Re: Re: Strange Manipulate+ContourPlot behavior
- To: mathgroup at smc.vnet.net
- Subject: [mg82204] Re: [mg82172] Re: Strange Manipulate+ContourPlot behavior
- From: DrMajorBob <drmajorbob at bigfoot.com>
- Date: Sun, 14 Oct 2007 06:21:11 -0400 (EDT)
- References: <fei3de$qn1$1@smc.vnet.net> <200710110416.AAA21323@smc.vnet.net> <16233724.1192265949170.JavaMail.root@m35>
- Reply-to: drmajorbob at bigfoot.com
With a fresh kernel in v6.0.1, none of THOSE results are Indeterminate:
e = (b (a - a p + b p))/(2 a b + (a - b)^2 p);
rules = {a -> 0, b -> 0, p -> 0};
Fold[ReplaceAll[#1, #2] &, e, #] & /@ Permutations[rules]
{1, 1, 0, 0, 1/2, 1/2}
$Version
"6.0 for Microsoft Windows (32-bit) (June 19, 2007)"
That's not to say some other evaluation order can't occur, of course.
Bobby
On Sat, 13 Oct 2007 03:05:49 -0500, Yaroslav Bulatov
<yaroslavvb at gmail.com> wrote:
> I suspect the discrepancy between "Manipulate[ContourPlot[e" and
> "Manipulate[ContourPlot[Evaluate[e" has to do with automatic
> simplification rules. The expression being plotted evaluates to 1, 0,
> 1/2 or Indeterminate depending on the order in which replacement rules
> are applied.
>
> e = (b (a - a p + b p))/(2 a b + (a - b)^2 p);
> rules = {a -> 0, b -> 0, p -> 0};
> Fold[ReplaceAll[#1, #2] &, e, #] & /@ Permutations[rules]
>
> Is it possible to disable automatic simplification rules to prevent
> expressions like a / (2 a) simplifying to 1/2?
>
>
> On Oct 12, 12:11 am, Syd Geraghty <sydgerag... at mac.com> wrote:
>> Not so weird...
>>
>> In[]:= a = 0; b = 0; p = 0; (b (a - a p + b p))/(-2 a b (-1 + p) +
>> a^2 p + b^2 p)
>>
>> Out[]:= Indeterminate
>>
>> Syd
>>
>> Syd Geraghty B.Sc., M.Sc.
>> sydgerag... at mac.com
>> San Jose, CA
>>
>> On Oct 10, 2007, at 9:16 PM, Will Robertson wrote:
>>
>> > On Oct 10, 5:44 pm, Yaroslav Bulatov <yarosla... at gmail.com> wrote:
>> >> Manipulate[
>> >> ContourPlot[(
>> >> b (a - a p + b p))/(-2 a b (-1 + p) + a^2 p + b^2 p), {a, 0, 1},
>> >> {b,
>> >> 0, 1}], {p, 0, 1}]
>>
>> > That is weird. Sticking in an Evaluate[] around the expression fixes
>> > the problem -- until the slider is first moved. I guess I'd just
>> > recommend avoiding p = 0 ... :)
>>
>> > Will
>
>
>
>
--
DrMajorBob at bigfoot.com
- References:
- Re: Strange Manipulate+ContourPlot behavior
- From: Will Robertson <wspr81@gmail.com>
- Re: Strange Manipulate+ContourPlot behavior