Using fry is convenient. Due to how we bootstrap factor, we can't use fry
right now in "core" vocabularies, like sequences, so you'll see a fair
amount of curry and compose which are more or less what fry is doing, just
not as clean syntax.
I love the REPL advent idea, go for it! And as you learn or hit roadblocks
or have questions, please share.
You can see how supremum-by is implemented by clicking on it in an
expression, or doing ``see``. It uses ``after?`` instead of ``>``, which
allows it to compare other kinds of objects with ``<=>`` when numbers are
not passed in.
Also, you might try using ``map-reduce`` which is a little faster than
using ``unclip-slice`` and ``reduce``, (and using ``_ bi@`` instead of ``[
_ call ] bi@`` to simplify):
: max-by ( seq quot -- result )
[ ] swap '[ [ _ bi@ > ] 2keep ? ] map-reduce ; inline
Best,
John.
On Mon, Dec 10, 2018 at 6:46 AM Alexander Ilin <[email protected]> wrote:
> Hey there!
>
> Sure, you can ask us here. I'm on sick leave, so got plenty of time.
> Coincidentally, working on my Factor hobby project.
>
> What you're looking for is called supremum-by.
>
> Terminology is a bit uncommon. Look up supremum and infimum in the help
> system.
>
> 10.12.2018, 17:06, "[email protected]" <[email protected]>:
> > Hello fellow concatenative fans.
> >
> > I'm working my way through this year's advent of code[0]. I'm doing it
> > in factor with a small twist - everything happening inside the REPL.
> > Makes it more fun and challenging (for me), although I don't have any
> > code to show for it in the end.
> >
> > I'm trying to find the right tools for the job since there's a lot of
> > words already available. Sometimes I reinvent something before I find
> > it, e.g. I wrote a last-n before finding last*.
> >
> > Would it be OK if I sometimes ask you for tips on what would be
> > idiomatic/simpler/shorter/faster than what I have(n't) found?
> >
> > The first problem I really couldn't find is picking a maximum from a
> > sequence based on a quotation pulling out the key. An example will say
> > it the best I guess:
> >
> >> { { 1 2 } { 1 3 } { -1 4 } } [ second ] max-by .
> >
> > { -1 4 }
> >
> > Here is my definition:
> >
> > : max-by ( seq quot -- result ) [ unclip-slice ] dip '[ [ [ _ call ] bi@
> >> ] 2keep ? ] reduce ; inline
> >
> > Using fry for such a general word would be a bit of an overkill but for
> > demonstration purposes it's OK I guess.
> >
> > Is there no standard word that already does something similar? Or is
> > there a way to write this with the existing sequence combinators that
> > it's so short there was no need to create this generic word?
> >
> > --
> > ------------
> > Peter Nagy
> > ------------
> >
> > _______________________________________________
> > Factor-talk mailing list
> > [email protected]
> > https://lists.sourceforge.net/lists/listinfo/factor-talk
>
> ---=====---
> Александр
>
>
>
> _______________________________________________
> Factor-talk mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/factor-talk
>
_______________________________________________
Factor-talk mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/factor-talk
