I would describe one as compositions and the other as array combinators (or array operators). In a Factor-like notation, the first (and also forks) would be dataflow combinators, although that applies just as easily to the second. The distinction between these two fits into a larger distinction between the two ways of automating things in J: using compositions or arrays. For example, you could say +&*: or +/@:*: to mean two similar operations, one on two arguments and one on an array. However, I haven't seen any material that clearly delineates these two methods or gives them any sort of terminology.
Marshall On Fri, Oct 21, 2011 at 6:43 PM, Alan Stebbens <a...@stebbens.org> wrote: > > Is there a better term, more respectable, or better known to the FP > > community, or more descriptive, or ... just better, than "slice and > dice"? > > Google calls their function "mapreduce", which is really f/. > > Others have called the f/ operation a "fold". > > see http://en.wikipedia.org/wiki/Fold_(higher-order_function) > > In fact, J is represented in the table about halfway down in the above > link. > > > > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
