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
HBHE: use proper quantization noise in the M2 #17037
Conversation
A new Pull Request was created by @mariadalfonso for CMSSW_9_0_X. It involves the following packages: RecoLocalCalo/HcalRecAlgos @cmsbuild, @cvuosalo, @slava77, @davidlange6 can you please review it and eventually sign? Thanks. cms-bot commands are listed here #13028 |
@mariadalfonso |
@cmsbuild please test |
The tests are being triggered in jenkins. |
if pre-2017 QIE8 reco can not be handled uniformly with 2017 with the new quantization uncertainties, I think it would be better to keep all QIE8/HPD reco the same in the same release. |
To have in the 2016 scenario as well I have to clone the function here I do think is better to have the fix for whole new 2017 detector (HB,HE) as we should aim to have the best hcal reco.
|
On 12/14/16 7:08 AM, mariadalfonso wrote:
@slava77 <https://github.com/slava77>
To have in the 2016 scenario as well I have to clone the function here
https://github.com/cms-sw/cmssw/pull/16951/files#diff-5bffd97385f18e7da1326160e40795c1R139
inside the
RecoLocalCalo/HcalRecAlgos/src/PulseShapeFitOOTPileupCorrection.cc
I do think is better to have the fix for whole new 2017 detector (HB,HE)
as we should aim to have the best hcal reco.
What is the pre-2017 use ?
1. comparison of the relVal ?
2. The "phase1" code can also be used for the run2 legacy re-reco and
in case we have the full detector with the HPD thanks to the switch
channelData.hasTimeInfo()
@igv4321 <https://github.com/igv4321>
We should have the best reco available everywhere a given detector is
active.
If the code is not flexible enough, it should be made so, and then the
reco applied.
…
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#17037 (comment)>, or
mute the thread
<https://github.com/notifications/unsubscribe-auth/AEdcbl5eubxAKgOVMTQhPToi2fr7WeKrks5rIAZggaJpZM4LM-fy>.
|
@slava77 |
On 12/14/16 7:20 AM, mariadalfonso wrote:
@slava77 <https://github.com/slava77>
ok, will change the 2016 scenario as well.
actually happy to do this for this and other fixes
Thank you very much.
…
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#17037 (comment)>, or
mute the thread
<https://github.com/notifications/unsubscribe-auth/AEdcbmY2hBvlss9krqStDmAyw-jRJ5_Oks5rIAkygaJpZM4LM-fy>.
|
tsDFcPerADC is actually the differential gain: the width of that particular ADC count in fC. The nominal noise is actually tsDFcPerADC(ip)/sqrt(12.0). The factor 1/sqrt(12) is probably too optimistic, so (in an email thread "Re: QIE slope") I suggested to make it configurable. |
Comparison job queued. |
Thanks for the reminder, Igor. |
Comparison is ready Alternative comparison was/were failed for workflow(s): |
Comparison job queued. |
@mariadalfonso |
For low energy recHit, it often happen as follow other repetitive empty events are also with This happen for the high energy (E>5 and chi2>100) |
@mariadalfonso @igv4321 @kpedro88 @abdoulline It looks like the errors decrease with this PR compared to the current CMSSW version. |
Thanks for the studies |
Slava, I do believe this implementation of ADC quantization error
estimate is the move in the right direction, as it's based on a real ADC
properties of QIE10 chip.
…On Thu, 22 Dec 2016, Slava Krutelyov wrote:
@mariadalfonso @igv4321 @kpedro88 @abdoulline
I will need some confirmation/clarification regarding behavior of M2 at high energy to
sign off on this PR.
It looks like the errors decrease with this PR compared to the current CMSSW version.
If it is appropriate and expected from first principles, we can move on, hoping that the
upcoming incorporation of shape variations or other updates will take care of the excess
chi2.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or mute the
thread.[AEx02imLl09-7D9dcnYRU_sI-x4Yccvnks5rKpLagaJpZM4LM-fy.gif]
|
On 12/22/16 1:17 PM, Salavat Abdullin wrote:
Slava, I do believe this implementation of ADC quantization error
estimate is the move in the right direction, as it's based on a real ADC
properties of QIE10 chip.
QIE11 (just a typo) ?
|
Right, mea cupla. Actually the only essential difference of QIE11 from
QIE10 is a programmable shunt (of the input current) which is taken into
account in the QIE11 calibration, other properties are very much the same.
…On Thu, 22 Dec 2016, Slava Krutelyov wrote:
On 12/22/16 1:17 PM, Salavat Abdullin wrote:
> Slava, I do believe this implementation of ADC quantization error
> estimate is the move in the right direction, as it's based on a real ADC
> properties of QIE10 chip.
QIE11 (just a typo) ?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or mute the
thread.[AEx02qv-WPD2Jq49GOwDGDvf7ntNsw7_ks5rKvACgaJpZM4LM-fy.gif]
|
I plotted here the relative error vs the input charge it's close enought to what I see in the slides of the electronic's guys see page22 |
+1
|
This pull request is fully signed and it will be integrated in one of the next CMSSW_9_0_X IBs (tests are also fine). This pull request requires discussion in the ORP meeting before it's merged. @slava77, @davidlange6, @smuzaffar |
+1 |
Hi.
Added proper quantization noise introduced in #16951
Validation test in
http://dalfonso.web.cern.ch/dalfonso/dalfonso_M2adcFIX.pdf
Fix only the HE for the 2017 scenario.
It affect both data and MC.