print "=" * 10, "EPOCH", i + 1, "=" * 10
# get the sets for training and testing
training_set_names = set_names[:]
test_set_names = [training_set_names.pop(i)]
training_set = fileutils.get_data_set(features_save_path,
training_set_names,
dictionary_size)
test_set = fileutils.get_data_set(features_save_path, test_set_names,
dictionary_size)
print "Training sets:", training_set_names
print "Test set:", test_set_names
# train and test
p.train(training_set)
p.test(test_set)
(precision, recall, f1) = p.get_results()
# prints the results and calc the avg results
print_results(precision, recall, f1)
avg_precision += precision / len(set_names)
avg_recall += recall / len(set_names)
avg_f1 += f1 / len(set_names)
print ""
print "=" * 10, "Average results", "=" * 10
print_results(avg_precision, avg_recall, avg_f1)
def train_and_perception():
p = Perception(2, f)
input_vecs, labels = get_training_dataset()
p.train(input_vecs, labels, 10, 0.1)
return p