Re: r23362 - /trunk/test_suite/unit_tests/_lib/_dispersion/test_ns_cpmg_2site_expanded.py -- May 23, 2014 - 09:42

Oh, that's right, the auto-analysis is using the default values if
grid_inc is set to None:

http://www.nmr-relax.com/api/3.2/auto_analyses.relax_disp-pysrc.html#Relax_disp.optimise

That's where the problem comes from!

Regards,

Edward



On 23 May 2014 09:33, Edward d'Auvergne <edward@xxxxxxxxxxxxx> wrote:
> Hi,
>
>
> > I have before tried to change the standard parameters for pA and kex.
> >
> > diff --git a/specific_analyses/relax_disp/parameter_object.py
> > b/specific_analyses/relax_disp/parameter_object.py
> >
> > -        self._add('pA', scope='spin', default=0.5, desc='The
> > population for state A', set='params', py_type=float,
> > grace_string='\\qp\\sA\\N\\Q', err=True, sim=True)
> > +        self._add('pA', scope='spin', default=0.95, desc='The
> > population for state A', set='params', py_type=float,
> > grace_string='\\qp\\sA\\N\\Q', err=True, sim=True)
> >
> > -        self._add('kex', scope='spin', default=10000.0, desc='The
> > exchange rate', set='params', py_type=float,
> > grace_string='\\qk\\sex\\N\\Q (rad.s\\S-1\\N)', err=True, sim=True)
> > +        self._add('kex', scope='spin', default=1000.0, desc='The
> > exchange rate', set='params', py_type=float,
> > grace_string='\\qk\\sex\\N\\Q (rad.s\\S-1\\N)', err=True, sim=True)
>
> These defaults are reasonable.  But to keep pA away from 1.0, I think
> I'd prefer 0.90.  I'll look and see what is happening to the system
> tests with this change.
>
>
> > But I skipped it, since many of your systemtests started to change a
> > lot of values?
>
> It shouldn't because I have tried to make sure that the value.set user
> function is always used to start at a reasonable position
> (substituting grid_search which is far too slow for the test suite).
> Therefore these defaults should almost never be used.  I might in the
> future completely eliminate the concept of the default value.
>
>
> > And how about the linear constraints for minimising?
> >
> > def linear_constraints(spins=None, scaling_matrix=None):
> >     """Set up the relaxation dispersion curve fitting linear
> > constraint matrices A and b.
> >
> >     Standard notation
> >     =================
> >
> >     The different constraints used within different models are::
> >
> >         0 <= R2 <= 200
> >         0 <= R2A <= 200
> >         0 <= R2B <= 200
> >         pC <= pB <= pA <= 1
> >         pA >= 0.85 (the skewed condition, pA >> pB)
> >         pB >= 0
> >         pC >= 0
> >         phi_ex >= 0
> >         phi_ex_B >= 0
> >         phi_ex_C >= 0
> >         padw2 >= 0
> >         dw >= 0
> >         0 <= kex <= 2e6
> >         0 <= k_AB <= 2e6
> >         0 <= kB <= 2e6
> >         0 <= kC <= 2e6
> >         tex >= 0
> >         k_AB >= 0
>
> If you look at the code, you'll see that this list is not complete:
>
> http://www.nmr-relax.com/api/3.2/specific_analyses.relax_disp.parameters-pysrc.html#linear_constraints
>
> Do you have a suggestion?  Do you see an error?
>
> Regards,
>
> Edward