Module Reference

This module reference extends the manual with a comprehensive overview of the currently available functionality, that is built into PyMVPA. However, instead of a full list including every single line of the PyMVPA code base, this reference limits itself to the relevant pieces of the application programming interface (API) that are of particular interest to users of this framework.

Each module in the package is documented by a general summary of its purpose and the list of classes and functions it provides.

Entry Point

mvpa2 MultiVariate Pattern Analysis Package Organization ==================== The mvpa package contains the following subpackages and modules: :group Algorithms: algorithms :group Anatomical Atlases: atlases :group Basic Data Structures: datasets :group Classifiers (supervised learners): clfs :group Feature Selections: featsel :group Mappers (usually unsupervised learners): mappers :group Measures: measures :group Miscellaneous: base misc support :group Unittests: tests :author: Michael Hanke, Yaroslav Halchenko, `Per B.

Basic Facilities

base Base functionality of PyMVPA
base.attributes Module with some special objects to be used as magic attributes with dedicated containers aka.
base.collections Module with some special objects to be used as magic attributes with dedicated containers aka.
base.config Registry-like monster
base.dochelpers Various helpers to improve docstrings and textual output
base.externals Helper to verify presence of external libraries and modules
base.hdf5 HDF5-based file IO for PyMVPA objects.
base.info Provide system and PyMVPA information useful while reporting bugs
base.learner Implementation of a common trainable processing object (Learner).
base.node Implementation of a common processing object (node).
base.param Parameter representation
base.report Creating simple PDF reports using reportlab
base.state Classes to control and store state information.
base.types Things concerned with types and type-checking in PyMVPA
base.verbosity Verbose output and debugging facility

Datasets: Input, Output, Storage and Preprocessing

base.dataset Multi-purpose dataset container with support for attributes.
datasets.base PyMVPA’s common Dataset container.
datasets.eventrelated Dataset for event-related samples.
datasets.eep Dataset that gets its samples from an EEP binary file
datasets.formats I/O helpers for some commonly used formats for datasets.
datasets.mri Dataset for magnetic resonance imaging (MRI) data.
datasets.miscfx Misc function performing operations on datasets.
datasets.sources.sklearn_data Wrapper for sklearn datasets/data generators.

Mappers: Data Transformations

mappers Algorithms for (reversible) data transformation.
mappers.base Basic, general purpose and meta mappers.
mappers.boxcar Data mapper
mappers.detrend Mapper for data detrending.
mappers.filters Filtering mappers.
mappers.flatten Data mapper
mappers.fx Transform data by applying a function along samples or feature axis.
mappers.fxy
mappers.lle Local Linear Embedding Data mapper.
mappers.mdp_adaptor MDP interface module.
mappers.procrustean Procrustean rotation mapper
mappers.projection
mappers.prototype Prototype-based Mapper.
mappers.shape Mappers that modify the shape of a dataset
mappers.slicing Mappers for Dataset slicing.
mappers.som Self-organizing map (SOM) mapper.
mappers.svd Singular-value decomposition mapper
mappers.wavelet Wavelet mappers
mappers.zscore Mapper for data normalization by Z-Scoring.

Generators: Repetitive Data Processing

generators Nodes that generate multiple datasets.
generators.base Generators for dataset resampling.
generators.partition
generators.permutation Generator nodes to permute datasets.
generators.resampling Generators for dataset resampling.
generators.splitters Generator nodes to split dataset into multiple parts.

Classifiers and Errors

clfs.base Base class for all XXX learners: classifiers and regressions.
clfs.meta Classes for meta classifiers – classifiers which use other classifiers
clfs.blr Bayesian Linear Regression (BLR).
clfs.enet Elastic-Net (ENET) regression classifier.
clfs.gda Gaussian Discriminant Analyses: LDA and QDA
clfs.glmnet GLM-Net (GLMNET) regression and classifier.
clfs.gnb Gaussian Naive Bayes Classifier Basic implementation of Gaussian Naive Bayes classifier.
clfs.gpr Gaussian Process Regression (GPR).
clfs.knn k-Nearest-Neighbour classifier.
clfs.lars Least angle regression (LARS).
clfs.model_selector Model selction.
clfs.plr Penalized logistic regression classifier.
clfs.ridge Ridge regression classifier.
clfs.similarity Similarity functions for prototype-based projection.
clfs.skl Classifiers provided by scikit-learn (skl) library
clfs.smlr Sparse Multinomial Logistic Regression classifier.
clfs.svm Importer for the available SVM and SVR machines.
clfs.sg Classifiers provided by shogun (sg) library
clfs.libsvmc Classifiers provied by LibSVM library
clfs.distance Distance functions to be used in kernels and elsewhere
clfs.similarity Similarity functions for prototype-based projection.
clfs.stats Estimator for classifier error distributions.
clfs.transerror Utility class to compute the transfer error of classifiers.
clfs.warehouse Collection of classifiers to ease the exploration.

Kernels

kernels Import helper for PyMVPA kernels/similarities and alike
kernels.base Base Kernel classes
kernels.libsvm PyMVPA LibSVM-based kernels These kernels do not currently have the ability to run the calculations, so they are not translateable to other kernel types.
kernels.np Kernels for Gaussian Process Regression and Classification.
kernels.sg PyMVPA shogun-based kernels Provides interface to kernels defined in shogun toolbox.

Measures: Searchlights and Sensitivties

measures.base Base classes for measures: algorithms that quantify properties of datasets.
measures.anova FeaturewiseMeasure performing a univariate ANOVA.
measures.corrcoef FeaturewiseMeasure of correlation with the labels.
measures.corrstability FeaturewiseMeasure of stability of labels across chunks based on correlation.
measures.ds Dissimilarity measure.
measures.irelief FeaturewiseMeasure performing multivariate Iterative RELIEF (I-RELIEF) algorithm.
measures.noiseperturbation This is a FeaturewiseMeasure that uses a scalar Measure and selective noise perturbation to compute a sensitivity map.
measures.pls PLS is not yet implemented
measures.adhocsearchlightbase An efficient implementation of searchlight for GNB.
measures.gnbsearchlight An efficient implementation of searchlight for GNB.
measures.nnsearchlight An efficient implementation of searchlight for M1NN.
measures.searchlight Implementation of the Searchlight algorithm
measures.statsmodels_adaptor

Feature Selection

featsel.base Feature selection base class and related stuff base classes and helpers.
featsel.ifs Incremental feature search (IFS).
featsel.rfe Recursive feature elimination.
featsel.helpers

Additional Algorithms

algorithms.hyperalignment Transformation of individual feature spaces into a common space

Miscellaneous

atlases Import helper for PyMVPA anatomical atlases
misc.args Helpers for arguments handling.
misc.attrmap Helper to map literal attribute to numerical ones (and back)
misc.cmdline Common functions and options definitions for command line Conventions: Every option (instance of optparse.Option) has prefix “opt”.
misc.data_generators Miscellaneous data generators for unittests and demos
misc.dcov Compute dcov/dcorr measures for independence testing
misc.errorfx Error functions helpers.
misc.exceptions Exception classes which might get thrown
misc.fx Misc.
misc.neighborhood Neighborhood objects
misc.sampleslookup Helper to map and validate samples’ origids into indices
misc.stats Little statistics helper
misc.support Support function – little helpers in everyday life
misc.transformers Simply functors that transform something.
misc.vproperty C++-like virtual properties

Testing

testing Helpers to unify/facilitate unittesting within PyMVPA
testing.clfs Provides clfs dictionary with instances of all available classifiers.
testing.datasets Provides convenience datasets for unittesting.
testing.tools A Collection of tools found useful in unittests.
testing.sweepargs(**kwargs) Decorator function to sweep over a given set of classifiers
tests Unit test interface for PyMVPA

Basic Plotting Utilities

misc.plot
misc.plot.base
misc.plot.erp
misc.plot.lightbox
misc.plot.topo

3rd-party Interfaces

misc.bv Import helper for Brain Voyager
misc.bv.base Tiny snippets to interface with FSL easily.
misc.fsl Import helper for FSL
misc.fsl.base Tiny snippets to interface with FSL easily.
misc.fsl.flobs Wrapper around FSLs halfcosbasis to generate HRF kernels
misc.fsl.melodic Wrapper around the output of MELODIC (part of FSL)
misc.io Import helper for IO helpers
misc.io.base Some little helper for reading (and writing) common formats from and to disk.
misc.io.hamster Helper for simple storage facility via cPickle and optionally zlib
misc.io.meg IO helper for MEG datasets.