bijx.MixtureStack

class bijx.MixtureStack

Bases: Distribution

Mixture of distributions of equal kind.

See also bijx.ScanChain for a related construct except for composing bijections instead of mixing distributions.

__init__(distributions, weights, *, rngs=None)

Parameters:

• distributions (Distribution)

• weights (Variable | Array | ndarray | Sequence[int | Any])

Methods

density(x, **kwargs)	Evaluate probability density at given points.
get_batch_shape(x)	Extract batch dimensions from a sample.
log_density(x)	Evaluate log probability density at given points.
sample([batch_shape, rng])	Generate samples from the distribution.

Attributes

stack
weights_log_normalized
weights_normalized

property weights_normalized

property weights_log_normalized

property stack

get_batch_shape(x)

Extract batch dimensions from a sample.

Parameters:

x – A sample from this distribution.

Returns:

Tuple representing the batch dimensions of the sample.

log_density(x)

Evaluate log probability density at given points.

Parameters:

• x – Points at which to evaluate density, with event dimensions matching the distribution’s event shape.

• **kwargs – Additional distribution-specific evaluation arguments.

Returns:

Log density values with batch dimensions matching input.

sample(batch_shape=(), rng=None)

Generate samples from the distribution.

Parameters:

• batch_shape – Shape of batch dimensions for vectorized sampling.

• rng – Random key for sampling, or None to use internal rngs.

• **kwargs – Additional distribution-specific sampling arguments.

Returns:

Tuple of (samples, log_densities) where samples have shape (*batch_shape, *event_shape) and log_densities have shape batch_shape.