Re: structure and interpretation of computer programs exercsie 1.3 in forth

From	"WJ" <w_a_x_man@yahoo.com>
Newsgroups	comp.lang.forth
Subject	Re: structure and interpretation of computer programs exercsie 1.3 in forth
Date	2013-06-14 15:37 +0000
Organization	A noiseless patient Spider
Message-ID	<kpfdap$eqo$1@dont-email.me> (permalink)
References	(8 earlier) <e9ea8ae4-4b51-42cd-9a44-eca5ba9a8683@10g2000yqo.googlegroups.com> <ae2d5ea7-b58a-4a22-b929-f5a1331208f1@k6g2000yqf.googlegroups.com> <7xobgsp5u3.fsf@ruckus.brouhaha.com> <c330752a-7a1b-4e74-8f06-76b15379758a@x3g2000yqo.googlegroups.com> <2243495.gS2lSoF1Fd@sunwukong.fritz.box>

Show all headers | View raw

Bernd Paysan wrote:

> Mark Wills wrote:
> 
> > On Jan 14, 7:43 am, Paul Rubin <no.em...@nospam.invalid> wrote:
> >> Mark Wills <forthfr...@gmail.com> writes:
> >> > I also spent a while looking at doing it without conditionals, and
> >> > eventually concluded that it couldn't be done.
> > > 
> >> : sq dup * ;
> >> : g { a b c -- n } a sq b sq + c sq + a b c min min sq - ;
> > 
> > Well, okay, I stand corrected! Nice job.
> 
> There are other ways:
> 
> : sq ( n -- n² ) dup * ;
> : g ( a b c -- n )  2dup max sq >r min max sq r> + ;
> 
> > The fact that it needs locals, and computes all the squares is a bit
> > of an "ouch"!
> 
> Yes.  The solution without locals and without squaring all three numbers 
> is shorter.

The solution without Forth and without using the return stack
is shorter, even when it is generalized to handle any number of
numbers.

Ruby:

g = proc{|x| a,b = x.sort.pop(2); a*a + b*b}
    ==>#<Proc:0x0297ea58@(irb):13>
g[ [5, 3, 33, 22, 333] ]
    ==>111978

Back to comp.lang.forth | Previous | Next | Find similar

Thread

Re: structure and interpretation of computer programs exercsie 1.3 in forth "WJ" <w_a_x_man@yahoo.com> - 2013-06-14 15:37 +0000

csiph-web