I feel similar feature is covered by @ (At) from Dyalog APL,
as I can think of the definition for monadic "structural
under" using APL could roughly be
r←(f sunder g)y;i
i←,g⍳⍴y
r←(f@i)y
(⌽sunder,) 3 3⍴⍳9
9 8 7
6 5 4
3 2 1
ldbeth
>>>>> In <cagvhrk3cxvmkykxooa1a2aktoaw5ns+yc2shbsx6mjnilrv...@mail.gmail.com>
>>>>> Henry Rich <henryhr...@gmail.com> wrote:
Henry> This seems like a useful feature and not difficult to implement. We can
Henry> discuss here what should be supported.
Henry> Morten's reservations (in the cited article) are reasonable. I note that
J
Henry> has semiduals, and that it would make sense to support structural under
for
Henry> dyads using semiduals.
Henry> Henry Rich
Henry> On Sat, Oct 14, 2023, 6:10 AM LdBeth <andp...@foxmail.com> wrote:
Viktor> Why does
Viktor> (|. &. ,) i.3 3
Viktor> result in a domain error (likewise with &.:), while
Viktor> ($ $ (|. @ ,) ) i.3 3
Viktor> doesn't?
ldb> The domain error is raised by ,^:_1 and as in J and APL
ldb> the inverse of ravel is not defined in present time.
ldb> This is not the first time people asking about this feature
ldb> and the APL community have given this a name "structural under"
ldb> https://www.dyalog.com/blog/2023/01/structural-vs-mathematical-under/
>> ... an important reason was that Roger Hui wasn’t a big fan of the
>> proposed extension known as “structural under”
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
