|View source on GitHub|
Create a random variable for Distribution.
tfp.edward2.Distribution( *args, **kwargs )
See Distribution for more details.
Original Docstring for Distribution
This is a private method for subclass use.
dtype: The type of the event samples.
Noneimplies no type-enforcement.
reparameterization_type: Instance of
tfd.FULLY_REPARAMETERIZED, then samples from the distribution are fully reparameterized, and straight-through gradients are supported. If
tfd.NOT_REPARAMETERIZED, then samples from the distribution are not fully reparameterized, and straight-through gradients are either partially unsupported or are not supported at all.
Truedistribution parameters are checked for validity despite possibly degrading runtime performance. When
Falseinvalid inputs may silently render incorrect outputs.
True, statistics (e.g., mean, mode, variance) use the value "
NaN" to indicate the result is undefined. When
False, an exception is raised if one or more of the statistic's batch members are undefined.
dictof parameters used to instantiate this
listof graph prerequisites of this
strname prefixed to Ops created by this class. Default: subclass name.
ValueError: if any member of graph_parents is
Noneor not a