def classifier_domainadaptationsvm_modular(fm_train_dna=traindna,fm_test_dna=testdna, \
label_train_dna=label_traindna, \
label_test_dna=label_testdna,fm_train_dna2=traindna2,fm_test_dna2=testdna2, \
label_train_dna2=label_traindna2,label_test_dna2=label_testdna2,C=1,degree=3):
feats_train = StringCharFeatures(fm_train_dna, DNA)
feats_test = StringCharFeatures(fm_test_dna, DNA)
kernel = WeightedDegreeStringKernel(feats_train, feats_train, degree)
labels = Labels(label_train_dna)
svm = SVMLight(C, kernel, labels)
svm.train()
#svm.io.set_loglevel(MSG_DEBUG)
#####################################
#print "obtaining DA SVM from previously trained SVM"
feats_train2 = StringCharFeatures(fm_train_dna, DNA)
feats_test2 = StringCharFeatures(fm_test_dna, DNA)
kernel2 = WeightedDegreeStringKernel(feats_train, feats_train, degree)
labels2 = Labels(label_train_dna)
# we regularize against the previously obtained solution
dasvm = DomainAdaptationSVM(C, kernel2, labels2, svm, 1.0)
dasvm.train()
out = dasvm.classify(feats_test2).get_labels()
return out #,dasvm TODO
#for i in xrange(len(out1)):
#
# try:
# assert(abs(out1[i]-out2[i])<= 0.001)
#
# except:
# print "(%.5f, %.5f)" % (out1[i], out2[i])
#
#print "comparing regularization vs vector of twos (DONE)\n"
#############################################
# check final output on test data using regtrick
#############################################
print "classifying with DA_SVM"
out_dasvm_libsvm = dasvm_libsvm.classify(feat_test).get_labels()
out_dasvm_liblinear = dasvm_liblinear.classify(feat_test).get_labels()
# compute manual output
svm_out = dasvm_manual_libsvm.classify(feat_test).get_labels()
pre_svm_out = presvm_libsvm.classify(feat_test).get_labels()
out_manual = svm_out + B * (pre_svm_out - presvm_liblinear.get_bias())
print "comparing final outputs"
# compare final outputs
for i in xrange(len(out_dasvm_libsvm)):
try:
assert (abs(out_dasvm_liblinear[i] - out_dasvm_libsvm[i]) <= 0.001)
except:
print "(%.5f, %.5f, %.5f)" % (out_dasvm_liblinear[i],
'AGCAGGAAGGGGGGGAGTC'] label_test_dna2 = numpy.array(5*[-1.0] + 5*[1.0]) C = 1.0 feats_train = StringCharFeatures(fm_train_dna, DNA) feats_test = StringCharFeatures(fm_test_dna, DNA) kernel = WeightedDegreeStringKernel(feats_train, feats_train, degree) labels = Labels(label_train_dna) svm = SVMLight(C, kernel, labels) svm.train() ##################################### print "obtaining DA SVM from previously trained SVM" feats_train2 = StringCharFeatures(fm_train_dna, DNA) feats_test2 = StringCharFeatures(fm_test_dna, DNA) kernel2 = WeightedDegreeStringKernel(feats_train, feats_train, degree) labels2 = Labels(label_train_dna) # we regularize against the previously obtained solution dasvm = DomainAdaptationSVM(C, kernel2, labels2, svm, 1.0) dasvm.train() out = dasvm.classify(feats_test2).get_labels() print out
#############################################
dat = run.get_eval_data()
dtest = dat["thaliana"]
subset_size = 5
test_examples = [i.example for i in dtest[0:subset_size]]
test_labels = [i.label for i in dtest[0:subset_size]]
feat_test = StringCharFeatures(DNA)
feat_test.set_features(test_examples)
print "classifying with DA_SVM"
out_dasvm = dasvm.classify(feat_test).get_labels()
print out_dasvm
svm_out = svm.classify(feat_test).get_labels()
pre_svm_out = presvm.classify(feat_test).get_labels()
total_out = svm_out + B*(pre_svm_out - presvm.get_bias())
print total_out
for i in xrange(len(out_dasvm)):
o1 = total_out[i]
