aeppl.mixture.MixtureRV#

class aeppl.mixture.MixtureRV(indices_end_idx, out_dtype, out_broadcastable)[source]#

A placeholder used to specify a log-likelihood for a mixture sub-graph.

Methods

MixtureRV.L_op(inputs, outputs, output_grads)

Construct a graph for the L-operator.

MixtureRV.R_op(inputs, eval_points)

Construct a graph for the R-operator.

MixtureRV.__init__(indices_end_idx, ...)

MixtureRV.add_tag_trace([user_line])

Add tag.trace to a node or variable.

MixtureRV.do_constant_folding(fgraph, node)

Determine whether or not constant folding should be performed for the given node.

MixtureRV.get_params(node)

Try to get parameters for the Op when Op.params_type is set to a ParamsType.

MixtureRV.grad(inputs, output_grads)

Construct a graph for the gradient with respect to each input variable.

MixtureRV.make_node(*inputs)

Construct an Apply node that represent the application of this operation to the given inputs.

MixtureRV.make_py_thunk(node, storage_map, ...)

Make a Python thunk.

MixtureRV.make_thunk(node, storage_map, ...)

Create a thunk.

MixtureRV.perform(node, inputs, outputs)

Calculate the function on the inputs and put the variables in the output storage.

MixtureRV.prepare_node(node, storage_map, ...)

Make any special modifications that the Op needs before doing Op.make_thunk.

Attributes

default_output

An int that specifies which output Op.__call__ should return.

destroy_map

A dict that maps output indices to the input indices upon which they operate in-place.

itypes

otypes

params_type

view_map

A dict that maps output indices to the input indices of which they are a view.