mvpa2.algorithms.hyperalignment.Hyperalignment

Inheritance diagram of Hyperalignment

class mvpa2.algorithms.hyperalignment.Hyperalignment(**kwargs)

...

Given a set of datasets (may be just data) provide mapping of features into a common space

Notes

Available conditional attributes:

  • choosen_ref_ds+: If ref_ds wasn’t provided, it gets choosen.
  • residual_errors: Residual error per each dataset at each level.

(Conditional attributes enabled by default suffixed with +)

Initialize instance of Hyperalignment

Parameters :

alignment :

The multidimensional transformation mapper. If None (default) an instance of ProcrusteanMapper is used. (Default: ProcrusteanMapper(scaling=True, reflection=True, reduction=True, oblique=False, oblique_rcond=-1))

level2_niter :

Number of 2nd level iterations. (Default: 1)

ref_ds :

Index of a dataset to use as a reference. If None, then dataset with maximal number of features is used. (Default: None)

zscore_all :

Z-score all datasets prior hyperalignment. Turn it off if zscoring is not desired or was already performed. If on, resultant mapping becomes a chain with ZScoreMapper. (Default: False)

zscore_common :

Z-score common space after each adjustment. (Default: True)

combiner1 :

How to update common space in the 1st loop. (Default: <function <lambda> at 0x94c056c>)

combiner2 :

How to combine all individual spaces to common space. (Default: <function <lambda> at 0x94c05a4>)

enable_ca : None or list of str

Names of the conditional attributes which should be enabled in addition to the default ones

disable_ca : None or list of str

Names of the conditional attributes which should be disabled

descr : str

Description of the instance

NeuroDebian

NITRC-listed