On 11/12/2014 10:25 PM, Serge-Étienne Parent wrote:
Hi Robert,
I compiled everything from the beginning. The test failed for tests/test_elasticity_small_strain.py and stops by a segmentation fault at tests/test_input_stokes_slip_bc.py, but the import return no error! Please find the debug info below.
So the other tests pass, right? The first example that fails uses a new feature (vtk probes) that is based on pyvtk. There might be some incompatibility or version problem - could you report the versions you use?
The other fails because of a problem in some versions of scipy.sparse [1].
r.
[1] https://code.google.com/p/pythonxy/issues/detail?id=747
Many**2 thanks!
Essi Parent Universite Laval, Quebec, Canada
*test_input_stokes_slip_bc.py*
test instance prepared (1 test(s)) ... solving ../examples/linear_elasticity/linear_elastic_probes.py... sfepy: reading mesh [line2, tri3, quad4, tetra4, hexa8] (/home/useraccount/Programmes/sfepy/meshes/3d/cylinder.mesh)... sfepy: ...done in 0.00 s sfepy: creating regions... sfepy: SomewhereTop sfepy: Right sfepy: Omega sfepy: Left sfepy: ...done in 0.02 s sfepy: equation "balance_of_forces": sfepy: dw_lin_elastic_iso.i.Omega(solid.lam, solid.mu, v, u) = 0 sfepy: using solvers: ts: no ts nls: newton ls: ls sfepy: updating variables... sfepy: ...done sfepy: setting up dof connectivities... sfepy: ...done in 0.00 s sfepy: matrix shape: (886, 886) sfepy: assembling matrix graph... sfepy: ...done in 0.01 s sfepy: matrix structural nonzeros: 31052 (3.96e-02% fill) sfepy: updating materials... sfepy: solid sfepy: ...done in 0.00 s sfepy: nls: iter: 0, residual: 4.778054e-03 (rel: 1.000000e+00) sfepy: rezidual: 0.00 [s] sfepy: solve: 0.01 [s] sfepy: matrix: 0.01 [s] sfepy: nls: iter: 1, residual: 7.452809e-18 (rel: 1.559800e-15) sfepy: equation "tmp": sfepy: ev_cauchy_strain.i.Omega( u ) sfepy: updating materials... sfepy: ...done in 0.00 s sfepy: equation "tmp": sfepy: ev_cauchy_stress.i.Omega( solid.D, u ) sfepy: updating materials... sfepy: solid sfepy: ...done in 0.00 s <type 'exceptions.AttributeError'> Traceback (most recent call last): File "run_tests.py", line 206, in <module> main() File "run_tests.py", line 197, in main run_tests(stats, dirname, [filename]) File "run_tests.py", line 131, in run_tests n_fail, n_total, test_time = run_test(conf_name, options) File "run_tests.py", line 99, in run_test ok, n_fail, n_total = test.run(options.debug) File "/home/useraccount/Programmes/sfepy/sfepy/base/testing.py", line 58, in run ret = test_method() File "/home/useraccount/Programmes/sfepy/tests/tests_basic.py", line 73, in test_input post_process_hook_final=self.post_process_hook_final) File "/home/useraccount/Programmes/sfepy/sfepy/applications/pde_solver_app.py",
useraccount@linux:~/Programmes/sfepy> python run_tests.py --debug tests/test_elasticity_small_strain.pypython run_tests.py --debug tests/test_input_linear_elastic_probes.py <<< directory: tests, test files: 0 <<< directory: , test files: 0 <<< directory: tests, test files: 1 <<< tests/test_input_linear_elastic_probes.py sfepy: left over: ['verbose', '__builtins__', '_filename', '__file__', '__package__', 'TestInput', 'input_name', '__name__', 'output_name', '__doc__'] sfepy: left over: ['lam', 'plt', 'Struct', '__builtins__', '__file__', 'deepcopy', 'fm', 'verbose', 'par_fun', 'mu', 'data_dir', 'nm', 'solid', '__name__', 'Probe', '__package__', 'stiffness_from_lame', '_filename', 'post_process', 'os', '__doc__'] line 55, in solve_pde return app(nls_status=nls_status) File "/home/useraccount/Programmes/sfepy/sfepy/applications/application.py", line 29, in call_basic return self.call(**kwargs) File "/home/useraccount/Programmes/sfepy/sfepy/applications/pde_solver_app.py", line 220, in call nls_status=nls_status) File "/home/useraccount/Programmes/sfepy/sfepy/solvers/ts_solvers.py", line 37, in __call__ file_per_var=None) File "/home/useraccount/Programmes/sfepy/sfepy/discrete/problem.py", line 725, in save_state out = post_process_hook(out, self, state, extend=extend) File "/home/useraccount/Programmes/sfepy/examples/linear_elasticity/linear_elastic_probes.py", line 99, in post_process probe = Probe(out, mesh) File "/home/useraccount/Programmes/sfepy/sfepy/postprocess/probes_vtk.py", line 34, in __init__ self.vtkprobe.SetSource(self.vtkdata) AttributeError: SetSource
*test_input_linear_elastic_probes.py*
useraccount@linux:~/Programmes/sfepy> python run_tests.py --debug tests/test_input_stokes_slip_bc.py <<< directory: tests, test files: 1 <<< tests/test_input_stokes_slip_bc.py sfepy: left over: ['verbose', '__builtins__', '_filename', '__file__', '__package__', 'TestInput', 'input_name', '__name__', 'output_name', '__doc__'] sfepy: left over: ['verbose', '__builtins__', 'dims', 'shape', '__file__', '__name__', 'nm', 'UserMeshIO', 'define_box_regions', 'gen_block_mesh', '__package__', 'get_u_d', '_filename', '__doc__', 'mesh_hook']
test instance prepared (1 test(s)) ... solving ../examples/navier_stokes/stokes_slip_bc.py... sfepy: reading mesh [user] (function:mesh_hook)... sfepy: ...done in 0.00 s sfepy: creating regions... sfepy: Right sfepy: Bottom sfepy: Far sfepy: Corners sfepy: Top sfepy: Omega sfepy: Near sfepy: Left sfepy: Gamma1_f sfepy: Edges_v sfepy: Gamma2_f sfepy: Gamma_f sfepy: Gamma_v sfepy: Inlet_f sfepy: ...done in 0.06 s sfepy: equation "balance": sfepy: dw_div_grad.5.Omega(m.nu, v, u) - dw_stokes.5.Omega(v, p) + dw_surface_dot.5.Gamma1_f(m.beta, v, u) + dw_surface_dot.5.Gamma2_f(m.beta, v, u) = + dw_surface_dot.5.Gamma1_f(m.beta, v, u_d) sfepy: equation "incompressibility": sfepy: dw_laplace.5.Omega(m.mu, q, p) + dw_stokes.5.Omega(u, q) = 0 sfepy: using solvers: ts: no ts nls: newton ls: ls_d sfepy: updating variables... sfepy: data of u_d set by get_u_d() sfepy: ...done sfepy: setting up dof connectivities... sfepy: ...done in 0.01 s sfepy: lcbc: walls sfepy: lcbc: edges Erreur de segmentation
Le mercredi 12 novembre 2014 15:14:41 UTC-5, Robert Cimrman a écrit :
Hi Serge-Étienne,
how did you compiled sfepy? We have recently fixed the documentation [1] w.r.t. the testing - you need to compile for the in-place build to run the tests in the sfepy directory.
r. [1] http://sfepy.org/doc-devel/installation.html#testing
On 11/12/2014 08:04 PM, Serge-Étienne Parent wrote:
Hi,
I tried to install SfePy on my openSUSE 13.2 64bit machine. I had a hard time installing PyTables, but I could handle this by installing it using /usr/bin/pip2.7 install tables
...after installing python-pip, hdf5-devel, python-devel and python-numpy-devel.
Everything else went smooth and SfePy compiled successfully. But most tests failed and importing in python did not work.