[Bug-gsl] gsl_spline bug

[Simon Ching]

Hello,

I have found a possible bug with gsl_spline +
cspline_periodic. For the set of 8 control
points of (0, 0), (45, hr2), (90, 1), (135, hr2),
(180, 0), (225, -hr2), (270, -1) and (315, -hr2)
where (hr2 = sqrt (2)/2) the intepolating
curve does not go back to the point (360, 0)

It also seems that the gsl_spline_eval_deriv and
gsl_spline_eval_deriv2 do not give correct results.

Attached please find spline_bug.c and
spline_bug.txt which are a sample bug-revealing
proram and its output repectively.

Regards,
Simon.

=====
Simon E Ching
-----------------------------------------------------------
                               email: address@hidden
                      http://www.geocities.com/simon_ching/

_______________________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com.hk address at http://mail.english.yahoo.com.hk