cma.fitness_functions

module documentation

versatile container for test objective functions.

For the time being this is probably best used like:

from cma.fitness_functions import ff

Class	`FitnessFunctions`	collection of objective functions.
Function	`elli`	unbound test function, needed to test multiprocessor, as long as the other test functions are defined within a class and only accessable via the class instance
Function	`sphere`	Undocumented
Variable	`binary_foffset`	default f-offset for binary functions at the optimum.
Variable	`binary_optimum_interval`	default interval where the optimum is assumed on binary functions.
Variable	`ff`	Undocumented
Class	`_coco_F_0`	return a "normalized" BBOB function, funID=1..24 when suite='bbob'.
Function	`_iqr`	Undocumented

def elli(x, cond=1000000.0): ¶

unbound test function, needed to test multiprocessor, as long as the other test functions are defined within a class and only accessable via the class instance

def sphere(x): ¶

Undocumented

binary_foffset = ¶

default f-offset for binary functions at the optimum.

binary_optimum_interval: tuple[float, ...] = ¶

default interval where the optimum is assumed on binary functions.

The interval is chosen such that the value from round(.) or floor(. + 1/2) or int(. + 1/2) is in the interval middle. This prevents some unexpected outcomes with algorithms that search on the continuous values. The most logical domain boundary values are now between [-0.499, 1.499] and [0, 1].

ff = ¶

Undocumented

def _iqr(x): ¶

Undocumented