The latenZy repository is a Python and MATLAB toolbox containing two non-parametric, binning-free methods for estimating the onset of neural spiking activity with high temporal precision: latenZy and latenZy2.
These methods leverage the cumulative distribution of spike times to detect time-locked changes in neural firing without relying on predefined time bins. Through an iterative process, they identify statistically significant changes in firing relative to the expected baseline activity—whether aligned to a single event or contrasting experimental conditions. This framework provides robust, data-driven estimates of when neural activity begins to change, without assuming any specific response pattern.
If you use latenZy in your work, please cite the paper: Haak R., Heimel J. A. (2025). LatenZy: non-parametric, binning-free estimation of latencies from neural spiking data. Journal of Neurophysiology. https://doi.org/10.1152/jn.00332.2025