Understand the limit drop at 500 GeV

[ram1123]

Hi @jialin-guo1 ,

I think for some reason, still, there is a sharp cut at 500 GeV:

You can check other plots here:

https://rasharma.web.cern.ch/rasharma/HighMassHiggsSearch/datacards_HIG_23_001_WithNewRangeFrom400/?match=mH5

[jialin-guo1]

there I checked signal Model plots. in resolved case at mass points 400 and 500. I didn't see a cut. so I am wondering if the file what you are using is not the latest one

[ram1123]

Thanks. Let me check.

[jialin-guo1]

hzz2l2q_Resolved_13TeV_xs_400.txt hzz2l2q_Resolved_13TeV_xs_500.txt

these are data card I got at 400 and 500 GeV. could you please cross check if the signal rates are the same as yours

[jialin-guo1]

+-----------------------------------+ | Running rc for mass 400 | +-----------------------------------+ <<< Combine >>>

random number generator seed is 123456 Will use a-priori expected background instead of a-posteriori one. method used is AsymptoticLimits SimNLL created with 12 channels, 0 generic constraints, 40 fast gaussian constraints, 0 fast poisson constraints, 0 fast group constraints, -- AsymptoticLimits ( CLs ) -- Expected 2.5%: r < 0.0006 Expected 16.0%: r < 0.0010 Expected 50.0%: r < 0.0017 Expected 84.0%: r < 0.0030 Expected 97.5%: r < 0.0048

this is the limit I got from 400GeV. it hard to read from your plot. could you please crosscheck it

[jialin-guo1]

-- AsymptoticLimits ( CLs ) -- Expected 2.5%: r < 0.0006 Expected 16.0%: r < 0.0011 Expected 50.0%: r < 0.0020 Expected 84.0%: r < 0.0034 Expected 97.5%: r < 0.0055

Done in 20.21 min (cpu), 20.23 min (real) +-----------------------------------+ | Running rc for mass 500 | +-----------------------------------+ <<< Combine >>>

random number generator seed is 123456 Will use a-priori expected background instead of a-posteriori one. method used is AsymptoticLimits SimNLL created with 12 channels, 0 generic constraints, 40 fast gaussian constraints, 0 fast poisson constraints, 0 fast group constraints,

-- AsymptoticLimits ( CLs ) -- Expected 2.5%: r < 0.0034 Expected 16.0%: r < 0.0047 Expected 50.0%: r < 0.0068 Expected 84.0%: r < 0.0099 Expected 97.5%: r < 0.0139

[jialin-guo1]

I further confirm that the rate in data card is exactly the same in SigEff root file. is it possible that the issue of limit dropping at 400-500 is come from backgrounds not from signal

[jialin-guo1]

this is root file of 1D template and vz distribution. the mass is start from 500GeV. that might be the reason you see a sharp cut

[jialin-guo1]

Hi @jialin-guo1 ,

I think for some reason, still, there is a sharp cut at 500 GeV:

You can check other plots here:

https://rasharma.web.cern.ch/rasharma/HighMassHiggsSearch/datacards_HIG_23_001_WithNewRangeFrom400/?match=mH5

Btw how to plot this such plot in CombineTool?

[jialin-guo1]

I got plot with wrong X axis how to setup it?

[jialin-guo1]

here I plot new plots with 1D templates bounds fixed

[ram1123]

@jialin-guo1 fixed by removing the mZZ cut from templates and efficiency computation. Here is the 2018 plot as a check

[ram1123]

The commit from Jialin: SHA - 436c308feeeb95c0bec8488a4d448397418e9820 fixes the issue