I won't pretend that I understand the issues here, but if it works predictably (even when hitting the vertex coordinate of a discontinuous space) then I'm all for it ;)
-j. On 11 June 2012 14:42, Anders Logg <[email protected]> wrote: > On Sun, Jun 10, 2012 at 01:21:57PM -0000, Marie Rognes wrote: > > Response to 4/ > > > > (Caveat: I've never used PointSource) > > > > How about adding a new measure: a point measure, corresponding to > evaluation at a given point, > > say > > > > dp > > > > Could that resolve the issues at hand? > > I think that would be neat. > > We set (formally) > > dp(\bar{x}) = \delta_{\bar{x}} dx > > and then the generated code would need to > > 1. Check whether point is in the current triangle (easily handled with > an appropriate code-snippet) > > 2. Evaluate the integrand at the point in question and add to the > local tensor > > This would require adding an addition integral to UFC: > > cell_integral > exterior_facet_integral > interior_facet_integral > dirac_integral / point_integral / ??? > > -- > Anders > > _______________________________________________ > Mailing list: https://launchpad.net/~dolfin > Post to : [email protected] > Unsubscribe : https://launchpad.net/~dolfin > More help : https://help.launchpad.net/ListHelp >
_______________________________________________ Mailing list: https://launchpad.net/~dolfin Post to : [email protected] Unsubscribe : https://launchpad.net/~dolfin More help : https://help.launchpad.net/ListHelp
