This page details the K factor extraction method for gg->H at 13 TeV for Run II offshell and high mass measurements.

Parameters for generation

These parameters are synchronized in CMS between MCFM and JHUGen MC generators for gluon fusion signal, bakground and signal+bkg. event generation. Please make sure to check the EW scheme of the particular generator so that the parameters below are passed without recalculation.

Parameter Value
H mass to normalize cross sections to N3LO value 125
W mass 80.399
Z mass 91.1876
H width (YR3 SM) 4.07
W width 2.085
Z width 2.4952
GFermi 1.16639e−5
alpha_EW(mZ) 1∕128=0.0078125
sin^2 theta_W 0.23119
v.e.v. 246.21846
Charm mass 1.275
Bottom mass 4.75
Top mass 173.2
Top width 2.0
Tau mass 1.7768
Tau width 0.002267 (eV)
Vud 0.97425
Vus 0.2253
Vub 0.0413
Vcd 0.225
Vcs 0.986
Vcb 0.0411
LO PDF NNPDF30_lo_as_0130
NLO PDF NNPDF30_nlo_as_0118 (central), NNPDF30_nlo_as_0117 (alpha_s(MZ) down), NNPDF30_nlo_as_0119 (alpha_s(MZ) up)
NNLO PDF NNPDF30_nnlo_as_0118 (central), NNPDF30_nnlo_as_0117 (alpha_s(MZ) down), NNPDF30_nnlo_as_0119 (alpha_s(MZ) up)

Additional cuts

No cuts other than those applied in HNNLO (no jet cuts by default).

Files needed for generating cross sections

  • submitMCFM.slurm.sh: SLURM batch submission script to run the actual job (ran by templateMCFM.sh. One would need to change this for LXBATCH. Typical running times are ~2-3 days for full NNLO calculation (need to run twice with different seeds for each mass to get better phase space statistics), ~1 day for NLO (can run just once) and <1 day for LO (can run just once).

  • br.sm1: Partial widths as in YR3 (not used in calculation of K factor)

  • brNarrowWidth.sm2: List of partial widths with total width set to 4.07 MeV (narrow-width cross sections to be used in K factor extraction, overall normalization comes from N3LO signal at 125 GeV)

  • br.sm2: List of partial widthswith total width set to SM value in YR3 up to 1 TeV (should not be used in K factor extraction, used as cross check)

  • higgsp.f: Need to replace src/Need/higgsp.f to read br.sm1/2 correctly

  • sethparams.f: Need to replace src/Need/sethparams.f to read br.sm1/2 correctly

  • mdata.f: Need to replace src/User/mdata.f to set input parameters as described above

Scripts to dump cross sections to a text file

  • getXsec.sh: Script to dump a single central PDF cross section

  • extractXsec.h: Header file that contains all the needed functions to turn cross sections into the K factors

Usage

  • Download HNNLO from Grazzini et al..
  • Replace higgsp.f, sethparams.f, and mdata.f. The mdata.f file provided contains the parameters listed above. If you need other parameters, you will need to change this file.
  • Compile.
  • You will need br.sm1, brNarrowWidth.sm2, and the template configurations provided to be able to run the code. Put these in the directories you run.
  • Run HNNLO using the submission scripts provided. See submitAll.sh for command line arguments you can use. To obtain K factors, you need to run the cross sections using the NarrowWidth specification in the script.
  • The following systematics are used:
    • QCD scale (central m4l/2, [m4l/4, m4l])
    • PDF scale (central m4l/2, [m4l/4, m4l])
    • alpha_s(MZ) variation (central 0.118, [0.1165, 0.1195] (use 0.117, 0.119 PDFs and multiply difference by 1.5).
    • PDF variation systematics from NNPDF 3.0 replicas (use dedicated bash scripts to submit the replicas).
    • The naming convention in the scripts for scale up and down systematics are Up vs Dn for LO (Nominal is the nominal m4l/2 central scale choice) and Nominal vs DnDn for NLO and NNLO (Dn is the nominal m4l/2 central scale choice)
  • Once jobs are run, collect the cross sections via getXsec.sh files. Then run the functions in extractXsec.h. Example usage cases are


// Collect cross sections and record them to intermediate ROOT files
extractXsec(order=0(LO).1(NLO)/2(NNLO) , syst=10-17(see the header file for 0-7))
extractXsec_replica_ratios(order=0(LO).1(NLO)/2(NNLO) , syst=10)

// Collect K factors
extractKfactor(order=0(LO).1(NLO)/2(NNLO) , syst=10-18(see the header file for 0-8))
collectXsec(order=0(LO).1(NLO)/2(NNLO) , syst=10)
collectKfactor(order=0(LO).1(NLO)/2(NNLO) , syst=10)

The final products are TSpline3 objects that can be used to evaluate the K factor. We suggest you use nominal LO cross section and NLO/LO and NNLO/LO K factors, all three of which are smooth, to find the NLO and NNLO cross sections.

Additional corrections

  • If one applies the same signal K factor on background, 10% additional uncertainty is expected from various studies on bkg K factor, which is only available partially at NLO. Interference between signal and bkg. is taken to have a variation squre-root of the signal one with no change in signal-bkg phase.
  • If you need to correct for N3LO cross section in signal, you need to apply x1.098946 to the K factor obtained. This factor obtained taken into account the different total width and BR values reported in YR4.

Current K factors and LO cross section

Topic attachments
I Attachment History Action Size Date Who Comment
Unknown file formatroot Kfactor_Collected_ggHZZ_2l2l_NLO_NNPDF_NarrowWidth_13TeV.root r1 manage 2011.4 K 2017-05-12 - 04:34 UlascanSarica  
Unknown file formatroot Kfactor_Collected_ggHZZ_2l2l_NNLO_NNPDF_NarrowWidth_13TeV.root r1 manage 2003.0 K 2017-05-12 - 04:34 UlascanSarica  
Unknown file formattemplate LO_NNPDF.template r1 manage 0.3 K 2017-05-11 - 20:04 UlascanSarica  
Unknown file formattemplate LO_NNPDF_replica.template r1 manage 0.4 K 2017-05-11 - 20:02 UlascanSarica  
Unknown file formattemplate NLO_NNPDF.template r1 manage 0.3 K 2017-05-11 - 20:04 UlascanSarica  
Unknown file formattemplate NLO_NNPDF_as117.template r1 manage 0.3 K 2017-05-11 - 20:02 UlascanSarica  
Unknown file formattemplate NLO_NNPDF_as119.template r1 manage 0.3 K 2017-05-11 - 20:02 UlascanSarica  
Unknown file formattemplate NLO_NNPDF_replica.template r1 manage 0.4 K 2017-05-11 - 20:02 UlascanSarica  
Unknown file formattemplate NNLO_NNPDF.template r1 manage 0.3 K 2017-05-11 - 20:04 UlascanSarica  
Unknown file formattemplate NNLO_NNPDF_as117.template r1 manage 0.3 K 2017-05-11 - 20:02 UlascanSarica  
Unknown file formattemplate NNLO_NNPDF_as119.template r1 manage 0.3 K 2017-05-11 - 20:02 UlascanSarica  
Unknown file formattemplate NNLO_NNPDF_replica.template r1 manage 0.4 K 2017-05-11 - 20:02 UlascanSarica  
Unknown file formatroot Xsec_Collected_ggHZZ_2l2l_LO_NNPDF_NarrowWidth_13TeV.root r1 manage 62.2 K 2017-05-12 - 04:34 UlascanSarica  
Unknown file formatsm1 br.sm1 r1 manage 20.8 K 2017-05-11 - 20:04 UlascanSarica  
Unknown file formatsm2 br.sm2 r1 manage 20.5 K 2017-05-11 - 20:04 UlascanSarica  
Unknown file formatsm2 brNarrowWidth.sm2 r1 manage 20.8 K 2017-05-11 - 20:02 UlascanSarica  
Header fileh extractXsec.h r1 manage 90.4 K 2017-05-11 - 20:39 UlascanSarica  
Unix shell scriptsh getXsec.sh r1 manage 1.3 K 2017-05-11 - 20:23 UlascanSarica  
Unix shell scriptsh getXsecAll.sh r1 manage 2.0 K 2017-05-11 - 20:23 UlascanSarica  
Unix shell scriptsh getXsecAllReplicas.sh r1 manage 0.3 K 2017-05-11 - 20:23 UlascanSarica  
Unix shell scriptsh getXsec_replicas.sh r1 manage 1.6 K 2017-05-11 - 20:23 UlascanSarica  
Unknown file formatf higgsp.f r1 manage 2.2 K 2017-05-11 - 20:04 UlascanSarica  
Unknown file formatf mdata.f r1 manage 4.5 K 2017-05-11 - 20:04 UlascanSarica  
Unknown file formatf sethparams.f r1 manage 2.3 K 2017-05-11 - 20:04 UlascanSarica  
Unix shell scriptsh submitMCFM.slurm.sh r1 manage 0.9 K 2017-05-11 - 20:02 UlascanSarica  
Unix shell scriptsh templateMCFM.sh r1 manage 2.1 K 2017-05-11 - 20:02 UlascanSarica  
Edit | Attach | Watch | Print version | History: r3 < r2 < r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r3 - 2017-05-15 - UlascanSarica
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    LHCPhysics All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback