mailRe: Stop and restart


Others Months | Index by Date | Thread Index
>>   [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Header


Content

Posted by Edward d'Auvergne on January 16, 2007 - 02:43:
Unfortunately it is not possible to stop relax at any point in time
and then restart it from that exact position.

The ability to suspend the program at any stage and then resume later
on is a very complex coding issue which relax has not even tried to
implement.  For example if you are in the middle of a large grid
search, there is the complex issue of the granularity of the
information saved about the current status of the program - which
function is currently being executed, what all the internal variables
are, how far down the function you are up to, etc - which for number
crunching would be a very large overhead.  There is also the issue of
how far down a script relax has read (which again is complex as the
script can include functions, classes, or any Python programming
semantics).  Also the script and where in the script needs to be
saved.

For the 'full_analysis.py' script, things are a little different.  For
one iteration of the script, many calculations are executed and saved.
If the script is interrupted, you'll have a lot of usable information
saved in the results files.  In that case what you need to do is
create a new script (or set of scripts) by copying parts of the
'full_analysis.py' script so that it runs parts of the analysis from
where you lost power.  For example if you were in the multi_model()
function of the script, then you just need to point it to the previous
iteration to get the diffusion tensor and optimise the model-free
models which weren't optimised, run the AIC model selection, and run
the final optimisation of all parameters.  I did want to make the
script flexible enough to stop and resume at any spot (with the
granularity at the level of the results files), but that would have
made the script far too complex.

I hope this helps,

Edward




On 1/16/07, Sebastien Morin <sebastien.morin.1@xxxxxxxxx> wrote:
Hi everyone

We just had a major electricity failure in Quebec City and Laval
University campus and this incident raised questions.

I wonder if it is possible to stop a relax run and then restart it
exactly where it was. I know it is possible with some programs...

If it is not possible, I guess I have to delete the output folders
created with the full_analysis.py script from the iteration that was
stopped and then start the iteration again...

Thanks !


Sébastien


_______________________________________________ relax (http://nmr-relax.com)

This is the relax-users mailing list
relax-users@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-users




Related Messages


Powered by MHonArc, Updated Tue Jan 16 07:40:13 2007