Skip to content
Projects
Groups
Snippets
Help
Loading...
Help
Support
Submit feedback
Contribute to GitLab
Sign in
Toggle navigation
S
ScalFMM
Project overview
Project overview
Details
Activity
Releases
Cycle Analytics
Repository
Repository
Files
Commits
Branches
Tags
Contributors
Graph
Compare
Charts
Issues
5
Issues
5
List
Boards
Labels
Milestones
Packages
Packages
Container Registry
Wiki
Wiki
Members
Members
Collapse sidebar
Close sidebar
Activity
Graph
Charts
Create a new issue
Commits
Issue Boards
Open sidebar
solverstack
ScalFMM
Commits
d3844968
Commit
d3844968
authored
Jun 23, 2016
by
COULAUD Olivier
Browse files
Options
Browse Files
Download
Email Patches
Plain Diff
Doc improvement
parent
6546b29c
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
29 additions
and
3 deletions
+29
-3
Src/Kernels/Interpolation/FInterpMatrixKernel_TensorialInteractions.hpp
...terpolation/FInterpMatrixKernel_TensorialInteractions.hpp
+29
-3
No files found.
Src/Kernels/Interpolation/FInterpMatrixKernel_TensorialInteractions.hpp
View file @
d3844968
...
...
@@ -45,16 +45,42 @@
* The table applyTab provides the indices in the reduced storage table
* corresponding to the application scheme depicted earlier.
*
*
PB:
BEWARE! Homogeneous matrix kernels do not support cell width extension
*
\warning
BEWARE! Homogeneous matrix kernels do not support cell width extension
* yet. Is it possible to find a reference width and a scale factor such that
* only 1 set of M2L opt can be used for all levels??
*
* The definition of the potential p and force f are extended to the case
* of tensorial interaction kernels:
*
*\f$ p_i(x) = K_{ip}(x,y)w_p(y),\f$ \f$ \forall i=1..NPOT, p=1..NPV\f$
*
* \f$f_{ik}= w_p(x)K_{ip,k}(x,y)w_p(y)\f$
*
* Since the interpolation scheme is such that
*
*\f$ p_i(x) \approx S^m(x) L^{m}_{ip}\f$
*
* \f$f_{ik}= w_p(x) \nabla_k S^m(x) L^{m}_{ip}\f$
*
* with
*
* \f$ L^{m}_{ip} = K^{mn}_{ip} S^n(y) w_p(y)\f$ (local expansion)
*
*\f$ M^{m}_{p} = S^n(y) w_p(y)\f$ (multipole expansion)
*
* then the multipole exp have NPV components and the local exp NPOT*NPV.
*
* NB1: Only the computation of forces requires that the sum over p is
* performed at L2P step. It could be done at M2L step for the potential.
*
* NB2: An efficient application of the matrix kernel is highly kernel
* dependent, we recommand overriding the P2M/M2L/L2P function of the kernel
* you are using in order to have opsimal performances + set your own NRHS/NLHS.*
*/
////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////
//
//
/
// Tensorial Matrix Kernels (NCMP>1)
//
// The definition of the potential p and force f are extended to the case
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment