def read_dataset():
dataset = sets.Ocr()
dataset = sets.OneHot(dataset.target, depth=2)(dataset, columns=['target'])
dataset['data'] = dataset.data.reshape(dataset.data.shape[:-2] +
(-1, )).astype(float)
train, test = sets.Split(0.66)(dataset)
return train, test
def test_split(dataset):
one, two = sets.Split(0.5)(dataset)
assert len(one) + len(two) == len(dataset)
data = np.concatenate((one.data, two.data))
target = np.concatenate((one.target, two.target))
assert (data == dataset.data).all()
assert (target == dataset.target).all()
def get_dataset():
"""Read dataset and flatten images."""
dataset = sets.Ocr()
dataset = sets.OneHot(dataset.target, depth=2)(dataset, columns=['target'])
dataset['data'] = dataset.data.reshape(dataset.data.shape[:-2] +
(-1, )).astype(float)
train, test = sets.Split(0.66)(dataset)
return train, test
def split(full, num):
"""returns the num-th partition of 10-fold cross validation (0 to 9)
full: sets.core.dataset.Dataset object
num: int
rtype: (Dataset, Dataset)"""
validation_size = len(full) // 10
data = full.data.tolist()
target = full.target.tolist()
length = full.length.tolist()
for _ in range(num * validation_size):
data.insert(0, data.pop())
target.insert(0, target.pop())
length.insert(0, length.pop())
data = np.array(data)
target = np.array(target)
length = np.array(length)
full = sets.core.dataset.Dataset(data=data, target=target)
full.__setitem__('length', length)
train, validation = sets.Split(0.9)(full)
return train, validation