The Picture output of alpha-g-pad-gain-calibration is a bit misleading, but I currently have no idea on how to make it better.
Currently I only show ~300 channels at random because PGFPlots is not able to keep up plotting all 18432 cumulative distributions at once. It would be better to display something else (maybe a distribution on the K.S. distance) where it could be shown that the calibration had a positive effect.
As I see it, the options are:
Do some custom weird things to fake plotting all distributions. For example removing points that are too close to other points.
Use something else to plot these.
Use a different metric. For example a distribution of the K.S. distances to the chosen pivot.
The
Pictureoutput ofalpha-g-pad-gain-calibrationis a bit misleading, but I currently have no idea on how to make it better.Currently I only show
~300channels at random becausePGFPlotsis not able to keep up plotting all18432cumulative distributions at once. It would be better to display something else (maybe a distribution on the K.S. distance) where it could be shown that the calibration had a positive effect.As I see it, the options are: