Okay, I'll take a look at this.

I can also take a look at the discrepancies between
SPSS and PSPP cdf's if you send me some syntax files
showing the problems.

-Jason

On Wed, Mar 09, 2005 at 10:09:36AM -0800, Ben Pfaff wrote:
> Jason Stover <[EMAIL PROTECTED]> writes:
>
> > I forgot about that. The modified gsl_cdf.h is attached.
>
> Thanks.  That took care of the compilation errors.
>
> I checked in expression code that uses your functions.  I put
> them in the PSPP source tree in a new library I called
> "gsl-extras", and I replaced the gsl_ prefix by gslextras_ to
> avoid conflicts should gsl in the future add those functions.
>
> There were a couple of warnings that concerned me:
>
> hypergeometric.c: In function `gslextras_cdf_hypergeometric_P':
> hypergeometric.c:86: warning: comparison of unsigned expression >= 0 is 
> always+true
> hypergeometric.c:107: warning: comparison of unsigned expression >= 0 is
+always true
>
> It seems to me like these could cause the relevant loops in
> gsl_cdf_hypergeometric_P to continue forever, if the other
> conditions fail to converge.  I didn't look closely enough at the
> math to figure out whether that could ever happen.
> --
> "If a person keeps faithfully busy each hour of the working day, he
>  can count on waking up some morning to find himself one of the
>  competent ones of his generation."
> --William James
>
>
> _______________________________________________
> pspp-dev mailing list
> [email protected]
> http://lists.gnu.org/mailman/listinfo/pspp-dev
> 
-- 
[EMAIL PROTECTED]
SDF Public Access UNIX System - http://sdf.lonestar.org


_______________________________________________
pspp-dev mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/pspp-dev

Reply via email to