spark.core.payloads

spark.core.payloads#

Classes#

`SparkPayload`	Abstract payload definition to validate exchanges between SparkModule's.
`SpikeArray`	Representation of a collection of spike events.
`ValueSparkPayload`	Abstract payload definition to single value payloads.
`CurrentArray`	Representation of a collection of currents.
`PotentialArray`	Representation of a collection of membrane potentials.
`BooleanMask`	Representation of an inhibitory boolean mask.
`IntegerMask`	Representation of an integer mask.
`FloatArray`	Representation of a float array.
`IntegerArray`	Representation of an integer array.

Module Contents#

class spark.core.payloads.SparkPayload[source]#

Bases: abc.ABC

Abstract payload definition to validate exchanges between SparkModule’s.

tree_flatten()[source]#

Return type:: tuple[tuple[jax.Array], None]

classmethod tree_unflatten(aux_data, children)[source]#

Return type:: Self

property shape: Any[source]#

Return type:: Any

property dtype: Any[source]#

Return type:: Any

class spark.core.payloads.SpikeArray(spikes, inhibition_mask=False, async_spikes=False)[source]#

Bases: SparkPayload

Representation of a collection of spike events.

Init:: spikes: jax.Array[bool], True if neuron spiked, False otherwise inhibition_mask: jax.Array[bool], True if neuron is inhibitory, False otherwise

The async_spikes flag is automatically set True by delay mechanisms that perform neuron-to-neuron specific delays. Note that when async_spikes is True the shape of the spikes changes from (origin_units,) to (origin_units, target_units). This is important when implementing new synaptic models, since fully valid synaptic models should be able to handle both cases.

Parameters:

spikes (jax.Array)
inhibition_mask (jax.Array)
async_spikes (bool)

async_spikes: bool = False[source]#

tree_flatten()[source]#

Return type:: tuple[tuple, tuple]

classmethod tree_unflatten(aux_data, children)[source]#

Return type:: Self

__jax_array__()[source]#

Return type:: jax.numpy.ndarray

__array__(dtype=None)[source]#

Return type:: numpy.ndarray

property spikes: jax.Array[source]#

Return type:: jax.Array

property inhibition_mask: jax.Array[source]#

Return type:: jax.Array

property value: jax.Array[source]#

Return type:: jax.Array

property shape: tuple[int, Ellipsis][source]#

Return type:: tuple[int, Ellipsis]

property dtype: jax.typing.DTypeLike[source]#

Return type:: jax.typing.DTypeLike

class spark.core.payloads.ValueSparkPayload[source]#

Bases: SparkPayload, abc.ABC

Abstract payload definition to single value payloads.

value: jax.numpy.ndarray[source]#

__jax_array__()[source]#

Return type:: jax.numpy.ndarray

__array__(dtype=None)[source]#

Return type:: numpy.ndarray

tree_flatten()[source]#

Return type:: tuple[tuple, tuple]

classmethod tree_unflatten(aux_data, children)[source]#

Return type:: Self

property shape: tuple[int, Ellipsis][source]#

Return type:: tuple[int, Ellipsis]

property dtype: jax.typing.DTypeLike[source]#

Return type:: jax.typing.DTypeLike

class spark.core.payloads.CurrentArray[source]#

Bases: ValueSparkPayload

Representation of a collection of currents.

class spark.core.payloads.PotentialArray[source]#

Bases: ValueSparkPayload

Representation of a collection of membrane potentials.

class spark.core.payloads.BooleanMask[source]#

Bases: ValueSparkPayload

Representation of an inhibitory boolean mask.

class spark.core.payloads.IntegerMask[source]#

Bases: ValueSparkPayload

Representation of an integer mask.

class spark.core.payloads.FloatArray[source]#

Bases: ValueSparkPayload

Representation of a float array.

class spark.core.payloads.IntegerArray[source]#

Bases: ValueSparkPayload

Representation of an integer array.