New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
HCAL reco speed up by David L. #6669
HCAL reco speed up by David L. #6669
Conversation
if(OffsetTemp + 25 >= (int)cipSize) | ||
C1 = CumulativeIdealPulse[cipSize-1]; | ||
else | ||
if( OffsetTemp >= 25) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I asked David if this is intended or an error.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think it's an error.
The old code had everything in terms of "OffsetTemp + 25". There were two ifs: one for at least cipSize, and one "between 0 and below cipSize". So, else if ( OffsetTemp >= 25)
should be else if ( OffsetTemp >= - 25)
It's a bit surprising that there is no large effect from this, maybe it's numerically small
A new Pull Request was created by @StoyanStoynev for CMSSW_7_4_X. HCAL reco speed up by David L. It involves the following packages: RecoLocalCalo/HcalRecAlgos @cmsbuild, @nclopezo, @StoyanStoynev, @slava77 can you please review it and eventually sign? Thanks. |
I ran on ttbar sample (wf 202, 200 events) including with igprof. From the time report I see |
I expect interested parties to comment/test/adopt (@lihux25 @violatingcp @abdoulline @bachtis ). |
@@ -62,6 +62,7 @@ void HcalSimpleRecAlgo::setpuCorrParams(bool iPedestalConstraint, bool iTimeCo | |||
double its3Chi2,double its4Chi2,double its345Chi2,double iChargeThreshold, int iFitTimes) { | |||
if( iPedestalConstraint ) assert ( iPedSig ); | |||
if( iTimeConstraint ) assert( iTimeSig ); | |||
std::cout << "ciao\n"; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
to be removed
I take the liberty to notify Igor. While I appreciate David's effort to make it 20% faster,
On Thu, 27 Nov 2014, StoyanStoynev wrote:
|
Hi Salavat, My understanding of the plans for HLT was that it will most likely be Please clarify. On 11/27/14, 11:59 AM, abdoulline wrote:
Vyacheslav (Slava) Krutelyov |
You're right, Slava. This is most plausible scenario. On Thu, 27 Nov 2014, Slava Krutelyov wrote:
|
@abdoulline Are you going to test or anyway make use of the PR? Our tests are with some limits so if you want us to sign off this PR better let us know you also approve it explicitly. In any case you have the code, you can just use it in your forthcoming (?) PR/s/. |
as far as I see, all the relevant PR test distributions come out green, Salavat On Mon, 1 Dec 2014, StoyanStoynev wrote:
|
+1 |
The plot above is for wf 202. |
This pull request is fully signed and it will be integrated in one of the next CMSSW_7_4_X IBs unless changes (tests are also fine). This pull request will be automatically merged. |
…00/HCAL_David HCAL reco speed up by David L.
This is a copy from https://github.com/davidlange6/cmssw/tree/hcal_speedups_141121 (davidlange6@477c4b2). For now the purpose of the PR is just to make the code available and ease the discussion. I am to comment more below but overall the code is a bit over 20% faster with minor regressions.