Configuration of the adaptive sampling.
The main setting is max_nlevels, which states the maximum number of levels that elements will be refined.
Type | Visibility | Attributes | Name | Initial | |||
---|---|---|---|---|---|---|---|
integer, | public | :: | max_nlevels | = | 0 |
Maximal number of levels by which any mesh element should be refined. A setting of 0 results in no sampling, and the original mesh elements will be used with the integral mean value (first degree of freedom). Higher levels provide a limit for the refinement of the mesh. Note, that even for large settings here, the overall mesh depth is restricted by the global limit, due to the available space in the integers representing the treeIDs. |
|
real(kind=rk), | public | :: | eps_osci |
Maximum allowed oscillation of the solution. For adaptive subsampling only. |
|||
integer, | public | :: | reduction_mode |
Method to use for the reduction. This may either be:
|
|||
logical, | public | :: | ignore_highmodes | = | .false. |
Indication whether to filter modes during refinement by ignoring all modes in the parent, that exceed the target polynomial degree of the child elements. This provides a simple lowpass filtering method if activated. Defaults to false. |
|
integer, | public | :: | dof_decrement | = | 1 |
Number of modes to cut off in each refinement. If the decrement mode for reduction is used, this setting will be used to cut off as many modes from the refined elements. |
|
real(kind=rk), | public | :: | dofReducFactor |
Factor to Reduce dofs for every sampling level. Can be used to avoid too drastic increase of memory consumption. For adaptive subsampling only. |
|||
logical, | public | :: | adaptiveDofReduction |
Indicator for the limitation of memory consumption. |
|||
integer, | public | :: | AbsUpperBoundLevel |
Absolute upper bound level to refine to. |
|||
type(ply_filter_element_type), | public | :: | filter_element |
Filtering the poylnomial modes during adaptive refinement. This filtering provides the possibility to change the applied filtering based on the polynomials and thereby attempting to capture discontinuities more sharply. |