Proble : maxima finds a primitive, but not a defnite integral.



On 2012-11-21, Raymond Toy <toy.raymond at gmail.com> wrote:

> I'm not 100% sure, but I think quad_qags does better because it
> handles the case of the infinite slope at zero better.  But I don't
> have the quadpack book to understand which cases should perform
> better.

If I remember correctly, QAGS computes the integral in pieces (applying
QAG to each piece) and adds them up, while QAG does one computation for
the whole thing. I speculate that QAGS is better for integrands which
have locally strange behavior, such as 1/(1 + sqrt(x)).

best

Robert Dodier