def test_multiclass_classifier(self, clf):
oldC = None
# XXX somewhat ugly way to force non-dataspecific C value.
# Otherwise multiclass libsvm builtin and our MultiClass would differ
# in results
if clf.params.has_key('C') and clf.params.C<0:
oldC = clf.params.C
clf.params.C = 1.0 # reset C to be 1
svm, svm2 = clf, clf.clone()
svm2.ca.enable(['training_stats'])
mclf = MulticlassClassifier(clf=svm,
enable_ca=['training_stats'])
svm2.train(datasets['uni2small'])
mclf.train(datasets['uni2small'])
s1 = str(mclf.ca.training_stats)
s2 = str(svm2.ca.training_stats)
self.failUnlessEqual(s1, s2,
msg="Multiclass clf should provide same results as built-in "
"libsvm's %s. Got %s and %s" % (svm2, s1, s2))
svm2.untrain()
self.failUnless(svm2.trained == False,
msg="Un-Trained SVM should be untrained")
self.failUnless(np.array([x.trained for x in mclf.clfs]).all(),
msg="Trained Boosted classifier should have all primary classifiers trained")
self.failUnless(mclf.trained,
msg="Trained Boosted classifier should be marked as trained")
mclf.untrain()
self.failUnless(not mclf.trained,
msg="UnTrained Boosted classifier should not be trained")
self.failUnless(not np.array([x.trained for x in mclf.clfs]).any(),
msg="UnTrained Boosted classifier should have no primary classifiers trained")
if oldC is not None:
clf.params.C = oldC
def test_multiclass_classifier(self, clf):
oldC = None
# XXX somewhat ugly way to force non-dataspecific C value.
# Otherwise multiclass libsvm builtin and our MultiClass would differ
# in results
if clf.params.has_key('C') and clf.params.C < 0:
oldC = clf.params.C
clf.params.C = 1.0 # reset C to be 1
svm, svm2 = clf, clf.clone()
svm2.ca.enable(['training_confusion'])
mclf = MulticlassClassifier(clf=svm, enable_ca=['training_confusion'])
svm2.train(datasets['uni2small_train'])
mclf.train(datasets['uni2small_train'])
s1 = str(mclf.ca.training_confusion)
s2 = str(svm2.ca.training_confusion)
self.failUnlessEqual(
s1,
s2,
msg="Multiclass clf should provide same results as built-in "
"libsvm's %s. Got %s and %s" % (svm2, s1, s2))
svm2.untrain()
self.failUnless(svm2.trained == False,
msg="Un-Trained SVM should be untrained")
self.failUnless(
np.array([x.trained for x in mclf.clfs]).all(),
msg=
"Trained Boosted classifier should have all primary classifiers trained"
)
self.failUnless(
mclf.trained,
msg="Trained Boosted classifier should be marked as trained")
mclf.untrain()
self.failUnless(
not mclf.trained,
msg="UnTrained Boosted classifier should not be trained")
self.failUnless(
not np.array([x.trained for x in mclf.clfs]).any(),
msg=
"UnTrained Boosted classifier should have no primary classifiers trained"
)
if oldC is not None:
clf.params.C = oldC
def __test_fspipeline_with_split_classifier(self, basic_clf):
#basic_clf = LinearNuSVMC()
multi_clf = MulticlassClassifier(clf=basic_clf)
#svm_weigths = LinearSVMWeights(svm)
# Proper RFE: aggregate sensitivities across multiple splits,
# but also due to multi class those need to be aggregated
# somehow. Transfer error here should be 'leave-1-out' error
# of split classifier itself
sclf = SplitClassifier(clf=basic_clf)
rfe = RFE(sensitivity_analyzer=sclf.get_sensitivity_analyzer(
enable_ca=["sensitivities"]),
transfer_error=trans_error,
feature_selector=FeatureSelectionPipeline([
FractionTailSelector(0.5),
FixedNElementTailSelector(1)
]),
train_clf=True)
# and we get sensitivity analyzer which works on splits and uses
# sensitivity
selected_features = rfe(self.dataset)
# - Nu-classifiers are turned off since for haxby DS default nu
# is an 'infisible' one
# - Python's SMLR is turned off for the duration of development
# since it is slow and results should be the same as of C version
#
clfswh += [
SMLR(lm=0.1, implementation="C", descr="SMLR(lm=0.1)"),
SMLR(lm=1.0, implementation="C", descr="SMLR(lm=1.0)"),
#SMLR(lm=10.0, implementation="C", descr="SMLR(lm=10.0)"),
#SMLR(lm=100.0, implementation="C", descr="SMLR(lm=100.0)"),
#SMLR(implementation="Python", descr="SMLR(Python)")
]
clfswh += \
[ MulticlassClassifier(clfswh['smlr'][0],
descr='Pairs+maxvote multiclass on ' + \
clfswh['smlr'][0].descr) ]
if externals.exists('libsvm'):
from mvpa.clfs import libsvmc as libsvm
clfswh._known_tags.union_update(libsvm.SVM._KNOWN_IMPLEMENTATIONS.keys())
clfswh += [
libsvm.SVM(descr="libsvm.LinSVM(C=def)", probability=1),
libsvm.SVM(C=-10.0, descr="libsvm.LinSVM(C=10*def)", probability=1),
libsvm.SVM(C=1.0, descr="libsvm.LinSVM(C=1)", probability=1),
libsvm.SVM(svm_impl='NU_SVC',
descr="libsvm.LinNuSVM(nu=def)",
probability=1)
]
clfswh += [
libsvm.SVM(kernel=RbfLSKernel(), descr="libsvm.RbfSVM()"),
def test_multiclass_classifier(self, clf):
oldC = None
# XXX somewhat ugly way to force non-dataspecific C value.
# Otherwise multiclass libsvm builtin and our MultiClass would differ
# in results
if clf.params.has_key('C') and clf.params.C<0:
oldC = clf.params.C
clf.params.C = 1.0 # reset C to be 1
svm, svm2 = clf, clf.clone()
svm2.ca.enable(['training_stats'])
mclf = MulticlassClassifier(clf=svm,
enable_ca=['training_stats'])
# with explicit MaximalVote with the conditional attributes
# enabled
mclf_mv = MulticlassClassifier(clf=svm,
combiner=MaximalVote(enable_ca=['estimates', 'predictions']),
enable_ca=['training_stats'])
for clf_ in svm2, mclf, mclf_mv:
clf_.train(datasets['uni2small'])
s1 = str(mclf.ca.training_stats)
s2 = str(svm2.ca.training_stats)
s3 = str(mclf_mv.ca.training_stats)
self.failUnlessEqual(s1, s2,
msg="Multiclass clf should provide same results as built-in "
"libsvm's %s. Got %s and %s" % (svm2, s1, s2))
self.failUnlessEqual(s1, s3,
msg="%s should have used maxvote resolver by default"
"so results should have been identical. Got %s and %s"
% (mclf, s1, s3))
assert_equal(len(mclf_mv.combiner.ca.estimates),
len(mclf_mv.combiner.ca.predictions))
# They should have came from assessing training_stats ca being
# enabled
# recompute accuracy on predictions for training_stats
training_acc = np.sum(mclf_mv.combiner.ca.predictions ==
datasets['uni2small'].targets) \
/ float(len(datasets['uni2small']))
# should match
assert_equal(mclf_mv.ca.training_stats.stats['ACC'], training_acc)
svm2.untrain()
self.failUnless(svm2.trained == False,
msg="Un-Trained SVM should be untrained")
self.failUnless(np.array([x.trained for x in mclf.clfs]).all(),
msg="Trained Boosted classifier should have all primary classifiers trained")
self.failUnless(mclf.trained,
msg="Trained Boosted classifier should be marked as trained")
mclf.untrain()
self.failUnless(not mclf.trained,
msg="UnTrained Boosted classifier should not be trained")
self.failUnless(not np.array([x.trained for x in mclf.clfs]).any(),
msg="UnTrained Boosted classifier should have no primary classifiers trained")
if oldC is not None:
clf.params.C = oldC
