Re: int(1/x^2,x,-1,1)
[Prev][Next][Index][Thread]
Re: int(1/x^2,x,-1,1)
Notrub wrote:
>
> Geoff H wrote:
> > So here's another test:
> > The '85 responses to both fnInt(1/x,x,0,1) and fnInt(1/x,x,-1,1) are
> > "ERROR 33 TOL NOT MET" (unless tol is cranked to about 1.9) and
> > "ERROR 02 DIV BY ZERO," respectively.
> >
> > Derive XM responds to INT(1/x,x,0,1) with infinity,
> > while INT(1/x,x,-1,1) gives -i*pi, which looks like an
> > application of Cauchy's residue theorem?!
> >
> > What does the '92 do?
>
> On the 92, nInt(1/x,x,-1,1) gives 0 with a warning: "Questionable
> accuracy." nInt(1/x,x,0,1) gives no answer (it actually returns the
> actual problem in symbolic form, again with the "Questionable accuracy"
> warning).
>
> Notrub
>
> --
> "We must view with profound respect the infinite capacity of the human
> mind to resist the introduction of useful knowledge."
>
> --Thomas Raynesford Lounsbury
> (1838-1915)
> mailto:notrub1@bangornews.infi.net
But if you use the symbolic integration, you get undef and (infinity)
respectively.
References: