MadGraph5_aMC@NLO

p p > t t~ [QCD] generation fail at specific mxx_min_pdg range

Asked by Marcos Miralles on 2023-07-28

Dear Experts,

We are running ttbar production using release 3.5.1. We are using the standard loop_sm model and the default param and run cards for this.

import model loop_sm
define p = g u c d s u~ c~ d~ s~ b b~
define j = p
generate p p > t t~ [QCD]
output ttbar-test_nlo
launch

The only exception is that we set

{6:640} = mxx_min_pdg ! inv. mass for any pair of (anti)particles

in the run_card.dat. This makes the run fail completely with the error attached at the bottom. There seems to be some strange problem with the minimum invariant mass in the range (590,690) that triggers this error since all of them fail. Others are ok. Our goal is to work out the differential ttbar cross-ections in m_ttbar, and later parametrise them using a SMEFT model.

We take the approach of using this variable because we want to set a different fixed scale for each bin (the choice of the scale, wether fixed -at any value- or dynamical doesn't affect the outcome, it still fails), that increases with m_ttbar. A possible work around is to generate at some other m_ttbar and then place the cuts at analysis level, but we would also want to understand why this problem arises in the first place.

It might be useful to mention that this only happens at NLO. At LO there are no issues.

Regards,
Marcos

INFO: Starting run
INFO: Compiling the code
INFO: Using built-in libraries for PDFs
INFO: For gauge cancellation, the width of 't' has been set to zero.
INFO: Compiling source...
INFO: ...done, continuing with P* directories
INFO: Compiling directories...
INFO: Compiling on 128 cores
INFO: Compiling P0_gg_ttx...
INFO: Compiling P0_uux_ttx...
INFO: Compiling P0_uxu_ttx...
INFO: P0_uux_ttx done.
INFO: P0_uxu_ttx done.
INFO: P0_gg_ttx done.
INFO: Checking test output:
INFO: P0_gg_ttx
INFO: Result for test_ME:
INFO: Output of the failing test:
Enter 0 to compute MC/MC(limit)
       1 to compute MC/ME(limit)
       2 to compute ME/ME(limit)
Enter xi_i, y_ij to be used in coll/soft tests
  Enter -2 to generate them randomly
Enter number of tests for soft and collinear limits
SPLIT TYPE USED: T F
A PDF is used, so alpha_s(MZ) is going to be modified
Old value of alpha_s from param_card: 0.11900000000000002
  ****************************************

       NNPDFDriver version 1.0.3
   Grid: NNPDF23nlo_as_0119_qed_mem0.grid
  ****************************************
New value of alpha_s from PDF nn23nlo: 0.11899999999999999
Give FKS configuration number ("0" loops over all)

=================================================

NEW FKS CONFIGURATION:
FKS configuration number is 1
FKS partons are: i= 5 j= 1
with PDGs: i= 21 j= 21

Enter graph number (iconfig), '0' loops over all graphs
Using random seed offsets: 1 , 1 , 0
with seed 33
Ranmar initialization seeds 11949 9408
tau_min 1 1 : 0.34600E+03 -- 0.64000E+03
tau_min 2 1 : 0.34600E+03 -- 0.64000E+03
tau_min 3 1 : 0.64000E+03 0.64000E+03 0.64000E+03
tau_min 4 1 : 0.64000E+03 0.64000E+03 0.64000E+03
tau_min 5 1 : 0.34600E+03 -- 0.64000E+03
tau_min 6 1 : 0.34600E+03 -- 0.64000E+03
tau_min 7 1 : 0.34600E+03 -- 0.64000E+03
tau_min 8 1 : 0.34600E+03 -- 0.64000E+03
BORN: keeping split order 1
counterterm S.O 1 QCD
BORN: keeping split order 1
counterterm S.O 2 QED
BORN: not keeping split order 1

REAL 1: keeping split order 1
Soft tests done for (Born) config 1
Failures: 100 0 100
Sum of all orders: Soft test 1 FAILED. Fraction of failures: 1.00
Split-order 2: QCD: 6; QED: 0; Soft test 1 FAILED. Fraction of failures: 1.00

Collinear tests done for (Born) config 1
Failures: 0 0 0
Sum of all orders: Collinear test 1 PASSED. Fraction of failures: 0.00
Split-order 2: QCD: 6; QED: 0; Collinear test 1 PASSED. Fraction of failures: 0.00

=================================================

NEW FKS CONFIGURATION:
FKS configuration number is 2
FKS partons are: i= 5 j= 2
with PDGs: i= 21 j= 21