Source Code for Module mvpa.tests.test_gnb

 1  # emacs: -*- mode: python; py-indent-offset: 4; indent-tabs-mode: nil -*- 
 2  # vi: set ft=python sts=4 ts=4 sw=4 et: 
 3  ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ## 
 4  # 
 5  #   See COPYING file distributed along with the PyMVPA package for the 
 6  #   copyright and license terms. 
 7  # 
 8  ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ### ## 
 9  """Unit tests for PyMVPA GNB classifier""" 
10   
11  from mvpa.clfs.gnb import GNB 
12  from tests_warehouse import * 
13   
14 -class GNBTests(unittest.TestCase): 
15   
16 -    def testGNB(self): 
17          gnb = GNB() 
18          gnb_nc = GNB(common_variance=False) 
19          gnb_n = GNB(normalize=True) 
20          gnb_n_nc = GNB(normalize=True, common_variance=False) 
21   
22          ds_tr = datasets['uni2medium_train'] 
23          ds_te = datasets['uni2medium_test'] 
24   
25          # Generic silly coverage just to assure that it works in all 
26          # possible scenarios: 
27          bools = (True, False) 
28          # There should be better way... heh 
29          for cv in bools:                # common_variance? 
30            for prior in ('uniform', 'laplacian_smoothing', 'ratio'): 
31              tp = None                   # predictions -- all above should 
32                                          # result in the same predictions 
33              for n in bools:             # normalized? 
34                for ls in bools:          # logspace? 
35                  for es in ((), ('values')): 
36                      gnb_ = GNB(common_variance=cv, 
37                                 prior=prior, 
38                                 normalize=n, 
39                                 logprob=ls, 
40                                 enable_states=es) 
41                      gnb_.train(ds_tr) 
42                      predictions = gnb_.predict(ds_te.samples) 
43                      if tp is None: 
44                          tp = predictions 
45                      self.failUnless((predictions == tp), 
46                                      msg="%s failed to reproduce predictions" % 
47                                      gnb_) 
48                      # if normalized -- check if values are such 
49                      if n and 'values' in es: 
50                          v = gnb_.values 
51                          if ls:          # in log space -- take exp ;) 
52                              v = N.exp(v) 
53                          d1 = N.sum(v, axis=1) - 1.0 
54                          self.failUnless(N.max(N.abs(d1)) < 1e-5) 
55   
56 -def suite(): 
57      return unittest.makeSuite(GNBTests) 
58   
59   
60  if __name__ == '__main__': 
61      import runner 
62