def test_plr(self): data = datasets['dumb2'] clf = PLR() clf.train(data) # prediction has to be perfect self.failUnless((clf.predict(data.samples) == data.targets).all())
def test_plr(self): data = datasets['dumb2'] clf = PLR() clf.train(data) # prediction has to be perfect self.assertTrue((clf.predict(data.samples) == data.targets).all())
def test_plr_state(self): data = datasets['dumb2'] clf = PLR() clf.train(data) clf.ca.enable('estimates') clf.ca.enable('predictions') p = clf.predict(data.samples) self.failUnless((p == clf.ca.predictions).all()) self.failUnless(np.array(clf.ca.estimates).shape == np.array(p).shape)
def test_plr_state(self): data = datasets['dumb2'] clf = PLR() clf.train(data) # Also get "sensitivity". Was introduced to check a bug with # processing dataset with numeric labels sa = clf.get_sensitivity_analyzer() sens = sa(data) clf.ca.enable('estimates') clf.ca.enable('predictions') p = clf.predict(data.samples) self.assertTrue((p == clf.ca.predictions).all()) self.assertTrue(np.array(clf.ca.estimates).shape == np.array(p).shape)
# lets remove multiclass label from it gprcb.__tags__.pop(gprcb.__tags__.index('multiclass')) clfswh += gprcb # and create a proper multiclass one clfswh += MulticlassClassifier(RegressionAsClassifier( GPR(kernel=GeneralizedLinearKernel())), descr="GPRCM(kernel='linear')") # BLR from mvpa2.clfs.blr import BLR clfswh += RegressionAsClassifier(BLR(descr="BLR()"), descr="BLR Classifier") #PLR from mvpa2.clfs.plr import PLR clfswh += PLR(descr="PLR()") if externals.exists('scipy'): clfswh += PLR(reduced=0.05, descr="PLR(reduced=0.01)") # SVM stuff if len(clfswh['linear', 'svm']) > 0: linearSVMC = clfswh['linear', 'svm', cfg.get('svm', 'backend', default='libsvm').lower()][0] # "Interesting" classifiers clfswh += \ FeatureSelectionClassifier( linearSVMC.clone(), SensitivityBasedFeatureSelection(