Re: [deal.II] Re: bug in program only with adaptive mesh refinement

2017-02-23 Thread Denis Davydov
> On 23 Feb 2017, at 17:08, Daniel Shapero wrote: > > The vectors you have in this equation P(u+h*v) =..., which of those have > constrains distributed and which zeroed? > If you assembly matrices with ConstraintMatrix.distribute_local_to_global() > the diagonal elements corresponding to con

[deal.II] Re: bug in program only with adaptive mesh refinement

2017-02-23 Thread Daniel Shapero
> > The vectors you have in this equation P(u+h*v) =..., which of those have > constrains distributed and which zeroed? > If you assembly matrices with ConstraintMatrix.distribute_local_to_global() > the diagonal elements corresponding to constrained DoFs > ... > Along the same lines: does yo

[deal.II] Re: bug in program only with adaptive mesh refinement

2017-02-22 Thread Denis Davydov
Hi Daniel, The vectors you have in this equation P(u+h*v) =..., which of those have constrains distributed and which zeroed? If you assembly matrices with ConstraintMatrix.distribute_local_to_global() the diagonal elements corresponding to constrained DoFs will be dummy (and positive) just to m