It looks like the idea is to defer the evaluation of the name 'windowsize'
until invocation of 'npixels'. So, for instance, if 'windowsize' were a
global variable that was assigned dynamically - at runtime - 'npixels'
takes the current value of that noun at that time? OK, I'm not sure that
differs from how I understand current behavior.
On Thu, Aug 23, 2018 at 10:22 PM Brian Schott <schott.br...@gmail.com>
wrote:
> Henry, I'm still a little lost.
> Is the final `npixels` a (monadic, in this case) verb which will take a
> noun argument,
> or is the current value of windowsize that noun.
>
> If the latter, and windowsize is that noun, then what is the value of y in
> the original expression `[x] ".@'name' y`?
>
> If the former (`npixels` is a verb -- and this is the one I'm leaning
> toward), then what is the y value that is its argument?
>
> TIA.
>
> On Thu, Aug 23, 2018 at 9:41 PM, Henry Rich <henryhr...@gmail.com> wrote:
>
> > Often you need for a verb to refer to the value of a name when the noun
> is
> > executed, rather than when it is defined. So, rather than
> >
> > npixels =: */ windowsize
> >
> > which would just be a noun, you write
> >
> > npixels =: 3 : '*/ windowsize'
> >
> > to defer taking the value of (windowsize) until (npixels) is executed.
> >
> > In tacit code, you used to write
> >
> > npixels =: */ @: ". @ ('windowsize'"_)
> >
> > to get the same result. Recently the syntax was added to allow
> >
> > npixels =: */ @: ". @ 'windowsize'
> >
> > All that still works. But the (".) verb starts the parser and executes
> > its y argument as a sentence, with a modest overhead that will matter if
> > the verbs have small ranks. Now, you can write
> >
> > npixels =: */ @: (". @ 'windowsize')
> >
> > [parentheses important! make sure you see why]. When (npixels) is
> > defined, this form will detect that the argument to (".) is a single
> name.
> > When (npixels) is executed, the name will be evaluated without going
> > through the overhead of starting a sentence.
> >
> > Henry Rich
> >
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
--
Devon McCormick, CFA
Quantitative Consultant
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
