response
A per-constraint assessment of this solution, keyed by response name: the estimate the constraint was judged on, the amount by which it is violated, the one-sided upper confidence limit, and whether the constraint can be declared feasible at that limit.
Prefer this to responseConstraintOneSidedIntervals when the results must be attributed to particular constraints. That function returns a positional list which silently omits any constraint whose response is missing from the estimates, so its indices do not reliably align with the problem's constraint list; this one is keyed and cannot be misattributed.
Constraints whose response is absent from the estimates are omitted, as they are there — no assessment is possible without an estimate.
Parameters
the overall confidence across all response constraints. The individual levels are adjusted so the tests jointly meet it.