[Octave-bug-tracker] [bug #51157] betainc.cc: Inaccurate double precision results

[Rik]

Update of bug #51157 (project octave):

        Operating System:                     Any => Microsoft Windows      

    _______________________________________________________

Follow-up Comment #3:

I checked the dispatch sequence and Octave is passing the arguments to the
dbetai.f function.  This function comes from slatec
(http://www.netlib.org/slatec/fnlib/).  I looked a bit more and the version of
the Fortran code used by Octave is up to date.  However, the last modification
was in 1992 so this is some very old code.  It appears that the Fortran code
itself is the source of the innaccuracies.

    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?51157>

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