Determine best matching distribution.
Can be used for ‘smelling’ the data, as well to choose a parametric distribution for data obtained from non-parametric testing (e.g. MCNullDist).
WiP: use with caution, API might change
Parameters : | data : np.ndarray
nsamples : int or None
loc : float or None
scale : float or None
test : str
distributions : None or list of str or tuple(str, dict)
**kwargs :
|
---|
Examples
>>> from mvpa2.clfs.stats import match_distribution
>>> data = np.random.normal(size=(1000,1));
>>> matches = match_distribution(
... data,
... distributions=['rdist',
... ('rdist', {'name':'rdist_fixed',
... 'loc': 0.0,
... 'args': (10,)})],
... nsamples=30, test='p-roc', p=0.05)