Metric to detect upward going action potentials

[emilymonroe95]

Is your feature request related to a problem? Please describe. When detecting negative going spikes, the peak offset metric works well for detecting other large negative deflections on the peak channel that are not the triggering event. This metric, though, does not help to detect large positive going action potentials. Trying to use peak offset to detect positive peaks results in almost every unit that having a peak offset value and is not useful for determining the relative amplitude of the positive going peak to the triggering negative event.

Describe the solution you'd like A metric that would let the user know if there were any peaks larger in "amplitude" than the triggering peak, whether they are positive or negative.

Other Solutions: It seems like pre-peak offset, the idea was to use the waveform misalignment function. Rhino's description was that this would basically take the amplitude of the waveform at the center index and divide it by the maximum waveform amplitude (positive or negative). If the peak was at time 0, this value would be 1. If there was a negative going offset peak, this value would be positive, but less than 1. If the highest amplitude were positive, this would result in a negative value. Therefore, the values would range from -integer to 1 and would be easy to threshold, theoretically. We should consider a different metric, similar to this, that would allow for detection of any offset, larger amplitude peak, whether it is positive or negative going.

Additional context @dpeg22 and I thought through this a bit, and he may be able to give some additional context to why peak offset was adopted rather than waveform misalignment/ my misunderstandings of the theoretical function versus how it was actually implemented.

See screenshots for examples