Devon, Yes, I think the creep problem is related to the different cardinality of the x and y variables. To see this I tried the following with my changed line.
'hist' plot ((8+]);*:) i.5 By the way, your example made me aware for the first time that hist is a valid type for plot. On Mon, Feb 9, 2015 at 10:53 AM, Devon McCormick <devon...@gmail.com> wrote: > My example is too complicated and uses my own verb to compute the histogram > buckets based on frequencies. When I try to simplify the example, it makes > me think that the problem I'm seeing is introduced by my own code - it's > some kind of "off-by-one" error. > > The simple example I came up with to illustrate the problem is this: > > 'hist;xlabel 0 1 2 3 4 5' plot i.5 > > but this mis-match between the labels and the data items is a specification > mis-match: there are six labels but only five data items. Anyway, this > looks like the problem I'm seeing, so the underlying cause is probably the > same. I think I probably define n+1 buckets for n data items. > > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
