Class LocallyActiveDataCoarsenPatchStrategy is an abstract base class that defines a Strategy pattern interface for operations that a user may supply for application-specific coarsening of locally-active data between two levels in an AMR patch hierarchy. This interface class is similar to the CoarsenPatchStrategy interface, except that it is used to treat locally-active patch data. A concrete subclass must define three member functions to perform the following tasks:
More...
#include <LocallyActiveDataCoarsenPatchStrategy.h>
template<int DIM>
class SAMRAI::xfer::LocallyActiveDataCoarsenPatchStrategy< DIM >
- define maximum stencil width for user-defined coarsen operations
- preprocess the coarsening
- postprocess the coarsening
Note that the preprocess member function is called before standard data coarsening using CoarsenOperators and the postprocessor member function is called afterwards.
- See also
- xfer::LocallyActiveDataCoarsenAlgorithm
-
xfer::LocallyActiveDataCoarsenSchedule
◆ LocallyActiveDataCoarsenPatchStrategy()
The constructor for the coarsen strategy does nothing interesting.
◆ ~LocallyActiveDataCoarsenPatchStrategy()
The virtual destructor for the coarsen strategy does nothing interesting.
◆ getCoarsenOpStencilWidth()
Return maximum stencil width needed over all user-defined data coarsening operations. This is needed to determine the correct coarsening data dependencies.
◆ preprocessCoarsen()
Perform user-defined coarsening operations. This member function is called before standard coarsening operations (expressed using concrete subclasses of the CoarsenOperator<DIM> base class).
The preprocess function should move data from the source components on the fine patch into the source components on the coarse patch in the specified coarse box region. Recall that the source components are specified in calls to the registerCoarsen() function in the CoarsenAlgorithm<DIM> class.
- Parameters
-
| coarse | Coarse patch containing destination data. |
| fine | Fine patch containing source data. |
| src_data_ids | List of integer patch data ids to preprocess. |
| coarse_box | Box region on coarse patch into which data is coarsened. |
| ratio | Integer vector containing ratio relating index space between coarse and fine patches. |
◆ postprocessCoarsen()
Perform user-defined coarsening operations. This member function is called after standard coarsening operations (expressed using concrete subclasses of the CoarsenOperator<DIM> base class).
The postprocess function should move data from the source components on the fine patch into the source components on the coarse patch in the specified coarse box region. Recall that the source components are specified in calls to the registerCoarsen() function in the CoarsenOperator<DIM> class.
- Parameters
-
| coarse | Coarse patch containing destination data. |
| fine | Fine patch containing source data. |
| src_data_ids | List of integer patch data ids to preprocess. |
| coarse_box | Box region on coarse patch into which data is copied. |
| ratio | Integer vector containing ratio |
The documentation for this class was generated from the following file: