Bug in moving average (Generate-->Moving Average...)
Posted: Fri Dec 18, 2009 3:50 pm
There is an apparent bug in the moving average command. I noticed this when trying to smooth some [profile] gravity data. For monotonically increasing data, the moving average seems to always be less than the original curve.
Easy way to test: Generate a straight line (say, for 0 <= x <=10), then run the moving average over the generated data: the moving average is always less than the original data (except for the case x=0).
Easy way to test: Generate a straight line (say, for 0 <= x <=10), then run the moving average over the generated data: the moving average is always less than the original data (except for the case x=0).