Issue with interpolating a gathered function

[nami]

Hi,

I have a minimal code below. I need to gather the function u to all processes and then interpolate it on the functionspaces in each process. From the print part you can see that the gathering is happening correctly. However, the interpolation gives me the error I have attached below.

from dolfin import *
from numpy import array
from mpi4py import MPI

mesh = UnitSquareMesh(2, 2)
V = FunctionSpace(mesh, "CG", 1)

u = interpolate(Expression("x[0]",degree=0), V)

u_vec = u.vector()
v_vec = Vector(MPI.COMM_SELF)
u_vec.gather(v_vec, array(range(V.dim())))
print(MPI.COMM_WORLD.Get_rank(), u_vec.get_local(),v_vec.get_local())

U = Function(V)
U = interpolate(v_vec,V)

An the error is:

'dolfin.cpp.la.Vector' object has no attribute '_cpp_object'

Any help would be much appreciated.

[dokken]

I am not certain I understand what you would like to achieve here. Note that the mesh is distributed over the set of processors. Therefore, there is only one function space (with global number of dofs=9) existing.

To create a mesh on each processor, see for instance:

Parallel for-loop with multiple solve-calls, MPI

It still seems like you haven’t read the previous post I linked to. You are in your example creating the mesh with the global MPI communicator, aka. a distributed mesh. What you could do is to use the comm_self communicator to solve one problem with one mesh per processor. Then you can gather the local vector (which is the array containing your solution, and post-process it). See the example below: from dolfin import * import numpy as np comm = MPI.comm_world size = comm.Get_size() rank = co…

Additionally, you should access the function vector directly, and not through interpolation, if getting arrays from another processor.