lava.lib.dnf.inputs.gauss_pattern

lava.lib.dnf.inputs.gauss_pattern.models

class lava.lib.dnf.inputs.gauss_pattern.models.GaussPatternProcessModel(proc_params=None)

Bases: PyLoihiProcessModel

PyLoihiProcessModel for GaussPatternProcess.

Implements the behavior of sending a gauss pattern asynchronously when a change is triggered.

a_out: PyOutPort = LavaPyType(cls=<class 'lava.magma.core.model.py.ports.PyOutPortVectorDense'>, d_type=<class 'float'>, precision=None)

changed: ndarray = LavaPyType(cls=<class 'numpy.ndarray'>, d_type=<class 'bool'>, precision=None)

implements_process: alias of GaussPattern

implements_protocol: alias of LoihiProtocol

null_pattern: ndarray = LavaPyType(cls=<class 'numpy.ndarray'>, d_type=<class 'float'>, precision=None)

pattern: ndarray = LavaPyType(cls=<class 'numpy.ndarray'>, d_type=<class 'float'>, precision=None)

required_resources: ty.List[ty.Type[AbstractResource]] = [<class 'lava.magma.core.resources.CPU'>]

run_spk()

Function that runs in Spiking Phase

Return type: None

lava.lib.dnf.inputs.gauss_pattern.process

class lava.lib.dnf.inputs.gauss_pattern.process.GaussPattern(**kwargs)

Bases: AbstractProcess

Gauss pattern generator Process.

This process generates Gauss patterns and send them through the OutPort a_out. It recomputes new patterns and sends them asynchronously only when one of the parameters amplitude, mean or stddev changes. Otherwise, sends an array full of numpy.nan.

Parameters:

shape: tuple(int): number of neurons per dimension, e.g. shape=(30, 40)
amplitude: float: amplitude of the Gauss pattern
mean: list(float) or float: mean of the Gauss pattern
stddev: list(float) or float: standard deviation of the Gauss pattern

property amplitude: Optional[ndarray]

Get value of the amplitude Var

Returns: amplitude
Return type: numpy.ndarray

property mean: Optional[ndarray]

Get value of the mean Var

Returns: mean
Return type: numpy.ndarray

property shape: Optional[ndarray]

Get value of the shape Var

Returns: shape
Return type: numpy.ndarray

property stddev: Optional[ndarray]

Get value of the stddev Var

Returns: stddev
Return type: numpy.ndarray