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.
- rngnumpy.random.Generator, optional
- Pseudorandom number generator used to perform resampling. - If - rngis passed by keyword to the initializer or the- rngattribute is used directly, types other than- numpy.random.Generatorare passed to- numpy.random.default_rngto instantiate a- Generatorbefore use. If- rngis already a- Generatorinstance, then the provided instance is used. Specify- rngfor repeatable behavior.- If this argument is passed by position, if - random_stateis passed by keyword into the initializer, or if the- random_stateattribute is used directly, legacy behavior for- random_stateapplies:- If - random_stateis None (or- numpy.random), the- numpy.random.RandomStatesingleton is used.
- If - random_stateis an int, a new- RandomStateinstance is used, seeded with- random_state.
- If - random_stateis already a- Generatoror- RandomStateinstance then that instance is used.
 - Changed in version 1.15.0: As part of the SPEC-007 transition from use of - numpy.random.RandomStateto- numpy.random.Generator, this attribute name was changed from- random_stateto- rng. For an interim period, both names will continue to work, although only one may be specified at a time. After the interim period, uses of- random_statewill emit warnings. The behavior of both- random_stateand- rngare outlined above, but only- rngshould be used in new code.