lava.lib.dnf.inputs.rate_code_spike_gen

lava.lib.dnf.inputs.rate_code_spike_gen.models

class lava.lib.dnf.inputs.rate_code_spike_gen.models.RateCodeSpikeGenProcessModel(proc_params)

Bases: PyLoihiProcessModel

PyLoihiProcessModel for SpikeGeneratorProcess.

Implements the behavior of a rate-coded spike input generator.

a_in: PyInPort = LavaPyType(cls=<class 'lava.magma.core.model.py.ports.PyInPortVectorDense'>, d_type=<class 'float'>, precision=None)

first_spike_times: ndarray = LavaPyType(cls=<class 'numpy.ndarray'>, d_type=<class 'int'>, precision=None)

implements_process

alias of RateCodeSpikeGen

implements_protocol

alias of LoihiProtocol

inter_spike_distances: ndarray = LavaPyType(cls=<class 'numpy.ndarray'>, d_type=<class 'int'>, precision=None)

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

min_spike_rate: 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

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

seed: ndarray = LavaPyType(cls=<class 'numpy.ndarray'>, d_type=<class 'int'>, precision=None)

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

lava.lib.dnf.inputs.rate_code_spike_gen.process

class lava.lib.dnf.inputs.rate_code_spike_gen.process.RateCodeSpikeGen(**kwargs)

Bases: AbstractProcess

Spike generator Process for rate-coded input.

This process generates spike trains based on patterns it receives through its InPort a_in. It interprets these patterns as spiking rates (rate coding).

Receives a new pattern through a_in only once and while and trigger state update upon receipt of new pattern. In other time steps, receives null patterns (array full of numpy.nan). Sends spike values through its OutPort s_out every time step.

Parameters:

shape: tuple(int)

number of neurons per dimension, e.g. shape=(30, 40)

min_spike_rate: float

minimum spike rate (neurons with rates below this value will never spike)

seed: int

seed used for computing first spike times everytime pattern changes