Mathematica 9 is now available
Services & Resources / Wolfram Forums
-----
 /
MathGroup Archive
2005
*January
*February
*March
*April
*May
*June
*July
*August
*September
*October
*November
*December
*Archive Index
*Ask about this page
*Print this page
*Give us feedback
*Sign up for the Wolfram Insider

MathGroup Archive 2005

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

Search the Archive

Re: What we get from (0.0*x), (0.0^x) and similar stuff

  • To: mathgroup at smc.vnet.net
  • Subject: [mg58821] Re: What we get from (0.0*x), (0.0^x) and similar stuff
  • From: "David W. Cantrell" <DWCantrell at sigmaxi.org>
  • Date: Wed, 20 Jul 2005 00:29:20 -0400 (EDT)
  • References: <dbidbr$bk3$1@smc.vnet.net>
  • Sender: owner-wri-mathgroup at wolfram.com

ted.ersek at tqci.net wrote:
> I think some of the code I sent in my last message on this subject was a
> bad idea. My current thoughts on this subject are summarized below.
> ------------------------------
>
> I like what I get from the next line. If a user wants to ignore the fact
> that certain values will give (0^0) or (0*Infinity) they can add
> definitions to do what they want. If a user wants (0^x) to return
> If[x==0,Indeterminate,1] they can add a definition to do that.

If you were thinking about my previous response when you said that,
please note that I was clearly talking about x^0, rather than 0^x,
when I mentioned If[x==0, Indeterminate, 1].

> In[1]:=
>  Clear[x,y];
>  {0^x,0.0^x,0.0*x*y}
>
> Out[2]=
>  {0^x, 0.x, 0. x y}

Surely you meant {0^x, 0.^x, 0. x y}.

> ------------------------------
> The next line should return the approximate number 0.

Should it? Given only that az is approximately zero and
-Infinity < x < Infinity, the product az*x could potentially be large.
I suggest that 0. x is the correct thing to return.

> In[3]:=
>  FullSimplify[ 0.0*x, -Infinity < x < Infinity ]
>
> Out[3]=
>  0. x
>
> -------------------------------
> (0^-2) and (0.0^-2) return ComplexInfinity.  Shouldn't the next line
> return {ComplexInfinity, ComplexInfinity}.

Yes and no, resp., IMO. See below.

> In[4]:=
>  FullSimplify[ {0^x,0.0^x}, -Infinity < x < 0 ]
>
> Out[4]=
>  {Infinity, Indeterminate}

Using version 5.1.0.0 for Windows, I get

In[10]:= FullSimplify[{0^x, 0.0^x}, -Infinity < x < 0]

Out[10]= {0^x, Indeterminate}

so I don't know how you got Infinity for 0^x when -Infinity < x < 0.
But in any event, Infinity is incorrect here. As you suggested
earlier, the best result would be ComplexInfinity. Returning 0^x
unchanged, as my version did, while not the best here, is at least
not wrong.

Messier is 0.0^x when -Infinity < x < 0. Indeterminate is the answer
given by both our versions, and I believe it is the correct answer
as well. Given only that az is approximately zero and -Infinity < x < 0,
the expression az^x could potentially be, say, close to 1 (taking |x| to
be very small). Thus, since |az^x| need not be large,
we should not return ComplexInfinity as you had suggested.

But we would have a different situation if the exponent were not
allowed to be close to 0. For example, given -Infinity < x < -1, it
might seem reasonable if 0.0^x simplified to ComplexInfinity,
although that is not what Mathematica does:

In[11]:= FullSimplify[0.0^x, -Infinity < x < -1]

Out[11]= Indeterminate

> -------------------------------
> Sqrt[0.0] and  (0.0^2.3) return the approximate number 0. Shouldn't the
> next line also return approximate zero.

No.  The reason is the same as that above, where I explained why
FullSimplify[0.0^x, -Infinity < x < 0] should not return ComplexInfinity.

But again we would have a different situation if the exponent were
not allowed to be close to 0. For example, given 1 < x < Infinity,
it might seem reasonable if 0.0^x simplified to 0., although that is
not what Mathematica does.

> In[5]:=
>  FullSimplify[ 0.0^x, 0 < x < Infinity]
>
> Out[5]=
>  Indeterminate

Regards,
David Cantrell


  • Prev by Date: Re: Re: "2p" in BenchmarkReport[] ?
  • Next by Date: "Many Digits" Friendly Competition 2005
  • Previous by thread: What we get from (0.0*x), (0.0^x) and similar stuff
  • Next by thread: Superscript Problem