JaxGaussianDistributionSample¶
- class probly.representation.distribution.jax_gaussian.JaxGaussianDistributionSample(array: D, sample_axis: int, weights: jax.Array | None = None)[source]¶
Bases:
GaussianDistributionSample[JaxGaussianDistribution],JaxSample[JaxGaussianDistribution]Sample type for empirical second-order Gaussian distributions.
- property T: ShapeResultT¶
Inverts the order of the axes of the underlying array.
- array¶
- astype(dtype: DTypeLike | None, copy: bool = False, device: Device | Sharding | None = None) Self[source]¶
Cast the underlying array to a new data type.
- Parameters:
dtype – The target data type.
copy – Whether to always return a copy.
device – The device the result should live on.
- Returns:
A new JaxSample with the cast array.
- block_until_ready() Self[source]¶
Block until the asynchronous computation of the underlying arrays has finished.
- Returns:
The sample itself.
- property dtype: DTypeLike¶
The data type of the underlying array.
- flatten(order: str = 'C') ShapeResultT[source]¶
Return a flattened version of the array.
- Parameters:
order – The read/write element order.
- Returns:
The flattened array.
- classmethod from_iterable(samples: Iterable[D], weights: Iterable[float] | None = None, sample_axis: SampleAxis = 'auto', dtype: DTypeLike | None = None) Self[source]¶
Create an JaxSample from a sequence of samples.
- Parameters:
samples – The predictions to create the sample from.
weights – Optional weights for the samples.
sample_axis – The dimension along which samples are organized.
dtype – Desired data type of the array.
- Returns:
The created JaxSample.
- classmethod from_sample(sample: Sample[D], sample_axis: SampleAxis = 'auto', dtype: DTypeLike | None = None) Self[source]¶
Create a new Sample from an existing Sample.
- Parameters:
sample – The sample to create the new sample from.
sample_axis – The dimension along which samples are organized.
- Returns:
The created Sample.
- item(*args: int) bool | int | float | complex[source]¶
Return the array as a Python scalar.
- Parameters:
*args – Optional index of the element to return.
- Returns:
The selected element as a Python scalar.
- property mH: ShapeResultT¶
The adjoint (conjugate) transposed version of the underlying array.
- property mT: ShapeResultT¶
The version of the underlying array with its last two axes swapped.
- move_sample_axis(new_sample_axis: int) JaxSample[D][source]¶
Return a new JaxSample with the sample dimension moved to new_sample_axis.
- Parameters:
new_sample_axis – The new sample dimension.
- Returns:
A new JaxSample with the sample dimension moved.
- pytree_children = ('array', 'weights')¶
- ravel(order: str = 'C') ShapeResultT[source]¶
Return a flattened version of the array.
- Parameters:
order – The read/write element order.
- Returns:
The flattened array.
- reshape(*args: int | Sequence[int], order: str = 'C') ShapeResultT[source]¶
Return a reshaped version of the array.
- Parameters:
*args – The target shape, either as a single sequence or as separate integers.
order – The read/write element order.
- Returns:
The reshaped array.
- sample_axis¶
- sample_space[source]¶
alias of
JaxGaussianDistribution
- property samples: D¶
Return an iterator over the samples.
- property sharding: Sharding¶
The sharding of the underlying array.
- squeeze(axis: int | Sequence[int] | None = None) ShapeResultT[source]¶
Return the array with axes of length one removed.
- Parameters:
axis – The axis or axes to remove. If omitted, all length-one axes are removed.
- Returns:
The squeezed array.
- stop_gradient() Self[source]¶
Return a copy detached from the autodiff graph.
Subclasses holding NumPy sidecar fields should override this to leave those fields untouched.
- Returns:
A copy through which gradients do not propagate.
- to_device(device: Device | Sharding, /, *, stream: int | Any | None = None) Self[source]¶
Move the underlying array to the specified device.
- Parameters:
device – The target device.
stream – not implemented, passing a non-None value will lead to an error.
- Returns:
A new JaxSample on the specified device.
- Raises:
NotImplementedError – If a stream is given.
- transpose(*args: int | Sequence[int] | None) ShapeResultT[source]¶
Return a transposed version of the array.
- Parameters:
*args – The permutation of the axes, either as a single sequence or as separate integers. If omitted, the axes are reversed.
- Returns:
The transposed array.
- tree_flatten() tuple[tuple[Any, ...], Any][source]¶
Split the object into pytree children and static auxiliary data.
Only fields declared in
pytree_childrenbecome children, regardless of their values. All other fields become static auxiliary data. Subclasses may override this together withtree_unflatten()to provide different transformation semantics. NumPy array children are rejected by default, including for native tree utilities; opaque eager mapping does not flatten.- Returns:
The children and the auxiliary data.
- Raises:
TypeError – If the schema, dynamic fields, or static metadata is unsupported.
ValueError – If child field names are repeated or unknown.
- tree_map_traverse = False¶
- classmethod tree_unflatten(aux_data: Any, children: tuple[Any, ...]) Self[source]¶
Rebuild an object from pytree children and auxiliary data.
Neither
__init__nor__post_init__is called: JAX may provide tracers, placeholders, or cotangents that violate primal-value invariants.- Parameters:
aux_data – The auxiliary data produced by
tree_flatten().children – The children produced by
tree_flatten().
- Returns:
The reconstructed object.
- view(dtype: DTypeLike | None = None, type: None = None) Self[source]¶
Return a bit-cast view of the array.
- Parameters:
dtype – The data type to reinterpret the underlying bytes as.
type – Unsupported by JAX, must be
None.
- Returns:
A copy with every array-valued field bit-cast to the given data type.
- Raises:
NotImplementedError – If
typeis not None.
- weights¶