Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Limit

20 views
Skip to first unread message

Thomas D. Dean

unread,
Mar 28, 2013, 12:41:53 PM3/28/13
to
> restart;
> about(f(x));
f(x):
nothing known about this object
> g:=(x)->x^2*f(x);
> limit(g(x),x=0);
0

Maple has to be making some assumptions about f(x) to arrive at this
conclusion. What?

limit(f(x)*g(x),x=a) = limit(f(x),x=a) * limit(g(x),x=a);

consider imit(f(x),x=a) = inifinity and limit(g(x),x=a) = 0, then,
limit(f(x)*g(x),x=a) is undefined.

Correct?

Tom Dean

Nasser M. Abbasi

unread,
Mar 28, 2013, 7:35:56 PM3/28/13
to
On 3/28/2013 11:41 AM, Thomas D. Dean wrote:
> > restart;
> > about(f(x));
> f(x):
> nothing known about this object
> > g:=(x)->x^2*f(x);
> > limit(g(x),x=0);
> 0
>
> Maple has to be making some assumptions about f(x) to arrive at this
> conclusion. What?
>

fyi, Mathematica v9.01, does not evaluate it

In[12]:= Limit[x*f[x], x -> 0]
Out[12]= Limit[x f[x], x -> 0]

and maxima 12.04 also does not evaluate this.

I think you are right, without knowing how the limit of f(x)
behaves as x->0 the answer seems strange. But Matlab 2013a,
which I tink uses mupad these days, gives zero also:

------------------------
EDU>> syms x f(x)
EDU>> limit(x*f(x),x,0)

ans =

0
----------------------

I think we need a math expert to come and sort out which
is correct and which is not. But I am voting for result
given by M and Maxima here and not Maple and Matlab.

> limit(f(x)*g(x),x=a) = limit(f(x),x=a) * limit(g(x),x=a);
>
> consider imit(f(x),x=a) = inifinity and limit(g(x),x=a) = 0, then,
> limit(f(x)*g(x),x=a) is undefined.
>
> Correct?
>
> Tom Dean
>

--Nasser

Joe Riel

unread,
Mar 29, 2013, 1:02:16 AM3/29/13
to
It's clear the result should be undefined; consider the case
when f(x) = k/x^2.

--
Joe Riel

Nasser M. Abbasi

unread,
Mar 29, 2013, 1:16:26 AM3/29/13
to
On 3/29/2013 12:02 AM, Joe Riel wrote:

>
> It's clear the result should be undefined; consider the case
> when f(x) = k/x^2.
>

Yes. sure. That would depend on what value K has.

In[8]:= Limit[x*k/x^2, x -> 0]
Out[8]= k Infinity

In[9]:= % /. k -> 0
Out[9]= Indeterminate

May be someone should send a bug report to Maplesoft on this.

--Nasser


A N Niel

unread,
Mar 29, 2013, 9:29:43 AM3/29/13
to
In article <hO2dnU_SuLxJ78nM...@megapath.net>, Thomas D.
Dean <tom...@speakeasy.org> wrote:

> > restart;
> > about(f(x));
> f(x):
> nothing known about this object
> > g:=(x)->x^2*f(x);
> > limit(g(x),x=0);
> 0
>
> Maple has to be making some assumptions about f(x) to arrive at this
> conclusion. What?
>

Maybe Maple is using this:

> restart;
> limit(f(x),x=0);
f(0)
0 new messages