|
VTK
9.7.20261014
|
A function space for basis functions defined on corner vertices of cells. More...
Functions | |
| bool VTKFILTERSCELLGRID_EXPORT | RegisterOperators () |
| Register basis-function operators for the "HGRAD" function space with vtkDGCell. | |
| bool VTKFILTERSCELLGRID_EXPORT | FixedOrderLagrangePoints (vtkDGCell::Shape shape, vtkStringToken basis, int nominalOrder, std::vector< std::vector< double > > &points) |
| Fetch parametric coordinates where Lagrange basis functions exactly interpolate a value. | |
| bool VTKFILTERSCELLGRID_EXPORT | FixedOrderGaussPoints (vtkDGCell::Shape shape, int nominalOrder, std::vector< std::vector< double > > &points) |
| Fetch parametric coordinates where the "G" (Gauss-point) basis functions interpolate. | |
A function space for basis functions defined on corner vertices of cells.
This function space is analogous to the traditional Lagrange shape functions.
| bool VTKFILTERSCELLGRID_EXPORT vtk::basis::hgrad::RegisterOperators | ( | ) |
Register basis-function operators for the "HGRAD" function space with vtkDGCell.
| bool VTKFILTERSCELLGRID_EXPORT vtk::basis::hgrad::FixedOrderLagrangePoints | ( | vtkDGCell::Shape | shape, |
| vtkStringToken | basis, | ||
| int | nominalOrder, | ||
| std::vector< std::vector< double > > & | points ) |
Fetch parametric coordinates where Lagrange basis functions exactly interpolate a value.
This only returns true when the shape, basis, and nominalOrder match a fixed-order basis function. You should prefer using the vtkLagrangePoints calculator as that class calls this method when appropriate but also handles arbitrary-order basis functions.
This function exists because programmatically generating parametric coordinates for serendipity and other finite elements would be difficult.
| bool VTKFILTERSCELLGRID_EXPORT vtk::basis::hgrad::FixedOrderGaussPoints | ( | vtkDGCell::Shape | shape, |
| int | nominalOrder, | ||
| std::vector< std::vector< double > > & | points ) |
Fetch parametric coordinates where the "G" (Gauss-point) basis functions interpolate.
The G bases are nodal on the Gauss points rather than on a uniform lattice, so they need their own point set. Returns true and fills points for the shapes whose G basis is nodal; returns false for the tetrahedron and pyramid, whose G bases are modal and so interpolate nowhere (see TestCellGridLagrangePoints).
Prefer the vtkLagrangePoints calculator, which calls this when appropriate.