-
Laurent Belcour authored
Forcing to use the interior-point-convex algorithm which produce more stable results. Still, it is not in the bounds. Commented the normalization in the rational function class, it seems to introduce some errors (weird).
c8b45837