Hello,
The mean error is calculated as being RMS/sqrt(no. of entries). My problem is that I need the mean error from a histogram which is obtained by subtraction from other histogram, and in some cases I have negative entries, and as a consequence the mean error is nan.
Is there a common accepted solution for this cases, or should I just disregard this histogram?
Thank you,
Angela.
p.s. I’m using root version 4.04/02.