Generating off-shell Higgs production in ggF and VBF

The syntax that you are looking for is likely:

generate p p > h > w+ w- j j $h $$ t t~ , w+ > ell+ vl, w- > ell- vl

Cheers,

Olivier

> On 9 Aug 2023, at 23:30, Taylor <email address hidden> wrote:
>
> New question #707594 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/707594
>
> Dear Experts,
>
> I am trying to produce a sample of events for an off-shell H (via VBF and ggF), along with its decay products of two W bosons. I would like to include the off-shell Higgs contributions. For this, I have attempted to change the BW cutoff from the standard value of 15 to 40 or so, but this only seems to widen the off-shell window, when I need the H pole masses to range from 125 all the way to 1000 GeV.
>
> I have also tried producing a gridpack for ggF and VBF that uses the notation `$h` to include off-shell H contributions, but this gave me an error:
>
> ```
> "generate p p > $h" with error:
> InvalidCmd : Empty or wrong format process, please try again.
> ```
> I am using MG5 2.9.13.
>
> Here is a more thorough example script for VBF production (using loop_sm-ckm_no_b_mass model and attempting to veto the tt~ background:
>
> ```
> import model loop_sm-ckm_no_b_mass
>
> #import model sm-ckm_no_b_mass
> #import model heft-no_b_mass
> #heft gave -99 as mww :/
> #switch to diagonal ckm matrix if needed for speed
> #import model sm-no_b_mass
>
> #!---define particles---#
>
> define ell+ = e+ mu+
> define ell- = e- mu-
>
> #!---define process---#
>
> generate p p > $h w+ w- j j, w+ > ell+ vl, w- > ell- vl $$ t t~
> output VBF_simple -nojpeg
>
> ```
>
> Is there a way to solve this problem?
>
> Looking forward to your feedback.
> Thanks so much!
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Hi Olivier,

Thank you very much for your response! Quick follow up question--

We are hoping to show a continuous distribution of the pole masses over which two W bosons can be reconstructed for an off-shell H, but this output only gives us an invariant mass histogram whose only spike is the mass of the SM H. In the customizecards.dat, we are attempting something like:

```
set param_card mass 25 125
launch
set param_card mass 25 130
launch
. . .
```

or even:

```
set param_card mass 25 scan: [ . . .]
```

But this doesn't provide the desired pole mass spectrum. What is the correct syntax for how to implement that here?

Thanks again, Olivier!

Hi,

Are you sure that you used the syntax suggested?
I have run the above command and looked at the plot of M(WW)
and this is a plot which has a peak at 180 and not that much before and slightly decreasing after.

So to me this looks quite reasonable.

Cheers,

Olivier

> On 10 Aug 2023, at 23:05, Taylor <email address hidden> wrote:
>
> Question #707594 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/707594
>
> Status: Answered => Open
>
> Taylor is still having a problem:
> Hi Olivier,
>
> Thank you very much for your response! Quick follow up question--
>
> We are hoping to show a continuous distribution of the pole masses over
> which two W bosons can be reconstructed for an off-shell H, but this
> output only gives us an invariant mass histogram whose only spike is the
> mass of the SM H. In the customizecards.dat, we are attempting something
> like:
>
> ```
> set param_card mass 25 125
> launch
> set param_card mass 25 130
> launch
> . . .
> ```
>
> or even:
>
> ```
> set param_card mass 25 scan: [ . . .]
> ```
>
> But this doesn't provide the desired pole mass spectrum. What is the
> correct syntax for how to implement that here?
>
> Thanks again, Olivier!
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Hi Olivier,

Thank you for your help! I just ran:

import model loop_sm-ckm_no_b_mass
generate p p > h > w+ w- j j $h $$ t t~ , w+ > l+ vl, w- > l- vl~
output VBF_simple -nojpeg

once more to check. If I plot the total distribution for M(WW), I am still getting the majority being -99, so most H are on-shell, and only around ~250 entries that are 125 GeV exactly. Unfortunately ours and yours do not match, but we wish they did...

Just in case it's useful, the VBF_simple/ code and cmsgrid_final.lhe file are accessible at: ~tcarnaha/public/HWW

Hi,

I guess that we do not check the same file or we do not agree on what you are looking for.

> I am still getting the majority being -99
I do not know what -99 means here. This is not part of the lhef convention.

> so most H are on-shell, and only around ~250 entries that are 125 GeV exactly

Given the small value of the Higgs width, if they are not at 125 exactly, then they are offshell...
(or what is your definition of the exactly?)

> ~tcarnaha/public/HWW

I only have access to CECI cluster. On which one (if any) did you put your data?

Cheers,

Olivier

Hi Olivier,

Ah my apologies. -99 is our convention for "does not appear in data," or "virtual," so when we Ntuplize, one off-shell W will register as -99 to give an on-shell H. For us, this means that the majority of the LHE file gives on-shell H instead of off-shell contributions.

In practice, we select the off-shell region to be $m_{WW} > 2m_{W}$ to allow for uncertainties. This may warrant a larger bwcutoff value. We are attempting to run $m_{WW}$ values all the way up to 1500 GeV in order to retrieve a continuous BW distribution from which we can extract the off-shell H width.

I apologize again; I do not have access to the CECI cluster; I was hoping to send you a picture of the plot, but there is no way to do that via this interface.

So I run the following command on the generated file:

[run_01]$ cat unweighted_events.lhe | grep "25 "
   25 1.250000e+02 # MH
DECAY 25 6.382339e-03 # WH
<weight id="70" MUR="1.0" MUF="1.0" PDF="247025" > PDF=247000 MemberID=25 </weight>

So they are no "H" onshell inside the events.

If I count the number of W+/W- onshell within the sample:

[run_01]$ cat unweighted_events.lhe | grep "24 2" | wc
   20434 265642 2922062

So I do have 434 events with a third "W+-" like in the events below (such events should typically be cut by an invariant mass cut on the two jet)

<event>
11 1 +1.7806774e-03 3.58998700e+02 7.54677100e-03 1.05898100e-01
        1 -1 0 0 501 0 +0.0000000000e+00 +0.0000000000e+00 +5.8964759810e+02 5.8964759810e+02 0.0000000000e+00 0.0000e+00 -1.0000e+00
       -2 -1 0 0 0 501 -0.0000000000e+00 -0.0000000000e+00 -5.4642828074e+01 5.4642828074e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
      -24 2 1 2 0 0 -7.2795435169e+01 +8.7202502223e+01 +1.7593947140e+02 2.2411340838e+02 7.9803950042e+01 0.0000e+00 0.0000e+00
       24 2 1 2 0 0 +3.6631379706e+01 -5.6458200508e+01 +2.7264073612e+02 2.9227448863e+02 8.1006291911e+01 0.0000e+00 0.0000e+00
      -24 2 1 2 0 0 +3.6164055463e+01 -3.0744301715e+01 +8.6424562505e+01 1.2790252917e+02 8.1466563494e+01 0.0000e+00 0.0000e+00
      -11 1 4 4 0 0 -1.3200266164e+01 -4.4712569265e+01 +1.8492140435e+02 1.9070759467e+02 0.0000000000e+00 0.0000e+00 1.0000e+00
       12 1 4 4 0 0 +4.9831645870e+01 -1.1745631243e+01 +8.7719331775e+01 1.0156689397e+02 0.0000000000e+00 0.0000e+00 -1.0000e+00
       13 1 5 5 0 0 -2.3763385374e+01 -1.4447074869e+01 +1.0294756946e+01 2.9654653552e+01 0.0000000000e+00 0.0000e+00 -1.0000e+00
      -14 1 5 5 0 0 +5.9927440837e+01 -1.6297226846e+01 +7.6129805559e+01 9.8247875614e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
        1 1 3 3 502 0 -3.7923748795e+01 +1.8154718641e+00 +9.1504008863e+01 9.9068109393e+01 0.0000000000e+00 0.0000e+00 -1.0000e+00
       -2 1 3 3 0 502 -3.4871686373e+01 +8.5387030359e+01 +8.4435462539e+01 1.2504529899e+02 0.0000000000e+00 0.0000e+00 1.0000e+00

Actually with the syntax that you use, you are forcing the two "W" to be onshell (up to bwcutoff)
and therefore you should always have two W written in your sample.

My advise is to run the same command as above since if you have the same result, then it is either your Ntuple which is wrong or your way to interpret it,

Cheers,

Olivier

> On 14 Aug 2023, at 23:00, Taylor <email address hidden> wrote:
>
> Question #707594 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/707594
>
> Status: Answered => Open
>
> Taylor is still having a problem:
> Hi Olivier,
>
> Ah my apologies. -99 is our convention for "does not appear in data," or
> "virtual," so when we Ntuplize, one off-shell W will register as -99 to
> give an on-shell H. For us, this means that the majority of the LHE file
> gives on-shell H instead of off-shell contributions.
>
> In practice, we select the off-shell region to be $m_{WW} > 2m_{W}$ to
> allow for uncertainties. This may warrant a larger bwcutoff value. We
> are attempting to run $m_{WW}$ values all the way up to 1500 GeV in
> order to retrieve a continuous BW distribution from which we can extract
> the off-shell H width.
>
> I apologize again; I do not have access to the CECI cluster; I was
> hoping to send you a picture of the plot, but there is no way to do that
> via this interface.
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Hey Olivier,

Thank you for this. Our results are as follows:

$ cat cmsgrid_final.lhe | grep "25 "

   25 1.250000e+02 # MH
DECAY 25 6.382339e-03 # WH
       25 2 1 2 0 0 -2.3526410512e+00 -1.0974160759e+01 -2.6516520345e+01 1.2826545780e+02 1.2499173848e+02 0.0000e+00 0.0000e+00
       25 2 1 2 0 0 -6.1949830954e+01 -3.7261660843e+02 -4.5149095776e+02 6.0178579453e+02 1.2498455555e+02 0.0000e+00 0.0000e+00

[. . .]

$ cat cmsgrid_final.lhe | grep "24 2" | wc

      1 5 29

Can you post the banner of that file?

Such that I can check all the input for that file.

Cheers,

Olivier

Absolutely; please let me know if you need more of the banner because it is quite a long header:

import model loop_sm-ckm_no_b_mass
INFO: This model [version 1.2] is provided by V. Hirschi (email: <email address hidden>). Please cite n/a
INFO: load particles
INFO: load vertices
INFO: Restrict model loop_sm-ckm_no_b_mass with file MG5_aMC_v2_9_13/models/loop_sm/restrict_ckm_no_b_mass.dat .
INFO: Run "set stdout_level DEBUG" before import for more information.
INFO: Change particles name to pass to MG5 convention
Defined multiparticle p = g u c d s u~ c~ d~ s~
Defined multiparticle j = g u c d s u~ c~ d~ s~
Defined multiparticle l+ = e+ mu+
Defined multiparticle l- = e- mu-
Defined multiparticle vl = ve vm vt
Defined multiparticle vl~ = ve~ vm~ vt~
Pass the definition of 'j' and 'p' to 5 flavour scheme.
Defined multiparticle all = g gh gh~ d u s c b d~ u~ s~ c~ b~ a ve vm vt e- mu- ve~ vm~ vt~ e+ mu+ t t~ z w+ h w- ta- ta+
define ell+ = e+ mu+
Defined multiparticle ell+ = e+ mu+
define ell- = e- mu-
Defined multiparticle ell- = e- mu-
generate p p > h > w+ w- j j $h $$ t t~ , w+ > ell+ vl, w- > ell- vl~
INFO: Checking for minimal orders which gives processes.
INFO: Please specify coupling orders to bypass this step.
INFO: Trying coupling order WEIGHTED<=6: WEIGTHED IS QCD+2*QED

This is the part of the log file, I was thinking to read the banner of the lhe file, so all the line before the firtst <event> flag within the lhef file.

Cheers,

Olivier

</slha>
<MGGenerationInfo>
# Number of Events : 1000
# Integrated weight (pb) : 0.0041487162944765475
</MGGenerationInfo>
</header>
<init>
2212 2212 6.500000e+03 6.500000e+03 0 0 263000 263000 -3 1
4.148716e-03 5.168894e-05 4.148716e-06 1
<generator name='MadGraph5_aMC@NLO' version='2.9.13'>please cite 1405.0301 </generator>
</init>
<event>
10 1 +4.1487163e-06 1.25846000e+02 7.54677100e-03 1.23682100e-01
        2 -1 0 0 502 0 +0.0000000000e+00 +0.0000000000e+00 +5.4660011797e+02 5.4660011797e+02 0.0000000000e+00 0.0000e+00 -1.0000e+00
        3 -1 0 0 501 0 -0.0000000000e+00 -0.0000000000e+00 -2.9526078554e+02 2.9526078554e+02 0.0000000000e+00 0.0000e+00 -1.0000e+00
       25 2 1 2 0 0 -2.3526410512e+00 -1.0974160759e+01 -2.6516520345e+01 1.2826545780e+02 1.2499173848e+02 0.0000e+00 0.0000e+00
       24 2 3 3 0 0 +2.8261052644e+01 +3.9590897630e+00 -1.8379105031e+01 8.8230150848e+01 8.1439588215e+01 0.0000e+00 0.0000e+00
      -13 1 4 4 0 0 +2.5357746673e+01 +3.8886294572e+01 +4.0819119866e+00 4.6602802784e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
       14 1 4 4 0 0 +2.9033059709e+00 -3.4927204809e+01 -2.2461017018e+01 4.1627348063e+01 0.0000000000e+00 0.0000e+00 -1.0000e+00
       13 1 3 3 0 0 -6.8802531262e+00 +2.3476241792e+00 -7.8175344555e+00 1.0675348581e+01 0.0000000000e+00 0.0000e+00 -1.0000e+00
      -14 1 3 3 0 0 -2.3733440569e+01 -1.7280874702e+01 -3.1988085837e-01 2.9359958370e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
        2 1 1 2 501 0 +1.0353548631e+01 -1.3231065359e+00 +4.9542363725e+02 4.9553357800e+02 0.0000000000e+00 0.0000e+00 -1.0000e+00
        3 1 1 2 502 0 -8.0009075799e+00 +1.2297267295e+01 -2.1756778448e+02 2.1806186771e+02 0.0000000000e+00 0.0000e+00 -1.0000e+00
<scales pt_clust_5="13000.00000" pt_clust_6="13000.00000" pt_clust_7="13000.00000" pt_clust_8="13000.00000" pt_clust_9="10.43775" pt_clust_10="14.63837"></scale\
s>

As I run, I am also getting this error:

Command "generate_events pilotrun" interrupted with error:
UnboundLocalError : local variable 'error' referenced before assignment
Please report this bug on https://bugs.launchpad.net/mg5amcnlo
More information is found in '/eos/home-f/fernanpe/genproductions/bin/MadGraph5_aMCatNLO/VBF_simple/VBF_simple_gridpack/work/processtmp/pilotrun_tag_1_debug.log'.
Please attach this file to your report.

And what is the content of the block <MG5ProcCard> ... </MG5ProcCard>
in the banner. (If you can put all the lines above the snipset that you quote it will help.

Concerning your new crash, can you copy/paste the log file (if you want to attach a file you can convert this question to a bug report)

Cheers,

Olivier

Hey Olivier, thank you for all of your help; it's much appreciated. I see what you're saying I believe, here is the banner between <MG5ProcCard> and </MG5ProcCard>:

<MG5ProcCard>
<![CDATA[
#************************************************************
#* MadGraph5_aMC@NLO *
#* *
#* * * *
#* * * * * *
#* * * * * 5 * * * * *
#* * * * * *
#* * * *
#* *
#* *
#* VERSION 2.9.13 2022-12-15 *
#* *
#* The MadGraph5_aMC@NLO Development Team - Find us at *
#* https://server06.fynu.ucl.ac.be/projects/madgraph *
#* *
#************************************************************
#* *
#* Command File for MadGraph5_aMC@NLO *
#* *
#* run as ./bin/mg5_aMC filename *
#* *
#************************************************************
set group_subprocesses Auto
set ignore_six_quark_processes False
set low_mem_multicore_nlo_generation False
set complex_mass_scheme False
set gauge unitary
set loop_optimized_output True
set loop_color_flows False
set max_npoint_for_channel 0
set default_unset_couplings 99
set max_t_for_channel 99
set zerowidth_tchannel True
import model loop_sm-ckm_no_b_mass
define p = g u c d s u~ c~ d~ s~
define j = g u c d s u~ c~ d~ s~
define l+ = e+ mu+
define l- = e- mu-
define vl = ve vm vt
define vl~ = ve~ vm~ vt~
define p = 21 2 4 1 3 -2 -4 -1 -3 5 -5 # pass to 5 flavors
define j = p
define ell+ = e+ mu+
define ell- = e- mu-
generate p p > h > w+ w- j j $h $$ t t~ , w+ > ell+ vl, w- > ell- vl~
output VBF_simple -nojpeg
]]>
</MG5ProcCard>

This make no sense.

My only advise is to remove that installation of MG5aMC and re-install a fresh one.
If this is a version installed by CMS, please report the problem to them in that case they likely have implemented a patch that is creating the issue.

Cheers,

Olivier

Okay, I can definitely try that; thank you, Olivier.
I just added the bug in the link: https://bugs.launchpad.net/mg5amcnlo/+bug/2031667