jaxsnn.event.from_nir_data.TimeGriddedData

class jaxsnn.event.from_nir_data.TimeGriddedData(data: numpy.ndarray, dt: float)

Bases: object

Either boolean entries indicate whether a binary event is present at a particular time step, or a real-valued signal provides the measurement of a quantity (e.g. the membrane potential).

datanp.ndarray, shape (n_samples, n_time_steps, n_neurons): Input data. For binary data the dtype should be bool.
dt: float: Time step size.

__init__(data: numpy.ndarray, dt: float) → None: Initialize self. See help(type(self)) for accurate signature.

Methods

`__init__`(data, dt)	Initialize self.
`to_event`(n_events[, time_shift])	Arguments

Attributes

`n_neurons`
`n_samples`
`n_time_steps`
`shape`
`t_max`

data: numpy.ndarray

dt: float

property n_neurons

property n_samples

property n_time_steps

property shape

property t_max

to_event(n_events: int, time_shift: float = 0.0) → nir.data_ir.graph.EventData

n_spikesint: Maximum number of events stored for each neuron.
time_shiftfloat, optional: Shift the event times by this value from the beginning of each time step. Must be in interval [0, dt). Default is 0.0.