switch to Pandora recipe with no fragment removal algorithms

[kpedro88]

This commit adds a modified Pandora recipe with no fragment removal algorithms, and sets it as the default recipe in the python configuration. This increases the speed of Pandora in high-pileup environments with little effect on the jet resolution, see attached slides:

attn: @lgray, @vandreev11, @pfs, @boudoul

[cmsbuild]

A new Pull Request was created by @kpedro88 (Kevin Pedro) for CMSSW_6_2_X_SLHC.

switch to Pandora recipe with no fragment removal algorithms

It involves the following packages:

RecoParticleFlow/PandoraTranslator

The following packages do not have a category, yet:

RecoParticleFlow/PandoraTranslator

@cmsbuild, @nclopezo can you please review it and eventually sign? Thanks.
@mmarionncern, @bachtis, @lgray this is something you requested to watch as well.
You can sign-off by replying to this message having '+1' in the first line of your reply.
You can reject by replying to this message having '-1' in the first line of your reply.
If you are a L2 or a release manager you can ask for tests by saying 'please test' in the first line of a comment.
@fratnikov, @mark-grimes you are the release manager for this.
You can merge this pull request by typing 'merge' in the first line of your comment.

[mark-grimes]

Do you want this as the default for all pileup scenarios? We could put something in the customisation that checks what pileup it is and decide which config to run.

[kpedro88]

I'm not sure what would be best at this moment. @lgray, @vandreev11, what do you think?

[lgray]

Busy right now, back to you later tonight.

(Sent from my Nexus 6)
On Apr 23, 2015 6:54 PM, "Kevin Pedro" notifications@github.com wrote:

I'm not sure what would be best at this moment. @lgray
https://github.com/lgray, @vandreev11 https://github.com/vandreev11,
what do you think?

—
Reply to this email directly or view it on GitHub
#8856 (comment).

[vandreev11]

On my opinion it should be the same for all PU scenarios. PU0 could be a special case, but it is better to avoid that for the sake of comparisons, drawing conclusions. I believe it is temporary measure, clustering in Pandora needs to be tuned yet for high pileup case.

[lgray]

Along the same lines as Valeri, since we're going to be doing a massive
overhaul of the clustering to begin with, perhaps it is better to keep
things configured the same with the stock pandora algorithms for all PU
scenarios.

We'll likely still be using some of the linear collider algorithms even
after we've finished developing specialized clustering/ID/PF, so it's good
to keep configurations straightforward and simple.

-L

(Sent from my Nexus 6)
On Apr 24, 2015 06:25, "vandreev11" notifications@github.com wrote:

On my opinion it should be the same for all PU scenarios. PU0 could be a
special case, but it is better to avoid that for the sake of comparisons,
drawing conclusions. I believe it is temporary measure, clustering in
Pandora needs to be tuned yet for high pileup case.

—
Reply to this email directly or view it on GitHub
#8856 (comment).

[mark-grimes]

merge

Okay, so that's everything for the next patch release right? Getting on a plane now, I'll put in the build request when I land.

[kpedro88]

Yes, that should be everything.