I try to make a comparison between the adaption with the command “adaptmesh” (that use the hessian method) and one that use the recovery-gradient method with zienchiewiz-zhu estimator (developed by myself).
But i don’t understand exactly what “err” parameter indicates in adaptmesh.
In the second case i set the “tol” parameter as the approximation of the H1-seminorm of the discretization error, and i would like to set the same also in the first case.
And if i want to set the number of elements (and not the error) during the adaptive procedure? Is it possible?
Thanks in advance for any help!