Improve documentation of BiunitSimplexSymmetricFormTransformation

Fix nodal bases for Lagrange dual space to avoid symmetry problems

Before, wrong: under symmetry transformation, a boundary dof could make to a linear combination of
boundary dofs and interior dofs

Fix nodal bases for Lagrange dual space to avoid symmetry problems

Before, wrong: under symmetry transformation, a boundary dof could make to a linear combination of
boundary dofs and interior dofs
Now, correct: under every symmetry transformation, boundary dofs and interior dofs map to boundary dofs and interior dofs, respectively

Why it was wrong before: the pushforward of the k-form basis at one point did not have the same boundary/interior decomposition
as the basis at the point it mapped to under symmetry

How it was fixed: choose a basis for the k-form that, when mapped to the equilateral simplex, looks appropriately symmetric.

show more ...

Extend test of lagrange dual space symmetries.

We solve a normal system where the solution is supposed to have 0 residual:
check to make sure it does.

Merge branch 'knepley/feature-snes-ex76-div-monitor' into 'master'

Knepley/feature snes ex76 div monitor

See merge request petsc/petsc!3580

FE: Allow moment integral for P0
- Eventually, we should allow moments for other dofs

Fix pullback matrix indexing in dspacelagrange.c

Thanks to: Yang Zongze

I think the incorrect indexing wasn't causing problems before because it doesn't matter if pNk == 1.
That's the case when the

Fix pullback matrix indexing in dspacelagrange.c

Thanks to: Yang Zongze

I think the incorrect indexing wasn't causing problems before because it doesn't matter if pNk == 1.
That's the case when the form degree we are pulling back is 0 or the boundary points' dimension. So
If we're only testing d <= 3, this should only have cause problems pulling a 1-form back onto a 2D face.
I will have to check to see if we have a regression test for this case.

show more ...

Merge branch 'jose/typos' into 'master'

Fix several typos in error strings

See merge request petsc/petsc!3141

Fix several typos in error strings

Merge branch 'dalcinl/fix-null' into 'master'

Fix 0 -> NULL for pointers

See merge request petsc/petsc!3086

DM: Replace 0 -> NULL for pointers

Merge branch 'tisaac/fix-bdm-documentation' into 'master'

Correct documentation about what happened to the BDM dual spaces

See merge request petsc/petsc!2962

Use PETSCDUALSPACEBDM macro when registring bdm

Merge branch 'knepley/fix-plex-hybrid-assembly' into 'master'

Knepley/fix plex hybrid assembly

See merge request petsc/petsc!2793

Plex+FE: Updates for hybrid integration
- Added support for hybrid Jacobians
- Added PetscDSHasBdJacobian()
- Added boundary Jacobian preconditioner
- Needed a path to integrate the hybrid Jacobian p

Plex+FE: Updates for hybrid integration
- Added support for hybrid Jacobians
- Added PetscDSHasBdJacobian()
- Added boundary Jacobian preconditioner
- Needed a path to integrate the hybrid Jacobian preconditioner
- Fixed indexing in hybrid element mat update
- Be careful about the difference between intrinsic dimension and embedding dimension for hybrid integration
- Need to use rectangular transformation of gradient in case they are different
- Fix use of embedding dimension
- Make sure FEGeom struct has the dimensions set
- Pointwise function derivatives are in the embedding space
- Use PetscArrayzero()
- Change TransitiveClosure() for hybrid cells
- We get the closures of the endcaps, and then the hybrid faces, this allows degenerate cells that we need in PyLith

show more ...

Merge branch 'jed/defined-nocpp' into 'master'

PetscDefined: test whether a configuration macro is defined without #ifdef

See merge request petsc/petsc!157

PetscDefined: remove many uses of #ifdef PETSC_USE_DEBUG

Merge remote-tracking branch 'origin/maint'

Merge branch 'jed/fix-real-constant/maint' into 'maint'

DMDT: fix PetscRealConstant(1.)

See merge request petsc/petsc!2672

DMDT: fix PetscRealConstant(1.)

Reported-by: Jose E. Roman <jroman@dsic.upv.es>

Merge branch 'maint'

Merge branch 'jed/fix-warnings/maint' into 'maint'

Jed/fix warnings/maint

See merge request petsc/petsc!2661

DMDT: fix dual-space guard

Reported-by: Jose Roman <jroman@dsic.upv.es>

Merge branch 'tisaac/feature-dual-space-rework' into 'master'

Rework PetscDualSpace

See merge request petsc/petsc!2399

Add some documentation of new data structures in dspacelagrance.c

Add more documentation in dspacelagrange.c

Add more documentation in dspacelagrange.c

Add more documentation in dspacelagrange.c

Fix errors revealed by gitlabci

Use PETSC_SMALL to determine when quadrature points are the same in PETSCDUALSPACELAGRANGE

Filter floating point output

Explicitly cast PetscBool to PetscInt before

Fix errors revealed by gitlabci

Use PETSC_SMALL to determine when quadrature points are the same in PETSCDUALSPACELAGRANGE

Filter floating point output

Explicitly cast PetscBool to PetscInt before printing

Do not construct vertex indices for simplex spaces in tensor cells

Right now the only way simplex spaces in tensor cells happens is discontinuous pressures
(typically in Q2P1). In this case there is no need to compute symmetries for the pressure
dofs.

show more ...