[bug #21953] Weird performance of grid search -- April 22, 2014 - 13:41

URL:
  <http://gna.org/bugs/?21953>

                 Summary: Weird performance of grid search
                 Project: relax
            Submitted by: tlinnet
            Submitted on: Tue 22 Apr 2014 11:41:51 AM UTC
                Category: relax's source code
Specific analysis category: Relaxation dispersion
                Priority: 5 - Normal
                Severity: 3 - Normal
                  Status: None
             Assigned to: tlinnet
         Originator Name: 
        Originator Email: 
             Open/Closed: Open
                 Release: Repository: trunk
         Discussion Lock: Any
        Operating System: 64-bit GNU/Linux

    _______________________________________________________

Details:

This bug has been discussed here:
http://thread.gmane.org/gmane.science.nmr.relax.devel/5302

The grid search is not finding suitable start values for parameters kex and
pA.

The solution could be:
grid_search_setup() function in the specific_analyses.relax_disp.optimisation
module.

The main problem would be the random guesses, as to what the lower and upper
grid search values should be.

For kex that the lower value is 1 and the upper value is 1e5.
Probably use 0 to 1e4 for kex instead, as the higher kex values should be rare
and easy to optimise from the lower values.

It could be that because most of the kex grid points are at very high values,
that most of the grid search is not doing anything.





    _______________________________________________________

Reply to this item at:

  <http://gna.org/bugs/?21953>

_______________________________________________
  Message sent via/by Gna!
  http://gna.org/