Hi,
In your testing, you should set the 'profile_flag' argument to True in
the 'relax' file. This will allow you to see the timings for all
function calls. For each optimisation, you use the profile printout
to determine if the optimisation actually worked.
Regards,
Edward
On 14 February 2014 10:25, Troels E. Linnet
<NO-REPLY.INVALID-ADDRESS@xxxxxxx> wrote:
URL:
<http://gna.org/support/?3118>
Summary: Optimization for faster reading of state files
Project: relax
Submitted by: tlinnet
Submitted on: Fri 14 Feb 2014 09:25:22 AM UTC
Category: None
Priority: 5 - Normal
Severity: 1 - Wish
Status: Postponed
Assigned to: None
Originator Email:
Open/Closed: Open
Discussion Lock: Any
Operating System: None
_______________________________________________________
Details:
This follows the discussion in:
http://thread.gmane.org/gmane.science.nmr.relax.devel/4999
I am storing different runs of R2eff values into state files.
Their file size is app. 5 MB each.
The loading time is about 20-30 seconds for a state file.
After turning of comprepression, the file size reached ~70 MB.
But with no considerable increase in reading speed.
The optimization should probably be found here:
----------
Converting between data structures in the relax data store and the XML.
These
are all the to_xml() and from_xml() functions you can find in the
'data_store'
package.
There have been zero optimisations here and, in some parts, the code is
likely
to be extremely inefficient (too much looping, inefficient ordering of loop
nesting, etc.).
This would be the best target for optimisations.
And the loops are where you will squeeze out the most.
To find the relevant functions/methods, just run:
$ grep -r "def to_xml\|def from_xml" . --exclude-dir=".svn"
To perform any optimisations, the test suite must be used to make sure that
nothing breaks.
If you notice a breakage in a save file you are using as a test for yourself
while the test suite passes, then a new
system test needs to be created using a heavily truncated save file.
_______________________________________________________
Reply to this item at:
<http://gna.org/support/?3118>
_______________________________________________
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
Re: [sr #3118] Optimization for faster reading of state files