problem with "==" notation for 4-fermion couplings in SMEFT

[Peter Galler]

Dear MadGraph Team,

I encountered an error that I can't resolve
while generating s-channel single top
in the SMEFT using only the dimension-six
squared term (NP==1). Specifically I ran
the following:

inside the interactive MG environment:

import model SMEFTsim_A_general_alphaScheme_UFO_v2
generate u d~ > t b~ / a h z NP==1
output single_top_schannel

in the command line shell (outside of MG):

I set all Wilson coefficients to zero except for
cqq3Abs1133 using the following commands:

sed -i 's/1\.000000e+00/0\.000000e+00/' single_top_schannel/Cards/param_card.dat
sed -i '/cqq3Abs1133/ s/0\.000000e+00/1\.000000e+00/' single_top_schannel/Cards/param_card.dat

Then I ran:

./single_top_schannel/bin/generate_events -f

The following output appears:

Generating 10000 events with run name run_01
survey run_01
INFO: compile directory
compile Source Directory
Using random number seed offset = 21
INFO: Running Survey
Creating Jobs
Working on SubProcesses
INFO: P1_udx_tbx
INFO: Idle: 1, Running: 0, Completed: 0 [ current time: 12h19 ]
STOP 1
rm: cannot remove ‘results.dat’: No such file or directory
ERROR DETECTED
INFO: Idle: 0, Running: 0, Completed: 1 [ 0.12s ]
INFO: Idle: 0, Running: 0, Completed: 1 [ 0.12s ]
INFO: End survey
refine 10000
Creating Jobs
INFO: Refine results to 10000
INFO: Generating 10000.0 unweigthed events.
Error when reading single_top_schannel/SubProcesses/P1_udx_tbx/G1.2/results.dat
Command "generate_events -f" interrupted with error:
ValueError : need more than 5 values to unpack
Please report this bug on https://bugs.launchpad.net/mg5amcnlo
More information is found in single_top_schannel/run_01_tag_1_debug.log'.
Please attach this file to your report.
quit
INFO:

I'm attaching the log file "run_01_tag_1_debug.log" as well as the log file
"single_top_schannel/SubProcesses/P1_udx_tbx/G1.2/log.txt". In the second
log file you can find the error message:
"Problem in the multi-channeling. All amp2 are zero but not the total matrix-element"
It seems as if this message is generated by the "DiscreteSampler module" written
by "Valentin Hirschi" so he might be the appropriate person to forward this
bug report to.

I'm using MadGraph5_aMC@NLO version 2.6.4 on Ubuntu 14.04.

I might add that this error seems to occur for 4-fermion operators only.
The "NP==1" notation also does not work for example for cqq1Abs1331 but it does
work for example for cuWAbs3x3. Generating the whole matrix element (sm + sm*eft + eft^2)
works for both 4-fermion and non-4-fermion operators.

Thanks, Peter

[Olivier Mattelaer (olivier-mattelaer)]

Hi,

Please use the restriction module idea to solve that issue.
https://answers.launchpad.net/mg5amcnlo/+faq/2312

Cheers,

Olivier

[Peter Galler (pgaller)]

Hi Olivier,
thanks for your suggestion.
I hope I understood your post correctly.
I generated a restriction module by modifying an existing one (restrict_massless.dat)
within the SMEFT model.
There all quark masses except the top quark mass are set to zero (not sure if this
is relevant). Additionally, I set all Wilson coefficients to zero except for cqq3Abs1133
which is set to one. Running MadGraph with the SMEFTsim model including the restriction
indeed produces only the four-fermion diagram but when generating the events I get the
same error as before. Maybe I configured the restriction file incorrectly. It's attached to this post. Or maybe I didn't get the point of your suggestion to use the restriction module.
Did you use a restriction file to solve this particular issue? If yes, could you send me that restriction file, please?
Thanks, Peter

[Olivier Mattelaer (olivier-mattelaer)]

Hi Peter,

Your model is not on the database of model:

> MG5_aMC> import model SMEFTsim_A_general_alphaScheme_UFO_v2
> DEBUG: no model with that name found online
> Error detected in "import model SMEFTsim_A_general_alphaScheme_UFO_v2"
> write debug file MG5_debug
> If you need help with this issue please contact us on https://answers.launchpad.net/mg5amcnlo
> UFOImportError : Path SMEFTsim_A_general_alphaScheme_UFO_v2 is not a valid pathname

So I'm not able to test.
I have quite similar model in the database:
SMEFTsim_A_MFV_alphaScheme_UFO_v2_1 SMEFTsim_A_U35_alphaScheme_UFO_v2_1 SMEFTsim_A_U35_MwScheme_UFO_v2_1/
SMEFTsim_A_MFV_MwScheme_UFO_v2_1 SMEFTsim_A_U35_MwScheme_UFO_v2_1 SMEFTsim_tb1/

But not that one.

But your problem is typical of any 4 point interaction that has all three point vanishing.
In that case we face a limitation of our PS integrator who crashes that way.
The only solution is to simplify the model to avoid to have ANY three-point Feynman diagram.

Cheers,

Olivier

> On 12 Dec 2018, at 01:18, Peter Galler <email address hidden> wrote:
>
> Hi Olivier,
> thanks for your suggestion.
> I hope I understood your post correctly.
> I generated a restriction module by modifying an existing one (restrict_massless.dat)
> within the SMEFT model.
> There all quark masses except the top quark mass are set to zero (not sure if this
> is relevant). Additionally, I set all Wilson coefficients to zero except for cqq3Abs1133
> which is set to one. Running MadGraph with the SMEFTsim model including the restriction
> indeed produces only the four-fermion diagram but when generating the events I get the
> same error as before. Maybe I configured the restriction file incorrectly. It's attached to this post. Or maybe I didn't get the point of your suggestion to use the restriction module.
> Did you use a restriction file to solve this particular issue? If yes, could you send me that restriction file, please?
> Thanks, Peter
>
> ** Attachment added: "restriction module"
> https://bugs.launchpad.net/mg5amcnlo/+bug/1807699/+attachment/5221507/+files/restrict_massless_cqq3Abs1133.dat
>
> --
> You received this bug notification because you are subscribed to
> MadGraph5_aMC@NLO.
> https://bugs.launchpad.net/bugs/1807699
>
> Title:
> problem with "==" notation for 4-fermion couplings in SMEFT
>
> Status in MadGraph5_aMC@NLO:
> Won't Fix
>
> Bug description:
> Dear MadGraph Team,
>
> I encountered an error that I can't resolve
> while generating s-channel single top
> in the SMEFT using only the dimension-six
> squared term (NP==1). Specifically I ran
> the following:
>
> inside the interactive MG environment:
>
> import model SMEFTsim_A_general_alphaScheme_UFO_v2
> generate u d~ > t b~ / a h z NP==1
> output single_top_schannel
>
> in the command line shell (outside of MG):
>
> I set all Wilson coefficients to zero except for
> cqq3Abs1133 using the following commands:
>
> sed -i 's/1\.000000e+00/0\.000000e+00/' single_top_schannel/Cards/param_card.dat
> sed -i '/cqq3Abs1133/ s/0\.000000e+00/1\.000000e+00/' single_top_schannel/...

[Peter Galler (pgaller)]

Hi Olivier,

thanks for your quick answer.
As I understand this is an inherent problem of the phase space integrator. I'm afraid that the "solution [...] to simplify the model to avoid to have ANY three-point Feynman diagram" is only viable using restrictions if one also removes all SM interactions from the SMEFT. I tried naively setting as, Gf and aEW to zero but that didn't work because these parameters are used to calculate others which then causes "division by zero". Hence, it seems to me one would have to generate an entirely new UFO file using FeynRules where the Lagrangian contains only the four-fermion interaction of interest. The other option is of course to go via |dim6|^2 = |sm+dim6|^2 - |sm|^2 - 2Re(sm*dim6cc). Now I just have to find out what is more practical for me.

Thanks, Olivier, for your help.

Best wishes,
Peter

[Olivier Mattelaer (olivier-mattelaer)]

Hi,

Let me be more precise, you do not need to have all three point to be zero.
But to not have the (vanishing) three points in the generated Feynman Diagram.

The restriction card should do exactly that job (remove all the vanishing three point interaction)
and leaving only a set of non vanishing three point (that madevent can integrate) or leaving only
four point interaction (That should be ok as well)

If that does not work, I would need your model to understand why.

Cheers,

Olivier

> On 12 Dec 2018, at 10:11, Peter Galler <email address hidden> wrote:
>
> Hi Olivier,
>
> thanks for your quick answer.
> As I understand this is an inherent problem of the phase space integrator. I'm afraid that the "solution [...] to simplify the model to avoid to have ANY three-point Feynman diagram" is only viable using restrictions if one also removes all SM interactions from the SMEFT. I tried naively setting as, Gf and aEW to zero but that didn't work because these parameters are used to calculate others which then causes "division by zero". Hence, it seems to me one would have to generate an entirely new UFO file using FeynRules where the Lagrangian contains only the four-fermion interaction of interest. The other option is of course to go via |dim6|^2 = |sm+dim6|^2 - |sm|^2 - 2Re(sm*dim6cc). Now I just have to find out what is more practical for me.
>
> Thanks, Olivier, for your help.
>
> Best wishes,
> Peter
>
> --
> You received this bug notification because you are subscribed to
> MadGraph5_aMC@NLO.
> https://bugs.launchpad.net/bugs/1807699
>
> Title:
> problem with "==" notation for 4-fermion couplings in SMEFT
>
> Status in MadGraph5_aMC@NLO:
> Won't Fix
>
> Bug description:
> Dear MadGraph Team,
>
> I encountered an error that I can't resolve
> while generating s-channel single top
> in the SMEFT using only the dimension-six
> squared term (NP==1). Specifically I ran
> the following:
>
> inside the interactive MG environment:
>
> import model SMEFTsim_A_general_alphaScheme_UFO_v2
> generate u d~ > t b~ / a h z NP==1
> output single_top_schannel
>
> in the command line shell (outside of MG):
>
> I set all Wilson coefficients to zero except for
> cqq3Abs1133 using the following commands:
>
> sed -i 's/1\.000000e+00/0\.000000e+00/' single_top_schannel/Cards/param_card.dat
> sed -i '/cqq3Abs1133/ s/0\.000000e+00/1\.000000e+00/' single_top_schannel/Cards/param_card.dat
>
> Then I ran:
>
> ./single_top_schannel/bin/generate_events -f
>
> The following output appears:
>
>
> Generating 10000 events with run name run_01
> survey run_01
> INFO: compile directory
> compile Source Directory
> Using random number seed offset = 21
> INFO: Running Survey
> Creating Jobs
> Working on SubProcesses
> INFO: P1_udx_tbx
> INFO: Idle: 1, Running: 0, Completed: 0 [ current time: 12h19 ]
> STOP 1
> rm: cannot remove ‘results.dat’: No such file or directory
> ERROR DETECTED
> INFO: Idle: 0, Running: 0, Completed: 1 [ 0.12s ]
> INFO: Idle: 0, Running: 0, Completed: 1 [ 0.12s ]
> INFO: End survey ...

[Peter Galler (pgaller)]

Hi,

so this time I used one of the models that you have available (by the way, the model I used originally can be found at https://feynrules.irmp.ucl.ac.be/wiki/SMEFT in the table at the bottom of that web page):
SMEFTsim_A_U35_alphaScheme_UFO_v2_1
Again, I modified the restriction file 'restrict_massless.dat' that comes with this model such that only the Wilson coefficient 'cqq3' is set to one. All other Wilson coefficients are set to zero. I attached this restriction file to this post. Then I run

import model SMEFTsim_A_U35_alphaScheme_UFO_v2_1-massless_cqq3
generate u d~ > t b~ / a h z NP==1

Indeed only one diagram with the four fermion interaction is generated. (That also worked for the model+restriction that I used before).

I saved the output in "single_top_schannel_test" and ran

./single_top_schannel_test/bin/generate_events -f

And I get the same error as before.

Cheers,
Peter

[Olivier Mattelaer (olivier-mattelaer)]

Hi,

I actually do not reproduce "the same error as before" but another one:

           7 > 6 -5 color 2 > 3 -3
 failed for ipartupdate. Please retry without MLM/default dynamical scale
STOP 3

Which is due to the fake treatment of the 4 point interaction that is not compatible with
the mlm merging/default dynamical scale choice.

I then change the dynamical scale choidce to "3" (HT/2)
and it went trough without any issue:

  === Results Summary for run: run_02 tag: tag_1 ===

     Cross-section : 36.83 +- 0.05626 pb
     Nb of events : 10000

INFO: Running Systematics computation
INFO: Idle: 1, Running: 3, Completed: 0 [ current time: 12h06 ]
INFO: Idle: 0, Running: 3, Completed: 1 [ 23.9s ]
INFO: # events generated with PDF: NNPDF23_lo_as_0130_qed (247000)
INFO: #Will Compute 145 weights per event.
INFO: #***************************************************************************
#
# original cross-section: 36.8179557177
# scale variation: +4.82% -4.47%
# central scheme variation: + 0% -6.72%
# PDF variation: +6.04% -6.04%
#
# dynamical scheme # 1 : 35.2076 +4.6% -4.29% # \sum ET
# dynamical scheme # 2 : 35.1704 +4.68% -4.36% # \sum\sqrt{m^2+pt^2}
# dynamical scheme # 3 : 36.818 +4.82% -4.47% # 0.5 \sum\sqrt{m^2+pt^2}
# dynamical scheme # 4 : 34.3422 +4.57% -4.26% # \sqrt{\hat s}
#***************************************************************************

On 12 Dec 2018, at 11:36, Peter Galler <<email address hidden><mailto:<email address hidden>>> wrote:

generate u d~ > t b~ / a h z NP==1

[Peter Galler (pgaller)]

Hi,

I really got the same error as before:

Generating 10000 events with run name run_03
survey run_03
INFO: compile directory
compile Source Directory
Using random number seed offset = 63
INFO: Running Survey
Creating Jobs
Working on SubProcesses
INFO: P1_qq_tq
INFO: Idle: 1, Running: 0, Completed: 0 [ current time: 11h42 ]
STOP 3
rm: cannot remove ‘results.dat’: No such file or directory
ERROR DETECTED
INFO: Idle: 0, Running: 0, Completed: 1 [ 0.17s ]
INFO: Idle: 0, Running: 0, Completed: 1 [ 0.17s ]
INFO: End survey
refine 10000
Creating Jobs
INFO: Refine results to 10000
INFO: Generating 10000.0 unweigthed events.
Error when reading single_top_schannel_test/SubProcesses/P1_qq_tq/G1/results.dat
ln: failed to create symbolic link ‘ME5_debug’: File exists
Command "generate_events " interrupted with error:
ValueError : need more than 5 values to unpack
Please report this bug on https://bugs.launchpad.net/mg5amcnlo
More information is found in 'single_top_schannel_test/run_03_tag_1_debug.log'.
Please attach this file to your report.
quit
INFO:

But when I use "3 = dynamical_scale_choice" in the run_card.dat it does work!! It also works with a fixed scale and it also works for the original model that I used "SMEFTsim_A_general_alphaScheme_UFO_v2" (when the restriction is included). I don't know why I didn't get the same error as you did. Anyway, it works now!

Thank you very much.

Cheers,
Peter