`LSHADETensor`

Module: opytimizer.optimizers.single_objective.evolutionary.lshade

Agnostic Tensorized L-SHADE Implementation (NumPy/CuPy).

Constructor

LSHADETensor(params: 'Dict' = None, MAX_NFE: 'int' = 100, H: 'int' = 100, p: 'float' = 0.11, f_arc: 'float' = 2.6)

Parameter	Type	Default	Description
`params`	`Dict`	`None`	—
`MAX_NFE`	`int`	`100`	—
`H`	`int`	`100`	—
`p`	`float`	`0.11`	—
`f_arc`	`float`	`2.6`	—

compile(self, space: '_SingleObjectiveTensorSpace', **kwargs)

Compiles additional information that is used by this optimizer.

This method is called before the optimization procedure and makes sure that the additional variable is available as a property.

Parameter	Type	Default	Description
`space`	`_SingleObjectiveTensorSpace`		—
`kwargs`			—

evaluate(self, space: '_SingleObjectiveTensorSpace', function: 'Function')

Evaluates the search space according to the objective function.

If you need a specific evaluate method, please re-implement it on child's class.

Also, note that function only accept arguments that are found on Opytimizer class.

Parameter	Type	Default	Description
`space`	`_SingleObjectiveTensorSpace`		A Space object that will be evaluated.
`function`	`Function`		A Function object serving as an objective function.

update(self, space: '_SingleObjectiveTensorSpace', function: 'Function')

Updates the agents' position array.

As each child has a different procedure of update, you will need to implement it directly on its class.

Parameter	Type	Default	Description
`space`	`_SingleObjectiveTensorSpace`		—
`function`	`Function`		—