[bug #23186] Error calculation of individual parameter "dw" from Monte-Carlo, is based on first spin. -- January 14, 2015 - 13:47

URL:
  <http://gna.org/bugs/?23186>

                 Summary: Error calculation of individual parameter "dw" from
Monte-Carlo, is based on first spin.
                 Project: relax
            Submitted by: tlinnet
            Submitted on: Wed 14 Jan 2015 12:47:37 PM UTC
                Category: None
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:

This was discovered in post:

http://thread.gmane.org/gmane.science.nmr.relax.user/1795

The problem is in:
specific_analyses.relax_disp/api.sim_return_param()

------------------
        # Convert the parameter index.
        if model_param:
            param_name, si, mi = param_index_to_param_info(index=index,
spins=spins)
            if not param_name in ['r2eff', 'i0']:
                si = 0
        else:
            param_name = aux_params[index - total_param_num]
            si = 0
            mi = 0

...............

        # Model and auxiliary parameters.
        else:
            sim_data = getattr(spins[si], param_name + "_sim")

------------------------------

Since si = 0, it takes dw values from initial spin, and uses this as error for
all spins.

This is not correct.





    _______________________________________________________

Reply to this item at:

  <http://gna.org/bugs/?23186>

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