Bo is correct. The classic textbook calculation would be as follows to symmetrically pad the shortened list. At least this calculation lines up the correct pairs, but instead of zeros there would be "na" symbols.
a,:0 0 0,7(+/%#)\a On Thu, Jun 10, 2010 at 2:38 AM, Bo Jacoby <[email protected]> wrote: > I don't think that the average calculation produces the zeros. > > a=.?20#100 > a > 89 65 84 30 49 30 60 87 69 28 31 81 7 42 23 45 95 20 23 87 > 7(+/%#)\a > 58.1429 57.8571 58.4286 50.4286 50.5714 55.1429 51.8571 49.2857 40.1429 > 36.7143 46.2857 44.7143 36.4286 47.8571 > > but the shorter row of data is padded in order to fit to the longer row of > data > > a,:7(+/%#)\a ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
