thanks Raul. 44 characters now.
}.|:1j1#a.{~((]+(i.25)|.~-)(98-33^96>]))3&u:'s'
e.g. testing, testing,. ...
}.|:1j1#a.{~((]+(i.25)|.~-)(98-33^96>]))3&u:'J'
k l m n o p q r s t u v w x y a b c d e f g h i
}.|:1j1#a.{~((]+(i.25)|.~-)(98-33^96>]))3&u:'J'
K L M N O P Q R S T U V W X Y A B C D E F G H I
NB. I introduced an issue in the previous I posted. This one appear
On 30 August 2012 10:45, Raul Miller <rauldmil...@gmail.com> wrote:
> On Thu, Aug 30, 2012 at 5:05 AM, Linda Alvord <lindaalv...@verizon.net>
> wrote:
> > This seems to be the shortest version at the moment.
> >
> > |:1j1#(25{.(u:,2#65 97+/i.26)(>:@i.}.[)])"0 's'
> >
> > If I apply Raul's explanation to the current shortest version, why
> doesn't
> > the following function work?
> >
> > f=: 13 :'|:1j1#(25{.x (>:@i.}.x)y)"0'
> > f=: 13 :'|:1j1#(25{.x (>:@i.}.x)y)"0'
> > (u:,2#65 97+/i.26) f 's'
> > |domain error: f
> > | |:1j1#(25{.x( >:@i.}.x)y)"0
>
> One issue here is >:@i. -- that was the left tine of a fork, but by
> replacing the right tine of that fork you no longer have a fork, you
> have a noun phrase.
>
> But if you replace that with (x >:@i. y) you will then get a syntax
> error...
>
> Here's a cleaned up version of f that is syntactically valid:
> f=: 13 :'|:1j1#(25{.((x>:@i.y)}.x))'
>
> > I didn't have much luck with Raul's definition either:
> >
> > 1j1#"1&|:(25{.]}.~1+i.~)&(u:,2#65 97+/i.26)"0 'L'
> > M N O P Q R S T U V W X Y Z A B C D E F G H I J K
> >
> > g=: 13 :'1j1#"1&|:(25{.y}.~1+i.~)&x"0'
> > (u:,2#65 97+/i.26) f 'L'
> > |syntax error: f
>
> First, you should be using g here, instead of f
>
> But the syntax error is because the result of the sentence you used
> for g's definition is a verb. It's a verb because of the trailing "0
> on your (25{.y}.~1+i.~)&x means that (25{.y}.~1+i.~)&x becomes the
> definition for a rank 0 verb. And the rest of the expression to the
> left of it is suitable to turn the whole thing into a fork. So if you
> had defined:
> g=: 2 :'1j1#"1&|:(25{.y}.~1+i.~)&x"0'
> it would have succeeded without error.
>
> But that's not the only problem. Another problem is that you have an
> expression which reads ()&x and this will also be generating a verb,
> So just removing the "0 is not enough to make your 13 : definition
> compute a noun result.
>
> And there are other problems... but working with the definition split
> out from the arguments seems to me to be tedious -- if everything is
> on one line, I can hit control-shift-uparrow to recall my previous
> line and then simplify my test until I understand what's going on.
> With things on two lines that process becomes much slower.
>
> Anyways, here's what seems to me to be a near minimal set of edits to
> your g to make it do what I am expecting you want it to do:
>
> g=: 2 :'1j1#"1&|:(25{.x}.~1+y i.~x)'
>
> Note that I changed an x to a y and I removed things that made your
> result want to be a verb and I gave your i.~ its explicit arguments.
>
> I hope this helps, but I'm thinking that you are making blind changes
> to full sentences instead of experimenting with small phrases and
> building those into full sentences. If you make blind changes (for
> example, putting y where you meant x, but also things like using a
> verb where you meant to use a noun) the results will tend to be
> confusing. To unravel the confusing stuff, you are going to need a
> good set of practices, for isolating the parts of the code which are
> behaving differently from what you want.
>
> Take care,
>
> --
> 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
