Re: [task #6847] The Bieri graphical user interface. -- March 05, 2010 - 10:58

Hi,

This is a very strange problem.  Can you try it again with a
completely unmodified version of the bieri_gui branch?  Running the
test suite with 'relax -x' passes perfectly.  But when I change line
157 from:

                print(("Simulation " + repr(i+1)))

to:

                rint(("Simulation " + repr(i+1)))

Then the test suite fails with 5 system test errors "NameError: global
name 'rint' is not defined".  This is a clear indication that
threading is causing the problem, and that the problem lies elsewhere.
 The "print 'Test'" replacement might just have been lucky and the
thread failure was avoided.  Have you made any changes to your local
copy of the branch?  Note that the test suite does not run any of the
tests in a thread.  I don't know if that is even possible.

Regards,

Edward



On 4 March 2010 23:12, Michael Bieri <michael.bieri@xxxxxx> wrote:
> Hi Edward
>
> It was the problem in line 157 in minimise.py. I just replaced it with
> 'print 'Test' (I know it is silly) and it works fine!
>
> i was also thinking to replace the threading, as with the other modules,
> we will be able to kill processes much easier (cancel button in controller).
>
> Cheers
> Michael
>
>
> Edward d'Auvergne schrieb:
> > Hi,
> >
> > For this problem, I think we need to first implement proper exception
> > handling in threads.  This will allow the errors to be properly
> > reported.  And this is up to the GUI as this runs the auto-analyses in
> > threaded mode.  According to the 'thread' module documentation
> > (http://docs.python.org/library/thread.html), "The threading module
> > provides an easier to use and higher-level threading API built on top
> > of this module".  I really think we should be using this instead, and
> > the thread handling should be more advanced (and exceptions should be
> > better handled).  Low level modules should only be touched by
> > hard-core coders who like pain - and 'thread' is one of those modules.
> >  I used it back before Gary wrote the multi-processor code, so I know
> > the pain.  I switched to the 'threading' module and, although this is
> > much easier, it can still be quite painful to run correctly.
> >
> > The threading module will require study
> > (http://docs.python.org/library/threading.html).  Threads are complex
> > beasts and need to be handled very carefully.  To help, there are many
> > tutorials on the Internet that can be found through Google
> > (http://www.google.com/search?hl=en&safe=off&q=python+threading+tutorial&aq=1&aqi=g10&aql=&oq=python+threading).
> >  Once threads are correctly handled, then these error messages will
> > make more sense.  Note that this traceback is not actually where the
> > problem occurred!  That line 157 of generic_fns/minimise.py is
> > perfectly fine.  Therefore until we switch to the 'threading' module
> > and have exceptions properly handled, we will never know what went
> > wrong.
> >
> > Regards,
> >
> > Edward
> >
> >
> >
> > On 4 March 2010 00:30, Michael Bieri <NO-REPLY.INVALID-ADDRESS@xxxxxxx> 
> > wrote:
> >
> > > Follow-up Comment #40, task #6847 (project relax):
> > >
> > > Hi Edward
> > >
> > > There seems to be a problem during dauvergne.py automatic model-free
> > > analysis; i.e. during the final calculation.
> > >
> > > I get the error message:
> > >
> > > relax> minimise(*args=('newton',), func_tol=1e-25, 
> > > max_iterations=10000000,
> > > constraints=True, scaling=True, verbosity=1)
> > >
> > >
> > > Over-fit spin deselection.
> > >
> > >
> > > Simulation 1Unhandled exception in thread started by
> > > Traceback (most recent call last):
> > >  File
> > > "/home/mbieri/LinuxBox/bieri_gui/gui_bieri/analyses/auto_model_free.py", 
> > > line
> > > 835, in execute_thread
> > >    dAuvergne_protocol(diff_model=global_model, mf_models=data.mf_models,
> > > local_tm_models=data.local_tm_models, pdb_file=data.structure_file,
> > > seq_args=data.seq_args, het_name=data.het_name, 
> > > relax_data=data.relax_data,
> > > unres=data.unres, exclude=data.exclude, bond_length=data.bond_length,
> > > csa=data.csa, hetnuc=data.hetnuc, proton=data.proton, 
> > > grid_inc=data.grid_inc,
> > > min_algor=data.min_algor, mc_num=data.mc_num, conv_loop=data.conv_loop)
> > >  File 
> > > "/home/mbieri/LinuxBox/bieri_gui/auto_analyses/dauvergne_protocol.py",
> > > line 395, in __init__
> > >    self.interpreter.minimise(self.min_algor)
> > >  File "/home/mbieri/LinuxBox/bieri_gui/prompt/minimisation.py", line 294, 
> > > in
> > > minimise
> > >    minimise.minimise(min_algor=min_algor, min_options=min_options,
> > > func_tol=func_tol, grad_tol=grad_tol, max_iterations=max_iterations,
> > > constraints=constraints, scaling=scaling, verbosity=verbosity)
> > >  File "/home/mbieri/LinuxBox/bieri_gui/generic_fns/minimise.py", line 157,
> > > in minimise
> > >    print(("Simulation " + repr(i+1)))
> > >
> > >
> > > Cheers
> > > Michael
> > >
> > >
> > > p.s.: I will work on the code next week, so reverting the auto_analysis
> > > changes.
> > >
> > >    _______________________________________________________
> > >
> > > Reply to this item at:
> > >
> > >  <http://gna.org/task/?6847>
> > >
> > > _______________________________________________
> > >  Nachricht geschickt von/durch Gna!
> > >  http://gna.org/
> > >
> > >
> > > _______________________________________________
> > > relax (http://nmr-relax.com)
> > >
> > > This is the relax-devel mailing list
> > > relax-devel@xxxxxxx
> > >
> > > To unsubscribe from this list, get a password
> > > reminder, or change your subscription options,
> > > visit the list information page at
> > > https://mail.gna.org/listinfo/relax-devel
>
> _______________________________________________
> relax (http://nmr-relax.com)
>
> This is the relax-devel mailing list
> relax-devel@xxxxxxx
>
> To unsubscribe from this list, get a password
> reminder, or change your subscription options,
> visit the list information page at
> https://mail.gna.org/listinfo/relax-devel