You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
stephane-caron
changed the title
Tighter cost and primal tolerances for Maros-Meszaros test set
Tighter tolerances for Maros-Meszaros test set
Jan 4, 2023
Primal: not all solvers allow for $\epsilon_{rel}$ in their contracts. To provide a wider yet fair comparison we currently set $\epsilon_{rel}=0$ in all settings, so that our conditions are $r_p, r_d, r_g \leq \epsilon_{abs}$.
Identify tighter values for the cost and primal tolerances.
See bodono/scs-python#63 (comment) for context.
The text was updated successfully, but these errors were encountered: