(self, X, Y, train_precent, seed=0)
| 51 | return Y |
| 52 | |
| 53 | def split_train_evaluate(self, X, Y, train_precent, seed=0): |
| 54 | state = numpy.random.get_state() |
| 55 | |
| 56 | training_size = int(train_precent * len(X)) |
| 57 | numpy.random.seed(seed) |
| 58 | shuffle_indices = numpy.random.permutation(numpy.arange(len(X))) |
| 59 | X_train = [X[shuffle_indices[i]] for i in range(training_size)] |
| 60 | Y_train = [Y[shuffle_indices[i]] for i in range(training_size)] |
| 61 | X_test = [X[shuffle_indices[i]] for i in range(training_size, len(X))] |
| 62 | Y_test = [Y[shuffle_indices[i]] for i in range(training_size, len(X))] |
| 63 | |
| 64 | self.train(X_train, Y_train, Y) |
| 65 | numpy.random.set_state(state) |
| 66 | return self.evaluate(X_test, Y_test) |
| 67 | |
| 68 | |
| 69 | def read_node_label(filename, skip_head=False): |
no test coverage detected