Question on the implementation of calc_waveform_metrics

[oaaij-gnahz]

Hi!

I was reading the implementation of calc_waveform_metrics, and on this line:
https://github.com/petersenpeter/CellExplorer/blob/master/calc_CellMetrics/calc_waveform_metrics.m#L24

It seems to be using a spline interpolation to adjust the sample timestamps of to-be-resampled waveform data.

However, it is simply interpolating the curve $y=x$ (the first two parameters passed to the function interp1 are the same timeWaveform).
I wonder in this case why would an interpolation or spline interpolation be needed? What is the consideration?

Thank you!

[petersenpeter]

Hi

Thanks for pointing this out. Indeed it was not necessary - it was a lazy implementation on my part. I am interpolating the waveforms using the same method a few lines below, so I had simply reused the line for the time-axis, which was not necessary. I altered the line to the following:
timeWaveform = timeWaveform(1):mean(diff(timeWaveform))/oversampling:timeWaveform(end);

I pushed the changes to GitHub.

[oaaij-gnahz]

Thanks for the clarification!