ex - use deal.II fe to build libCEED Basis#1827
Conversation
peterrum
left a comment
peterrum
left a comment
There was a problem hiding this comment.
@jeremylt Thank you for working for this! The 1D information you need is inside of https://www.dealii.org/developer/doxygen/deal.II/structinternal_1_1MatrixFreeFunctions_1_1ShapeInfo.html => https://www.dealii.org/developer/doxygen/deal.II/structinternal_1_1MatrixFreeFunctions_1_1UnivariateShapeData.html (shape_values, shape_gradients, quadrature). Do you want to give it a try else I would do it tomorrow or on Thursday! One question: who is responsible for cleaning up after new?
|
Thanks for that info - I can give it a try. Re |
2a5c0a4 to
571e8cf
Compare
| for (const auto q : quadrature.get_tensor_basis()[0].get_points()) | ||
| q_ref_1d.push_back(q(0)); |
There was a problem hiding this comment.
This is fine, but, the same 1D data should be in: https://www.dealii.org/developer/doxygen/deal.II/structinternal_1_1MatrixFreeFunctions_1_1UnivariateShapeData.html#aeb5b85f1120f2957d39b51e2d348aedb
134d7d7 to
c311271
Compare
Co-authored-by: Peter Munch <peterrmuench@gmail.com>
c311271 to
bec2b28
Compare
2 participants
@peterrum I don't know if this is how to best approach this, but have tried to use the deal.II fe object to build the libCEED Basis object. (I apologize for my non-ideomatic C++). Is this a reasonable approach?