NDI.FUN.PROBE.IMPORT.KILOSORT.MEANWAVEFORM - amplitude-weighted mean waveform for a cluster
MEANWF = NDI.FUN.PROBE.IMPORT.KILOSORT.MEANWAVEFORM(CID, SPIKE_CLUSTERS, ...
SPIKE_TEMPLATES, AMPLITUDES, TEMPLATES, WINV)
Computes the mean waveform (NumSamples x NumChannels) for curated cluster CID.
Because a curated cluster may span several kilosort templates (after merges), the
waveform is computed as the AMPLITUDE-WEIGHTED AVERAGE of every template that
contributes spikes to the cluster: each contributing template is weighted by the
sum of the spike amplitudes assigned to it within this cluster. The result is then
scaled by the cluster's mean spike amplitude so the waveform has a meaningful
magnitude, and, if an inverse whitening matrix WINV is provided, un-whitened into
(approximately) physical units.
Inputs:
CID - the cluster id to compute
SPIKE_CLUSTERS - cluster id of every spike
SPIKE_TEMPLATES - template id (0-based) of every spike
AMPLITUDES - amplitude of every spike
TEMPLATES - nTemplates x nSamples x nChannels template shapes
WINV - inverse whitening matrix (nChannels x nChannels) or []
See also: NDI.FUN.PROBE.IMPORT.KILOSORT.PROBE