Thank you for these helpful explanations and pointers. I did have a feeling that "combinations" was the wrong term. The informal term "odometer" nicely captures the intuition here and the examples in the Odometers Essay provide excellent examples of how to do this sort of thing in J.
It is also interesting to note how J's various vector-processing constructs often permit several alternative solutions to the same problem, as seen here. And I now see that the "Antibase" dictionary entry did actually already have everything I needed - I should have run the example in stages, to see the intermediate results of invoking Antibase on Base. Thank you for this help and for making the J language available. J is one of the rare languages that is truly a pleasure to work with. Roger Hui wrote: What you are generating are usually not called "combinations" as order is not relevant in combinations, nor are duplications permitted in a combination. Different ways of generating the desired result are described in http://www.jsoftware.com/jwiki/Essays/Odometer *> As a side note - I am curious if there is a more compact way of representing* *> the repetitive phrase:* *> (i. 6);(i. 6);(i. 6) * i.&.> 6 6 6 3$<i.6 *> Upon first reading, it wasn't clear to me from the definition of dyad* *> Antibase provided in the Vocabulary here:* *> **http://www.jsoftware.com/help/dictionary/d402.htm* <http://www.jsoftware.com/help/dictionary/d402.htm> *> that its x argument could be an integer array each of whose * *> elements is the desired base. * The first example in the dictionary entry for the dyad #: uses an integer array left argument. ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
