bessel.lisp error codes and noisy quad_qagi



On Nov. 11, 2011 I wrote
-----------------------

> (%i5) quad_qagi(realpart(bessel_k(2,%i*x)),x,1,inf);
[snip]

>(%o5) [-311.0051190974822,70.21683385597015,5985,5]
>
>Is there a *quieter way* I can  use quad_qagi for this
>calculation?
>
>The asserted Mathematica answer to this numerical integral
>is   -1.45355 + 1.50855*%i
---------------------------
Since only the real part is being evaluated, the asserted Mathematica
answer is  -1.45355.

Ted