Re: [bug #22017] LinAlgError, for all numerical CPMG models. -- May 07, 2014 - 18:34

Hi Edward.

Some test print shows this:
In lib/ns_cpmg_2site_3d.py

        print R*tcp[i]
[[ NaN  NaN  NaN  NaN  NaN  NaN  NaN]
 [ NaN  NaN  NaN  NaN  NaN  NaN  NaN]
 [ NaN  NaN  NaN  NaN  NaN  NaN  NaN]
 [ NaN  NaN  NaN  NaN  NaN  NaN  NaN]
 [ NaN  NaN  NaN  NaN  NaN  NaN  NaN]
 [ NaN  NaN  NaN  NaN  NaN  NaN  NaN]
 [ NaN  NaN  NaN  NaN  NaN  NaN  NaN]]

Error message:
 Rexpo = matrix_exponential(R*tcp[i])
  File 
"/sbinlab2/tlinnet/software/NMR-relax/relax_trunk/lib/linear_algebra/matrix_exponential.py",
line 46, in matrix_exponential
    W, V = eig(A)
  File "/usr/lib64/python2.6/site-packages/numpy/linalg/linalg.py",
line 999, in eig
    _assertFinite(a)
  File "/usr/lib64/python2.6/site-packages/numpy/linalg/linalg.py",
line 144, in _assertFinite
    raise LinAlgError, "Array must not contain infs or NaNs"
LinAlgError: Array must not contain infs or NaNs



2014-05-07 17:14 GMT+02:00 Edward d'Auvergne <edward@xxxxxxxxxxxxx>:
> Hi,
>
> This is part of the parameter checking required in the target function
> code.  Something is producing a NaN, and this needs to be caught.
> You'll see my attempts at catching these conditions by running the
> following from the relax base directory:
>
> $ grep "if " lib/dispersion/* -A1
>
> Obviously something is not caught when it should be.  You can use the
> --traceback and --numpy-raise relax options to better debug this.
> Once you find the position in the code where the NaN (or Inf) arises,
> you can then add such an 'if' condition to fix the bug.
>
> Regards,
>
> Edward
>
>
>
> On 7 May 2014 17:07, Troels E. Linnet <NO-REPLY.INVALID-ADDRESS@xxxxxxx> 
> wrote:
> > URL:
> >   <http://gna.org/bugs/?22017>
> >
> >                  Summary: LinAlgError, for all numerical CPMG models.
> >                  Project: relax
> >             Submitted by: tlinnet
> >             Submitted on: Wed 07 May 2014 03:07:49 PM UTC
> >                 Category: relax's source code
> > Specific analysis category: Relaxation dispersion
> >                 Priority: 7 - High
> >                 Severity: 4 - Important
> >                   Status: None
> >              Assigned to: None
> >          Originator Name:
> >         Originator Email:
> >              Open/Closed: Open
> >                  Release: Repository: trunk
> >          Discussion Lock: Any
> >         Operating System: All systems
> >
> >     _______________________________________________________
> >
> > Details:
> >
> > Hi Edward.
> >
> > The attached script, will make relax trip.
> >
> > I am reading in fake R2eff points, and then do a back calculation with:
> >
> >     sfrq_1 = 599.8908617*1E6
> >     ncycs_1 = [2, 4, 8, 10, 20, 30, 40, 60]
> >     sfrq_2 = 499.8908617*1E6
> >     ncycs_2 = [2, 4, 8, 10, 30, 35, 40, 50]
> >
> > With the following dynamic parameters
> > ['Ala', 1, 'N', {'r2': {r20_key_1:10, r20_key_2:11.5}, 'r2a': 
> > {r20_key_1:10,
> > r20_key_2:11.5}, 'r2b': {r20_key_1:10, r20_key_2:11.5}, 'kex': 1000, 'pA':
> > 0.99, 'dw': 2} ]
> >
> > #################3
> > relax_disp.r2eff_read_spin(id='CPMG_599.9', spin_id=':1@N',
> > file='CPMG_599.9_1_N.txt', dir=None, disp_point_col=1, offset_col=None,
> > data_col=2, error_col=3, sep=None)
> > Opening the file 'CPMG_599.9_1_N.txt' for reading.
> > The following R2eff/R1rho data has been loaded into the relax data store:
> >
> > # R2eff_key                                 Disp_point              R2eff
> >              R2eff_error
> > sq_cpmg_599.89086170_0.000_33.333             33.333333333333336
> > 1.000000000000000       0.100000000000000
> > sq_cpmg_599.89086170_0.000_66.667             66.666666666666671
> > 1.000000000000000       0.100000000000000
> > sq_cpmg_599.89086170_0.000_133.333           133.333333333333343
> > 1.000000000000000       0.100000000000000
> > sq_cpmg_599.89086170_0.000_166.667           166.666666666666686
> > 1.000000000000000       0.100000000000000
> > sq_cpmg_599.89086170_0.000_333.333           333.333333333333371
> > 1.000000000000000       0.100000000000000
> > sq_cpmg_599.89086170_0.000_500.000           500.000000000000000
> > 1.000000000000000       0.100000000000000
> > sq_cpmg_599.89086170_0.000_666.667           666.666666666666742
> > 1.000000000000000       0.100000000000000
> > sq_cpmg_599.89086170_0.000_1000.000         1000.000000000000000
> > 1.000000000000000       0.100000000000000
> > Traceback (most recent call last):
> >   File 
> > "/sbinlab2/tlinnet/software/NMR-relax/relax_trunk/multi/processor.py",
> > line 494, in run
> >     self.callback.init_master(self)
> >   File 
> > "/sbinlab2/tlinnet/software/NMR-relax/relax_trunk/multi/__init__.py",
> > line 318, in default_init_master
> >     self.master.run()
> >   File "/sbinlab2/tlinnet/software/NMR-relax/relax_trunk/relax.py", line 
> > 194,
> > in run
> >     self.interpreter.run(self.script_file)
> >   File
> > "/sbinlab2/tlinnet/software/NMR-relax/relax_trunk/prompt/interpreter.py", 
> > line
> > 275, in run
> >     return run_script(intro=self.__intro_string, local=locals(),
> > script_file=script_file, show_script=self.__show_script,
> > raise_relax_error=self.__raise_relax_error)
> >   File
> > "/sbinlab2/tlinnet/software/NMR-relax/relax_trunk/prompt/interpreter.py", 
> > line
> > 569, in run_script
> >     return console.interact(intro, local, script_file,
> > show_script=show_script, raise_relax_error=raise_relax_error)
> >   File
> > "/sbinlab2/tlinnet/software/NMR-relax/relax_trunk/prompt/interpreter.py", 
> > line
> > 468, in interact_script
> >     exec_script(script_file, local)
> >   File
> > "/sbinlab2/tlinnet/software/NMR-relax/relax_trunk/prompt/interpreter.py", 
> > line
> > 347, in exec_script
> >     runpy.run_module(module, globals)
> >   File "/usr/lib64/python2.6/runpy.py", line 140, in run_module
> >     fname, loader, pkg_name)
> >   File "/usr/lib64/python2.6/runpy.py", line 34, in _run_code
> >     exec code in run_globals
> >   File
> > "/sbinlab2/tlinnet/software/NMR-relax/relax_trunk/test_suite/system_tests/scripts/relax_disp/cpmg_synthetic.py",
> > line 270, in <module>
> >     r2effs = optimisation.back_calc_r2eff(spin=cur_spin, 
> > spin_id=cur_spin_id)
> >   File
> > "/sbinlab2/tlinnet/software/NMR-relax/relax_trunk/specific_analyses/relax_disp/optimisation.py",
> > line 188, in back_calc_r2eff
> >     chi2 = model.func(param_vector)
> >   File
> > "/sbinlab2/tlinnet/software/NMR-relax/relax_trunk/target_functions/relax_disp.py",
> > line 1419, in func_ns_cpmg_2site_3D
> >     return self.calc_ns_cpmg_2site_3D_chi2(R20A=R20, R20B=R20, dw=dw, 
> > pA=pA,
> > kex=kex)
> >   File
> > "/sbinlab2/tlinnet/software/NMR-relax/relax_trunk/target_functions/relax_disp.py",
> > line 542, in calc_ns_cpmg_2site_3D_chi2
> >     r2eff_ns_cpmg_2site_3D(r180x=self.r180x, M0=self.M0, 
> > r20a=R20A[r20_index],
> > r20b=R20B[r20_index], pA=pA, pB=pB, dw=dw_frq, k_AB=k_AB, k_BA=k_BA,
> > inv_tcpmg=self.inv_relax_times[0][mi], tcp=self.tau_cpmg[0][mi],
> > back_calc=self.back_calc[0][si][mi][0],
> > num_points=self.num_disp_points[0][si][mi][0], power=self.power[0][mi])
> >   File
> > "/sbinlab2/tlinnet/software/NMR-relax/relax_trunk/lib/dispersion/ns_cpmg_2site_3d.py",
> > line 115, in r2eff_ns_cpmg_2site_3D
> >     Rexpo = matrix_exponential(R*tcp[i])
> >   File
> > "/sbinlab2/tlinnet/software/NMR-relax/relax_trunk/lib/linear_algebra/matrix_exponential.py",
> > line 46, in matrix_exponential
> >     W, V = eig(A)
> >   File "/usr/lib64/python2.6/site-packages/numpy/linalg/linalg.py", line 
> > 999,
> > in eig
> >     _assertFinite(a)
> >   File "/usr/lib64/python2.6/site-packages/numpy/linalg/linalg.py", line 
> > 144,
> > in _assertFinite
> >     raise LinAlgError, "Array must not contain infs or NaNs"
> > LinAlgError: Array must not contain infs or NaNs
> > #########################
> >
> >
> >
> >
> >
> >
> >     _______________________________________________________
> >
> > File Attachments:
> >
> >
> > -------------------------------------------------------
> > Date: Wed 07 May 2014 03:07:49 PM UTC  Name: cpmg_synthetic.py  Size: 16kB
> > By: tlinnet
> >
> > <http://gna.org/bugs/download.php?file_id=20667>
> >
> >     _______________________________________________________
> >
> > Reply to this item at:
> >
> >   <http://gna.org/bugs/?22017>
> >
> > _______________________________________________
> >   Message sent via/by Gna!
> >   http://gna.org/
> >
> >
> > _______________________________________________
> > relax (http://www.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