# neupy.datasets.digits module

neupy.datasets.digits.make_digits(n_samples=100, noise_level=0.1, mode='flip')[source]

Returns discrete digits dataset.

Parameters: n_samples : int Number of samples. Defaults to 100. noise_level : float Defines level of a discrete noise added to the images. Noise level defines probability for the pixel to be removed. Value should be in [0, 1) range. Defaults to 0.1. mode : {remove, flip} This option allow to specify how additional noise will modify each image. flip - Per every randomly selected pixel function flips binary value. 1 -> 0 and 0 -> 1. remove - Per every randomly selected pixel function checks if value equal to 1 if it’s true that it gets replaced with 0. tuple Tuple contains two values. First one is a matrix with shape (n_samples, 24). Second one is a vector that contains labels for each row. Each digit can be transformed into (6, 4) binary image.

Examples

>>> from neupy import datasets, utils
>>>
>>> utils.reproducible()
>>>
>>> digits, labels = datasets.make_digits(noise_level=0.15)
>>> digit, label = digits, labels
>>>
>>> label
5
>>>
>>> digit.reshape((6, 4))
array([[0, 0, 1, 1],
[1, 0, 0, 0],
[0, 1, 1, 0],
[0, 0, 0, 0],
[0, 0, 0, 1],
[1, 1, 1, 0]], dtype=uint8)

neupy.datasets.digits.load_digits()[source]

Returns dataset that contains discrete digits.

Returns: tuple Tuple contains two values. First one is a matrix with shape (10, 24). Second one is a vector that contains labels for each row. Each digit can be transformed into (6, 4) binary image.