On Thu, Oct 14, 2010 at 3:19 PM, Justin Zamora <jus...@zamora.com> wrote: > On Thu, Oct 14, 2010 at 2:26 PM, Joe Marshall <jmarsh...@alum.mit.edu> wrote: >> Here's a fast way to get the leftmost digit of a number: >> ... >> This version is slightly slower for small numbers, but much much better >> for large ones: >> (defun leftmost-digit (n base) >> (if (> base n) >> n >> (do* ((i base next) >> (next (* base base) (* i i))) >> ((> next n) (leftmost-digit (floor n i) base))))) > > I tried a comparison of your method to the string-conversion method > and was surprised that the string-comparison version ran 3 times > faster! I think I implemented your method correctly (at least it > appears to produce correct results). Any insights?
Interesting. Here is the thread that it appeared on first: http://groups.google.com/group/comp.lang.lisp/browse_thread/thread/ed990c2efd24373e/e6ada2b307534225?hl=en&ie=UTF-8&q=group:comp.lang.lisp+first-digit&pli=1 One person there reported: ``Joe's iterative solution was 26x faster than my string-based version and allocates no memory. '' So here are some possible explanations: 1) number->string might be written in C. 2) generic arithmetic might be causing too much overhead 3) Something that ought to be compiled is not being compiled. -- ~jrm _________________________________________________ For list-related administrative tasks: http://lists.racket-lang.org/listinfo/users
