This content refers to the previous stable release of PyMVPA. Please visit www.pymvpa.org for the most recent version of PyMVPA and its documentation.

measures.splitmeasure

Module: measures.splitmeasure

Inheritance diagram for mvpa.measures.splitmeasure:

This is a FeaturewiseDatasetMeasure that uses another FeaturewiseDatasetMeasure and runs it multiple times on differents splits of a Dataset.

Classes

SplitFeaturewiseMeasure

class mvpa.measures.splitmeasure.SplitFeaturewiseMeasure(sensana, splitter=<class 'mvpa.datasets.splitters.NoneSplitter'>, combiner=<function FirstAxisMean at 0x4892de8>, **kwargs)

Bases: mvpa.measures.base.FeaturewiseDatasetMeasure

This is a FeaturewiseDatasetMeasure that uses another FeaturewiseDatasetMeasure and runs it multiple times on differents splits of a Dataset.

When called with a Dataset it returns the mean sensitivity maps of all data splits.

Additonally this class supports the State interface. Several postprocessing functions can be specififed to the constructor. The results of the functions specified in the postproc dictionary will be available via their respective keywords.

Note

Available state variables:

• base_sensitivities: Stores basic sensitivities if the sensitivity relies on combining multiple ones
• maps: To store maps per each split
• null_prob+: State variable
• null_t: State variable
• raw_results: Computed results before applying any transformation algorithm

(States enabled by default are listed with +)

See also

Please refer to the documentation of the base class for more information:

FeaturewiseDatasetMeasure

Cheap initialization.

Parameters:

• sensana (FeaturewiseDatasetMeasure) – that shall be run on the Dataset splits.
• splitter (Splitter) – used to split the Dataset. By convention the first dataset in the tuple returned by the splitter on each iteration is used to compute the sensitivity map.
• combiner – This functor will be called on an array of sensitivity maps and the result will be returned by __call__(). The result of a combiner must be an 1d ndarray.
• enable_states (None or list of basestring) – Names of the state variables which should be enabled additionally to default ones
• disable_states (None or list of basestring) – Names of the state variables which should be disabled
• transformer (Functor) – This functor is called in __call__() to perform a final processing step on the to be returned dataset measure. If None, nothing is called
• null_dist (instance of distribution estimator) – The estimated distribution is used to assign a probability for a certain value of the computed measure.

TScoredFeaturewiseMeasure

class mvpa.measures.splitmeasure.TScoredFeaturewiseMeasure(sensana, splitter, noise_level=0.0, **kwargs)

Bases: mvpa.measures.splitmeasure.SplitFeaturewiseMeasure

SplitFeaturewiseMeasure computing featurewise t-score of sensitivities across splits.

Note

Available state variables:

• base_sensitivities: Stores basic sensitivities if the sensitivity relies on combining multiple ones
• maps: To store maps per each split
• null_prob+: State variable
• null_t: State variable
• raw_results: Computed results before applying any transformation algorithm

(States enabled by default are listed with +)

See also

Please refer to the documentation of the base class for more information:

SplitFeaturewiseMeasure

Cheap initialization.

Parameters:

• sensana (SensitivityAnalyzer) – that shall be run on the Dataset splits.
• splitter (Splitter) – used to split the Dataset. By convention the first dataset in the tuple returned by the splitter on each iteration is used to compute the sensitivity map.
• noise_level (float) – Theoretical output of the respective SensitivityAnalyzer for a pure noise pattern. For most algorithms this is probably zero, hence the default.
• enable_states (None or list of basestring) – Names of the state variables which should be enabled additionally to default ones
• disable_states (None or list of basestring) – Names of the state variables which should be disabled
• combiner – This functor will be called on an array of sensitivity maps and the result will be returned by __call__(). The result of a combiner must be an 1d ndarray.
• transformer (Functor) – This functor is called in __call__() to perform a final processing step on the to be returned dataset measure. If None, nothing is called
• null_dist (instance of distribution estimator) – The estimated distribution is used to assign a probability for a certain value of the computed measure.