January 2019 - firedrake - imperial.ac.uk

Interpolating/projecting between meshes
by Sentz, Peter 31 Jan '19

31 Jan '19

Hello, In Fenics/Dolfin, I can interpolate between different meshes, even with less basic elements like Raviart-Thomas. For example: from fenics import * mesh1 = UnitSquareMesh(1,1) mesh2 = UnitSquareMesh(2,2) V1 = FunctionSpace(mesh1,'RT',1) V2 = FunctionSpace(mesh2,'RT',1) u = Expression(('x[0] + 1','x[1]'),degree = 1) U = interpolate(u,V1) Uf = interpolate(U,V2) Interpolating the function 'U' that is defined on one mesh can be moved to another in a fairly simple matter. Trying something similar in Firedrake with 'project' instead of 'interpolate' results in an error message. A different error appears for nodal FE functions as well, when using 'interpolate'. However, in that case, I have developed a work-around using mesh.coordinates.dat.data. However, I can't see how to extend this to nodal DG or Raviart-Thomas FE functions. Can you access the coordinates of the degrees of freedom of an arbitrary function space? Barring that, is there some mechanism to move between meshes if I use a MeshHierarchy, for example? Thanks, Peter Sentz

1 0

Question about saddle_point_systems.py demo
by Justin Chang 30 Jan '19

30 Jan '19

Hi guys, (This might actually be a PETSc question, but I know there are some PETSc folks that also peruse through this mailing list) I'm looking again through the Preconditioning saddle-point systems <https://firedrakeproject.org/demos/saddle_point_systems.py.html> demo, one example shown was providing the scour complement approximation for the mixed poisson problem. I see that the bilinear form for the discontinuous laplacian (plus the flux mass term) is set as the preconditioning Mat for KSPSetOperators() and that these PETSc options are utilized: parameters = { "ksp_type": "gmres", "ksp_rtol": 1e-8, "pc_type": "fieldsplit", "pc_fieldsplit_type": "schur", "pc_fieldsplit_schur_fact_type": "full", "pc_fieldsplit_off_diag_use_amat": True, "fieldsplit_0_ksp_type": "preonly", "fieldsplit_0_pc_type": "ilu", "fieldsplit_1_ksp_type": "preonly", "fieldsplit_1_pc_type": "hypre" } This raises a couple questions from myself: 1) Why is "pc_fieldsplit_off_diag_use_amat": True necessary? In the aP term, I see a dot(sigma, tau)*dx term in the bilinear, so wouldn't that actually be the amat? 2) I don't see -pc_fieldsplit_schur_precondition set anywhere meaning the default is the a11 matrix. In the original bilinear form, a11 would be zero but because you passed in aP, a11 now pick up the laplacian term? 3) Can this easily have been done instead using "pc_fieldsplit_schur_precondition": "user" and the user would provide a python context that builds only the schur complement approximation and applies the action? Thanks, Justin

2 2

heat equation demo
by Shawn W Walker 29 Jan '19

29 Jan '19

Hello. Is there a demo on solving the linear heat equation with time-dependent RHS (bulk term) and time-dependent BCs? I'm having some trouble figuring this out. -Shawn

3 2

Firedrake Installation on Cluster
by Shrimali, Bhavesh 28 Jan '19

28 Jan '19

Hello, I have been trying to install firedrake on our campus cluster, however I’ve had little success so far. I am not from a CS background so mostly have been googling stuff and trying different things. I am trying to make use of the modules available on our cluster, and use the PETSc already installed on the cluster. export I_MPI_CC=gcc export I_MPI_CXX=g++ export I_MPI_F90=gfortran export I_MPI_FC="$I_MPI_F90" export I_MPI_F77="$I_MPI_F90" export PETSC_CONFIGURE_OPTIONS=--download-fblaslapack export PETSC_CONFIGURE_OPTIONS=--download-scalapack export PETSC_CONFIGURE_OPTIONS=--download-mumps export PETSC_CONFIGURE_OPTIONS=--download-superlu_dist export PETSC_CONFIGURE_OPTIONS=--download-hypre curl -O https://raw.githubusercontent.com/firedrakeproject/firedrake/master/scripts… python3 firedrake-install --no-package-manager --honour-pythonpath --honour-petsc-dir This throws exception at the following point (Also attached separately as firedrakeErrortxt) Removing existing h5py installations Installing h5py/ Installing libspatialindex Cloning libspatialindex Failed to clone libspatialindex using ssh, falling back to https. Successfully cloned repository libspatialindex. Checking out branch master Successfully checked out branch master Installing petsc4py/ Traceback (most recent call last): File "firedrake-install", line 1145, in <module> install(p+"/") File "firedrake-install", line 642, in install run_pip_install(["--ignore-installed", package]) File "firedrake-install", line 600, in run_pip_install check_call(pipinstall + pipargs) File "firedrake-install", line 438, in check_call log.debug(subprocess.check_output(arguments, stderr=subprocess.STDOUT, env=env).decode()) File "/usr/local/python/3.7.0/lib/python3.7/subprocess.py", line 376, in check_output **kwargs).stdout File "/usr/local/python/3.7.0/lib/python3.7/subprocess.py", line 468, in run output=stdout, stderr=stderr) subprocess.CalledProcessError: Command '['/projects/meca/bshrima2/packages/firedrake/bin/python', '-m', 'pip', 'install', '--no-binary', 'mpi4py', '--no-deps', '-vvv', '--ignore-installed', 'petsc4py/']' returned non-zero exit status 1. The complete log file is also attached along with this email. Please let me know if you can be of some help, as I have received little-to-no help from the people managing the cluster. Also let me know if this is due to an error on my side while running the installer. Thank you --Bhavesh Sent from Mail<https://go.microsoft.com/fwlink/?LinkId=550986> for Windows 10

2 4

Re: [firedrake] Failed Firedrake Installation
by Lawrence Mitchell 16 Jan '19

16 Jan '19

Hi Anthony, > On 16 Jan 2019, at 12:29, Webster, Anthony <anthony.webster15(a)imperial.ac.uk> wrote: > > Hello Prof. Mitchell, (That's definitely more title than I have!) > I am on clustor2 so I shouldn't have a quota, that was the reason I moved from clustor1 to clustor2 in the first place. > > The installation is in the home directory (the default directory when I log in). OK, so we need to try and figure out more details of what failed. Since the "pip install" of petsc blows away the configure.log that will provide us with relevant information, we'll need to do some debugging manually. The (failed) installation will have left a firedrake/src/petsc directory around. So can you please: cd firedrake/src/petsc and run: ./configure --prefix=/home/clustor2/ma/a/aw3315/firedrake/src/firedrake/scripts/firedrake/lib/python3.5/site-packages/petsc --PETSC_ARCH=arch-python-linux-x86_64 --with-shared-libraries=1 --with-debugging=0 --with-c2html=0 --with-cc=mpicc --with-cxx=mpicxx --with-fc=mpif90 --download-eigen=/home/clustor2/ma/a/aw3315/firedrake/src/firedrake/scripts/firedrake/src/eigen-3.3.3.tgz --with-fortran-bindings=0 --download-chaco --download-metis --download-parmetis --download-scalapack --download-hypre --download-mumps --with-zlib --download-netcdf --download-hdf5 --download-pnetcdf --download-exodusii This should fail in the same way, and when it does, please post the resulting configure.log Thanks, Lawrence

1 0

Failed Firedrake Installation
by Webster, Anthony 16 Jan '19

16 Jan '19

Hello Sir or Madam, I have a failed Firedrake installation on the Maths computing cluster. Log is attached. Regards, Anthony Webster

2 1

SLEPc + mixed
by Cotter, Colin J 15 Jan '19

15 Jan '19

Dear all, I'd like to find the maximum eigenvalue from a mixed generalised eigenvalue problem. Has anyone tried this? I can follow the QG slepc example but I would need to assemble monolithic PETSc matrices. Is there a way to assemble monolithic PETSc matrices for mixed systems? cheers --cjc

3 3

Re: [firedrake] question about installing tests
by Ham, David A 08 Jan '19

08 Jan '19

Ah, misunderstanding. firedrake-status is a script, you should run it, not run more on it. Please type firedrake-status at the command line and post the output. Regards, David On 08/01/2019, 14:18, "Shawn W Walker" <walker(a)lsu.edu> wrote: Hello Lawrence. When I type " more firedrake-status" I get what is at the bottom. I'll need to rerun the tests and keep the output. Most things were fine, but there were a few F's. The platform is linux. I'll have to find out what kind of linux. -Shawn chaos1:~/firedrake/bin> more firedrake-status #!/U1/accounts/walker/firedrake/bin/python # EASY-INSTALL-DEV-SCRIPT: 'firedrake==0.13.0+2409.g011aa013','firedrake-status' __requires__ = 'firedrake==0.13.0+2409.g011aa013' __import__('pkg_resources').require('firedrake==0.13.0+2409.g011aa013') __file__ = '/U1/accounts/walker/firedrake/src/firedrake/scripts/firedrake-status ' with open(__file__) as f: exec(compile(f.read(), __file__, 'exec')) ==================================================== -----Original Message----- From: Lawrence Mitchell [mailto:wence@gmx.li] Sent: Tuesday, January 8, 2019 5:23 AM To: Shawn W Walker <walker(a)lsu.edu> Cc: firedrake(a)imperial.ac.uk Subject: Re: [firedrake] question about installing tests Dear Shawn, > On 7 Jan 2019, at 20:16, Shawn W Walker <walker(a)lsu.edu> wrote: > > Hello. > > I’ve been installing FireDrake as well as the tests. While installing the tests/demos, it sometimes prints an “F” (meaning fail I presume). Is this a problem? I have seen some occasional failures of the demos (but only when run as tests). But I thought I had fixed recently. Can you provide any more details? Particularly, which test fails (and are there any further details). Can you also provide the output of "firedrake-status"? A failing test is a problem, but we need to debug a little to figure out what is going on. It's possible that the installation was successful nonetheless, and we just have a bug in one of the demos, or there might be something that needs to be fixed. Thanks, Lawrence

2 1

question about installing tests
by Shawn W Walker 08 Jan '19

08 Jan '19

Hello. I've been installing FireDrake as well as the tests. While installing the tests/demos, it sometimes prints an "F" (meaning fail I presume). Is this a problem? -SW ---------------------------------------------------------------- Shawn W. Walker Associate Professor Department of Mathematics Louisiana State University Lockett Hall 210, Baton Rouge, LA 70803 http://www.math.lsu.edu/~walker/ ----------------------------------------------------------------

3 2