Postby albertc » 21 May 2009
Sorry, my mistake. I found out that the nested WAverage DOES indeed work, albeit VERY slowly. The code for WAverage is very inefficient so when the smoothing lengths are large, it takes a long time for the calculations to complete.
In the end, I got around this issue by writing a fast execution weighted moving average algorithm. Since a WMA is a convolution, the concept for the code is similar to that of SummationFC.