[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Pure (side-effect-free) calls into c/c++?
From: |
Linus Björnstam |
Subject: |
Re: Pure (side-effect-free) calls into c/c++? |
Date: |
Sun, 12 Jan 2020 11:35:30 +0100 |
User-agent: |
Cyrus-JMAP/3.1.7-740-g7d9d84e-fmstable-20200109v1 |
Thanks!
I haven't used this macro in a billion years and wrote it as a comfort thingie
for the first 100 project Euler problems.
This thread got me to start thinking about how to memoize "smarter" and having
a macro that allows you to trade a bit of speed for being able to specifically
memoize the last n or the most common n arguments.
--
Linus Björnstam
On Sun, 12 Jan 2020, at 04:03, Christopher Lam wrote:
> I can add a contribution! The good thing about memoize is it's simple
> to create. You forgot a catch however: if the memoized return-val is #f
> then your memoizer
> https://hg.sr.ht/~bjoli/misc/browse/default/memoize.scm will not
> recognise that #f is a valid cached return-val and will call the lambda
> again. (FWIW I shudder think what a *fast* memoizer would do).
>
> Here's how I did mine:
>
> (define (memoize f)
> (let ((h (make-hash-table)))
> (lambda args
> (cond
> ((hash-ref h args) => car)
> (else (let ((res (apply f args)))
> (hash-set! h args (list res))
> res))))))
>
> (define-syntax-rule (lambda/macro args body ...)
> (memoize (lambda args body ...)))
>
> (define-syntax-rule (define/macro (f . args) body ...)
> (define f lambda/macro args body ...))
>
>
>
> On Sat, 11 Jan 2020 at 17:40, Linus Björnstam
> <address@hidden> wrote:
> > I have a macro called lambda/memo and define/memo for these situations:
> > https://hg.sr.ht/~bjoli/misc/browse/default/memoize.scm
> >
> > If the function gets called with a gazillion different arguments the
> > memoizatiin hash gets large, and there are no mechanisms to stop that from
> > happening. It also lacks a fast path for single argument functions.
> >
> > You can disregard the repo license. Use that function is you like, if you
> > like to.
> >
> >
> > --
> > Linus Björnstam
> >
> > On Fri, 10 Jan 2020, at 23:36, Linas Vepstas wrote:
> > > So, I've got lots of C code wrapped up in guile, and I'd like to declare
> > > many of these functions to be pure functions, side-effect-free, thus
> > > hopefully garnering some optimizations. Is this possible? How would I do
> > > it? A cursory google-search reveals no clues.
> > >
> > > To recap, I've got functions f and g that call into c++, but are pure
> > (i.e.
> > > always return the same value for the same arguments). I've got
> > > user-written code that looks like this:
> > >
> > > (define (foo x)
> > > (g (f 42) (f x) (f 43))
> > >
> > > and from what I can tell, `f` is getting called three times whenever the
> > > user calls `foo`. I could tell the user to re-write their code to cache,
> > > manually: viz:
> > >
> > > (define c42 (f 42))
> > > (define c43 (f 43))
> > > (define (foo x) (g c42 (f x) c43))
> > >
> > > but asking the users to do this is .. cumbersome. And barely worth it:
> > `f`
> > > takes under maybe 10 microseconds to run; so most simple-minded caching
> > > stunts don't pay off. But since `foo` is called millions/billions of
> > times,
> > > I'm motivated to find something spiffy.
> > >
> > > Ideas? suggestions?
> > >
> > > -- Linas
> > > --
> > > cassette tapes - analog TV - film cameras - you
> > >
> >
Re: Pure (side-effect-free) calls into c/c++?, Linus Björnstam, 2020/01/11