Re: [Maxima-discuss] limit(tan(x), x, %pi2) (Was Re: 1/0, 0/0, -1/0, -0 implementation)

[Richard F. <fa...@be...>]

On 1/14/2016 11:24 AM, Barton Willis wrote:
>
>
> > So for example limit(sqrt(x),x,0,minus)=>0, though sqrt(x) is not 
> defined over the negative reals and so I suppose in principle this 
> should be UND.
>
>
>
> Sorry, in my other post on this, I missed the limit from the left. The 
> function x in [0,\inf) --> sqrt(x) does not have a limit from the left 
> toward zero.  This is because 0 isn't a limit point of (-inf,0] 
> intersect [0, inf).
>
>
> --Barton
>
>
>
I think we have to consider that Maxima is perfectly happy to evaluate 
sqrt(x) for negative reals,
and so we could convert the limit(sqrt(x),x,0,minus)
to  limit(sqrt(-x),x,0,plus)
to limit(%i*sqrt(x),0,plus)
to %i* limit(sqrt(x),x,0,plus)
to %i* 0

to 0.

This seems different in nature from
limit(log(x),x,0,minus)
which ends up with  %i*%pi   hanging out...