sciope.sampling package

Submodules

sciope.sampling.maximin_sampling module

Maximin space-filling sampling algorithm Ranks monte-carlo samples such that minimum distance between them is maximized

class sciope.sampling.maximin_sampling.MaximinSampling(xmin, xmax, use_logger=False)

Bases: sciope.sampling.sampling_base.SamplingBase

Algorithm: 1. Generate MC candidate samples 2. Compute pairwise distance between existing samples and candidates 3. Select new samples that maximize the minimum distance

Key reference: Johnson, Mark E., Leslie M. Moore, and Donald Ylvisaker. “Minimax and maximin distance designs.” Journal of statistical planning and inference 26.2 (1990): 131-148.

select_point(x)

Get top ranked candidate according to maximin sampling to add to current samples x

xvector or array-like

existing design to which a new point must be added

dask.delayed

select_points(x, n)

Get ‘n’ top ranked candidates according to maximin sampling to add to current samples x

xvector or array-like

existing design to which a new point must be added

ninteger

number of new samples to be selected

dask.delayed

sciope.sampling.sampling_base module

Sequential Sampling Base Class

class sciope.sampling.sampling_base.SamplingBase(name, xmin, xmax, use_logger=False)

Bases: object

Base class for sequential sampling. Must not be used directly! Each sampling algorithm must implement the methods described herein:

• SamplingBase.select_point()

• SamplingBase.select_points(n)

The following variables are available to derived classes: *

abstract select_point(x)

Sub-classable method for selecting one new point to X. Each derived class must implement

xvector or array-like

existing design to which a new point must be added

abstract select_points(x, n)

Sub-classable method for selecting ‘n’ new points to X. Each derived class must implement.

xvector or array-like

existing design to which a new point must be added

ninteger

number of new points to be sampled

Module contents