Mailman 3 April 2010 - SfePy - {{ block.super }

Initial basic read support for MED file format (salome platform and pythonocc)
by Logan Sorenson 03 May '10

03 May '10

Hi All, Please find attached a patch to sfepy/fem/meshio.py for basic read support for MED format meshes. I'm attaching the patch here since github seems to be down in the US. MED is the native mesh format for the very powerful open-source meshing tools salome [1] and pythonocc [2]. Actually, it is very incomplete, but seems to work with the provided example mesh (Mesh_1.med). I only tried loading the mesh with postproc.py, so I haven't run any calculations on it yet (I assume that should work ok). So far, it only supports 3D 8-node hexahedral elements, so I suppose it is more of an example of reading the format than an actual final production code. :) I hope it can be generalized to work with more MED meshes. I found some French documentation on MED in the source packages at [1]; I used google docs to translate it to English, so if someone wants a copy of that, let me know. I'm looking for a better English version of these docs. Best regards, Logan [1] http://www.salome-platform.org [2] http://www.pythonocc.org -- You received this message because you are subscribed to the Google Groups "sfepy-devel" group. To post to this group, send email to sfepy...(a)googlegroups.com. To unsubscribe from this group, send email to sfepy-devel...(a)googlegroups.com. For more options, visit this group at http://groups.google.com/group/sfepy-devel?hl=en.

2 12

docstring standard + a bit of git
by Robert Cimrman 29 Apr '10

29 Apr '10

Hi All, The developer guide [1] now contains info on our doctring standard + info on how to introduce yourself properly to git: Introduce yourself to git and make (optionally) some handy aliases either in .gitconfig in your home directory (global setting for all your git projects), or directly in .git/config in the repository: [user] email = ma...(a)mail.org name = Name Surname [color] ui = auto interactive = true [alias] ci = commit di = diff --color-words st = status co = checkout r. [1] http://docs.sfepy.org/doc-devel/developer_guide.html -- You received this message because you are subscribed to the Google Groups "sfepy-devel" group. To post to this group, send email to sfepy...(a)googlegroups.com. To unsubscribe from this group, send email to sfepy-devel...(a)googlegroups.com. For more options, visit this group at http://groups.google.com/group/sfepy-devel?hl=en.

1 0

refactored integrals and quadratures
by Robert Cimrman 28 Apr '10

28 Apr '10

Hi, To address issue 125 [0], I have refactored integrals and quadratures, see [1]. Now it remains to write some tests into tests/test_quadratures.py, and add more quadratures. it resulted in: - much simpler code - quadratures can now be easily added - integral 'family' is just an arbitrary name given by user [2] was updated, too - quadratures.py, intergrals.py were added to sfepy.fem package section. Bug reports welcome! r. [0] http://code.google.com/p/sfepy/issues/detail?id=125 [1] http://github.com/rc/sfepy/ [2] http://docs.sfepy.org/doc-devel/developer_guide.html#module-index -- You received this message because you are subscribed to the Google Groups "sfepy-devel" group. To post to this group, send email to sfepy...(a)googlegroups.com. To unsubscribe from this group, send email to sfepy-devel...(a)googlegroups.com. For more options, visit this group at http://groups.google.com/group/sfepy-devel?hl=en.

1 0

new debug()
by Robert Cimrman 28 Apr '10

28 Apr '10

FYI: the debug() function [1] was significantly improved - it uses IPython/ipdb if those are available, so you get colors, tab completion and all that fancy stuff while debugging. r. [1] http://docs.sfepy.org/doc-devel/src/sfepy/base/base.html -- You received this message because you are subscribed to the Google Groups "sfepy-devel" group. To post to this group, send email to sfepy...(a)googlegroups.com. To unsubscribe from this group, send email to sfepy-devel...(a)googlegroups.com. For more options, visit this group at http://groups.google.com/group/sfepy-devel?hl=en.

1 0

SfePy beginner's tutorial on wiki
by Andre Smit 28 Apr '10

28 Apr '10

http://code.google.com/p/sfepy/wiki/Primer Rough draft - please edit as you see fit. Comments and criticism welcome :) Andre -- You received this message because you are subscribed to the Google Groups "sfepy-devel" group. To post to this group, send email to sfepy...(a)googlegroups.com. To unsubscribe from this group, send email to sfepy-devel...(a)googlegroups.com. For more options, visit this group at http://groups.google.com/group/sfepy-devel?hl=en.

2 3

new section in docs: how to contribute
by Robert Cimrman 27 Apr '10

27 Apr '10

Hi, I have written some notes summarizing how to contribute to our project, see [1]. Updates to the text from other sides are welcome! r. [1] http://docs.sfepy.org/doc-devel/developer_guide.html -- You received this message because you are subscribed to the Google Groups "sfepy-devel" group. To post to this group, send email to sfepy...(a)googlegroups.com. To unsubscribe from this group, send email to sfepy-devel...(a)googlegroups.com. For more options, visit this group at http://groups.google.com/group/sfepy-devel?hl=en.

3 12

newby
by D-rod 21 Apr '10

21 Apr '10

Dear sfepy-devel group, I have read with interest the home website of the sfepy project. I would like to give it a try, however I have a question about mesh generation. I am downloading Mayavi2 at the moment. As far as I know it is the best front end to generate and visualise VTK files. However, I would like to know if you have a better suggestion for a 'free' mesh generator software. regards, David -- You received this message because you are subscribed to the Google Groups "sfepy-devel" group. To post to this group, send email to sfepy...(a)googlegroups.com. To unsubscribe from this group, send email to sfepy-devel...(a)googlegroups.com. For more options, visit this group at http://groups.google.com/group/sfepy-devel?hl=en.

3 2

Extract Jacobian of a specific equation
by Dominique 21 Apr '10

21 Apr '10

Hi Robert, >From a given configuration file, I think I am able to extract the Jacobian (called the "tangent matrix" in the code) and the right-hand side as follows: def get_problem(fName=__file__): """ Read the configuration specified in the file specified in argument (the present file by default) and return an initialized problem constructed from it along with an initial state vector. """ from sfepy.base.conf import ProblemConf, get_standard_keywords from sfepy.fem import eval_term_op, ProblemDefinition required, other = get_standard_keywords() # Use this file as the input file to gather configuration. conf = ProblemConf.from_file(fName, required, other) # Create a ProblemDefinition instance from current configuration. problem = ProblemDefinition.from_conf(conf, init_variables=True, init_equations=True) state = problem.create_state_vector() problem.init_variables(state) problem.time_update(None) return (problem, state) def get_jacobian(problem, state=None): """ Return the Jacobian matrix of the system in SciPy CSR format. Input arguments are as returned by `get_problem()`. """ # Instantiate a problem evaluator. evaluator = problem.get_evaluator(mtx=problem.mtx_a) if state is None: state = problem.create_state_vector() jac = evaluator.eval_tangent_matrix(state) return jac Am I missing anything here (e.g., I'm not sure the BCs and ICs are taken into account)? Now I'd like to extract the contribution of a particular equation to this Jacobian. How can I do that? This is what I've tried but it seems to return the whole Jacobian: def get_equation_jacobian(problem, equation, state=None): """ Return the Jacobian matrix of a specific equation in SciPy CSR format. """ from sfepy.fem.evaluate import assemble_matrix evaluator = problem.get_evaluator(mtx=problem.mtx_a) if state is None: state = problem.create_state_vector() problem.variables.data_from_state(state) jac = problem.mtx_a jac = assemble_matrix(jac, equation, problem.variables, problem.materials, **evaluator.data) return jac In fact, this function returns None, but problem.mtx_a contains the entire Jacobian. I suppose I should specify a different matrix when I instantiate the evaluator, but where can I discover its size? Thanks! Dominique -- You received this message because you are subscribed to the Google Groups "sfepy-devel" group. To post to this group, send email to sfepy...(a)googlegroups.com. To unsubscribe from this group, send email to sfepy-devel...(a)googlegroups.com. For more options, visit this group at http://groups.google.com/group/sfepy-devel?hl=en.

2 5

sfepy-issues mailing list
by Robert Cimrman 21 Apr '10

21 Apr '10

Just a little reminder: there is also the sfepy-issues mailing list, which reports automatically any changes of the issues (new issues, updates, comments etc.). Often the comments can help you solve some problem, so you may want to join that list too, if you read this one - it complements the information given here. r. -- You received this message because you are subscribed to the Google Groups "sfepy-devel" group. To post to this group, send email to sfepy...(a)googlegroups.com. To unsubscribe from this group, send email to sfepy-devel...(a)googlegroups.com. For more options, visit this group at http://groups.google.com/group/sfepy-devel?hl=en.

1 0

Material not found
by Dominique 20 Apr '10

20 Apr '10

Hi there, I'm experimenting with a mixed finite element formulation of the Poisson equation to become more familiar with the terms in SfePy. The equations are int( v . grad(p) ) + int ( f p ) = 0, int( c u div(q) ) = int( v . q ), where u is a scalar unknown, v is a vector unknown, p is a scalar test function and q is a vector test function. I'm getting the error message: IndexError: material(s) "v" not found! My code is as follows: material_1 = { 'name' : 'coef', 'region' : 'Omega', 'values' : {'val' : 1.0}, } material_2 = { 'name' : 'source_term', 'region' : 'Omega', 'mode' : 'function', 'function' : 'source_term_func', } field_1 = { 'name' : 'temperature', 'dim' : (1,1), # First component = # degrees of freedom per node 'flags' : (), 'domain' : 'Omega', 'bases' : {'Omega' : '2_3_P1'} } field_2 = { 'name' : 'grad_temperature', 'dim' : (2,1), 'flags' : (), 'domain' : 'Omega', 'bases' : {'Omega' : '2_3_P2'} } variable_1 = { 'name' : 'u', 'kind' : 'unknown field', 'field' : 'temperature', 'order' : 0, # Appears first in the global vector of unknowns. } variable_2 = { 'name' : 'p', 'kind' : 'test field', 'field' : 'temperature', 'dual' : 'u', # This is a test function corresponding to u. } variable_3 = { 'name' : 'v', 'kind' : 'unknown field', 'field' : 'grad_temperature', 'order' : 1, } variable_4 = { 'name' : 'q', 'kind' : 'test field', 'field' : 'grad_temperature', 'dual' : 'v', # This is a test function corresponding to v. } integral_1 = { 'name' : 'i1', 'kind' : 'v', 'quadrature' : 'gauss_o2_d2', } equations = { 'Temperature1' : """dw_stokes.i1.Omega( v, p ) = dw_volume_lvf.i1.Omega( source_term.val, p )""", 'Temperature2' : """dw_stokes.i1.Omega( q, u ) = dw_volume_lvf.i1.Omega( v, q )""", } The whole log is: sfepy: left over: ['source_term_func', 'data_dir', 'nm', '_filename', '__package__', '__doc__', '__builtins__', 'ebc_sin', '__file__', '__name__', 'ebc_sin2', 'amplitude'] sfepy: reading mesh (/Users/dpo/local/src/sfepy/meshes/2d/ square_unit_tri.mesh)... sfepy: ...done in 0.00 s /Users/dpo/local/lib/python2.6/site-packages/numpy/lib/utils.py:140: DeprecationWarning: `unique1d` is deprecated! warnings.warn(depdoc, DeprecationWarning) sfepy: setting up domain edges... sfepy: ...done in 0.00 s sfepy: creating regions... sfepy: Omega sfepy: Gamma_Top sfepy: Gamma_Bottom sfepy: ...done in 0.01 s sfepy: equation "Temperature2": sfepy: dw_stokes.i1.Omega( q, u ) = dw_volume_lvf.i1.Omega( v, q ) sfepy: equation "Temperature1": sfepy: dw_stokes.i1.Omega( p, v ) = dw_volume_lvf.i1.Omega( source_term.val, p ) Traceback (most recent call last): File "/Users/dpo/local/src/sfepy/simple.py", line 106, in <module> main() File "/Users/dpo/local/src/sfepy/simple.py", line 99, in main app = SimpleApp( conf, options, output_prefix ) File "/Users/dpo/local/src/sfepy/sfepy/applications/simple_app.py", line 50, in __init__ **kwargs ) File "/Users/dpo/local/src/sfepy/sfepy/fem/problemDef.py", line 88, in from_conf obj.set_equations( conf.equations ) File "/Users/dpo/local/src/sfepy/sfepy/fem/problemDef.py", line 168, in set_equations self.materials, user ) File "/Users/dpo/local/src/sfepy/sfepy/fem/equations.py", line 142, in setup_terms eq.setup_terms(regions, variables, materials, self.caches, user) File "/Users/dpo/local/src/sfepy/sfepy/fem/equations.py", line 310, in setup_terms self.check_term_args( variables, materials, user ) File "/Users/dpo/local/src/sfepy/sfepy/fem/equations.py", line 260, in check_term_args term.check_args(variables, materials, user) File "/Users/dpo/local/src/sfepy/sfepy/terms/terms.py", line 330, in check_args 'material(s) "%s" not found!') File "/Users/dpo/local/src/sfepy/sfepy/base/base.py", line 780, in check_names raise IndexError(msg % missing) IndexError: material(s) "v" not found! What am I doing wrong? Thanks! Dominique -- You received this message because you are subscribed to the Google Groups "sfepy-devel" group. To post to this group, send email to sfepy...(a)googlegroups.com. To unsubscribe from this group, send email to sfepy-devel...(a)googlegroups.com. For more options, visit this group at http://groups.google.com/group/sfepy-devel?hl=en.

2 5

SfePy April 2010