Hi Paolo,
Welcome to the relax mailing lists. I've looked at your suggestion,
and it might work. Unfortunately I wrote this mfout parsing code a
long, long time ago, and it is not very flexible. There is now the
pystarlib project (http://code.google.com/p/pystarlib/) which would be
a much better way of processing these STAR formatted mfout files.
This is included with the bmrblib project
(http://gna.org/projects/bmrblib/) which will be bundled with relax in
the future. However, considering that Art Palmer has stopped
development of his program Modelfree for over a decade (it's been in
maintenance mode since 04/15/98), this solution would probably be too
much effort.
As Seb said, would you be able to share some of your data? If you
could attach an mfout file with and without simulations to the bug
report (http://gna.org/bugs/?18391), that would be much appreciated.
This only needs to have data for 1 or 2 spins (in the mfdata, etc
files), and this truncated data can be randomised if total secrecy is
needed. I would then be able to add both files to the relax test
suite and add checks to see if relax can read the files and data
correctly. This will guarantee that relax will be able to read these
mfout file for eternity.
Another advantage of adding this to the test suite will be that I can
compare the reading of different mfout files. I have an old file from
2002 generated by Modelfree 4.15 which relax 1.3 can read perfectly as
the code currently is. Which version of Modelfree have you used?
Maybe this is a version difference.
Cheers,
Edward
P. S. If you insist on using Modelfree4, I would recommend a few
source code changes so that you obtain much better results. Firstly
you need to massively increase the precision by the following change:
diff -r modelfree/fglob.F modelfree_orig/fglob.F
49c49
< parameter (maxit=10000000,convrg=1E-25)
---
parameter (maxit=50,convrg=0.0001)
These numbers are much more reasonable and match the default precision
of relax. This low precision was because this code was originally
written to work on this http://en.wikipedia.org/wiki/VAX, without
taking a lifetime to complete. The high precision calculations are no
longer a problem for modern machines.
On 22 July 2011 12:18, Sébastien Morin <sebastien.morin.1@xxxxxxxxx> wrote:
Dear Paolo,
First of all, welcome to relax.
Second, thanks for submitting this bug report as well as a solution.
Have you tested the solution with data without MC simulations performed ?
Could you share a part of your dataset (possibly randomized) such that we
can reproduce the bug and include the data into the test suite for the bug
never to come back again, please ?
Thanks !
Séb :)
On 11-07-21 2:42 PM, Paolo Calligari wrote:
URL:
<http://gna.org/bugs/?18391>
Summary: Palmer extract. Error in chi2 read when 'sims'
is
different from 0
Project: relax
Submitted by: pcalligari
Submitted on: Thu 21 Jul 2011 12:42:45 PM GMT
Category: relax's source code
Severity: 3 - Normal
Priority: 5 - Normal
Status: None
Privacy: Public
Assigned to: None
Originator Name:
Originator Email:
Open/Closed: Open
Discussion Lock: Any
Release: 1.3.10
Operating System: All systems
_______________________________________________________
Details:
When trying to read palmer's modelfree output performed with MC
simulations,
the module palmer.extract(dir) fails.
Looking to the code, I'd say that line 731 in generic_fns/palmer.py
should be changed from
row = split(mfout_lines[chi2_pos + 2 + 22*pos])
to
row = split(mfout_lines[chi2_pos + 2 + 41*pos])
I've tried this change and now it works fine.
cheers,
Paolo
_______________________________________________________
Reply to this item at:
<http://gna.org/bugs/?18391>
_______________________________________________
Message sent via/by Gna!
http://gna.org/
_______________________________________________
relax (http://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
--
Sébastien Morin, Ph.D.
Postdoctoral Fellow, S. Grzesiek NMR Laboratory
Department of Structural Biology
Biozentrum, Universität Basel
Klingelbergstrasse 70
4056 Basel
Switzerland
_______________________________________________
relax (http://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