In that case, if your originally supplied list was D,
L=: ;{.D
P=: }.D
As for the extra boxing, consider this alternative:
;(C.~ <)~&.> ::]/|. D
Thanks,
--
Raul
On Sat, May 30, 2020 at 5:53 PM Hauke Rehr wrote:
>
> the verb would need to apply to L;P
> (in my current implementation s
the verb would need to apply to L;P
(in my current implementation since
that’s what the mentioned iteration
spits out)
I don’t understand why the boxed pairs
are boxed again in your last solution
but I guess I’ll figure that one out
I never learned ::, that might be the reason
maybe I should have
Hmm..
In the current J902 release, if your 50 element list was unboxed in L
and your pairs were boxes in P, you could do
;C.&.(a:`>) ::]/|. L;P
or
L ] F..(C. ::]) P
(Caution: lately I've been getting spelling errors, copying and
pasting from email into a J session -- this is because when
As I said, y is a boxed list the first item of which
contains some vector V which needs permutation.
All the other items contain a list of two values,
that’s why I said they represent transpositions.
I could as well have built their product and
then applied that permutation to the vector.
I only
Could you please give a sample of your use of your swap.
I can't understand your description or the input(s).
Thanks,
On Sat, May 30, 2020 at 3:27 AM Hauke Rehr wrote:
[snip]
My last use case was a boxed list consisting of
> a 1-d array and a set of transpositions I produced.
> I wrote (I bet
For something like your 'swap', having example arguments helps, so:
(i.3) swap 'abcd'
bcad
So your reduction swap&.>/@:|. suggests you were working with
permutation indices, possibly of the same length (though if that were
the case, there wouldn't have been any need for boxing -- but the
other
As with natural languages, it doesn’t get you fluent
gaining a large thesaurus only.
If you learn the grammar (there are two of them in J:
the set of production rules called a grammar; and the
natural language grammar terms used; here I am mostly
concerned with the latter) with the words, you’ll
b
