Re: Diophantine matters

clicliclic@xxxxxxxxxx wrote:

But so far I haven't seen an explicit statement of the PSLQ algorithmn
on the internet - either as mathematical formulae or as computer code.
In particular, P. Zimmerman's implementation using GMP at

<http://www.loria.fr/~zimmerma/free/pslq-1.0.c>

seems no longer available :(. Could this algorithm (or perhaps the LLL
one) be implemented in one afternoon, or is this a much more complicated
affair?

Looking at the Mathematica notebook available at
http://library.wolfram.com/infocenter/MathSource/4263/
it appears that this algorithm is not terribly complicated.

It does require some basic linear algebra operations.

- Daniel
.

Relevant Pages

  • Re: Diophantine matters
    ... on the internet - either as mathematical formulae or as computer code. ... P. Zimmerman's implementation using GMP at ... Could this algorithm (or perhaps the LLL ...
    (sci.math.symbolic)
  • Re: Public domain algorithm for arbitrary-precision fast division?
    ... David N. Williams wrote: ... >> I'm looking for an algorithm for doing arbitrary-precision ... the fastest method uses fast multiplication by two steps. ... I've never used it, but I must say that from the outside, GMP ...
    (sci.math.num-analysis)
  • Re: Public domain algorithm for arbitrary-precision fast division?
    ... >> I'm looking for an algorithm for doing arbitrary-precision ... the fastest method uses fast multiplication by two steps. ... on some machines it is faster ... I've never used it, but I must say that from the outside, GMP ...
    (sci.math.num-analysis)
  • Re: Fast Binary-to-Decimal Conversion Algorithms?
    ... >> about he needs the kind of stuff that they use in gmp. ... Notice that the ADD-3 to TENS suddenly is ... i.e. a register long enough to hold the value. ... It's a nice algorithm and you have a good writeup, ...
    (comp.programming)
  • Re: Dividing long numbers
    ... something else) seemed to be the big bottleneck. ... to me like I was using the right approach, only that my algorithm was ... You might, as an alternative, use GMP. ...
    (sci.crypt)