def test_multi_dimensional_numpy(self):
actual_encoded, encoding = one_hot_encoding(
self.multi_dimensional_numpy)
assert len(self.multi_dimensional) == len(actual_encoded)
assert type(actual_encoded) == np.ndarray
actual = actual_encoded
expected = None
for _ in range(NUMBER_OF_RANDOM_TEST):
index = randint(0, len(actual_encoded) - 1)
actual = actual_encoded[index]
expected = self.multi_dimensional[index]
assert type(actual) == np.ndarray
assert len(actual) == 5
index = randint(0, len(actual) - 1)
actual = actual[index]
expected = expected[index]
assert type(actual) == np.ndarray
assert len(actual) == 2
index = randint(0, len(actual) - 1)
actual = actual[index]
expected = expected[index]
assert type(actual) == np.ndarray
assert len(actual) == (N_ELEMENTS * 1000 / (5 * 4 * 2))
for _ in range(NUMBER_OF_RANDOM_TEST):
index = randint(0, len(actual) - 1)
actual_element = actual[index]
expected_element = encoding[expected[index]]
assert type(actual_element) == np.ndarray
assert len(actual_element) == N_ELEMENTS
assert abs(sum(actual_element) - 1) < EPSILON
assert (actual_element == expected_element).all()
def test_one_dimension(self):
actual, encoding = one_hot_encoding(self.one_dimension)
assert len(self.one_dimension) == len(actual)
assert type(actual) == list
for _ in range(NUMBER_OF_RANDOM_TEST):
index = randint(0, len(actual) - 1)
actual_element = actual[index]
expected = encoding[self.one_dimension[index]]
assert type(actual_element) == list
assert len(actual_element) == N_ELEMENTS
assert abs(sum(actual_element) - 1) < EPSILON
assert (expected == actual_element).all()
def test_one_dimension_numpy(self):
actual, encoding = one_hot_encoding(self.one_dimension_numpy)
assert len(self.one_dimension_numpy) == len(actual)
assert actual.shape == (len(actual), N_ELEMENTS)
assert type(actual) == np.ndarray
for _ in range(NUMBER_OF_RANDOM_TEST):
index = randint(0, len(actual) - 1)
actual_element = actual[index]
expected = encoding[self.one_dimension_numpy[index]]
assert type(actual_element) == np.ndarray
assert len(actual_element) == N_ELEMENTS
assert abs(actual_element.sum() - 1) < EPSILON
assert (expected == actual_element).all()
# Initialize network
network = NeuralNetwork(4, [6], 3, [tanh, sigmoid], LR)
filename = "network"
type_net = "Neural"
k_fold = ""
if args.normalize:
network = NormalizedNetwork(4, [6], 3, [tanh, sigmoid], LR)
type_net = "Normalized"
filename = type_net.lower()
# iris dataset
dataset = import_data("../../data/iris.data")
labels, encoding = one_hot_encoding(dataset[-1])
classes = list(encoding.keys())
dataset = dataset[0:-1]
# Define Trainer
trainer = StandardTrainer(dataset, labels.T, TRAIN_SIZE)
k = 1
if args.cross_validation is not None:
k = args.cross_validation
k_fold = "_{}fold".format(k)
trainer = KFoldTrainer(k, 2, dataset, labels.T)
fig = plt.figure(figsize=FIG_SIZE)
fig.subplots_adjust(wspace=0.3)
ax = fig.add_subplot(121)
def format_axes(title: str, scale: int, ax: Axes) -> None:
"""
Format axes of a matplotlib graph
:param title: Title
:param scale: Scale
:param ax: Axes of matplotlib
:return: None, modify ax
"""
ax.set_title("{}\n".format(title), fontsize=TITLE_SIZE)
ax.set_aspect(aspect=str(6 * scale))
ax.set_xlabel("Indexes", fontsize=FONT_SIZE)
ax.grid()
if __name__ == '__main__':
fig, axes = plt.subplots(3, 1, figsize=FIG_SIZE)
dataset = import_data("../../data/iris.data")
one_hot_dataset, encoding = one_hot_encoding(dataset[4])
axes[0].imshow(one_hot_dataset.T, cmap='Greys')
format_axes("Labels of dataset, size: {}".format(one_hot_dataset.shape[0]),
one_hot_dataset.shape[0] / 150, axes[0])
train_set, test_set = split_set(one_hot_dataset.T, 0.6)
axes[1].imshow(train_set, cmap='Greys')
format_axes("Train set, size: {}".format(train_set.shape[1]),
train_set.shape[1] / 150, axes[1])
axes[2].imshow(test_set, cmap='Greys')
format_axes("Test set, size: {}".format(test_set.shape[1]),
test_set.shape[1] / 150, axes[2])
plt.savefig("../results/labels_of_dataset.png")
def train_evaluate(architecture: dict, dataset_name: str) -> NeuralNetwork:
"""
Train and evaluate a Network
:param architecture: Architecture of NeuralNetwork (above)
:param dataset_name: Dataset to use
:return: Trained Neural Network
"""
# import dataset
dataset = import_data("../data/{}.data".format(dataset_name))
dataset = oversample(dataset)
more_info = "(oversample)"
labels, encoding = one_hot_encoding(dataset[-1])
classes = list(encoding.keys())
dataset = np.delete(dataset, -1, 0)
dataset = np.delete(dataset, [0], 0)
# Initialize network
logging.info("Input size: {}\tOutput size: {}".format(
dataset.shape[0], len(encoding)))
network = NeuralNetwork(dataset.shape[0], architecture["INTERNAL_LAYERS"],
len(encoding), architecture["ACT_FUNCS"],
architecture["LR"])
# Define Trainer
trainer = StandardTrainer(dataset, labels.T, TRAIN_SIZE)
fig = plt.figure(figsize=FIG_SIZE)
fig.subplots_adjust(wspace=0.3)
ax = fig.add_subplot(121)
ax2 = ax.twinx()
ax3 = fig.add_subplot(122)
trained, (learn, costs) = trainer.train(network,
epochs=EPOCHS,
repeat=True)
prediction = trainer.evaluate(trained)
c_m = confusion_matrix(prediction, trainer.get_labels())
line = ax.plot(learn, label="Learning Curve", linewidth=2.5, c="b")
line2 = ax2.plot(costs, label="MSE", linestyle="--", linewidth=2.5, c="r")
lines = line + line2
ax.set_ylabel("Learning Curve", fontsize=FONT_SIZE)
ax.set_xlabel("Epochs", fontsize=FONT_SIZE)
ax.set_title("Network on {}\n".format(dataset_name), fontsize=TITLE_SIZE)
ax.grid()
ax2.set_ylabel("Cost", fontsize=FONT_SIZE)
ax2.grid()
labels = [l.get_label() for l in lines]
ax2.legend(lines, labels, fontsize=FONT_SIZE, loc="center right")
show_matrix(
ax3, c_m,
(classes, ["Predicted\n{}".format(a_class) for a_class in classes]),
"Confusion Matrix of Test Set\n", FONT_SIZE, TITLE_SIZE)
measures = {
"accuracy": accuracy(c_m),
"precision": precision(c_m),
"recall": recall(c_m),
"f1_score": f1_score(c_m)
}
print("Summary on {}:\n".format(dataset))
report_results(c_m)
plt.savefig("../results/Network_on_{}{}.png".format(
dataset_name, more_info))
return trained
else:
path_dataset = args.dataset + ".data"
# import dataset
dataset = import_data("../data/{}".format(path_dataset), sep=args.sep, header=args.header)
more_info = ""
if args.oversample:
dataset = oversample(dataset, label=args.labels)
more_info = "(oversampled)"
if args.undersample:
dataset = undersample(dataset, label=args.labels)
more_info = "(undersampled)"
labels, encoding = one_hot_encoding(dataset[args.labels])
classes = list(encoding.keys())
dataset = np.delete(dataset, args.labels, 0)
encodings = list()
first = True
for i in reversed(args.categorical):
temp1, temp2 = one_hot_encoding(dataset[i])
dataset = np.concatenate((np.delete(dataset, i, 0), temp1.copy().T))
encodings.append(temp2.copy())
del temp1
del temp2
dataset = np.delete(dataset, args.exclude, 0)
from useful.preprocess_dataset import import_data, one_hot_encoding
from useful.results import confusion_matrix, show_matrix, annotate
from useful.results import accuracy, precision, recall, f1_score
N = int(1e5)
X_MIN = Y_MIN = -50
X_MAX = Y_MAX = 50
FIG_SIZE = (28, 12)
FONT_SIZE = 20
TITLE_SIZE = 30
np.random.seed(2)
if __name__ == '__main__':
dataset = import_data("../../data/iris.data")[4]
classes = np.unique(dataset)
labels, _ = one_hot_encoding(dataset)
prediction, _ = one_hot_encoding(
np.random.choice(["a", "b", "c"], size=labels.shape[0], replace=True))
matrix = confusion_matrix(prediction.T, labels.T)
fig, (ax1, ax2) = plt.subplots(1, 2, figsize=FIG_SIZE)
show_matrix(ax1, matrix, ([classes[0], classes[1], classes[2]], [
"Predicted\n" + classes[0], "Predicted\n" + classes[1],
"Predicted\n" + classes[2]
]), "Confusion matrix of a iris dataset\n", FONT_SIZE, TITLE_SIZE)
measures = np.zeros((3, 4))
ax2.matshow(measures, cmap="Greys")
to_show = np.zeros((3, 4))
to_show[0][0] = round(accuracy(matrix), 4)
to_show[1][0] = np.nan
to_show[2][0] = np.nan
_precision = precision(matrix)