On 10 April 2013 14:53, Andrew McRae <email address hidden> wrote:
> >From eyeballing the .msh file, there are occasionally variations in the
> last decimal place - O(10^-15) for an O(1) number.
>
The code in PeriodicBoundaryComputation.cpp tests for matching
entities using DOLFIN_EPS, which is equal to 3.0e-16 (I don't know how
this number for DOLFIN_EPS came about).
On 10 April 2013 14:53, Andrew McRae <email address hidden> wrote:
> >From eyeballing the .msh file, there are occasionally variations in the
> last decimal place - O(10^-15) for an O(1) number.
>
The code in PeriodicBoundar yComputation. cpp tests for matching
entities using DOLFIN_EPS, which is equal to 3.0e-16 (I don't know how
this number for DOLFIN_EPS came about).
Garth
> -- /bugs.launchpad .net/bugs/ 1167036 /bugs.launchpad .net/dolfin/ +bug/1167036/ +subscriptions
> You received this bug notification because you are a member of DOLFIN
> Core Team, which is subscribed to DOLFIN.
> https:/
>
> Title:
> dolfin-convert/gmsh incompatible with periodic boundaries
>
> To manage notifications about this bug go to:
> https:/