PermutationMethod#
- class scipy.stats.PermutationMethod(n_resamples=9999, batch=None, random_state=None, *, rng=None)[source]#
Configuration information for a permutation hypothesis test.
Instances of this class can be passed into the method parameter of some hypothesis test functions to perform a permutation version of the hypothesis tests.
- Attributes:
- n_resamplesint, optional
The number of resamples to perform. Default is 9999.
- batchint, optional
The number of resamples to process in each vectorized call to the statistic. Batch sizes >>1 tend to be faster when the statistic is vectorized, but memory usage scales linearly with the batch size. Default is
None
, which processes all resamples in a single batch.- rng
numpy.random.Generator
, optional Pseudorandom number generator used to perform resampling.
If
rng
is passed by keyword to the initializer or therng
attribute is used directly, types other thannumpy.random.Generator
are passed tonumpy.random.default_rng
to instantiate aGenerator
before use. Ifrng
is already aGenerator
instance, then the provided instance is used. Specifyrng
for repeatable behavior.If this argument is passed by position, if
random_state
is passed by keyword into the initializer, or if therandom_state
attribute is used directly, legacy behavior forrandom_state
applies:If
random_state
is None (ornumpy.random
), thenumpy.random.RandomState
singleton is used.If
random_state
is an int, a newRandomState
instance is used, seeded withrandom_state
.If
random_state
is already aGenerator
orRandomState
instance then that instance is used.
Changed in version 1.15.0: As part of the SPEC-007 transition from use of
numpy.random.RandomState
tonumpy.random.Generator
, this attribute name was changed fromrandom_state
torng
. For an interim period, both names will continue to work, although only one may be specified at a time. After the interim period, uses ofrandom_state
will emit warnings. The behavior of bothrandom_state
andrng
are outlined above, but onlyrng
should be used in new code.