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.

misc.transformers

Module: misc.transformers

Inheritance diagram for mvpa.misc.transformers:

Simply functors that transform something.

Classes

DistPValue

class mvpa.misc.transformers.DistPValue(sd=0, distribution='rdist', fpp=None, nbins=400, **kwargs)

Bases: mvpa.misc.state.ClassWithCollections

Converts values into p-values under vague and non-scientific assumptions

Note

Available state variables:

• nulldist_number+: Number of features within the estimated null-distribution
• positives_recovered+: Number of features considered to be positives and which were recovered

(States enabled by default are listed with +)

See also

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

ClassWithCollections

L2-Norm the values, convert them to p-values of a given distribution.

Parameters:

• sd (int) – Samples dimension (if len(x.shape)>1) on which to operate
• distribution (string) – Which distribution to use. Known are: ‘rdist’ (later normal should be there as well)
• fpp (float) – At what p-value (both tails) if not None, to control for false positives. It would iteratively prune the tails (tentative real positives) until empirical p-value becomes less or equal to numerical.
• nbins (int) – Number of bins for the iterative pruning of positives
• 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

WARNING: Highly experimental/slow/etc: no theoretical grounds have been presented in any paper, nor proven

OverAxis

class mvpa.misc.transformers.OverAxis(transformer, axis=None)

Bases: object

Helper to apply transformer over specific axis

Initialize transformer wrapper with an axis.

Parameters:

• transformer – A callable to be used
• axis (None or int) – If None – apply transformer across all the data. If some int – over that axis

Functions

mvpa.misc.transformers.Absolute(x)

Returns the elementwise absolute of any argument.

Parameters:x (scalar | sequence) –

mvpa.misc.transformers.FirstAxisMean(x)

Mean computed along the first axis.

mvpa.misc.transformers.FirstAxisSumNotZero(x)

Sum computed over first axis of whether the values are not equal to zero.

mvpa.misc.transformers.GrandMean(x)

Just what the name suggests.

mvpa.misc.transformers.Identity(x)

Return whatever it was called with.

mvpa.misc.transformers.L1Normed(x, norm=1.0, reverse=False)

Norm the values so that L_1 norm (sum|x|) becomes norm

mvpa.misc.transformers.L2Normed(x, norm=1.0, reverse=False)

Norm the values so that regular vector norm becomes norm

More verbose: Norm that the sum of the squared elements of the returned vector becomes norm.

mvpa.misc.transformers.OneMinus(x)

Returns elementwise ‘1 - x’ of any argument.

mvpa.misc.transformers.RankOrder(x, reverse=False)

Rank-order by value. Highest gets 0

mvpa.misc.transformers.ReverseRankOrder(x)

Convinience functor

mvpa.misc.transformers.SecondAxisMaxOfAbs(x)

Max of absolute values along the 2nd axis

mvpa.misc.transformers.SecondAxisMean(x)

Mean across 2nd axis

Use cases:

• to combine multiple sensitivities to get sense about mean sensitivity across splits

mvpa.misc.transformers.SecondAxisSumOfAbs(x)

Sum of absolute values along the 2nd axis

Use cases:

• to combine multiple sensitivities to get sense about what features are most influential