Package mvpa :: Package tests :: Module tests_warehouse
[hide private]
[frames] | no frames]

Module tests_warehouse

source code

Miscelaneous functions/datasets to be used in the unit tests
Functions [hide private]
 
sweepargs(**kwargs)
Decorator function to sweep over a given set of classifiers
source code
Variables [hide private]
  unittest
  N
  _all_states_enabled = False
  snr_scale = cfg.getAsDType('tests', 'snr scale', float, defaul...
  specs = {'large': {'perlabel': 99, 'nchunks': 11, 'nfeatures':...
  nonbogus_pool = [0, 1, 3, 5]
  datasets = {}
  _dsinv = dumbFeatureDataset()

Imports: environ, traceback, sys, cfg, OddEvenSplitter, MaskedDataset, Classifier, ClassWithCollections, chirpLinear, noisy_2d_fx, multipleChunks, linear_awgn, Dataset, dumbFeatureDataset, wr1996, normalFeatureDataset, pureMultivariateSignal, normalFeatureDataset__, getMVPattern, sinModulated, dumbFeatureBinaryDataset, debug


Function Details [hide private]

sweepargs(**kwargs)

source code 

Decorator function to sweep over a given set of classifiers

Often some unittest method can be ran on multiple classifiers. So this decorator aims to do that

Parameters:
  • clfs (list of Classifier) - List of classifiers to run method on

Variables Details [hide private]

snr_scale

Value:
cfg.getAsDType('tests', 'snr scale', float, default= 1.0)

specs

Value:
{'large': {'perlabel': 99, 'nchunks': 11, 'nfeatures': 20, 'snr': 8* s\
nr_scale}, 'medium': {'perlabel': 24, 'nchunks': 6, 'nfeatures': 14, '\
snr': 8* snr_scale}, 'small': {'perlabel': 12, 'nchunks': 4, 'nfeature\
s': 6, 'snr': 14* snr_scale}}