def test(model, data, labels, test_size, data_type, method, color):
train_data, test_data, train_labels, test_labels = train_test_split(data, labels, test_size=test_size)
model.fit(train_data, train_labels)
accuracy = model.score(test_data, test_labels)
print("accuracy with original data: %0.2f " % accuracy)
data_new=copy.deepcopy(data)
train_data, test_data, train_labels, test_labels = train_test_split(util.shuffle_data(data_new), labels, test_size=test_size)
model.fit(train_data, train_labels)
accuracy = model.score(test_data, test_labels)
print("accuracy with shuffled data: %0.2f " % accuracy)
labels_new = copy.deepcopy(labels)
train_data, test_data, train_labels, test_labels = train_test_split(data, util.shuffle_labels(labels_new), test_size=test_size)
model.fit(train_data, train_labels)
accuracy = model.score(test_data, test_labels)
print("accuracy with shuffled labels: %0.2f " % accuracy)
#train with real; test with shuffled
train_data, test_data, train_labels, test_labels = train_test_split(data, labels, test_size=test_size)
model.fit(train_data, train_labels)
accuracy = model.score(util.shuffle_data(test_data), test_labels)
print("accuracy - real training data + shuffled test data: %0.2f " % accuracy)
#train with shuffled data; test with real stuff
train_data, test_data, train_labels, test_labels = train_test_split(data, labels, test_size=test_size)
model.fit(util.shuffle_data(train_data), train_labels)
accuracy = model.score(test_data, test_labels)
print("accuracy - shuffled training data + real test data: %0.2f " % accuracy)
output_dir = sys.argv[2]
util.remove_old_output_files(output_dir)
sys.stdout = open(util.generate_output_filename(output_dir), "w")
lines = util.readFile(input_file)
threegrams = create_keys(3)
twograms = create_keys(2)
proteins = []
accuracies = []
for line in lines:
if line:
print("-----------------------------------------")
print(line[0], line[1], line[2], line[3])
print("start processing ", line[2], line[3])
file1 = line[0]
file2 = line[1]
analysis_type = line[2]
protein = line[3]
util.start_roc_plot(analysis_type + '-' + protein)
data, labels = get_data_from_files(file1, file2)
accuracy = run(data, labels, analysis_type + '-' + protein)
accuracies.append(accuracy)
proteins.append(protein)
print("&&& &&& &&& &&& &&& &&& &&& &&& &&&")
accuracy = run(util.shuffle_data(data), labels, analysis_type + '-' + protein + '_shuffle')
util.save_plot(output_dir, analysis_type + '-' + protein, 'svg', 600)
print("done processing ", line[2], line[3])
else:
util.plot_accuracies(accuracies, proteins, analysis_type, output_dir)
proteins = []
accuracies = []
print("\n\n")
