Re: [bug #22461] NS R1rho 2-site_fit_r1 has extremely high chi2 value in systemtest Relax_disp.test_r1rho_kjaergaard_missing_r1 -- August 22, 2014 - 09:25

Hi Troels,

I've had a quick look at this bug and it is caused by some assumptions
that do not hold for this data set.  I've partly fixed the bug
(http://article.gmane.org/gmane.science.nmr.relax.scm/22943), in that
the maximum relaxation time is taken instead of the first value of
0.0.  For exponential curves, it shouldn't really matter what the time
is as R2eff will be correctly estimated independently of this (for
sanity, it might be worth testing though).  The problem lies in the
return_r2eff_arrays() dispersion function.  For the data set, when
printing out exp_type, frq, offset, point, ei, mi, oi, di, and
relax_times we see:

- R1rho 799777399.1 118.078 431.0 0 0 0 0 [0.0, 0.04, 0.1, 0.2]
- R1rho 799777399.1 118.078 651.2 0 0 0 1 [0.0, 0.04, 0.1, 0.2, 0.4]
RelaxError: ...

In this function, have a look at how relax_times is set up.  This is a
list of lists with the dimensions [Ei][Mi].  But from the above
printout, it is obvious that we have different relaxation times per
offset and dispersion point!  It is rather logical that such a case
would be possible - users can measure whatever they like, and they
usually come up with very strange combinations.  So the solution is to
extend relax_times to have the full dimensions [Ei][Mi][Oi][Di].  From
memory, we have discussed this exact issue before, but nothing was
done about it (I can't find that conversation in the archives though).
This should not be too hard to implement in this function as well as
handling the new dimensions in the dispersion target function class.
I'll let you do it as it's your data set and your code that is
affected.  I hope it'll be easy enough to use the extended structure
in your super fast lib.dispersion modules!  You already have coded the
[Ei][Mi] to [Ei][Si][Mi][Oi][Di] conversion in the target function
class __init__() method, so a [Ei][Mi][Oi][Di] to [Ei][Si][Mi][Oi][Di]
conversion should be trivial for you.

Cheers,

Edward



On 12 August 2014 18:01, Troels E. Linnet
<NO-REPLY.INVALID-ADDRESS@xxxxxxx> wrote:
> Follow-up Comment #2, bug #22461 (project relax):
>
> This error can be triggered by:
>
> relax -s Relax_disp.test_r1rho_kjaergaard_missing_r1 --numpy-raise -d
>
> The error is:
> .....
>   File "/Users/tlinnet/software/R1_fitting/target_functions/relax_disp.py",
> line 314, in __init__
>     self.inv_relax_times[ei, si, mi, :] = 1.0 / relax_time
> FloatingPointError: divide by zero encountered in double_scalars
>
> The time reported is: [[ 0.]]
> This is a division error with 0.0.
>
> This is strange, since the print of time is:
> in: specific_analyses/relax_disp/optimisation.py, line 561
> print self.relax_times : [[ 0.]]
> print cdp.relax_time_list : [0.0, 0.04, 0.1, 0.14, 0.2, 0.4]
>
> The problem resides in:
> specific_analyses/relax_disp/data.py in function 'return_r2eff_arrays()'.
>
> The relax times is expected to have indices [ei][mi].
>
> A "print ei, mi, oi, di, relax_time" in line 4392 in data.py shows:
> 0 0 0 0 0.0
> 0 0 0 0 0.04
> 0 0 0 0 0.1
> 0 0 0 0 0.2
> 0 0 0 1 0.0
> 0 0 0 1 0.04
> 0 0 0 1 0.1
> 0 0 0 1 0.2
> 0 0 0 1 0.4
> 0 0 0 2 0.0
> 0 0 0 2 0.04
> 0 0 0 2 0.1
> 0 0 0 2 0.2
> 0 0 0 2 0.4
> 0 0 0 2 0.0
> 0 0 0 2 0.04
> 0 0 0 2 0.1
> 0 0 0 2 0.2
> 0 0 0 2 0.4
> 0 0 0 2 0.0
> 0 0 0 2 0.04
> 0 0 0 2 0.1
> 0 0 0 2 0.2
> 0 0 0 2 0.4
> 0 0 0 3 0.0
> 0 0 0 3 0.04
> 0 0 0 3 0.1
> 0 0 0 3 0.2
> 0 0 0 3 0.4
> 0 0 0 4 0.0
> 0 0 0 4 0.04
> 0 0 0 4 0.1
> 0 0 0 4 0.2
> 0 0 0 4 0.4
> 0 0 0 4 0.0
> 0 0 0 4 0.04
> 0 0 0 4 0.1
> 0 0 0 4 0.2
> 0 0 0 4 0.4
> 0 0 0 4 0.0
> 0 0 0 4 0.04
> 0 0 0 4 0.1
> 0 0 0 4 0.2
> 0 0 0 4 0.4
> 0 0 0 4 0.0
> 0 0 0 4 0.04
> 0 0 0 4 0.1
> 0 0 0 4 0.2
> 0 0 0 4 0.4
> 0 0 0 4 0.0
> 0 0 0 4 0.04
> 0 0 0 4 0.1
> 0 0 0 4 0.2
> 0 0 0 4 0.4
> 0 0 0 4 0.0
> 0 0 0 4 0.04
> 0 0 0 4 0.1
> 0 0 0 4 0.2
> 0 0 0 4 0.4
> 0 0 0 5 0.0
> 0 0 0 5 0.04
> 0 0 0 5 0.1
> 0 0 0 5 0.14
> 0 0 0 5 0.2
> 0 0 0 5 0.0
> 0 0 0 5 0.04
> 0 0 0 5 0.1
> 0 0 0 5 0.14
> 0 0 0 5 0.2
> 0 0 1 0 0.0
> 0 0 1 0 0.04
> 0 0 1 0 0.1
> 0 0 1 0 0.2
> 0 0 1 0 0.4
> 0 0 1 0 0.0
> 0 0 1 0 0.04
> 0 0 1 0 0.1
> 0 0 1 0 0.2
> 0 0 1 0 0.4
> 0 0 1 0 0.0
> 0 0 1 0 0.04
> 0 0 1 0 0.1
> 0 0 1 0 0.2
> 0 0 1 0 0.4
> 0 0 1 0 0.0
> 0 0 1 0 0.04
> 0 0 1 0 0.1
> 0 0 1 0 0.2
> 0 0 1 0 0.4
> 0 0 1 0 0.0
> 0 0 1 0 0.04
> 0 0 1 0 0.1
> 0 0 1 0 0.2
> 0 0 1 0 0.4
> 0 0 1 0 0.0
> 0 0 1 0 0.04
> 0 0 1 0 0.1
> 0 0 1 0 0.2
> 0 0 1 0 0.4
> 0 0 2 0 0.0
> 0 0 2 0 0.04
> 0 0 2 0 0.1
> 0 0 2 0 0.2
> 0 0 2 0 0.4
> 0 0 2 0 0.0
> 0 0 2 0 0.04
> 0 0 2 0 0.1
> 0 0 2 0 0.2
> 0 0 2 0 0.4
> 0 0 2 0 0.0
> 0 0 2 0 0.04
> 0 0 2 0 0.1
> 0 0 2 0 0.2
> 0 0 2 0 0.4
> 0 0 2 1 0.0
> 0 0 2 1 0.04
> 0 0 2 1 0.1
> 0 0 2 1 0.2
> 0 0 2 1 0.4
> 0 0 2 1 0.0
> 0 0 2 1 0.04
> 0 0 2 1 0.1
> 0 0 2 1 0.2
> 0 0 2 1 0.4
> 0 0 2 1 0.0
> 0 0 2 1 0.04
> 0 0 2 1 0.1
> 0 0 2 1 0.2
> 0 0 2 1 0.4
> 0 0 2 1 0.0
> 0 0 2 1 0.04
> 0 0 2 1 0.1
> 0 0 2 1 0.2
> 0 0 2 1 0.4
> 0 0 2 1 0.0
> 0 0 2 1 0.04
> 0 0 2 1 0.1
> 0 0 2 1 0.2
> 0 0 2 1 0.4
> 0 0 2 1 0.0
> 0 0 2 1 0.04
> 0 0 2 1 0.1
> 0 0 2 1 0.2
> 0 0 2 1 0.4
> 0 0 2 2 0.0
> 0 0 2 2 0.04
> 0 0 2 2 0.1
> 0 0 2 2 0.14
> 0 0 2 2 0.2
> 0 0 2 2 0.0
> 0 0 2 2 0.04
> 0 0 2 2 0.1
> 0 0 2 2 0.14
> 0 0 2 2 0.2
> 0 0 3 0 0.0
> 0 0 3 0 0.04
> 0 0 3 0 0.1
> 0 0 3 0 0.2
> 0 0 3 0 0.4
> 0 0 3 0 0.0
> 0 0 3 0 0.04
> 0 0 3 0 0.1
> 0 0 3 0 0.2
> 0 0 3 0 0.4
> 0 0 3 0 0.0
> 0 0 3 0 0.04
> 0 0 3 0 0.1
> 0 0 3 0 0.2
> 0 0 3 0 0.4
> 0 0 3 1 0.0
> 0 0 3 1 0.04
> 0 0 3 1 0.1
> 0 0 3 1 0.2
> 0 0 3 1 0.4
> 0 0 3 1 0.0
> 0 0 3 1 0.04
> 0 0 3 1 0.1
> 0 0 3 1 0.2
> 0 0 3 1 0.4
> 0 0 3 1 0.0
> 0 0 3 1 0.04
> 0 0 3 1 0.1
> 0 0 3 1 0.2
> 0 0 3 1 0.4
> 0 0 3 1 0.0
> 0 0 3 1 0.04
> 0 0 3 1 0.1
> 0 0 3 1 0.2
> 0 0 3 1 0.4
> 0 0 3 1 0.0
> 0 0 3 1 0.04
> 0 0 3 1 0.1
> 0 0 3 1 0.2
> 0 0 3 1 0.4
> 0 0 3 1 0.0
> 0 0 3 1 0.04
> 0 0 3 1 0.1
> 0 0 3 1 0.2
> 0 0 3 1 0.4
> 0 0 4 0 0.0
> 0 0 4 0 0.04
> 0 0 4 0 0.1
> 0 0 4 0 0.2
> 0 0 4 0 0.4
> 0 0 4 0 0.0
> 0 0 4 0 0.04
> 0 0 4 0 0.1
> 0 0 4 0 0.2
> 0 0 4 0 0.4
> 0 0 4 0 0.0
> 0 0 4 0 0.04
> 0 0 4 0 0.1
> 0 0 4 0 0.2
> 0 0 4 0 0.4
> 0 0 4 0 0.0
> 0 0 4 0 0.04
> 0 0 4 0 0.1
> 0 0 4 0 0.2
> 0 0 4 0 0.4
> 0 0 4 0 0.0
> 0 0 4 0 0.04
> 0 0 4 0 0.1
> 0 0 4 0 0.2
> 0 0 4 0 0.4
> 0 0 4 0 0.0
> 0 0 4 0 0.04
> 0 0 4 0 0.1
> 0 0 4 0 0.2
> 0 0 4 0 0.4
> 0 0 5 0 0.0
> 0 0 5 0 0.04
> 0 0 5 0 0.1
> 0 0 5 0 0.2
> 0 0 5 0 0.4
> 0 0 5 0 0.0
> 0 0 5 0 0.04
> 0 0 5 0 0.1
> 0 0 5 0 0.2
> 0 0 5 0 0.4
> 0 0 5 0 0.0
> 0 0 5 0 0.04
> 0 0 5 0 0.1
> 0 0 5 0 0.2
> 0 0 5 0 0.4
> 0 0 5 0 0.0
> 0 0 5 0 0.04
> 0 0 5 0 0.1
> 0 0 5 0 0.2
> 0 0 5 0 0.4
> 0 0 5 0 0.0
> 0 0 5 0 0.04
> 0 0 5 0 0.1
> 0 0 5 0 0.2
> 0 0 5 0 0.4
> 0 0 5 0 0.0
> 0 0 5 0 0.04
> 0 0 5 0 0.1
> 0 0 5 0 0.2
> 0 0 5 0 0.4
>
>     _______________________________________________________
>
> Reply to this item at:
>
>   <http://gna.org/bugs/?22461>
>
> _______________________________________________
>   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