Threshold Based Representation Algorithm

threshold_based_representation(signal: ndarray, factor: float | list[float]) → ndarray[source]

Perform Threshold-Based Representation (TBR) encoding on the input signal.

This function takes a continuous signal and converts it into a spike train using a fixed threshold based on the signal’s variations. A spike is generated when the variation exceeds the computed threshold.

Refer to the Threshold-Based Representation (TBR) Encoding for a detailed explanation of the TBR encoding algorithm.

Code Example:

import numpy as np
from spikify.encoding.temporal.contrast import threshold_based_representation
signal = np.array([0.1, 0.3, 0.4, 0.2, 0.5, 0.6])
factor = 0.5
encoded_signal = threshold_based_representation(signal, factor)
Parameters:

  • signal (numpy.ndarray) – The input signal to be encoded. This should be a numpy ndarray.

  • factor (float | list[float]) – The factor value (γ) that controls the noise-reduction threshold. Can be a float or a list of floats.

Returns:

A numpy array representing the encoded spike train.

Return type:

numpy.ndarray

Raises:

  • ValueError – If the input signal is empty.

  • TypeError – If the signal is not a numpy ndarray.