Package mvpa :: Package clfs :: Module warehouse
[hide private]
[frames] | no frames]

Module warehouse

source code

Collection of classifiers to ease the exploration.
Classes [hide private]
  Warehouse
Class to keep known instantiated classifiers
Variables [hide private]
  _KNOWN_INTERNALS = ['knn', 'binary', 'svm', 'linear', 'smlr', ...
  clfs = Warehouse(known_tags= _KNOWN_INTERNALS)
  regrs = Warehouse(known_tags= _KNOWN_INTERNALS)
  bad_classifiers = ['mpd', 'gpbt', 'gmnp', 'svrlight', 'krr',]
  linearSVMC = clfs ['linear', 'svm', cfg.get('svm', 'backend', ...
  rfesvm_split = SplitClassifier(linearSVMC)
  rfesvm = LinearCSVMC()

Imports: Set, operator, OddEvenSplitter, FeatureSelectionClassifier, SplitClassifier, MulticlassClassifier, SMLR, kNN, KernelLinear, KernelSquaredExponential, TransferError, externals, cfg, OneWayAnova, Absolute, RFE, SMLRWeights, FractionTailSelector, FixedNElementTailSelector, RangeElementSelector, FixedErrorThresholdStopCrit, ConfusionBasedError, SensitivityBasedFeatureSelection, libsvm, sg, lars, LARS, GPR, BLR


Variables Details [hide private]

_KNOWN_INTERNALS

Value:
['knn', 'binary', 'svm', 'linear', 'smlr', 'does_feature_selection', '\
has_sensitivity', 'multiclass', 'non-linear', 'kernel-based', 'lars', \
'regression', 'libsvm', 'sg', 'meta', 'retrainable', 'gpr', 'notrain2p\
redict', 'ridge', 'blr', 'gnpp']

linearSVMC

Value:
clfs ['linear', 'svm', cfg.get('svm', 'backend', default= 'libsvm').lo\
wer()] [0]