mxnet.gluon.probability.distributions.normal¶

Normal distribution

Classes

Normal([loc, scale, validate_args])

Create a Normal distribution object.

class mxnet.gluon.probability.distributions.normal.Normal(loc=0.0, scale=1.0, validate_args=None)[source]¶

Create a Normal distribution object.

Parameters:

loc (Tensor or scalar, default 0) – mean of the distribution.
scale (Tensor or scalar, default 1) – standard deviation of the distribution

broadcast_to(batch_shape)[source]¶

Returns a new distribution instance with parameters expanded to batch_shape. This method calls numpy.broadcast_to on the parameters.

Parameters:: batch_shape (Tuple) – The batch shape of the desired distribution.

cdf(value)[source]¶: Returns the cumulative density/mass function evaluated at value.

entropy()[source]¶: Return the entropy of a distribution. The entropy of distributions in exponential families could be computed by: H(P) = F(theta) - <theta, F(theta)’> - E_p[k(x)]

icdf(value)[source]¶: Returns the inverse cumulative density/mass function evaluated at value.

log_prob(value)[source]¶

Compute the log likelihood of value.

sample(size=None)[source]¶

Generate samples of size from the normal distribution parameterized by self._loc and self._scale

Parameters:: size (Tuple, Scalar, or None) – Size of samples to be generated. If size=None, the output shape will be broadcast(loc, scale).shape
Returns:: Samples from Normal distribution.
Return type:: Tensor

sample_n(size=None)[source]¶

Generate samples of (batch_size + broadcast(loc, scale).shape) from the normal distribution parameterized by self._loc and self._scale

Parameters:: size (Tuple, Scalar, or None) – Size of independent batch to be generated from the distribution.
Returns:: Samples from Normal distribution.
Return type:: Tensor