[Getfem-users] Warning: Inefficient addition of element in rsvector

[David Danan]

Dear Getfem users,

i wanted to solve a simple laplacian with pure Neumann boundary condition on a cube. 
In order to do so, since the problem is ill-posed, i added the following constraint on the unknown u
 int_omega u dx=0
via a Lagrangian multiplier.

In the code, the left-hand side is defined by
  Model.add_fem_variable("u", mf_u);
  getfem::add_linear_generic_assembly_brick(model, mim, "(Grad_u.Grad_Test_u)");
  model.add_fixed_size_variable("lambda", 1);
  getfem::add_linear_generic_assembly_brick(model, mim, "(lambda*Test_u)");
  getfem::add_linear_generic_assembly_brick(model, mim, "(u*Test_lambda)");

If i am not mistaken, it should correspond to the left hand side described in
https://fenicsproject.org/olddocs/dolfin/1.3.0/python/demo/documented/neumann-poisson/python/documentation.html
for instance
 
As a result, the mean value is relatively small in comparison with u (while not exactly 0). 

However, when i run the program, i get the following warning:
Level 2 Warning in ../../src/gmm/gmm_vector.h, line 592: Inefficient addition of element in rsvector with **** non-zero entries

Did i do something wrong? If i didn't, is it possible to disable this warning?

Thanks in advance,
David.