Services & Resources / Wolfram Forums / MathGroup Archive
-----

MathGroup Archive 2009

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Change in NSolve algorithm

  • To: mathgroup at smc.vnet.net
  • Subject: [mg98616] Re: Change in NSolve algorithm
  • From: ADL <alberto.dilullo at tiscali.it>
  • Date: Tue, 14 Apr 2009 06:17:54 -0400 (EDT)
  • References: <grpif7$lnr$1@smc.vnet.net> <grs6as$qie$1@smc.vnet.net>

Thank you for the clarification between MachinePrecision and
$MachinePrecision behaviors, which was not clear to me before.

Then, perhaps, a possible "communication" improvement for Mathematica
developers might be to substitute/alias the keyword MachinePrecision
with a new keyword, e.g. FPU, which would more easily convey the
difference between:

    NSolve[eqns, vars, $MachinePrecision]
    NSolve[eqns, vars, FPU]

Anyway, thanks again

ADL


On 13 Apr, 09:30, Szabolcs Horv=E1t <szhor... at gmail.com> wrote:
> Yes,  WorkingPrecision -> $MachinePrecision will use Mathematica's
> arbitrary precision arithmetic with $MachinePrecision number of digits
> (around 16 on most machines).  WorkingPrecision -> MachinePrecision wil=
l
> actually use the FPU to do the calculations.



  • Prev by Date: Re: Best practices of coding using Mathematica?
  • Next by Date: Re: UNDO and Mathematica - how useless
  • Previous by thread: Re: Change in NSolve algorithm
  • Next by thread: Re: Help with ScriptSource in GUIKit