Mercurial > repos > public > sbplib_julia
comparison src/SbpOperators/volumeops/laplace/laplace.jl @ 922:0bf5952c240d feature/laplace_opset
Review: Add review comment regarding restructuring of Laplace
| author | Vidar Stiernström <vidar.stiernstrom@it.uu.se> |
|---|---|
| date | Sun, 30 Jan 2022 13:00:18 +0100 |
| parents | 86776d06b883 |
| children | 12e8e431b43c |
comparison
equal
deleted
inserted
replaced
| 872:6a4d36eccf39 | 922:0bf5952c240d |
|---|---|
| 1 export Laplace | 1 export Laplace |
| 2 export laplace | 2 export laplace |
| 3 # REVIEW: Makes more sense to me to have the exports at the top of the file. | 3 # REVIEW: Makes more sense to me to have the exports at the top of the file. |
| 4 # Might as well start fixing that. | 4 # Might as well start fixing that. |
| 5 | |
| 6 # REVIEW: | |
| 7 # Design discussions has led to attempt a restructuring of Laplace to a more | |
| 8 # minimal type, holding the tensor mapping and a stencil set. This allows | |
| 9 # construction of associated tensor mappings, e.g. boundary operators, based on the | |
| 10 # stencil set while keeping the type simpler. | |
| 5 | 11 |
| 6 # REVIEW: The style of name `Laplace` might clash with other concepts. When | 12 # REVIEW: The style of name `Laplace` might clash with other concepts. When |
| 7 # thinking about implementing the variable second derivative I think I will | 13 # thinking about implementing the variable second derivative I think I will |
| 8 # have to create it as a full TM for the full dimensional problem instead of | 14 # have to create it as a full TM for the full dimensional problem instead of |
| 9 # building it as a 1D operator and then use that with outer products. The | 15 # building it as a 1D operator and then use that with outer products. The |
