Hmm... well... for example, if you want to take the implementation at http://www.volkerschatz.com/science/hastensor.html I think that the rank operator would be roughly described as a constraint on zips.
Really, though, the "problem" is not that the text is "too original" but that there are just so many flavors of jargon which get used to describe this kind of abstraction. (And, each carries with it a variety of other assumptions and issues.) Which, perhaps, is related to why math errors tend to creep into wikipedia writeups on math: everyone wants to use their own favorite way of describing things and as a result you tend to get revisions presenting different partially unrelated abstractions that only make sense when you go read the other works and figure out where the mistakes and missing pieces are. That said, I think it's fair to say that the this concept of "rank" was inspired by the concept of tensor operator rank (where, if I understand right, rank 0 is a scalar operator, rank 1 is a vector operator, rank 2 is a matrix operator, etc.) So maybe if you could find someone with a math background who also has a background in tensor operations, they could point you at some good, relevant references?? Anyways, ... just keep at it, I guess? Thanks, -- Raul On Wed, Sep 5, 2018 at 4:05 PM Devon McCormick <devon...@gmail.com> wrote: > > Thanks for all the responses. > > Raul: you are right as array programming without rank also lets you avoid > loops; rank refines how one slices an array. > > What I'm looking for are works I can cite on the Wikipedia page. As it > stands, it's too free-form and and too original which is not what the > Wikipedia guidelines specify. I will resort to things like the J site or a > Mathematica help page but I'm concerned about the potential drawbacks of > citing a web-page that may change or disappear as opposed to a published > work. Also, now that I look at the Mathematica page for "map", it does not > seem to be a good fit as their map looks very vectorish, not generally > array-oriented, so it's not clear that rank applies. I may have to remove > these claims unless I find a more apt source. > > Pascal: thanks for pointing out the problem with /. I'll fix that. > > Part of the problem is that most other languages' references to "array" > really mean "numeric vector", at least in my experience. > > > > On Wed, Sep 5, 2018 at 2:52 PM 'Pascal Jasmin' via Programming < > programm...@jsoftware.com> wrote: > > > the wikipage mislabels the / operator as map as well (it is in fact a > > reduce operator) > > > > The functional map function replaces a "for each" loop. Rank is very > > closely tied to this, with the addition of specifying the size/slice/scope > > of the "each element" group > > > > > > On Wednesday, September 5, 2018, 1:56:33 p.m. EDT, Cliff Reiter < > > reit...@lafayette.edu> wrote: > > > > > > > > > > > > From Mathematica help: > > > > Map (Built-in Symbol) > > Map[f, expr] or f /@ expr applies f to each element on the first level > > in expr. Map[f, expr, levelspec] applies f to parts of expr specified by > > levelspec. Map[f] represents\[Ellipsis] > > > > > > On 9/5/2018 10:58 AM, devonmcc wrote: > > > I am trying to bring the Wikipedia page on "rank (J programming > > language)" into compliance with the guidelines by inserting more inline > > references but am stumped by that page's initial assertion that rank is a > > generalization of looping, as well as the subsequent claims about rank > > generalizing "map" and "maplist". > > > Searches for a word like "rank" bring up many irrelevant hits even when > > included with programming-specific terms. Also, the best reference I have > > found so far - www.ccs.neu.edu/home/jrslepak/typed-j.pdf - uses language > > like "...a form of rank...which renders much of a program's control > > structure implicit by lifting base operators to higher-dimensional array > > structures." I will have to include this in the references to use it but > > was hoping to find an older document, preferably one already in the > > references, but have been stymied by the lack of explicit reference to > > avoiding loops by using rank because it seems so obvious that it is only > > implied. > > > In any case, a reference to the assertions about "map" and such would > > also be appreciated. > > > Thanks, > > > Devon > > > Sent from my Verizon, Samsung Galaxy smartphone > > > ---------------------------------------------------------------------- > > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > > -- > > Devon McCormick, CFA > > Quantitative Consultant > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
