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

Re: Floor doesn't compute in some cases

• To: mathgroup at smc.vnet.net
• Subject: [mg83302] Re: Floor doesn't compute in some cases
• From: Fabian <hasslerf at phys.ethz.ch>
• Date: Sat, 17 Nov 2007 05:18:27 -0500 (EST)
• References: <fhjqkk$3gq$1@smc.vnet.net>
• Reply-to: Fabian <NOSPAM_fahasch at yahoo.de>

On Fri, 16 Nov 2007 10:15:48 +0000 (UTC), Valeri Astanoff
<astanoff at gmail.com> wrote:

> Floor[Log[32]/Log[2]]
>
> Floor::meprec: Internal precision limit $MaxExtraPrecision = 50.`
> reached while evaluating Floor[Log[32]/Log[2]]. >>
>
> Floor[Log[32]/Log[2]]
>
> $Version
>
> 6.0 for Microsoft Windows (32-bit) (June 19, 2007)
>
> I don't see the reason why...

Mathematica tries to evaluate Log[32]/Log[2] numerically to
decide which integer to round to. As the exact value (obtained by
FullSimplify[Log[32]/Log[2]] is 5, the numerical error is always to big
to decide weather the output should be 5 or 4.

Cheers, Fabian