def load_train_data():
util.download_and_extract(
"./dataset/MNIST", "http://deeplearning.net/data/mnist/mnist.pkl.gz")
train, valid, test = util.unpickle("./dataset/MNIST/mnist.pkl")
X_train = np.concatenate([train[0], valid[0]]).reshape([-1, 28, 28])
Y_train = np.concatenate([train[1], valid[1]]).T
return X_train, Y_train, util.class_encoding(Y_train)
def load_test_data():
util.download_and_extract(
"./dataset/MNIST", "http://deeplearning.net/data/mnist/mnist.pkl.gz")
train, valid, test = util.unpickle("./dataset/MNIST/mnist.pkl")
X_test = test[0].reshape([-1, 28, 28])
Y_test = test[1].T
return X_test, Y_test, util.class_encoding(Y_test)
def load_test_data():
util.download_and_extract(
"./dataset/CIFAR-10",
"https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz")
data_dict = util.unpickle(
"./dataset/CIFAR-10/cifar-10-batches-py/test_batch")
X_test = (np.array(data_dict[b'data'], dtype=float) / 255.0).reshape(
[-1, 3, 32, 32]).transpose([0, 2, 3, 1])
Y_test = np.array(data_dict[b'labels'], dtype=int)
return X_test, Y_test, util.class_encoding(Y_test)
def load_train_data():
util.kaggle_and_extract("./dataset/FRUITS", "moltean/fruits")
keys = [
str(x) for x in sorted(
Path('./dataset/FRUITS/fruits-360/Training').glob('*'))
]
X_train = np.asarray(
sorted(Path('./dataset/FRUITS/fruits-360/Training').glob("**/*.jpg")))
Y_train = np.asarray(
[keys.index('/'.join(str(x).split('/')[:-1])) for x in X_train])
return X_train, Y_train, util.class_encoding(Y_train)
def load_train_data():
util.download_and_extract(
"./dataset/CIFAR-10",
"https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz")
X_train = np.zeros(shape=[50000, 32, 32, 3], dtype=float)
Y_train = np.zeros(shape=[50000], dtype=int)
begin = 0
for i in range(1, 6):
data_dict = util.unpickle(
"./dataset/CIFAR-10/cifar-10-batches-py/data_batch_{}".format(i))
x_raw = (np.array(data_dict[b'data'], dtype=float) / 255.0).reshape(
[-1, 3, 32, 32]).transpose([0, 2, 3, 1])
y_raw = np.array(data_dict[b'labels'], dtype=int)
num = len(x_raw)
end = begin + num
X_train[begin:end, :] = x_raw
Y_train[begin:end] = y_raw
begin = end
return X_train, Y_train, util.class_encoding(Y_train)
def load_test_data():
util.kaggle_and_extract("./dataset/IRIS", "uciml/iris")
raw = list(csv.reader(open('./dataset/IRIS/Iris.csv', 'r')))[1:]
X_test = np.asarray([np.asarray([x for x in r[:-1]]) for r in raw])
Y_test = np.asarray([np.asarray(get_key(r[-1])) for r in raw])
return X_test, Y_test, util.class_encoding(Y_test)
