Create a random variable for MultivariateStudentTLinearOperator.

See MultivariateStudentTLinearOperator for more details.


Original Docstring for Distribution

Construct Multivariate Student's t-distribution on R^k.

The batch_shape is the broadcast shape between df, loc and scale arguments.

The event_shape is given by last dimension of the matrix implied by scale. The last dimension of loc must broadcast with this.

Additional leading dimensions (if any) will index batches.

df A positive floating-point Tensor. Has shape [B1, ..., Bb] where `b

= 0. </td> </tr><tr> <td>loc</td> <td> Floating-pointTensor. Has shape[B1, ..., Bb, k]wherekis the event size. </td> </tr><tr> <td>scale</td> <td> Instance ofLinearOperatorwith a floatingdtypeand shape[B1, ..., Bb, k, k]. </td> </tr><tr> <td>validate_args</td> <td> Pythonbool, defaultFalse. Whether to validate input with asserts. Ifvalidate_argsisFalse, and the inputs are invalid, correct behavior is not guaranteed. </td> </tr><tr> <td>allow_nan_stats</td> <td> Pythonbool, defaultTrue. IfFalse, raise an exception if a statistic (e.g. mean/variance/etc...) is undefined for any batch member IfTrue, batch members with valid parameters leading to undefined statistics will return NaN for this statistic. </td> </tr><tr> <td>name`

The name to give Ops created by the initializer.

TypeError if not scale.dtype.is_floating.
ValueError if not scale.is_non_singular.