twinlab.sampling.LatinHypercube

class twinlab.sampling.LatinHypercube(scramble=True, optimization='random-cd')

A sampling strategy that uses Latin Hypercube Sampling.

Parameters:

• scramble (bool, optional) – Whether to scramble the samples within sub-cubes. The default value is True.

• optimization (Union[str, None], optional) –

  The optimization method to use for generating the samples. Options are:

  • None: No optimization is performed once the intial samples are generated.

  • "random-cd": Randomly permute the columns of the matrix in order to lower the centred discrepancy of the generated samples.

  • "lloyd": Perturb the samples using a modified Lloyd-Max algorithm. The process converges to equally spaced samples.

  The default is "random-cd".

__init__(scramble=True, optimization='random-cd')

Methods

__init__([scramble, optimization])
to_json()