bases

jaxns.framework.bases

Module Contents

class BaseAbstractPrior(name=None)[source]

Bases: jaxns.framework.abc.AbstractPrior

The base prior class with public methods.

Parameters:: name (Optional[str]) –

property dtype[source]: The dtype of the prior random variable in X-space.

property base_shape: Tuple[int, Ellipsis][source]

The base shape of the prior random variable in U-space.

Return type:: Tuple[int, Ellipsis]

property base_ndims[source]: The number of dimensions of the prior random variable in U-space.

property shape: Tuple[int, Ellipsis][source]

The shape of the prior random variable in X-space.

Return type:: Tuple[int, Ellipsis]

__repr__()[source]: Return repr(self).

forward(U)[source]

The forward transformation from U-space to X-space.

Parameters:: U (jaxns.internals.types.UType) – U-space representation
Returns:: X-space representation
Return type:: jaxns.internals.types.RandomVariableType

inverse(X)[source]

The inverse transformation from X-space to U-space.

Parameters:: X (jaxns.internals.types.RandomVariableType) – X-space representation
Returns:: U-space representation
Return type:: jaxns.internals.types.UType

log_prob(X)[source]

The log probability of the prior.

Parameters:: X (jaxns.internals.types.RandomVariableType) – X-space representation
Returns:: log probability of the prior
Return type:: jaxns.internals.types.MeasureType

PriorModelGen[source]

PriorModelType[source]

class BaseAbstractModel(prior_model, log_likelihood)[source]

Bases: jaxns.framework.abc.AbstractModel

The base model class with public methods.

Parameters:

prior_model (PriorModelType) –
log_likelihood (jaxns.internals.types.LikelihoodType) –

property prior_model: PriorModelType[source]

The prior model.

Return type:: PriorModelType

property log_likelihood: jaxns.internals.types.LikelihoodType[source]

The log likelihood function.

Returns:: log likelihood function
Return type:: jaxns.internals.types.LikelihoodType

property U_placeholder: jaxns.internals.types.UType[source]

A placeholder for U-space sample.

Return type:: jaxns.internals.types.UType

property X_placeholder: jaxns.internals.types.XType[source]

A placeholder for X-space sample.

Return type:: jaxns.internals.types.XType

property U_ndims: int[source]

The prior dimensionality.

Return type:: int

class BaseAbstractDistribution[source]

Bases: jaxns.framework.abc.AbstractDistribution

The base distribution class with public methods.

property dtype[source]: The dtype of the distribution, in X-space.