benchmarks

[Sammo]

There's a coding error in the loops benchmark. The benchmark code is (in part):

Code Select Expand

(for (a 0 n)
   (for (b 0 n)
      (for (c 0 n))  <-- misplaced paren
         (for (d 0 n)
            (for (e 0 n)
               (for (f 0 n)
                  (inc 'x) )))))It should be

Code Select Expand

(for (a 0 n)
   (for (b 0 n)
      (for (c 0 n)
         (for (d 0 n)
            (for (e 0 n)
               (for (f 0 n)
                  (inc 'x) )))))) <-- moved paren


[Lutz]

Oops, this will change the the benchmark time, I will rerun the benchmark for this and change the numbers ASAP



This was important, thankyou



Lutz

[Lutz]

I replaced the 'for' with 'dotimes', which is also a little bit faster, because it does not need an initializer.



Lutz

[HPW]

Code Select Expand


(define (loop1 n)
(set 'x 0)
(for (a 1 n)
   (for (b 1 n)
      (for (c 1 n)
         (for (d 1 n)
            (for (e 1 n)
               (for (f 1 n)
                  (inc 'x) ))))))
(println x)
)

(define (loop2 n)
(set 'x 0)
(dotimes (a n)
(dotimes (b n)
(dotimes (c n)
(dotimes (d n)
(dotimes (e n)
(dotimes (f n)
(inc 'x)))))))
(println x)
)


> (time (loop1 20))

64000000

21625

> (time (loop2 20))

64000000

22359

>

For me 'for' is a little bit faster?

[Sammo]

On my 500 MHz laptop, 'dotimes' wins consistently.

Code Select Expand

> (time (loop1 20)) ; for
64000000
51028
> (time (loop2 20)) ; dotimes
64000000
49876

[HPW]

> (time (loop1 20))

64000000

21610

> (time (loop2 20))

64000000

22297

> (time (loop1 20))

64000000

21390

> (time (loop2 20))

64000000

22172

>

1.8 GHZ Desktop P4