Francesc Verdugo

Provide the option to chose one or another for the global dofs. For the local dofs "node major" is already enough

I have realized that removing `testargs` is not as easy as expected. Thanks to `testargs`, we can provide a default implementation of `return_value` for the gradient and the gradient of...

I remove this issue from the milestone v0.15.0. I do not close it since the doc for `testargs` and `return_value` can be added anyway at any time.

The is no specific reason. We can fix this for the next (breaking) release

Hi @bhaveshshrimali `S(F) = ∇(ψ)(F)` This does not work since the AD machinery is for functions taking `Point` i.e. `VectorValue` as argument, not 2nd order tensors. To auto generate a...

BTW, `μ` and `λ ` are not defined in the MWE. Can you update the code with some values?

> A side question: how are you computing the individual time for the jacobian compilation? I explicitly call `jac`. ```julia u = zero(U) v = get_cell_shapefuns(V) du = get_cell_shapefuns_trial(U) j...

Hi @rveltz , nice that you have interested in the project. I don't really understand what you mean by "parameter dependent". Are you willing to include state variables in your...

@rveltz As you say, in the example I have provided, we recompute from scratch everything (e.g. the jacopian). However, this can be easily improved by properly using the API of...

Provably you also need `dF_dp(u,p)` (jacobian wrt `p`). I think computing this is also possible. In any case, I believe that all what you need can be done now, but...