To expand on greg heil's suggestion, and to try and interpret what you suggested, it sounds like you want something like this:
original=: <"0 i.3 1000 roots=: %:each original logs=: ^. each original But you also talked about amending all of the values, so maybe you want original=: roots or original=: logs Though, obviously, if you do both of those updates to the 'original', one of them would overwrite the other. So, if that's not what you want, perhaps you could provide a concrete example of what you want using a smaller array? (It's straightforward to scale up an approach in size, but examples which fit in an email message need to be small enough to fit in an email message). Thanks, -- Raul On Fri, May 14, 2021 at 11:13 PM HH PackRat <hhpack...@gmail.com> wrote: > > Hello again! > > I'm thinking, for example, of a table of boxed data in its transposed > horizontal position that has, say, 3 rows of, say, 1000 *boxed* values > each. My goal is to amend *all* of the boxed values to their square > roots in one version of what I want to do and *all* of the original > boxed values to their logarithmic equivalents (maybe multiplied by a > constant, say, 100) in a second version . > > Using what I know from past answers to my questions, I'm sure I could > use "for" loops with repeated amends to accomplish this, but that > seems to be working against J's efficiency in dealing with large > quantities of data at once. > > How can I do such a mass "amend" of entire tables in one fell swoop in > J? Please show EXPLICIT code rather than tacit code. > > Thanks for any and all help! > > Harvey > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
