I have a question about where you guys put the files you are working on with SfePy. For SciPy and NumPy, I am used to having an svn directory that is only for the scipy/numpy code. Once I do "sudo python setup.py install", I leave that directory and don't do any actual work in it. SfePy doesn't seem to work that way. The thing about that that is tricky for me is that I like to use Unison to sychronize my laptop and desktop. They have different processors, so I don't want to syncronize any build directories (or anything else that would get typically modified by mercurial), but I do want to sync my own stuff (meshes and input files and such).
Do you put all your personal work in folders that are underneath the main SfePy directory? Or with enough paths added to PYTHONPATH, can I put my SfePy related stuff in any directory?
Thanks,
Ryan
Ryan Krauss wrote:
I have a question about where you guys put the files you are working on with SfePy. For SciPy and NumPy, I am used to having an svn directory that is only for the scipy/numpy code. Once I do "sudo python setup.py install", I leave that directory and don't do any actual work in it. SfePy doesn't seem to work that way. The thing about that that is tricky for me is that I like to use Unison to sychronize my laptop and desktop. They have different processors, so I don't want to syncronize any build directories (or anything else that would get typically modified by mercurial), but I do want to sync my own stuff (meshes and input files and such).
I put my own stuff under input and database. You can make your own dir there (out of the repo control, if you do not wish to share your initial versions of inputs).
Do you put all your personal work in folders that are underneath the main SfePy directory? Or with enough paths added to PYTHONPATH, can I put my SfePy related stuff in any directory?
But you could put it anywhere assuming you set the PYTHONPATH as you said, and also the PATH to be able to run the scripts like simple.py.
Also note that default output directories are 'output', and 'output-tests' - you should try how this works outside of the sfepy tree. :) I will happily fix any issues you may have with.
r.
Robert Cimrman wrote:
Ryan Krauss wrote:
I have a question about where you guys put the files you are working on with SfePy. For SciPy and NumPy, I am used to having an svn directory that is only for the scipy/numpy code. Once I do "sudo python setup.py install", I leave that directory and don't do any actual work in it. SfePy doesn't seem to work that way. The thing about that that is tricky for me is that I like to use Unison to sychronize my laptop and desktop. They have different processors, so I don't want to syncronize any build directories (or anything else that would get typically modified by mercurial), but I do want to sync my own stuff (meshes and input files and such).
I put my own stuff under input and database. You can make your own dir there (out of the repo control, if you do not wish to share your initial versions of inputs).
A little explanation why sfepy works this way: I like the fact that when I change some Python code I can just run it to see new results without installing anything. It is really annoying to have to: 1. change a single letter, 2. 'python setup.py install' 3. run your code
Do you put all your personal work in folders that are underneath the main SfePy directory? Or with enough paths added to PYTHONPATH, can I put my SfePy related stuff in any directory?
But you could put it anywhere assuming you set the PYTHONPATH as you said, and also the PATH to be able to run the scripts like simple.py.
Also note that default output directories are 'output', and 'output-tests' - you should try how this works outside of the sfepy tree. :) I will happily fix any issues you may have with.
On Wed, Jul 9, 2008 at 3:54 PM, Robert Cimrman <cimr...@ntc.zcu.cz> wrote:
Robert Cimrman wrote:
Ryan Krauss wrote:
I have a question about where you guys put the files you are working on with SfePy. For SciPy and NumPy, I am used to having an svn directory that is only for the scipy/numpy code. Once I do "sudo python setup.py install", I leave that directory and don't do any actual work in it. SfePy doesn't seem to work that way. The thing about that that is tricky for me is that I like to use Unison to sychronize my laptop and desktop. They have different processors, so I don't want to syncronize any build directories (or anything else that would get typically modified by mercurial), but I do want to sync my own stuff (meshes and input files and such).
I put my own stuff under input and database. You can make your own dir there (out of the repo control, if you do not wish to share your initial versions of inputs).
A little explanation why sfepy works this way: I like the fact that when I change some Python code I can just run it to see new results without installing anything. It is really annoying to have to: 1. change a single letter, 2. 'python setup.py install' 3. run your code
Exactly, SymPy works in exactly the same way.
But in SymPy we of course support and test setup.py install, because this is the standard way of installing things, so we should make sure sfepy can do it too in the next release.
Also this will be necessary when I create a debian package for it, later when the API stabilizes and sfepy becomes more mature.
Ondrej
I can totally understand not wanting to reinstall every time you change code that you are developing. But I like to separate code that I am developing that I think of as utility code or general code from specific uses of my code applied to my research.
In trying to run my from_Robert.py file from my research directory I get this: In [3]: run simple.py from_Robert.py sfepy: warning: other missing: ['functions', 'modules', 'epbc_[0-9]+|epbcs', 'lcbc_[0-9]+|lcbcs', 'nbc_[0-9]+|nbcs'] sfepy: warning: left over: ['tractionLoad'] sfepy: reading mesh (first_mesh.1.node)... nodes: 100% |############################################| Time: 00:00:00 elements: 100% |#########################################| Time: 00:00:00 sfepy: ...done in 0.01 s
<type 'exceptions.IOError'> Traceback (most recent call last)
/home/ryan/siue/Research/SfePy/first_problem/simple.py in <module>() 107 108 if __name__ == '__main__': --> 109 main() 110 111
/home/ryan/siue/Research/SfePy/first_problem/simple.py in main() 100 setOutputPrefix( opts.outputPrefix ) 101 --> 102 dpb, vecDP, data = solveDirect( conf, options ) 103 104 if hasattr( opts, 'postProcessHookFinal' ): # User postprocessing.
/home/ryan/svn/sfepy/sfepy/solvers/generic.py in solveDirect(conf, options) 217 return None, None, None 218 --> 219 pb = ProblemDefinition.fromConf( conf ) 220 if hasattr( conf.options, 'ts' ): 221 ##
/home/ryan/svn/sfepy/sfepy/fem/problemDef.py in fromConf(conf, initFields, initVariables, initEquations, initSolvers) 34 mesh = Mesh.fromFile( conf.fileName_mesh ) 35 ---> 36 domain = Domain.fromMesh( mesh, 'eldesc' ) 37 domain.setupGroups() 38 domain.fixElementOrientation()
/home/ryan/svn/sfepy/sfepy/fem/domain.py in fromMesh(mesh, componentDir) 169 170 if not geomEls.find( desc ): --> 171 geomEls.append( read( GeomElement, desc ) ) 172 173 interps = {}
/home/ryan/svn/sfepy/sfepy/base/reader.py in __call__(self, objClass, name) 18 19 aux = {} ---> 20 execfile( fileName, {}, aux ) 21 22 obj = objClass()
<type 'exceptions.IOError'>: [Errno 2] No such file or directory: 'eldesc/3_4.py' WARNING: Failure executing file: <simple.py>
What is in 'eldesc'? Is that a directory I should put on my path? Is it something I might modify for specific problems and should therefore copy to my research folder?
Thanks,
Ryan
On Wed, Jul 9, 2008 at 9:05 AM, Ondrej Certik <ond...@certik.cz> wrote:
On Wed, Jul 9, 2008 at 3:54 PM, Robert Cimrman <cimr...@ntc.zcu.cz> wrote:
Robert Cimrman wrote:
Ryan Krauss wrote:
I have a question about where you guys put the files you are working on with SfePy. For SciPy and NumPy, I am used to having an svn directory that is only for the scipy/numpy code. Once I do "sudo python setup.py install", I leave that directory and don't do any actual work in it. SfePy doesn't seem to work that way. The thing about that that is tricky for me is that I like to use Unison to sychronize my laptop and desktop. They have different processors, so I don't want to syncronize any build directories (or anything else that would get typically modified by mercurial), but I do want to sync my own stuff (meshes and input files and such).
I put my own stuff under input and database. You can make your own dir there (out of the repo control, if you do not wish to share your initial versions of inputs).
A little explanation why sfepy works this way: I like the fact that when I change some Python code I can just run it to see new results without installing anything. It is really annoying to have to: 1. change a single letter, 2. 'python setup.py install' 3. run your code
Exactly, SymPy works in exactly the same way.
But in SymPy we of course support and test setup.py install, because this is the standard way of installing things, so we should make sure sfepy can do it too in the next release.
Also this will be necessary when I create a debian package for it, later when the API stabilizes and sfepy becomes more mature.
Ondrej
FYI, adding the absolute path to sfey/eldesc to my PATH and PYTHONPATH in .bashrc did not solve this problem. I think the code is looking for a relative path to the files.
Ryan
On Wed, Jul 9, 2008 at 10:56 AM, Ryan Krauss <ryan...@gmail.com> wrote:
I can totally understand not wanting to reinstall every time you change code that you are developing. But I like to separate code that I am developing that I think of as utility code or general code from specific uses of my code applied to my research.
In trying to run my from_Robert.py file from my research directory I get this: In [3]: run simple.py from_Robert.py sfepy: warning: other missing: ['functions', 'modules', 'epbc_[0-9]+|epbcs', 'lcbc_[0-9]+|lcbcs', 'nbc_[0-9]+|nbcs'] sfepy: warning: left over: ['tractionLoad'] sfepy: reading mesh (first_mesh.1.node)... nodes: 100% |############################################| Time: 00:00:00 elements: 100% |#########################################| Time: 00:00:00 sfepy: ...done in 0.01 s
<type 'exceptions.IOError'> Traceback (most recent call last)
/home/ryan/siue/Research/SfePy/first_problem/simple.py in <module>() 107 108 if __name__ == '__main__': --> 109 main() 110 111
/home/ryan/siue/Research/SfePy/first_problem/simple.py in main() 100 setOutputPrefix( opts.outputPrefix ) 101 --> 102 dpb, vecDP, data = solveDirect( conf, options ) 103 104 if hasattr( opts, 'postProcessHookFinal' ): # User postprocessing.
/home/ryan/svn/sfepy/sfepy/solvers/generic.py in solveDirect(conf, options) 217 return None, None, None 218 --> 219 pb = ProblemDefinition.fromConf( conf ) 220 if hasattr( conf.options, 'ts' ): 221 ##
/home/ryan/svn/sfepy/sfepy/fem/problemDef.py in fromConf(conf, initFields, initVariables, initEquations, initSolvers) 34 mesh = Mesh.fromFile( conf.fileName_mesh ) 35 ---> 36 domain = Domain.fromMesh( mesh, 'eldesc' ) 37 domain.setupGroups() 38 domain.fixElementOrientation()
/home/ryan/svn/sfepy/sfepy/fem/domain.py in fromMesh(mesh, componentDir) 169 170 if not geomEls.find( desc ): --> 171 geomEls.append( read( GeomElement, desc ) ) 172 173 interps = {}
/home/ryan/svn/sfepy/sfepy/base/reader.py in __call__(self, objClass, name) 18 19 aux = {} ---> 20 execfile( fileName, {}, aux ) 21 22 obj = objClass()
<type 'exceptions.IOError'>: [Errno 2] No such file or directory: 'eldesc/3_4.py' WARNING: Failure executing file: <simple.py>
What is in 'eldesc'? Is that a directory I should put on my path? Is it something I might modify for specific problems and should therefore copy to my research folder?
Thanks,
Ryan
On Wed, Jul 9, 2008 at 9:05 AM, Ondrej Certik <ond...@certik.cz> wrote:
On Wed, Jul 9, 2008 at 3:54 PM, Robert Cimrman <cimr...@ntc.zcu.cz> wrote:
Robert Cimrman wrote:
Ryan Krauss wrote:
I have a question about where you guys put the files you are working on with SfePy. For SciPy and NumPy, I am used to having an svn directory that is only for the scipy/numpy code. Once I do "sudo python setup.py install", I leave that directory and don't do any actual work in it. SfePy doesn't seem to work that way. The thing about that that is tricky for me is that I like to use Unison to sychronize my laptop and desktop. They have different processors, so I don't want to syncronize any build directories (or anything else that would get typically modified by mercurial), but I do want to sync my own stuff (meshes and input files and such).
I put my own stuff under input and database. You can make your own dir there (out of the repo control, if you do not wish to share your initial versions of inputs).
A little explanation why sfepy works this way: I like the fact that when I change some Python code I can just run it to see new results without installing anything. It is really annoying to have to: 1. change a single letter, 2. 'python setup.py install' 3. run your code
Exactly, SymPy works in exactly the same way.
But in SymPy we of course support and test setup.py install, because this is the standard way of installing things, so we should make sure sfepy can do it too in the next release.
Also this will be necessary when I create a debian package for it, later when the API stabilizes and sfepy becomes more mature.
Ondrej
Ryan Krauss wrote:
FYI, adding the absolute path to sfey/eldesc to my PATH and PYTHONPATH in .bashrc did not solve this problem. I think the code is looking for a relative path to the files.
This is a bug or, say, legacy feature. 'eldesc' directory contains definition of finite elements and approximations and is very old. Originally I thought to keep it outside of the sources as people might define their own elements. Now all this stuff is ripe for a redesign. I will fix this, so be patient for the moment.
r.
Robert Cimrman wrote:
Ryan Krauss wrote:
FYI, adding the absolute path to sfey/eldesc to my PATH and PYTHONPATH in .bashrc did not solve this problem. I think the code is looking for a relative path to the files.
This is a bug or, say, legacy feature. 'eldesc' directory contains definition of finite elements and approximations and is very old. Originally I thought to keep it outside of the sources as people might define their own elements. Now all this stuff is ripe for a redesign. I will fix this, so be patient for the moment.
ok, done:
export PYTHONPATH=$PYTHONPATH:~/projects/sfepy export PATH=$PATH:~/projects/sfepy
simple.py <something.py>
works now, pull the changeset.
r.
Cool. Thanks. I will try this after lunch.
Ryan
On Wed, Jul 9, 2008 at 11:33 AM, Robert Cimrman <cimr...@ntc.zcu.cz> wrote:
Robert Cimrman wrote:
Ryan Krauss wrote:
FYI, adding the absolute path to sfey/eldesc to my PATH and PYTHONPATH in .bashrc did not solve this problem. I think the code is looking for a relative path to the files.
This is a bug or, say, legacy feature. 'eldesc' directory contains definition of finite elements and approximations and is very old. Originally I thought to keep it outside of the sources as people might define their own elements. Now all this stuff is ripe for a redesign. I will fix this, so be patient for the moment.
ok, done:
export PYTHONPATH=$PYTHONPATH:~/projects/sfepy export PATH=$PATH:~/projects/sfepy
simple.py <something.py>
works now, pull the changeset.
r.
Have fun. How many hours ahead of me are you? I am in US central time zone.
Ryan
On Wed, Jul 9, 2008 at 12:11 PM, Robert Cimrman <cimr...@ntc.zcu.cz> wrote:
Ryan Krauss wrote:
Cool. Thanks. I will try this after lunch.
So see you tomorrow, we (me and Ondrej, who is visiting at my place today) go get some evening beer. :)
r.
OK, I need to stop working on this and focus on class prep anyways, but I am stuck for today. I can't quite make this work. I am not sure what is in the ~/projects/sfepy folder and whether or not that or something else is my problem. I believe I have pulled the lates stuff from hg (once you learn more about hg and get .hgrc set up correctly, it really is nice. I am now handling my conflicts with emacs and ediff - I really like it - but I still haven't read the HG tutorial :). Here is the error message:
In [1]: run simple.py from_Robert.py sfepy: warning: other missing: ['functions', 'modules', 'epbc_[0-9]+|epbcs', 'lcbc_[0-9]+|lcbcs', 'nbc_[0-9]+|nbcs'] sfepy: warning: left over: ['tractionLoad'] sfepy: reading mesh (first_mesh.1.node)... nodes: 100% |############################################| Time: 00:00:00 elements: 100% |#########################################| Time: 00:00:00 sfepy: ...done in 0.03 s sfepy: setting up domain edges... sfepy: ...done in 0.00 s sfepy: setting up domain faces... sfepy: ...done in 0.01 s sfepy: creating regions... sfepy: leaf Top region_2 sfepy: leaf Omega region_1000 sfepy: leaf Bottom region_1 sfepy: ...done in 0.04 s sfepy: equation "balance_of_forces": sfepy: dw_lin_elastic_iso.i1.Omega( solid.lame, v, u ) = dw_surface_ltr.isurf.Top( traction.val, v ) sfepy: describing geometries... sfepy: ...done in 0.01 s sfepy: setting up dof connectivities... sfepy: ...done in 0.00 s sfepy: using solvers: ts: ts nls: newton ls: ls
<type 'exceptions.TypeError'> Traceback (most recent call last)
/home/ryan/siue/Research/SfePy/first_problem/simple.py in <module>() 107 108 if __name__ == '__main__': --> 109 main() 110 111
/home/ryan/siue/Research/SfePy/first_problem/simple.py in main() 100 setOutputPrefix( opts.outputPrefix ) 101 --> 102 dpb, vecDP, data = solveDirect( conf, options ) 103 104 if hasattr( opts, 'postProcessHookFinal' ): # User postprocessing.
/home/ryan/svn/sfepy/sfepy/solvers/generic.py in solveDirect(conf, options) 220 # Time-dependent problem. 221 state, data = solveEvolutionaryOP( pb, options, --> 222 postProcessHook = postProcessHook ) 223 else: 224 ##
/home/ryan/svn/sfepy/sfepy/solvers/generic.py in solveEvolutionaryOP(problem, options, saveResults, returnHistory, postProcessHook) 143 data = {} 144 timeSolver = problem.getTimeSolver( stepFun = timeStepFunction, --> 145 stepArgs = (problem, data) ) 146 147 ofnTrunk, suffix, isSave = prepareSaveData( timeSolver.ts,
/home/ryan/svn/sfepy/sfepy/fem/problemDef.py in getTimeSolver(self, tsConf, **kwargs) 483 'you must set time-stepping solver!' ) 484 --> 485 return Solver.anyFromConf( tsConf, **kwargs ) 486 487
/home/ryan/siue/Research/SfePy/first_problem/sfepy/solvers/__init__.py in anyFromConf(conf, **kwargs) 34 # 23.10.2007, c 35 def anyFromConf( conf, **kwargs ): ---> 36 return solverTable[conf.kind]( conf, **kwargs ) 37 insertStaticMethod( Solver, anyFromConf ) 38 del anyFromConf
/home/ryan/svn/sfepy/sfepy/solvers/ts.py in __init__(self, conf, **kwargs) 114 self.ts = TimeStepper.fromConf( conf ) 115 nd = self.ts.nDigit --> 116 format = '====== time %%e (step %%%dd of %%%dd) =====' % (nd, nd) 117 118 self.format = format
<type 'exceptions.TypeError'>: int argument required WARNING: Failure executing file: <simple.py>
My main hg directory is ~/svn/sfepy and that is in my PATH and PYTHONPATH. I am running the script simply.py from this directory: /home/ryan/siue/Research/SfePy/first_problem
which contains only the following: a.lua first_mesh.1.node from_Robert.py notes.pyp first_mesh.1.ele first_mesh.1.vtk from_Robert.py~ output/ first_mesh.1.face first_mesh.poly from_Robert.pyc simple.py*
(I created the output dir myself and copied results there from an earlier run from ~/svn/sfepy).
Do I need to set this up slightly differently or do I have some other problem (i.e. something not directory related)?
Ryan On Wed, Jul 9, 2008 at 11:33 AM, Robert Cimrman <cimr...@ntc.zcu.cz> wrote:
Robert Cimrman wrote:
Ryan Krauss wrote:
FYI, adding the absolute path to sfey/eldesc to my PATH and PYTHONPATH in .bashrc did not solve this problem. I think the code is looking for a relative path to the files.
This is a bug or, say, legacy feature. 'eldesc' directory contains definition of finite elements and approximations and is very old. Originally I thought to keep it outside of the sources as people might define their own elements. Now all this stuff is ripe for a redesign. I will fix this, so be patient for the moment.
ok, done:
export PYTHONPATH=$PYTHONPATH:~/projects/sfepy export PATH=$PATH:~/projects/sfepy
simple.py <something.py>
works now, pull the changeset.
r.
I get the same error trying to run my script from ~/svn/sfepy, so it probably isn't directory related - but I do think it was caused by my latest hg pull.
On Wed, Jul 9, 2008 at 1:03 PM, Ryan Krauss <ryan...@gmail.com> wrote:
OK, I need to stop working on this and focus on class prep anyways, but I am stuck for today. I can't quite make this work. I am not sure what is in the ~/projects/sfepy folder and whether or not that or something else is my problem. I believe I have pulled the lates stuff from hg (once you learn more about hg and get .hgrc set up correctly, it really is nice. I am now handling my conflicts with emacs and ediff - I really like it - but I still haven't read the HG tutorial :). Here is the error message:
In [1]: run simple.py from_Robert.py sfepy: warning: other missing: ['functions', 'modules', 'epbc_[0-9]+|epbcs', 'lcbc_[0-9]+|lcbcs', 'nbc_[0-9]+|nbcs'] sfepy: warning: left over: ['tractionLoad'] sfepy: reading mesh (first_mesh.1.node)... nodes: 100% |############################################| Time: 00:00:00 elements: 100% |#########################################| Time: 00:00:00 sfepy: ...done in 0.03 s sfepy: setting up domain edges... sfepy: ...done in 0.00 s sfepy: setting up domain faces... sfepy: ...done in 0.01 s sfepy: creating regions... sfepy: leaf Top region_2 sfepy: leaf Omega region_1000 sfepy: leaf Bottom region_1 sfepy: ...done in 0.04 s sfepy: equation "balance_of_forces": sfepy: dw_lin_elastic_iso.i1.Omega( solid.lame, v, u ) = dw_surface_ltr.isurf.Top( traction.val, v ) sfepy: describing geometries... sfepy: ...done in 0.01 s sfepy: setting up dof connectivities... sfepy: ...done in 0.00 s sfepy: using solvers: ts: ts nls: newton ls: ls
<type 'exceptions.TypeError'> Traceback (most recent call last)
/home/ryan/siue/Research/SfePy/first_problem/simple.py in <module>() 107 108 if __name__ == '__main__': --> 109 main() 110 111
/home/ryan/siue/Research/SfePy/first_problem/simple.py in main() 100 setOutputPrefix( opts.outputPrefix ) 101 --> 102 dpb, vecDP, data = solveDirect( conf, options ) 103 104 if hasattr( opts, 'postProcessHookFinal' ): # User postprocessing.
/home/ryan/svn/sfepy/sfepy/solvers/generic.py in solveDirect(conf, options) 220 # Time-dependent problem. 221 state, data = solveEvolutionaryOP( pb, options, --> 222 postProcessHook = postProcessHook ) 223 else: 224 ##
/home/ryan/svn/sfepy/sfepy/solvers/generic.py in solveEvolutionaryOP(problem, options, saveResults, returnHistory, postProcessHook) 143 data = {} 144 timeSolver = problem.getTimeSolver( stepFun = timeStepFunction, --> 145 stepArgs = (problem, data) ) 146 147 ofnTrunk, suffix, isSave = prepareSaveData( timeSolver.ts,
/home/ryan/svn/sfepy/sfepy/fem/problemDef.py in getTimeSolver(self, tsConf, **kwargs) 483 'you must set time-stepping solver!' ) 484 --> 485 return Solver.anyFromConf( tsConf, **kwargs ) 486 487
/home/ryan/siue/Research/SfePy/first_problem/sfepy/solvers/__init__.py in anyFromConf(conf, **kwargs) 34 # 23.10.2007, c 35 def anyFromConf( conf, **kwargs ): ---> 36 return solverTable[conf.kind]( conf, **kwargs ) 37 insertStaticMethod( Solver, anyFromConf ) 38 del anyFromConf
/home/ryan/svn/sfepy/sfepy/solvers/ts.py in __init__(self, conf, **kwargs) 114 self.ts = TimeStepper.fromConf( conf ) 115 nd = self.ts.nDigit --> 116 format = '====== time %%e (step %%%dd of %%%dd) =====' % (nd, nd) 117 118 self.format = format
<type 'exceptions.TypeError'>: int argument required WARNING: Failure executing file: <simple.py>
My main hg directory is ~/svn/sfepy and that is in my PATH and PYTHONPATH. I am running the script simply.py from this directory: /home/ryan/siue/Research/SfePy/first_problem
which contains only the following: a.lua first_mesh.1.node from_Robert.py notes.pyp first_mesh.1.ele first_mesh.1.vtk from_Robert.py~ output/ first_mesh.1.face first_mesh.poly from_Robert.pyc simple.py*
(I created the output dir myself and copied results there from an earlier run from ~/svn/sfepy).
Do I need to set this up slightly differently or do I have some other problem (i.e. something not directory related)?
Ryan On Wed, Jul 9, 2008 at 11:33 AM, Robert Cimrman <cimr...@ntc.zcu.cz> wrote:
Robert Cimrman wrote:
Ryan Krauss wrote:
FYI, adding the absolute path to sfey/eldesc to my PATH and PYTHONPATH in .bashrc did not solve this problem. I think the code is looking for a relative path to the files.
This is a bug or, say, legacy feature. 'eldesc' directory contains definition of finite elements and approximations and is very old. Originally I thought to keep it outside of the sources as people might define their own elements. Now all this stuff is ripe for a redesign. I will fix this, so be patient for the moment.
ok, done:
export PYTHONPATH=$PYTHONPATH:~/projects/sfepy export PATH=$PATH:~/projects/sfepy
simple.py <something.py>
works now, pull the changeset.
r.
FYI, I can run the same script from my ~/svn/sfepy directory on my laptop where I haven't pulled the latest changes. So, it seems like something in the changes caused the error.
Ryan
On Wed, Jul 9, 2008 at 1:05 PM, Ryan Krauss <ryan...@gmail.com> wrote:
I get the same error trying to run my script from ~/svn/sfepy, so it probably isn't directory related - but I do think it was caused by my latest hg pull.
On Wed, Jul 9, 2008 at 1:03 PM, Ryan Krauss <ryan...@gmail.com> wrote:
OK, I need to stop working on this and focus on class prep anyways, but I am stuck for today. I can't quite make this work. I am not sure what is in the ~/projects/sfepy folder and whether or not that or something else is my problem. I believe I have pulled the lates stuff from hg (once you learn more about hg and get .hgrc set up correctly, it really is nice. I am now handling my conflicts with emacs and ediff - I really like it - but I still haven't read the HG tutorial :). Here is the error message:
In [1]: run simple.py from_Robert.py sfepy: warning: other missing: ['functions', 'modules', 'epbc_[0-9]+|epbcs', 'lcbc_[0-9]+|lcbcs', 'nbc_[0-9]+|nbcs'] sfepy: warning: left over: ['tractionLoad'] sfepy: reading mesh (first_mesh.1.node)... nodes: 100% |############################################| Time: 00:00:00 elements: 100% |#########################################| Time: 00:00:00 sfepy: ...done in 0.03 s sfepy: setting up domain edges... sfepy: ...done in 0.00 s sfepy: setting up domain faces... sfepy: ...done in 0.01 s sfepy: creating regions... sfepy: leaf Top region_2 sfepy: leaf Omega region_1000 sfepy: leaf Bottom region_1 sfepy: ...done in 0.04 s sfepy: equation "balance_of_forces": sfepy: dw_lin_elastic_iso.i1.Omega( solid.lame, v, u ) = dw_surface_ltr.isurf.Top( traction.val, v ) sfepy: describing geometries... sfepy: ...done in 0.01 s sfepy: setting up dof connectivities... sfepy: ...done in 0.00 s sfepy: using solvers: ts: ts nls: newton ls: ls
<type 'exceptions.TypeError'> Traceback (most recent call last)
/home/ryan/siue/Research/SfePy/first_problem/simple.py in <module>() 107 108 if __name__ == '__main__': --> 109 main() 110 111
/home/ryan/siue/Research/SfePy/first_problem/simple.py in main() 100 setOutputPrefix( opts.outputPrefix ) 101 --> 102 dpb, vecDP, data = solveDirect( conf, options ) 103 104 if hasattr( opts, 'postProcessHookFinal' ): # User postprocessing.
/home/ryan/svn/sfepy/sfepy/solvers/generic.py in solveDirect(conf, options) 220 # Time-dependent problem. 221 state, data = solveEvolutionaryOP( pb, options, --> 222 postProcessHook = postProcessHook ) 223 else: 224 ##
/home/ryan/svn/sfepy/sfepy/solvers/generic.py in solveEvolutionaryOP(problem, options, saveResults, returnHistory, postProcessHook) 143 data = {} 144 timeSolver = problem.getTimeSolver( stepFun = timeStepFunction, --> 145 stepArgs = (problem, data) ) 146 147 ofnTrunk, suffix, isSave = prepareSaveData( timeSolver.ts,
/home/ryan/svn/sfepy/sfepy/fem/problemDef.py in getTimeSolver(self, tsConf, **kwargs) 483 'you must set time-stepping solver!' ) 484 --> 485 return Solver.anyFromConf( tsConf, **kwargs ) 486 487
/home/ryan/siue/Research/SfePy/first_problem/sfepy/solvers/__init__.py in anyFromConf(conf, **kwargs) 34 # 23.10.2007, c 35 def anyFromConf( conf, **kwargs ): ---> 36 return solverTable[conf.kind]( conf, **kwargs ) 37 insertStaticMethod( Solver, anyFromConf ) 38 del anyFromConf
/home/ryan/svn/sfepy/sfepy/solvers/ts.py in __init__(self, conf, **kwargs) 114 self.ts = TimeStepper.fromConf( conf ) 115 nd = self.ts.nDigit --> 116 format = '====== time %%e (step %%%dd of %%%dd) =====' % (nd, nd) 117 118 self.format = format
<type 'exceptions.TypeError'>: int argument required WARNING: Failure executing file: <simple.py>
My main hg directory is ~/svn/sfepy and that is in my PATH and PYTHONPATH. I am running the script simply.py from this directory: /home/ryan/siue/Research/SfePy/first_problem
which contains only the following: a.lua first_mesh.1.node from_Robert.py notes.pyp first_mesh.1.ele first_mesh.1.vtk from_Robert.py~ output/ first_mesh.1.face first_mesh.poly from_Robert.pyc simple.py*
(I created the output dir myself and copied results there from an earlier run from ~/svn/sfepy).
Do I need to set this up slightly differently or do I have some other problem (i.e. something not directory related)?
Ryan On Wed, Jul 9, 2008 at 11:33 AM, Robert Cimrman <cimr...@ntc.zcu.cz> wrote:
Robert Cimrman wrote:
Ryan Krauss wrote:
FYI, adding the absolute path to sfey/eldesc to my PATH and PYTHONPATH in .bashrc did not solve this problem. I think the code is looking for a relative path to the files.
This is a bug or, say, legacy feature. 'eldesc' directory contains definition of finite elements and approximations and is very old. Originally I thought to keep it outside of the sources as people might define their own elements. Now all this stuff is ripe for a redesign. I will fix this, so be patient for the moment.
ok, done:
export PYTHONPATH=$PYTHONPATH:~/projects/sfepy export PATH=$PATH:~/projects/sfepy
simple.py <something.py>
works now, pull the changeset.
r.
Just writing from my cell - you can use "hg bisect" if something was working and it stopped, that way you can easily determine which patch broke something. Ondrej
On 7/9/08, Ryan Krauss <ryan...@gmail.com> wrote:
I get the same error trying to run my script from ~/svn/sfepy, so it probably isn't directory related - but I do think it was caused by my latest hg pull.
On Wed, Jul 9, 2008 at 1:03 PM, Ryan Krauss <ryan...@gmail.com> wrote:
OK, I need to stop working on this and focus on class prep anyways, but I am stuck for today. I can't quite make this work. I am not sure what is in the ~/projects/sfepy folder and whether or not that or something else is my problem. I believe I have pulled the lates stuff from hg (once you learn more about hg and get .hgrc set up correctly, it really is nice. I am now handling my conflicts with emacs and ediff - I really like it - but I still haven't read the HG tutorial :). Here is the error message:
In [1]: run simple.py from_Robert.py sfepy: warning: other missing: ['functions', 'modules', 'epbc_[0-9]+|epbcs', 'lcbc_[0-9]+|lcbcs', 'nbc_[0-9]+|nbcs'] sfepy: warning: left over: ['tractionLoad'] sfepy: reading mesh (first_mesh.1.node)... nodes: 100% |############################################| Time: 00:00:00 elements: 100% |#########################################| Time: 00:00:00 sfepy: ...done in 0.03 s sfepy: setting up domain edges... sfepy: ...done in 0.00 s sfepy: setting up domain faces... sfepy: ...done in 0.01 s sfepy: creating regions... sfepy: leaf Top region_2 sfepy: leaf Omega region_1000 sfepy: leaf Bottom region_1 sfepy: ...done in 0.04 s sfepy: equation "balance_of_forces": sfepy: dw_lin_elastic_iso.i1.Omega( solid.lame, v, u ) = dw_surface_ltr.isurf.Top( traction.val, v ) sfepy: describing geometries... sfepy: ...done in 0.01 s sfepy: setting up dof connectivities... sfepy: ...done in 0.00 s sfepy: using solvers: ts: ts nls: newton ls: ls
<type 'exceptions.TypeError'> Traceback (most recent call last)
/home/ryan/siue/Research/SfePy/first_problem/simple.py in <module>() 107 108 if __name__ == '__main__': --> 109 main() 110 111
/home/ryan/siue/Research/SfePy/first_problem/simple.py in main() 100 setOutputPrefix( opts.outputPrefix ) 101 --> 102 dpb, vecDP, data = solveDirect( conf, options ) 103 104 if hasattr( opts, 'postProcessHookFinal' ): # User postprocessing.
/home/ryan/svn/sfepy/sfepy/solvers/generic.py in solveDirect(conf, options) 220 # Time-dependent problem. 221 state, data = solveEvolutionaryOP( pb, options, --> 222 postProcessHook = postProcessHook ) 223 else: 224 ##
/home/ryan/svn/sfepy/sfepy/solvers/generic.py in solveEvolutionaryOP(problem, options, saveResults, returnHistory, postProcessHook) 143 data = {} 144 timeSolver = problem.getTimeSolver( stepFun = timeStepFunction, --> 145 stepArgs = (problem, data) ) 146 147 ofnTrunk, suffix, isSave = prepareSaveData( timeSolver.ts,
/home/ryan/svn/sfepy/sfepy/fem/problemDef.py in getTimeSolver(self, tsConf, **kwargs) 483 'you must set time-stepping solver!' ) 484 --> 485 return Solver.anyFromConf( tsConf, **kwargs ) 486 487
/home/ryan/siue/Research/SfePy/first_problem/sfepy/solvers/__init__.py in anyFromConf(conf, **kwargs) 34 # 23.10.2007, c 35 def anyFromConf( conf, **kwargs ): ---> 36 return solverTable[conf.kind]( conf, **kwargs ) 37 insertStaticMethod( Solver, anyFromConf ) 38 del anyFromConf
/home/ryan/svn/sfepy/sfepy/solvers/ts.py in __init__(self, conf, **kwargs) 114 self.ts = TimeStepper.fromConf( conf ) 115 nd = self.ts.nDigit --> 116 format = '====== time %%e (step %%%dd of %%%dd) =====' % (nd, nd) 117 118 self.format = format
<type 'exceptions.TypeError'>: int argument required WARNING: Failure executing file: <simple.py>
My main hg directory is ~/svn/sfepy and that is in my PATH and PYTHONPATH. I am running the script simply.py from this directory: /home/ryan/siue/Research/SfePy/first_problem
which contains only the following: a.lua first_mesh.1.node from_Robert.py notes.pyp first_mesh.1.ele first_mesh.1.vtk from_Robert.py~ output/ first_mesh.1.face first_mesh.poly from_Robert.pyc simple.py*
(I created the output dir myself and copied results there from an earlier run from ~/svn/sfepy).
Do I need to set this up slightly differently or do I have some other problem (i.e. something not directory related)?
Ryan On Wed, Jul 9, 2008 at 11:33 AM, Robert Cimrman <cimr...@ntc.zcu.cz> wrote:
Robert Cimrman wrote:
Ryan Krauss wrote:
FYI, adding the absolute path to sfey/eldesc to my PATH and PYTHONPATH in .bashrc did not solve this problem. I think the code is looking for a relative path to the files.
This is a bug or, say, legacy feature. 'eldesc' directory contains definition of finite elements and approximations and is very old. Originally I thought to keep it outside of the sources as people might define their own elements. Now all this stuff is ripe for a redesign. I will fix this, so be patient for the moment.
ok, done:
export PYTHONPATH=$PYTHONPATH:~/projects/sfepy export PATH=$PATH:~/projects/sfepy
simple.py <something.py>
works now, pull the changeset.
r.
OK, here is the output of my hg bisect session (I followed this tutorial: http://www.selenic.com/mercurial/wiki/index.cgi/BisectExtension):
ryan@am2:~/svn/sfepy$ hg bisect init ryan@am2:~/svn/sfepy$ hg bisect bad ryan@am2:~/svn/sfepy$ hg bisect good 05eb71d35ccb Testing changeset 264:7fcce8121f04 (2 changesets remaining, ~1 tests) 3 files updated, 0 files merged, 0 files removed, 0 files unresolved ryan@am2:~/svn/sfepy$ hg bisect good The first bad revision is: changeset: 265:c93532ab3400 tag: tip user: Robert Cimrman <cimr...@ntc.zcu.cz> date: Wed Jul 09 18:52:19 2008 +0200 summary: added getPrintInfo()
I was expecting it to encourage me to test 264.7fcce8121f04 since I suspect that 265.c93532ab3400 might really be the problem (Robert's comments mention something about printing and it is a print-type statement that seems to cause the problem), but hg bisect didn't ask me to test anymore after I marked one as good.
Ryan
On Wed, Jul 9, 2008 at 1:14 PM, Ondrej Certik <ond...@certik.cz> wrote:
Just writing from my cell - you can use "hg bisect" if something was working and it stopped, that way you can easily determine which patch broke something. Ondrej
On 7/9/08, Ryan Krauss <ryan...@gmail.com> wrote:
I get the same error trying to run my script from ~/svn/sfepy, so it probably isn't directory related - but I do think it was caused by my latest hg pull.
On Wed, Jul 9, 2008 at 1:03 PM, Ryan Krauss <ryan...@gmail.com> wrote:
OK, I need to stop working on this and focus on class prep anyways, but I am stuck for today. I can't quite make this work. I am not sure what is in the ~/projects/sfepy folder and whether or not that or something else is my problem. I believe I have pulled the lates stuff from hg (once you learn more about hg and get .hgrc set up correctly, it really is nice. I am now handling my conflicts with emacs and ediff - I really like it - but I still haven't read the HG tutorial :). Here is the error message:
In [1]: run simple.py from_Robert.py sfepy: warning: other missing: ['functions', 'modules', 'epbc_[0-9]+|epbcs', 'lcbc_[0-9]+|lcbcs', 'nbc_[0-9]+|nbcs'] sfepy: warning: left over: ['tractionLoad'] sfepy: reading mesh (first_mesh.1.node)... nodes: 100% |############################################| Time: 00:00:00 elements: 100% |#########################################| Time: 00:00:00 sfepy: ...done in 0.03 s sfepy: setting up domain edges... sfepy: ...done in 0.00 s sfepy: setting up domain faces... sfepy: ...done in 0.01 s sfepy: creating regions... sfepy: leaf Top region_2 sfepy: leaf Omega region_1000 sfepy: leaf Bottom region_1 sfepy: ...done in 0.04 s sfepy: equation "balance_of_forces": sfepy: dw_lin_elastic_iso.i1.Omega( solid.lame, v, u ) = dw_surface_ltr.isurf.Top( traction.val, v ) sfepy: describing geometries... sfepy: ...done in 0.01 s sfepy: setting up dof connectivities... sfepy: ...done in 0.00 s sfepy: using solvers: ts: ts nls: newton ls: ls
<type 'exceptions.TypeError'> Traceback (most recent call last)
/home/ryan/siue/Research/SfePy/first_problem/simple.py in <module>() 107 108 if __name__ == '__main__': --> 109 main() 110 111
/home/ryan/siue/Research/SfePy/first_problem/simple.py in main() 100 setOutputPrefix( opts.outputPrefix ) 101 --> 102 dpb, vecDP, data = solveDirect( conf, options ) 103 104 if hasattr( opts, 'postProcessHookFinal' ): # User postprocessing.
/home/ryan/svn/sfepy/sfepy/solvers/generic.py in solveDirect(conf, options) 220 # Time-dependent problem. 221 state, data = solveEvolutionaryOP( pb, options, --> 222 postProcessHook = postProcessHook ) 223 else: 224 ##
/home/ryan/svn/sfepy/sfepy/solvers/generic.py in solveEvolutionaryOP(problem, options, saveResults, returnHistory, postProcessHook) 143 data = {} 144 timeSolver = problem.getTimeSolver( stepFun = timeStepFunction, --> 145 stepArgs = (problem, data) ) 146 147 ofnTrunk, suffix, isSave = prepareSaveData( timeSolver.ts,
/home/ryan/svn/sfepy/sfepy/fem/problemDef.py in getTimeSolver(self, tsConf, **kwargs) 483 'you must set time-stepping solver!' ) 484 --> 485 return Solver.anyFromConf( tsConf, **kwargs ) 486 487
/home/ryan/siue/Research/SfePy/first_problem/sfepy/solvers/__init__.py in anyFromConf(conf, **kwargs) 34 # 23.10.2007, c 35 def anyFromConf( conf, **kwargs ): ---> 36 return solverTable[conf.kind]( conf, **kwargs ) 37 insertStaticMethod( Solver, anyFromConf ) 38 del anyFromConf
/home/ryan/svn/sfepy/sfepy/solvers/ts.py in __init__(self, conf, **kwargs) 114 self.ts = TimeStepper.fromConf( conf ) 115 nd = self.ts.nDigit --> 116 format = '====== time %%e (step %%%dd of %%%dd) =====' % (nd, nd) 117 118 self.format = format
<type 'exceptions.TypeError'>: int argument required WARNING: Failure executing file: <simple.py>
My main hg directory is ~/svn/sfepy and that is in my PATH and PYTHONPATH. I am running the script simply.py from this directory: /home/ryan/siue/Research/SfePy/first_problem
which contains only the following: a.lua first_mesh.1.node from_Robert.py notes.pyp first_mesh.1.ele first_mesh.1.vtk from_Robert.py~ output/ first_mesh.1.face first_mesh.poly from_Robert.pyc simple.py*
(I created the output dir myself and copied results there from an earlier run from ~/svn/sfepy).
Do I need to set this up slightly differently or do I have some other problem (i.e. something not directory related)?
Ryan On Wed, Jul 9, 2008 at 11:33 AM, Robert Cimrman <cimr...@ntc.zcu.cz> wrote:
Robert Cimrman wrote:
Ryan Krauss wrote:
FYI, adding the absolute path to sfey/eldesc to my PATH and PYTHONPATH in .bashrc did not solve this problem. I think the code is looking for a relative path to the files.
This is a bug or, say, legacy feature. 'eldesc' directory contains definition of finite elements and approximations and is very old. Originally I thought to keep it outside of the sources as people might define their own elements. Now all this stuff is ripe for a redesign. I will fix this, so be patient for the moment.
ok, done:
export PYTHONPATH=$PYTHONPATH:~/projects/sfepy export PATH=$PATH:~/projects/sfepy
simple.py <something.py>
works now, pull the changeset.
r.
On Wed, Jul 9, 2008 at 8:47 PM, Ryan Krauss <ryan...@gmail.com> wrote:
OK, here is the output of my hg bisect session (I followed this tutorial: http://www.selenic.com/mercurial/wiki/index.cgi/BisectExtension):
ryan@am2:~/svn/sfepy$ hg bisect init ryan@am2:~/svn/sfepy$ hg bisect bad ryan@am2:~/svn/sfepy$ hg bisect good 05eb71d35ccb Testing changeset 264:7fcce8121f04 (2 changesets remaining, ~1 tests) 3 files updated, 0 files merged, 0 files removed, 0 files unresolved ryan@am2:~/svn/sfepy$ hg bisect good The first bad revision is: changeset: 265:c93532ab3400 tag: tip user: Robert Cimrman <cimr...@ntc.zcu.cz> date: Wed Jul 09 18:52:19 2008 +0200 summary: added getPrintInfo()
I was expecting it to encourage me to test 264.7fcce8121f04 since I suspect that 265.c93532ab3400 might really be the problem (Robert's
It didn't test 264? That's weird, as how can it be sure it's 265 then... I tried hg bisect now and it tests 264 in my case. Maybe you overlooked?
comments mention something about printing and it is a print-type statement that seems to cause the problem), but hg bisect didn't ask me to test anymore after I marked one as good.
Anyway, so just use 264 and put all your patches against this revision. When Roberts comes to work tomorrow, he will fix it and then you can merge.
Ondrej
Sounds good.
On Wed, Jul 9, 2008 at 3:20 PM, Ondrej Certik <ond...@certik.cz> wrote:
On Wed, Jul 9, 2008 at 8:47 PM, Ryan Krauss <ryan...@gmail.com> wrote:
OK, here is the output of my hg bisect session (I followed this tutorial: http://www.selenic.com/mercurial/wiki/index.cgi/BisectExtension):
ryan@am2:~/svn/sfepy$ hg bisect init ryan@am2:~/svn/sfepy$ hg bisect bad ryan@am2:~/svn/sfepy$ hg bisect good 05eb71d35ccb Testing changeset 264:7fcce8121f04 (2 changesets remaining, ~1 tests) 3 files updated, 0 files merged, 0 files removed, 0 files unresolved ryan@am2:~/svn/sfepy$ hg bisect good The first bad revision is: changeset: 265:c93532ab3400 tag: tip user: Robert Cimrman <cimr...@ntc.zcu.cz> date: Wed Jul 09 18:52:19 2008 +0200 summary: added getPrintInfo()
I was expecting it to encourage me to test 264.7fcce8121f04 since I suspect that 265.c93532ab3400 might really be the problem (Robert's
It didn't test 264? That's weird, as how can it be sure it's 265 then... I tried hg bisect now and it tests 264 in my case. Maybe you overlooked?
comments mention something about printing and it is a print-type statement that seems to cause the problem), but hg bisect didn't ask me to test anymore after I marked one as good.
Anyway, so just use 264 and put all your patches against this revision. When Roberts comes to work tomorrow, he will fix it and then you can merge.
Ondrej
Ryan Krauss wrote:
OK, I need to stop working on this and focus on class prep anyways, but I am stuck for today. I can't quite make this work. I am not sure what is in the ~/projects/sfepy folder and whether or not that or something else is my problem. I believe I have pulled the lates stuff from hg (once you learn more about hg and get .hgrc set up correctly, it really is nice. I am now handling my conflicts with emacs and ediff - I really like it - but I still haven't read the HG tutorial :). Here is the error message:
Oh, sorry, I just pasted what I have used to set the paths. You have to replace '~/projects/sfepy' by your location of sfepy sources, i.e. probably '~/svn/sfepy'. You did that anyway, so it is ok :)
In [1]: run simple.py from_Robert.py sfepy: warning: other missing: ['functions', 'modules', 'epbc_[0-9]+|epbcs', 'lcbc_[0-9]+|lcbcs', 'nbc_[0-9]+|nbcs'] sfepy: warning: left over: ['tractionLoad'] sfepy: reading mesh (first_mesh.1.node)... nodes: 100% |############################################| Time: 00:00:00 ... /home/ryan/svn/sfepy/sfepy/solvers/ts.py in __init__(self, conf, **kwargs) 114 self.ts = TimeStepper.fromConf( conf ) 115 nd = self.ts.nDigit --> 116 format = '====== time %%e (step %%%dd of %%%dd) =====' % (nd, nd) 117 118 self.format = format
<type 'exceptions.TypeError'>: int argument required WARNING: Failure executing file: <simple.py>
My main hg directory is ~/svn/sfepy and that is in my PATH and PYTHONPATH. I am running the script simply.py from this directory: /home/ryan/siue/Research/SfePy/first_problem
which contains only the following: a.lua first_mesh.1.node from_Robert.py notes.pyp first_mesh.1.ele first_mesh.1.vtk from_Robert.py~ output/ first_mesh.1.face first_mesh.poly from_Robert.pyc simple.py*
(I created the output dir myself and copied results there from an earlier run from ~/svn/sfepy).
Do I need to set this up slightly differently or do I have some other problem (i.e. something not directory related)?
No, it was not directory related, it was a stupid bug introduced by me, which is hopefully fixed now.
r.
It seems to work perfectly. For whatever reason, the annimation in paraview just shows a red cube (i.e. all values are 0). When I look at the vtk output files, they aren't all zeros. I saw this yesterday when trying to run the debian version, but this is the binary download from kitware that worked yesterday. Very strange. It also incorrectly displays the results from my ~/svn/sfepy dir which were correct yesterday.
If someone could verify that the attached output files create a cube of many colors in their paraview, then we have fixed the directory issues with SfePy and I have a paraview problem of some kind.
Thanks,
Ryan
On Thu, Jul 10, 2008 at 2:58 AM, Robert Cimrman <cimr...@ntc.zcu.cz> wrote:
Ryan Krauss wrote:
OK, I need to stop working on this and focus on class prep anyways, but I am stuck for today. I can't quite make this work. I am not sure what is in the ~/projects/sfepy folder and whether or not that or something else is my problem. I believe I have pulled the lates stuff from hg (once you learn more about hg and get .hgrc set up correctly, it really is nice. I am now handling my conflicts with emacs and ediff - I really like it - but I still haven't read the HG tutorial :). Here is the error message:
Oh, sorry, I just pasted what I have used to set the paths. You have to replace '~/projects/sfepy' by your location of sfepy sources, i.e. probably '~/svn/sfepy'. You did that anyway, so it is ok :)
In [1]: run simple.py from_Robert.py sfepy: warning: other missing: ['functions', 'modules', 'epbc_[0-9]+|epbcs', 'lcbc_[0-9]+|lcbcs', 'nbc_[0-9]+|nbcs'] sfepy: warning: left over: ['tractionLoad'] sfepy: reading mesh (first_mesh.1.node)... nodes: 100% |############################################| Time: 00:00:00 ... /home/ryan/svn/sfepy/sfepy/solvers/ts.py in __init__(self, conf, **kwargs) 114 self.ts = TimeStepper.fromConf( conf ) 115 nd = self.ts.nDigit --> 116 format = '====== time %%e (step %%%dd of %%%dd) =====' % (nd, nd) 117 118 self.format = format
<type 'exceptions.TypeError'>: int argument required WARNING: Failure executing file: <simple.py>
My main hg directory is ~/svn/sfepy and that is in my PATH and PYTHONPATH. I am running the script simply.py from this directory: /home/ryan/siue/Research/SfePy/first_problem
which contains only the following: a.lua first_mesh.1.node from_Robert.py notes.pyp first_mesh.1.ele first_mesh.1.vtk from_Robert.py~ output/ first_mesh.1.face first_mesh.poly from_Robert.pyc simple.py*
(I created the output dir myself and copied results there from an earlier run from ~/svn/sfepy).
Do I need to set this up slightly differently or do I have some other problem (i.e. something not directory related)?
No, it was not directory related, it was a stupid bug introduced by me, which is hopefully fixed now.
r.
Ryan Krauss wrote:
It seems to work perfectly. For whatever reason, the annimation in paraview just shows a red cube (i.e. all values are 0). When I look at the vtk output files, they aren't all zeros. I saw this yesterday when trying to run the debian version, but this is the binary download from kitware that worked yesterday. Very strange. It also incorrectly displays the results from my ~/svn/sfepy dir which were correct yesterday.
If someone could verify that the attached output files create a cube of many colors in their paraview, then we have fixed the directory issues with SfePy and I have a paraview problem of some kind.
It seems ok, just the first file in the tar.gz seems wrong (too big, binary data). Maybe just press 'rescale to data range' in the 'display' tab in object inspector.
r
Perfect. Rescaling to data range did the trick. Thanks Robert.
On Thu, Jul 10, 2008 at 8:37 AM, Robert Cimrman <cimr...@ntc.zcu.cz> wrote:
Ryan Krauss wrote:
It seems to work perfectly. For whatever reason, the annimation in paraview just shows a red cube (i.e. all values are 0). When I look at the vtk output files, they aren't all zeros. I saw this yesterday when trying to run the debian version, but this is the binary download from kitware that worked yesterday. Very strange. It also incorrectly displays the results from my ~/svn/sfepy dir which were correct yesterday.
If someone could verify that the attached output files create a cube of many colors in their paraview, then we have fixed the directory issues with SfePy and I have a paraview problem of some kind.
It seems ok, just the first file in the tar.gz seems wrong (too big, binary data). Maybe just press 'rescale to data range' in the 'display' tab in object inspector.
r
Ryan Krauss wrote:
Perfect. Rescaling to data range did the trick. Thanks Robert.
So if it works even from another directory, maybe you could write how to set it up at http://code.google.com/p/sfepy/wiki/Documentation under Installation notes?
HTH, r.
On Thu, Jul 10, 2008 at 3:29 PM, Ryan Krauss <ryan...@gmail.com> wrote:
It seems to work perfectly. For whatever reason, the annimation in paraview just shows a red cube (i.e. all values are 0). When I look at the vtk output files, they aren't all zeros. I saw this yesterday when trying to run the debian version, but this is the binary download from kitware that worked yesterday. Very strange. It also incorrectly displays the results from my ~/svn/sfepy dir which were correct yesterday.
If you get some free time, I'd be interested if the ubuntu version of paraview works. It's basically just the debian package but recompiled to ubuntu.
Ondrej
I don't mind doing that. Should I be able to add something to source.lst and then apt-get install?
This: https://launchpad.net/ubuntu/intrepid/+source/paraview/3.2.2-1 doesn't appear to be a valid repo and it doesn't contain a .deb file. I can easily download and test the tarball if that is what you are asking.
Ryan
On Thu, Jul 10, 2008 at 9:03 AM, Ondrej Certik <ond...@certik.cz> wrote:
On Thu, Jul 10, 2008 at 3:29 PM, Ryan Krauss <ryan...@gmail.com> wrote:
It seems to work perfectly. For whatever reason, the annimation in paraview just shows a red cube (i.e. all values are 0). When I look at the vtk output files, they aren't all zeros. I saw this yesterday when trying to run the debian version, but this is the binary download from kitware that worked yesterday. Very strange. It also incorrectly displays the results from my ~/svn/sfepy dir which were correct yesterday.
If you get some free time, I'd be interested if the ubuntu version of paraview works. It's basically just the debian package but recompiled to ubuntu.
Ondrej
On Thu, Jul 10, 2008 at 4:24 PM, Ryan Krauss <ryan...@gmail.com> wrote:
I don't mind doing that. Should I be able to add something to source.lst and then apt-get install?
This: https://launchpad.net/ubuntu/intrepid/+source/paraview/3.2.2-1 doesn't appear to be a valid repo and it doesn't contain a .deb file. I can easily download and test the tarball if that is what you are asking.
As I wrote, either download the .deb by hand for your architecture, e.g.:
You just need to add sources for the intrepid, e.g. this should do the job:
deb http://archive.ubuntu.com/ubuntu intrepid main restricted universe
Or you can download the binary by hand from here:
https://launchpad.net/ubuntu/intrepid/i386/paraview/3.2.2-1
e.g.: http://launchpadlibrarian.net/15828310/paraview_3.2.2-1_i386.deb
Ondrej
On Thu, Jul 10, 2008 at 5:06 PM, Ondrej Certik <ond...@certik.cz> wrote:
On Thu, Jul 10, 2008 at 4:24 PM, Ryan Krauss <ryan...@gmail.com> wrote:
I don't mind doing that. Should I be able to add something to source.lst and then apt-get install?
This: https://launchpad.net/ubuntu/intrepid/+source/paraview/3.2.2-1 doesn't appear to be a valid repo and it doesn't contain a .deb file. I can easily download and test the tarball if that is what you are asking.
As I wrote, either download the .deb by hand for your architecture, e.g.:
You just need to add sources for the intrepid, e.g. this should do the job:
deb http://archive.ubuntu.com/ubuntu intrepid main restricted universe
And install with
apt-get install -t intrepid paraview
Ondrej
So, that lead to a depenency issue:
Couldn't configure pre-depend libc6 for findutils, probably a dependency cycle.
Ryan
On Thu, Jul 10, 2008 at 10:07 AM, Ondrej Certik <ond...@certik.cz> wrote:
On Thu, Jul 10, 2008 at 5:06 PM, Ondrej Certik <ond...@certik.cz> wrote:
On Thu, Jul 10, 2008 at 4:24 PM, Ryan Krauss <ryan...@gmail.com> wrote:
I don't mind doing that. Should I be able to add something to source.lst and then apt-get install?
This: https://launchpad.net/ubuntu/intrepid/+source/paraview/3.2.2-1 doesn't appear to be a valid repo and it doesn't contain a .deb file. I can easily download and test the tarball if that is what you are asking.
As I wrote, either download the .deb by hand for your architecture, e.g.:
You just need to add sources for the intrepid, e.g. this should do the job:
deb http://archive.ubuntu.com/ubuntu intrepid main restricted universe
And install with
apt-get install -t intrepid paraview
Ondrej
I have dependency issues with the deb as well.
On Thu, Jul 10, 2008 at 10:44 AM, Ondrej Certik <ond...@certik.cz> wrote:
On Thu, Jul 10, 2008 at 5:42 PM, Ryan Krauss <ryan...@gmail.com> wrote:
So, that lead to a depenency issue:
Couldn't configure pre-depend libc6 for findutils, probably a dependency cycle.
Ok. I'll try it myself in ubuntu in virtualbox, if I find some free time.
Ondrej
I will try a source install using the tarball from paraview.org and following these instructions: http://ubuntuforums.org/showthread.php?t=260698
I have the dependency listed there installed an am downloading the source now.
By the way, I am still running Ubuntu 7.10 - I won't have time to upgrade until my class is over in a week.
Ryan
On Thu, Jul 10, 2008 at 10:51 AM, Ryan Krauss <ryan...@gmail.com> wrote:
I have dependency issues with the deb as well.
On Thu, Jul 10, 2008 at 10:44 AM, Ondrej Certik <ond...@certik.cz> wrote:
On Thu, Jul 10, 2008 at 5:42 PM, Ryan Krauss <ryan...@gmail.com> wrote:
So, that lead to a depenency issue:
Couldn't configure pre-depend libc6 for findutils, probably a dependency cycle.
Ok. I'll try it myself in ubuntu in virtualbox, if I find some free time.
Ondrej
FYI, the source install seems to be working correctly.
Ryan
On Thu, Jul 10, 2008 at 10:59 AM, Ryan Krauss <ryan...@gmail.com> wrote:
I will try a source install using the tarball from paraview.org and following these instructions: http://ubuntuforums.org/showthread.php?t=260698
I have the dependency listed there installed an am downloading the source now.
By the way, I am still running Ubuntu 7.10 - I won't have time to upgrade until my class is over in a week.
Ryan
On Thu, Jul 10, 2008 at 10:51 AM, Ryan Krauss <ryan...@gmail.com> wrote:
I have dependency issues with the deb as well.
On Thu, Jul 10, 2008 at 10:44 AM, Ondrej Certik <ond...@certik.cz> wrote:
On Thu, Jul 10, 2008 at 5:42 PM, Ryan Krauss <ryan...@gmail.com> wrote:
So, that lead to a depenency issue:
Couldn't configure pre-depend libc6 for findutils, probably a dependency cycle.
Ok. I'll try it myself in ubuntu in virtualbox, if I find some free time.
Ondrej
participants (3)
-
Ondrej Certik -
Robert Cimrman -
Ryan Krauss