MadGraph5_aMC@NLO

Different cross section of w1j_fxfx between MG5 v331 and MG5 v265

Asked by xiao jie

Dear experts,

We generated the W1Jet FXFX process with both MG5 v331 and MG5 v265. ~10% discrepancy was found on the cross-sections,
v265: Total cross section: 3.324e+04 +- 2.0e+02 pb
v331: Total cross section: 3.061e+04 +- 1.7e+02 pb

The proc_card.dat we used is:
-----------------------------------------------------------------------------
import model loop_sm-ckm_no_b_mass
define p = 21 2 4 1 3 -2 -4 -1 -3 5 -5
define j = p
define ell+ = e+ mu+ ta+
define ell- = e- mu- ta-
define vl = ve vm vt
define vl~ = ve~ vm~ vt~
generate p p > ell+ vl j $$ t t~ h [QCD] @0
add process p p > ell- vl~ j $$ t t~ h [QCD] @1
-----------------------------------------------------------------------------

We also set "3 = ickkw" in the run_card.dat.

It will be great if you know what causes this discrepancy? Please let me know.

Thank you,
Jie

The restrict card restrict_ckm_no_b_mass.dat:
######################################################################
## PARAM_CARD AUTOMATICALY GENERATED BY THE UFO #####################
######################################################################

###################################
### INFORMATION FOR LOOP
####################################
Block loop
 1 9.118800e+01 # MU_R

###################################
## INFORMATION FOR SMINPUTS
###################################
Block SMINPUTS
 1 1.325070e+02 # aEWM1
 2 1.166390e-05 # Gf
 3 1.180000e-01 # aS

###################################
## INFORMATION FOR MASS
###################################
Block MASS
 4 0.000000e+00 # MC
 5 0.000000e+00 # MB
 6 1.730000e+02 # MT
 11 0.000000e+00 # Me
 13 0.000000e+00 # MM
 15 1.777000e+00 # MTA
 23 9.118800e+01 # MZ
 25 1.250000e+02 # MH

###################################
## INFORMATION FOR DECAY
###################################
DECAY 6 1.491500E+00
DECAY 15 0.000000e+00
DECAY 23 2.441404e+00
DECAY 24 2.047600e+00
DECAY 25 6.38233934e-03

###################################
## INFORMATION FOR WOLFENSTEIN
###################################
Block Wolfenstein
 1 2.253000e-01 # lamWS
 2 8.080000e-01 # AWS
 3 1.320000e-01 # rhoWS
 4 3.410000e-01 # etaWS

###################################
## INFORMATION FOR YUKAWA
###################################
Block YUKAWA
 4 0.000000e+00 # ymc
 5 0.000000e+00 # ymb
 6 1.730000e+02 # ymt
 11 0.000000e+00 # yme
 13 0.000000e+00 # ymm
 15 1.777000e+00 # ymtau

Question information

Language:
English Edit question
Status:
Answered
For:
MadGraph5_aMC@NLO Edit question
Assignee:
Rikkert Frederix Edit question
Last query:
Last reply:
Revision history for this message
Rikkert Frederix (frederix) said : 		#1

Dear Jie,

There has been an update to the FxFx merging prescription in v3.3.0. This requires the use of an updated pythia jet-matching user hook, see here: http://amcatnlo.web.cern.ch/amcatnlo/FxFx_merging.htm. Did you use these consistently?

Also, there seems to be a typo in your process definition: you are twice including the +1j process.

best,
Rikkert

Revision history for this message
xiao jie (talalal) said : 		#2

Dear Rikkert,

Thanks for pointing out the link. It's very useful if we want to use the latest MG v331.

Firstly, let me clarify several things:

1. The process I am testing, is the w plus only one additional jet. So it's not included twice.

2. I only set "3 = ickkw" in the run_card, but I didn't really run pythia8. It was turned off.

3. According to your suggestion, I installed pythia8, and replaced JetMatching.h from the link. Other settings were the same as previous ("3 = ickkw", turned pythia8 off). The cross-section was: 3.014e+04 +- 1.7e+02 pb. Which is still different from v265.

So, do you mean the cross-sections can be different, but after turning on pythia8. After matching, the cross-sections should be consistent between v331 and v265.

Please let me know!

Thanks,
Jie

Revision history for this message
Rikkert Frederix (frederix) said : 		#3

Dear Jie,

For consistency, you also have to include the 0-jet sample.

The cross section before showering is non-physical when doing FxFx merging. The rejecting rates might be different in v331 and v265. It should be that the cross sections (for the sum of the 0j+1j samples) after shower (i.e., after rejection of events), is similar between v331 and v265. If it is not, then there might indeed be a bug.

best,
Rikkert

Revision history for this message
xiao jie (talalal) said : 		#4

Dear Rikkert,

We did a further study.

Still the same process, w+1j (since in CMS the jet-binned samples are validated with the inclusive sample, I think the process should be ok).

I use Pythia8 v306 compiled with the jet-matching user hook you mentioned. Then the cross-sections are as follows:

For v 331,
[genval-xsec] xsec_before[pb] xsec_match[pb] accepted[%]
[0j] 18210.00 +/- 46.53 4985.72 +/- 29.15 27.38
[1j] 13580.00 +/- 39.99 3698.51 +/- 24.87 27.24
[total] 31790.00 +/- 61.36 8684.19 +/- 38.31 27.32

For v265,
[genval-xsec] xsec_before[pb] xsec_match[pb] accepted[%]
[0j] 19870.00 +/- 53.10 5140.19 +/- 30.55 25.87
[1j] 14770.00 +/- 44.53 3807.60 +/- 25.96 25.78
[total] 34640.00 +/- 69.30 8947.76 +/- 40.09 25.83

The cross-section of v331 (8684.19 +/- 38.31 pb) after matching is still different from it of v265 (8947.76 +/- 40.09 pb). Do you think this is normal?

Best regards,
Jie

Revision history for this message
Rikkert Frederix (frederix) said : 		#5

Hello Jie,

I think this is normal. The differences are at the <3% level, which should be well within the NLO uncertainties.

best,
Rikkert

Can you help with this problem?

Provide an answer of your own, or ask xiao jie for more information if necessary.

To post a message you must log in.