HyperGP.executor Module

Contents

HyperGP.executor Module#

executor(progs, input, pset, cashset=None)[source]#

Execute the GP programs

Parameters:

  • progs (list) – The GP programs need to be executed.

  • input (HyperGP.Tensor or numpy.ndarray or list) – shape = (feature size, …)

  • pset (HyperGP.PrimitiveSet) – used for the progs

  • cashset (CashManager)

Returns:

The outputs generated by the progs with the given inputs. records: The outputs of the record nodes.

Return type:

outputs

Examples

>>> from HyperGP import executor, PrimitiveSet, Population
>>> import numpy as np

>>> # Generae the primitive set
>>> pset = PrimitiveSet(
        input_arity=1,
        primitive_set=[
        ('add', HyperGP.add, 2),
        ('sub', HyperGP.sub, 2),
        ('mul', HyperGP.mul, 2),
        ('div', HyperGP.div, 2),
    ])

>>> # Generate the input data
>>> input_array = Tensor(np.random.uniform(0, 10, size=(1, 10000)))

>>> # Initialize the population
>>> pop = Population()
>>> pstates = ProgBuildStates(pset=pset, depth_rg=[2, 6], len_limit=100)
>>> pop.initPop(pop_size=pop_size, prog_paras=pstates)

>>> # Execute the individuals
>>> output, _ = executor(pop.states['progs'].indivs, input_array, pset)
Contents