OpenEuphoria

RE: contest #2

• Posted by bensler at mail.com Mar 17, 2002
• 387 views

My contest#2 routine is pure EU. No memory addressing.

I'll give you one hint :) I used Derek/Eumans hash function.

If I explain how my data is stored, I'll give away my method :P

Chris


tone.skoda at siol.net wrote:
> If I run that test (Duron700 128MB RAM) I get these results:
> 
> 2.2 - file load time
> 23.43 - matching patterns
> 25.63 - total
> 0.000586 - average
> 
> Are you even storing all your data in sequences, or are you poking it in
> memory?
> I organize data in two non-nested (1-level nested?) sequences, or better
> said two sequences which have sequences which have numbers.
> 
> ----- Original Message -----
> From: <bensler at mail.com>
> To: "EUforum" <EUforum at topica.com>
> Sent: Sunday, March 17, 2002 4:03 PM
> Subject: RE: contest #2
> 
> 
> > Here is my complete test, using Derek's last list.
> >
> > It takes a total of 14.77 seconds to run this on my machine.
> >
> > <TEST>
> > sequence words
> >
> > atom t,t2   t=time()
> > include match_pattern.e
> >
> > t2 = time()-t
> > ? t2
> >
> > for i = 1 to 1000 do
> >    words = match_pattern({1,1,2,3,4})
> >    words = match_pattern({'E', 1, 2, 3, 1, 3})
> >    words = match_pattern({1,'E',2,3,4,'S'})
> >    words = match_pattern({1,1})
> >    words = match_pattern({1,'R',2,'R',3})
> >    words = match_pattern({1})
> >    words = match_pattern({1,2,3})
> >    words = match_pattern({4,'I','E'})
> >    words = match_pattern({1,'E',2, 'E', 3, 'E'})
> >    words = match_pattern({1,2,2,3,'S'})
> >    words = match_pattern("RABBIT")
> >    words = match_pattern("R" & {1,1,2})
> >    words = match_pattern({'T',1,2})
> >    words = match_pattern({1,2,3,'S'})
> >    words = match_pattern({1,1,2,2})
> >    words = match_pattern({'B','E','A','R'})
> >    words = match_pattern({1,2,3,4,'E','D'})
> >    words = match_pattern({1,2,3,4,5,6,7})
> >    words = match_pattern({1,2,3,1,4,3,3,2,5,3})
> >    words = match_pattern({1,2,3,4,5,4,3,2,1})
> >    words = match_pattern({1,2,'X'})
> >    words = match_pattern({'M',1,2,3,4,5})
> >    words = match_pattern({1,2,3,4,5,6,7,8,9,10})
> >    words = match_pattern({'E',1,1,2})
> >    words = match_pattern({1,2,2,1,3})
> >    words = match_pattern({1,1,2})
> >    words = match_pattern({1,2,1})
> >    words = match_pattern({'M',1,2,1,'M'})
> >    words = match_pattern("MARTIN")
> >    words = match_pattern({1,2,'X',2,1})
> >    words = match_pattern({1,2,3,'B',3,4})
> >    words = match_pattern({1,2,'M',2,1})
> >    words = match_pattern({9,8,7,6,7,8,9})
> >    words = match_pattern({'E',8,7,6,7,8,9})
> >    words = match_pattern({9,'E',7,6,7,8,9})
> >    words = match_pattern({9,8,'E',6,7,8,9})
> >    words = match_pattern({9,8,7,'E',7,8,9})
> >    words = match_pattern({9,8,7,6,'E',8,9})
> >    words = match_pattern({9,8,7,6,7,'E',9})
> >    words = match_pattern({9,8,7,6,7,8,'E'})
> > end for
> >
> > ? time()-t-t2
> > ? time()-t
> > ? (time()-t-t2)/1000/40
> >
> > for i = 1 to length(words) do
> > --   puts(1,words[i]&"\n") -- uncomment this, to list results
> > end for
> > while get_key()=-1 do end while
> > <END TEST>
> >
> > C. K. Lester wrote:
> > > Are we still using this list for benchmarking?
> > >
> > > {1,1,2,3,4},
> > > {'E', 1, 2, 3, 1, 3},
> > > {1,'E',2,3,4,'S'},
> > > {1,1},
> > > {1},
> > > {1,2,3},
> > > {4,'I','E'},
> > > {1,'E',2, 'E', 3, 'E'},
> > > {1,2,2,3,'S'},
> > > "RABBIT",
> > > "R" & {1,1,2},
<snip>