mindspore.nn.probability.bijector.GumbelCDF

class mindspore.nn.probability.bijector.GumbelCDF(loc=0.0, scale=1.0, name='GumbelCDF')

GumbelCDF Bijector. This Bijector performs the operation:

\[Y = \exp(-\exp(\frac{-(X - loc)}{scale}))\]

Parameters

• loc (float, list, numpy.ndarray, Tensor) – The location. Default: 0.0.

• scale (float, list, numpy.ndarray, Tensor) – The scale. Default: 1.0.

• name (str) – The name of the Bijector. Default: ‘GumbelCDF’.

Note

scale must be greater than zero. For inverse and inverse_log_jacobian, input should be in range of (0, 1). The dtype of loc and scale must be float. If loc, scale are passed in as numpy.ndarray or tensor, they have to have the same dtype otherwise an error will be raised.

Raises

TypeError – When the dtype of loc or scale is not float, or when the dtype of loc and scale is not same.

Supported Platforms:

Ascend GPU

Examples

>>> import mindspore
>>> import mindspore.nn as nn
>>> import mindspore.nn.probability.bijector as msb
>>> from mindspore import Tensor
>>>
>>> # To initialize a GumbelCDF bijector of loc 1.0, and scale 2.0.
>>> gumbel_cdf = msb.GumbelCDF(1.0, 2.0)
>>> # To use a GumbelCDF bijector in a network.
>>> x = Tensor([1, 2, 3], dtype=mindspore.float32)
>>> y = Tensor([0.1, 0.2, 0.3], dtype=mindspore.float32)
>>> ans1 = gumbel_cdf.forward(x)
>>> print(ans1.shape)
(3,)
>>> ans2 = gumbel_cdf.inverse(y)
>>> print(ans2.shape)
(3,)
>>> ans3 = gumbel_cdf.forward_log_jacobian(x)
>>> print(ans3.shape)
(3,)
>>> ans4 = gumbel_cdf.inverse_log_jacobian(y)
>>> print(ans4.shape)
(3,)

property loc

Return the loc parameter of the bijector.

Returns

Tensor, the loc parameter of the bijector.

property scale

Return the scale parameter of the bijector.

Returns

Tensor, the scale parameter of the bijector.

forward(value)

forward mapping, compute the value after mapping as \(Y = g(X)\).

Parameters

• value (Tensor) - the value to compute.

Returns

Tensor, the value to compute.

forward_log_jacobian(value)

compute the log value after mapping, namely \(\log(dg(x) / dx)\).

Parameters

• value (Tensor) - the value to compute.

Returns

Tensor, the log value of forward mapping.

inverse(value)

Inverse mapping, compute the value after inverse mapping as \(X = g(X)\).

Parameters

• value (Tensor) - the value of output after mapping.

Returns

Tensor, the value of output after mapping.

inverse_log_jacobian(value)

Compute the log value of the inverse mapping, namely \(\log(dg^{-1}(x) / dx)\).

Parameters

• value (Tensor) - the value of output after mapping.

Returns

Tensor, the log value of the inverse mapping.