a=. 2 3 4 5 6 7 8 9 10 +/ .(*"4)&i. 7 8 9 10 b=. ((*/2 3 4 5),6,(*/7 8 9 10)) +/ .(*"1)&i. */7 8 9 10 a-:b 0
That said, it might be interesting to replace +/ with +/@, -- Raul On Mon, May 30, 2022 at 2:25 AM Elijah Stone <elro...@elronnd.net> wrote: > > That's reinforcing my point--there are no more than three significant axes > there. > > a=. 2 3 4 5 6 7 8 9 10 +/ .(*"4)&i. 7 8 9 10 > b=. ((*/2 3 4 5),6,(*/7 8 9 10)) +/ .(*"1)&i. */7 8 9 10 > a -:&, b > 1 > > > On Mon, 30 May 2022, Raul Miller wrote: > > > On Mon, May 30, 2022 at 1:31 AM Elijah Stone <elro...@elronnd.net> wrote: > >> Rank is about projecting a 2- or 3-dimensional structure onto > >> multidimensional arrays. > > > > Those are common cases, but the concepts behind the notation do > > support operations like: > > > > $2 3 4 5 6 7 8 9 10 +/ .(*"4)&i. 7 8 9 10 > > 2 3 4 5 7 8 9 10 > > > > And, +/ there could be replaced with something rather different... > > > > That said, data sets with high dimensionality tend to run into machine > > limitations (some of which I believe sparse arrays were meant to > > address). > > > > Thanks, > > > > -- > > Raul > > ---------------------------------------------------------------------- > > 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
