameters: - true_pareto_front: A list of tuples representing the true Pareto front. - approximate_pareto_front: A list of tuples representing the approximate Pareto front.

Returns: - The inverted generational distance metric value.

Run state

class OMADS._analytics._metadata.MadsState

Bases: object

A data class for the MADS session state and meta data

fk_frame_center: int = 0

uk_frame_center: int = 0

ncache_hits: int = 0

ordered_frame_centers: Tuple[int, int] = (-1, -1)

h_max: float | None = inf

is_phase_one: bool = False

last_success: SUCCESS_TYPES = 1

stop_reason: STOP_TYPE = 1

class OMADS._analytics._metadata.MadsStatistics

Bases: object

A class for the MADS statistical metrics