[seqfan] Re: A085361
charles.greathouse at case.edu
Fri Jul 13 23:42:14 CEST 2012
Well my GP script does a thousand digits in a second on my old
machine, so it shouldn't be that hard. The real problem is that
Mathematica is trying to use too much precision, since it sees the
exact number 1 and tries to give a number with the desired number of
digits after that subtraction. Possibly just using -1.0 in place of -1
would make it run at a reasonable speed.
Sidenote: I don't think there's anything wrong with a program
returning a number for a cons sequence, nor do I see a problem with a
tabl or tabf program returning a table. Of course when your language
has a nice way to convert between forms it's probably best to do so,
as in this case.
Case Western Reserve University
On Fri, Jul 13, 2012 at 5:18 PM, Harvey P. Dale <hpd1 at nyu.edu> wrote:
> There appear to be two problems with the Mma program for
> this sequence:
> 1. It would only provide a decimal number rather than the terms of
> the sequence, so it needs to be wrapped inside RealDigits, and
> 2. It fails even to provide a decimal number because it quickly
> exhausts the default value of $MaxExtraPrecision.
> I tried to deal with the latter by increasing $MaxExtraPrecision to
> 100000, but had to abort the program after it ran a LONG time.
> Seqfan Mailing list - http://list.seqfan.eu/
More information about the SeqFan