[Bug-gsl] [bug #53451] gsl_sf_ellint_Pcomp( k, n, mode ) returns NaN if mode < -1

[Dennis Amelunxen]

URL:
  <http://savannah.gnu.org/bugs/?53451>

                 Summary: gsl_sf_ellint_Pcomp( k, n, mode ) returns NaN if
mode < -1
                 Project: GNU Scientific Library
            Submitted by: damelunx
            Submitted on: Sat 24 Mar 2018 10:00:17 AM UTC
                Category: None
                Severity: 3 - Normal
        Operating System: Linux
                  Status: None
             Assigned to: None
             Open/Closed: Open
                 Release: 2.1
         Discussion Lock: Any

    _______________________________________________________

Details:

Hi there,

I am using 

gsl_sf_ellint_Pcomp( k, n, mode )

from the GSL, and only get NaNs when using mode<-1 and k>1.3.
But the integral is defined, compare for example with the result from
WolframAlpha:

https://www.wolframalpha.com/input/?i=EllipticPi(0.1,1.3,0.25)

gives 1.43517..., which is also the output of

gsl_sf_ellint_Pcomp(1.3,0.5,-0.1)

But while

https://www.wolframalpha.com/input/?i=EllipticPi(1.1,1.3,0.25)

gives 4.72112..., the output of

gsl_sf_ellint_Pcomp(1.3,0.5,-1.1)

is NaN.

I'm using GSL in R (via the wrapper package
https://cran.r-project.org/web/packages/gsl/index.html), so I don't see any
error messages.

Best,
Dennis





    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?53451>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/