I detected a problem in the way that the mean is computed: instead of the running average it is more correct to compute the mean of the samples. The standard deviation was under estimated.
It does not change the results much but it can make the difference in edge cases.
I detected a problem in the way that the mean is computed: instead of the running average it is more correct to compute the mean of the samples. The standard deviation was under estimated.
It does not change the results much but it can make the difference in edge cases.