ML Community Day is November 9! Join us for updates from TensorFlow, JAX, and more Learn more

tfp.experimental.tangent_spaces.FullSpace

Tangent space of M for distributions supported on all of R^n.

Inherits From: TangentSpace

This subclass covers full-rank distributions on n-dimensional manifolds. In this common case we can take the basis to be the standard basis for R^n, so we need not explicitly represent it at all.

Methods

transform_coordinatewise

View source

Same as transform_dimension_preserving, for a coordinatewise f.

Default falls back to transform_dimension_preserving, which may be overridden in subclasses.

Args
x same as in transform_dimension_preserving.
f same as in transform_dimension_preserving.

Returns
density A Tensor representing the density correction of f at x
space A TangentSpace representing the tangent to fM at f(x)

Raises
NotImplementedError if the TangentSpace subclass does not implement transform_dimension_preserving.

transform_dimension_preserving

View source

Same as transform_general, assuming f goes from R^n to R^n.

Default falls back to transform_general, which may be overridden in subclasses.

Args
x same as in transform_general.
f same as in transform_general.

Returns
density A Tensor representing the density correction of f at x
space A TangentSpace representing the tangent to fM at f(x)

Raises
NotImplementedError if the TangentSpace subclass does not implement transform_general.

transform_general

View source

If the bijector is weird, fall back to the general case.

transform_projection

View source

Same as transform_general, with f a projection on some coordinates.

Default falls back to transform_general, which may be overridden in subclasses.

Args
x same as in transform_general.
f same as in transform_general.

Returns
density A Tensor representing the density correction of f at x
space A TangentSpace representing the tangent to fM at f(x)

Raises
NotImplementedError if the TangentSpace subclass does not implement transform_general.