For info, I've modified the .pep.xml parsing into a single step, and on
a very small test dataset ASAPRatioProteinRatioParser run-time is down
from 19.8s to 0.8s. I'd expect the speed-up on a large dataset to be
higher due to the compound effects of more protein groups + bigger .pep.xml.
Will tidy up the code and test thoroughly next week, and then share it.
Cheers,
DT
Brian Pratt wrote:
Nothing to add! Dave has the right idea for a performance fix, and Natalie is
correct about TPP being addicted to Boost already.
Brian
On Thu, Feb 4, 2010 at 2:37 PM, Natalie Tasman
<[email protected]<mailto:[email protected]>> wrote:
Hi Dave, Brian,
Just jumping into comment on Boost. Yes, it is welcomed and in fact already
used in the TPP (as well as the TPP-included ProteoWizard project); however,
because the Boost API and process of building Boost libraries have not been
particularly stable, we've found it necessary to fix the version of Boost that
we work with. Our current version is 1.39.0. As long as you can test against
that, you're contributes would be fine-- and no doubt very welcomed.
-Natalie
On 2/4/10 2:26 PM, Dave Trudgian wrote:
Brian,
Yup. I just discovered this too, as per other post. On our servers it's not
disk-bound, as the 1.6GB .pep.xml is fully cached, but the continued rpeated
slows things down.
R.E. solutions for this, is Boost code welcomed in the main TPP tools? I think
I can re-write using a single pass parse of the .pep.xml into a
Boost.MultiIndex hash of structs/classes containing the required peptide info.
DT
Brian Pratt wrote:
Looking at the code I can see where this would easily become diskbound for
large data sets - it reads and rereads the same pepXML files repeatedly, but
the effect is probably masked by disk cacheing up to a certain point. Somebody
would need to write some logic for cacheing the file contents to speed this up
properly.
Brian
On Thu, Feb 4, 2010 at 12:09 PM, Jake W <[email protected]<mailto:[email protected]>
<mailto:[email protected]<mailto:[email protected]>>> wrote:
I've seen the same thing. On datasets where
ASAPRatioPeptideRatioParser completes in a few minutes,
ASAPRatioProteinRatioParser can take an hour or so. This is on a
Windows machine running TPP ver. 4.3.0.
Jake
--
You received this message because you are subscribed to the Google
Groups "spctools-discuss" group.
To post to this group, send email to
[email protected]<mailto:[email protected]>
<mailto:[email protected]<mailto:[email protected]>>.
To unsubscribe from this group, send email to
[email protected]<mailto:spctools-discuss%[email protected]>
<mailto:spctools-discuss%[email protected]<mailto:spctools-discuss%[email protected]>>.
For more options, visit this group at
http://groups.google.com/group/spctools-discuss?hl=en.
--
You received this message because you are subscribed to the Google Groups
"spctools-discuss" group.
To post to this group, send email to
[email protected]<mailto:[email protected]>.
To unsubscribe from this group, send email to
[email protected]<mailto:spctools-discuss%[email protected]>.
For more options, visit this group at
http://groups.google.com/group/spctools-discuss?hl=en.
--
You received this message because you are subscribed to the Google Groups
"spctools-discuss" group.
To post to this group, send email to
[email protected]<mailto:[email protected]>.
To unsubscribe from this group, send email to
[email protected]<mailto:spctools-discuss%[email protected]>.
For more options, visit this group at
http://groups.google.com/group/spctools-discuss?hl=en.
--
You received this message because you are subscribed to the Google Groups
"spctools-discuss" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/spctools-discuss?hl=en.
--
Dr. David Trudgian
Bioinformatician in Proteomics
University of Oxford
Mon-Thu: CCMP, Roosevelt Drive
Tel: (+44) (01865 2)87784
Friday : Dunn School of Pathology, S. Parks Rd.
Tel: (+44) (01865 2)75557
--
You received this message because you are subscribed to the Google Groups
"spctools-discuss" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/spctools-discuss?hl=en.
