13 : '(1) x}y'  or  13 : '1 x}y'  can not be made tacit because they use a
function argument (x) as an operand.  Similarly 13 : '+/"y x' or 13 : '+/"x
y' can not be made tacit either.

Why not, you might ask?  The sequence for an operator is that you _first_
derive a function, and _then_ the function is applied to function arguments
x and/or y .  When you say x} or +/"x  or +/"y , at that point you don't
yet have the function argument.



On Fri, Jul 26, 2019 at 9:40 PM Devon McCormick <[email protected]> wrote:

> I am intrigued by tacit programming but agree with Henry that you should
> drink the Kool-Aid with slow, measured sips.
>
> I continue to rely heavily on "13 : .." (to the dismay of tacit die-hards)
> but still puzzle over why certain expressions cause it to give up, e.g.
>
>    13 : '(1) x}y'
> 4 : '(1) x}y'
>
> Yet this works
>    x=. 2 4 6
>    y=. i.10
>    (1) x}y
> 0 1 1 3 1 5 1 7 8 9
>
> Whereas these do not:
>    x ((1) [}]) y
> 0 1 2 3 4 5 6 7 8 9
>    x ((1)& [}]) y
> 0 6 2 3 4 5 6 7 8 9
>
>
> On Fri, Jul 26, 2019 at 11:53 PM Henry Rich <[email protected]> wrote:
>
> > Yes, you have to think differently.
> >
> > Please, don't make a fetish of tacit programming.  The important thing
> > about becoming a J programmer is training yourself to look at the whole
> > problem rather than working with atoms.  That is, getting over the loop
> > mindset.
> >
> > Problem: You have an array a and you want to apply a linear function
> > y=3x+1 to every atom.  How do you do it?
> >
> > 0. (3 * a) + 1
> > Bravo! You didn't have to loop.  Full marks.
> >
> > 1. 1 + 3 * a
> > Piu bravo!  You are thinking right to left.
> >
> > 2. (1 + 3&*) a
> > Bravissimo!  You have written a little bit of tacit code.  You could
> > also have written (1 + 3 * ]) a
> >
> > 3. 1 3 p. a
> > Congratulations, you have reached master level.  Let the interpreter
> > sweat the details.
> >
> >
> > My very first J program was a model of texture processing for a
> > computer-graphics system.  I imbibed the tacit liquor too deeply, and
> > wrote the whole thing as a sequence of tacit verbs, 50 of them.  It
> > worked.  It was absolutely unmaintainable.  Don't go there.
> >
> > Use tacit verbs for problems whose spec is mathematically unchangeable.
> > For everything else, use explicit verbs.
> >
> > Henry Rich
> >
> >
> >
> >
> >
> > On 7/26/2019 7:20 PM, The3DSquare Josh Yang wrote:
> > > I was wondering if you guys have any resources for learning tacit
> > > programming in J. I've been struggling with tacit programming, probably
> > > because all the other programming languages are mainly explicit.
> > Something
> > > like "Tacit programming for the explicit programmer" would be great
> > since I
> > > often fail to derive the tacit equivalent of a series of explicit
> > > eeaxpressions (especially when there are mutable variables).
> > >
> > > I know I need to think differently, but having some guidance would help
> > > greatly.
> > > ----------------------------------------------------------------------
> > > For information about J forums see http://www.jsoftware.com/forums.htm
> >
> >
> > ---
> > This email has been checked for viruses by AVG.
> > https://www.avg.com
> >
> > ----------------------------------------------------------------------
> > 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
>
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to