return classifier
if __name__ == "__main__":
if len(sys.argv) != 4:
print >> sys.stderr, "Usage: classify.py [TAB_FILE] [classifier: tree, bayes, svm, logreg] [number to filter, 0 -> no filtering]"
sys.exit(1)
data = proj_utils.load_data(sys.argv[1])
type = sys.argv[2]
features = int(sys.argv[3])
train_data, test_data = proj_utils.partition_data(data)
model = train_classifier(train_data, type, features)
train_CA, train_results = proj_utils.test_classifier(model, train_data)
test_CA, test_results = proj_utils.test_classifier(model, test_data)
#print "Train Accuracy: %f, Test Accuracy: %f" % (train_CA, test_CA)
train_stats = proj_utils.get_stats(train_results)
test_stats = proj_utils.get_stats(test_results)
print "Train:\n%s" % str(train_stats)
print "\nTest:\n%s" % str(test_stats)
if features != 0:
print "Features selected:\n%s" % str(model.domain)
filename = '%s%sresults%s%s_filtered_%s.txt' % (
os.path.dirname(__file__), os.sep, os.sep, type, str(features)
) #+ type + '_filtered_' + str(features) + '.txt'
else:
return classifier
if __name__ == "__main__":
if len(sys.argv) != 4:
print >> sys.stderr, "Usage: classify.py [TAB_FILE] [classifier: tree, bayes, svm, logreg] [number to filter, 0 -> no filtering]"
sys.exit(1)
data = proj_utils.load_data(sys.argv[1])
type = sys.argv[2]
features = int(sys.argv[3])
train_data, test_data = proj_utils.partition_data(data)
model = train_classifier(train_data, type, features)
train_CA, train_results = proj_utils.test_classifier(model, train_data)
test_CA, test_results = proj_utils.test_classifier(model, test_data)
#print "Train Accuracy: %f, Test Accuracy: %f" % (train_CA, test_CA)
train_stats = proj_utils.get_stats(train_results)
test_stats = proj_utils.get_stats(test_results)
print "Train:\n%s" % str(train_stats)
print "\nTest:\n%s" % str(test_stats)
if features != 0:
print "Features selected:\n%s" % str(model.domain)
filename = '%s%sresults%s%s_filtered_%s.txt' % (os.path.dirname(__file__), os.sep, os.sep, type, str(features)) #+ type + '_filtered_' + str(features) + '.txt'
else:
filename = '%s%sresults%s%s.txt' % (os.path.dirname(__file__), os.sep, os.sep, type)
print filename