Problem decaying NLO single T' with madspin - "No amplitudes generated from process"

MadSpin is limited to sequence of two body decay. We do have "special" mode where those limitations are shifted (but with lower level of precision) but here it seems that you did not activate any of those mode and therefore your madspin card is not valid.

Cheers,

Olivier

Hi Olivier

Thanks for your prompt reply ; can you give such a decay example/reference how could I make it to work?

edit - even by doing

decay tp > w+ b

still gives the same error

Hi,

So we can not do it with offshell effect.
so you have to either change the spinmode line by hand in the madspin card.
or use the following script
import model VLQ_v4_4FNS_UFO-3rdL_noCKM
set complex_mass_scheme True
define ferm = ve vm vt ve~ vm~ vt~ u c d s b u~ c~ d~ s~ b~ e- mu- ta\
- e+ mu+ ta+
define tt = t t~
define bb = b b~
define ww = w+ w-
define tpp = tp tp~
define bpp = bp bp~
define xx = x x~
define yy = y y~
generate p p > tp b~ j / bpp xx yy [QCD] QED=1 QCD=1 VLQ=1
add process p p > tp~ b j / bpp xx yy [QCD] QED=1 QCD=1 VLQ=1
output
launch
madspin=none
# specify the decay for the final state particles
decay tp > ferm ferm bb / Z h
decay tp~ > ferm ferm bb / Z h

so the two mode for the "madspin=XXXX" mode or for the equivalent spinmode in MadSpin
are described here: https://cp3.irmp.ucl.ac.be/projects/madgraph/wiki/MadSpin

set spinmode [none|onshell|madspin][default:madspin]
    madspin default mode with full-spin correlation and off-shell effect
    none(new in 2.3) No spin-correlation between production and decay. No off-shell effects. However this mode allows to handle three-body decay and loop induced processes. Due to the lack of spin-correlation it is mainly intended for scalar particle (and in particular the Higgs).
    onshell (new in 2.5.3) Keep the full spin-correlation but keep the particle exactly on-shell. Supports three-body decay but not loop-induced processes (in either production or decay). Intended for particle with small width without any two body decay. This mode requires f2py program to be available on your machine.

The "onshell" method is based on reweighting module which has limitation too. In particular, your process might requires 3.4.1 version (so not the LTS version) because the reweighting can not handle the ambiguity of identical particle in production and decay in previous version. Note that the ambiguity is taken by doing the average over all possible assignment which should be ok only if the width of the decaying particle is very narrow.

Cheers,

Olivier

Hi again and thanks

I ve tried the spinmode none but then

Defined multiparticle bb = b b~
set spinmode none
decay tp > ferm ferm bb / Z h
decay tp~ > ferm ferm bb / Z h
launch

INFO: Will use seed 369503446
Command "decay_events run_01" interrupted with error:
InvalidCmd : The bridge mode of MadSpin does not support event files where events do not *all* share the same set of final state particles to be decayed. One workaround is to force the final cross-section manually.

Oh I did not remember that limitation.

I know that one reason was related to the issue of the cross-section since you have plenty of special case when you are allowing such type of decay and covering all cases is complicated.

I do not know if the onshell method does have the same limitation (I would guess),
but if it is you would need to do two different generation for
 p p > tp b~ j / bpp xx yy [QCD] QED=1 QCD=1 VLQ=1
and for
 p p > tp~ b j / bpp xx yy [QCD] QED=1 QCD=1 VLQ=1

so no add process

Cheers,

Olivier

> On 18 Jan 2023, at 21:50, Alexis Kalogeropoulos <email address hidden> wrote:
>
> Question #704460 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/704460
>
> Status: Answered => Open
>
> Alexis Kalogeropoulos is still having a problem:
> Hi again and thanks
>
> I ve tried the spinmode none but then
>
> Defined multiparticle bb = b b~
> set spinmode none
> decay tp > ferm ferm bb / Z h
> decay tp~ > ferm ferm bb / Z h
> launch
>
> INFO: Will use seed 369503446
> Command "decay_events run_01" interrupted with error:
> InvalidCmd : The bridge mode of MadSpin does not support event files where events do not *all* share the same set of final state particles to be decayed. One workaround is to force the final cross-section manually.
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Hi Oliver

Thanks, I ve tried what you suggested. I have generated only one process, and then using a madspin card like

set max_weight_ps_point 400 # number of PS to estimate the maximum for each event
#

 set max_running_process 10
 set spinmode none

define ferm = ve vm vt ve~ vm~ vt~ u c d s b u~ c~ d~ s~ b~ e- mu- ta- e+ mu+ ta+

decay tp > w+ b / Z h
launch

still that fails, and looking at the output, what seems strange, is that although I have generated in my lhe 5000, when launching madspin, I do see a line

INFO: generate 5000 decay event for particle tp
Now generating 6000 events with random seed None and granularity 1

and then it fails (so why trying to generate 6000 events? where does it get that from?)

WRITE GRIDCARD /afs/cern.ch/work/a/alkaloge/Gridpacks/VLQ/genproductions/bin/MadGraph5_aMCatNLO/TestGrid/MG5_aMC_v2_6_3_2/bW_t_NLOv263_QCD1_VLQ1_tp/madspingrid/decay_6000006_0/madevent
Traceback (most recent call last):
  File "./madevent/bin/gridrun", line 86, in <module>
    cmd_line = cmd_interface.GridPackCmd(me_dir=root_path, nb_event=args[0], seed=args[1], gran=args[2])
  File "/afs/cern.ch/work/a/alkaloge/Gridpacks/VLQ/genproductions/bin/MadGraph5_aMCatNLO/TestGrid/MG5_aMC_v2_6_3_2/bW_t_NLOv263_QCD1_VLQ1_tp/madspingrid/decay_6000006_0/madevent/bin/internal/madevent_interface.py", line 6310, in __init__
    self.write_gridcard(nb_event, seed, gran) # set readonly on True if needed
  File "/afs/cern.ch/work/a/alkaloge/Gridpacks/VLQ/genproductions/bin/MadGraph5_aMCatNLO/TestGrid/MG5_aMC_v2_6_3_2/bW_t_NLOv263_QCD1_VLQ1_tp/madspingrid/decay_6000006_0/madevent/bin/internal/madevent_interface.py", line 6361, in write_gridcard
    gridpackcard['gseed'] = seed
  File "/afs/cern.ch/work/a/alkaloge/Gridpacks/VLQ/genproductions/bin/MadGraph5_aMCatNLO/TestGrid/MG5_aMC_v2_6_3_2/bW_t_NLOv263_QCD1_VLQ1_tp/madspingrid/decay_6000006_0/madevent/bin/internal/banner.py", line 1197, in __setitem__
    value = self.format_variable(value, targettype, name=name)
  File "/afs/cern.ch/work/a/alkaloge/Gridpacks/VLQ/genproductions/bin/MadGraph5_aMCatNLO/TestGrid/MG5_aMC_v2_6_3_2/bW_t_NLOv263_QCD1_VLQ1_tp/madspingrid/decay_6000006_0/madevent/bin/internal/banner.py", line 1347, in format_variable
    raise InvalidCmd, "%s can not be mapped to an integer" % value
internal.InvalidCmd: None can not be mapped to an integer
mv: cannot stat './Events/GridRun_None/unweighted_events.lhe.gz': No such file or directory

Any ideas?

Seems to be a bug related to the seed, I see that you are using a non supported version of MG5aMC.
So my advice is to upgrade to the Long Term Stable version where that particular issue is likely fixed.
The version number for the LTS is 2.9.13 (with bug fix only since 2.9.4 release so a bit less than two years).

For the 5000 going to 6000, it is a margin of security.

Cheers,

Olivier

Thanks Olivier - I was trying 2.6.3 because that was the version the authors used in the paper, so we were using that as a reference.

That said, and given that it is practically impossible to reproduce their results, I then wonder how it was possible for them to

a) produce and decay lhe with both the p p > tp b~ j and p p > tp~ b j processes as we see that madspin does not like that
b) even if producing one leg at the time, as you said, the bug with the seed should also be a problem for them...

anyway, this is not for your to answer, I am just putting here all info for future reference. I am moving to 2.9.13 as you suggested and will update the post later on.

Thanks for your patience and support

Update - Moving to 2.9.13 process/diagrams are built ok, but then trying to generate events:

python2.7 ./bin/aMCatNLO
>launch
...
WARNING: Fail to compile the Subprocesses
Command "launch" interrupted with error:
AttributeError : 'list' object has no attribute 'clear'

looking inside the debug

Traceback (most recent call last):
  File "/afs/cern.ch/work/a/alkaloge/Gridpacks/VLQ/genproductions/bin/MadGraph5_aMCatNLO/TestGrid/MG5_aMC_v2_9_13/bW_t_NLOv2913_QCD1_VLQ1_tp/bin/internal/extended_cmd.py", line 1541, in onecmd
    return self.onecmd_orig(line, **opt)
  File "/afs/cern.ch/work/a/alkaloge/Gridpacks/VLQ/genproductions/bin/MadGraph5_aMCatNLO/TestGrid/MG5_aMC_v2_9_13/bW_t_NLOv2913_QCD1_VLQ1_tp/bin/internal/extended_cmd.py", line 1490, in onecmd_orig
    return func(arg, **opt)
  File "/afs/cern.ch/work/a/alkaloge/Gridpacks/VLQ/genproductions/bin/MadGraph5_aMCatNLO/TestGrid/MG5_aMC_v2_9_13/bW_t_NLOv2913_QCD1_VLQ1_tp/bin/internal/amcatnlo_run_interface.py", line 1698, in do_launch
    self.compile(mode, options)
  File "/afs/cern.ch/work/a/alkaloge/Gridpacks/VLQ/genproductions/bin/MadGraph5_aMCatNLO/TestGrid/MG5_aMC_v2_9_13/bW_t_NLOv2913_QCD1_VLQ1_tp/bin/internal/amcatnlo_run_interface.py", line 5252, in compile
    compile_cluster.remove()
  File "/afs/cern.ch/work/a/alkaloge/Gridpacks/VLQ/genproductions/bin/MadGraph5_aMCatNLO/TestGrid/MG5_aMC_v2_9_13/bW_t_NLOv2913_QCD1_VLQ1_tp/bin/internal/cluster.py", line 755, in remove
    self.demons.clear()
AttributeError: 'list' object has no attribute 'clear'

any hints?

Hi,

Yes you are not the first one to report that one.
The issue is that we do not validate the code for python2.7 anymore.

Lookslike in the last bug fix, we use a syntax which is only valid from python3.
The patch is very easy replace the line

755 from bin/internal/cluster.py
from
   self.demons.clear()
to
   self.demons[:] = []

This will be part of 2.9.14.
As far as I know this is the only line which is problematic with python2.7 (if we do not count all the feature which are python3 only obviously). But since we do not test for it anymore this is obviously not certain.

Cheers,

Olivier

> On 20 Jan 2023, at 13:15, Alexis Kalogeropoulos <email address hidden> wrote:
>
> Question #704460 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/704460
>
> Status: Answered => Open
>
> Alexis Kalogeropoulos is still having a problem:
> Update - Moving to 2.9.13 process/diagrams are built ok, but then trying
> to generate events:
>
> python2.7 ./bin/aMCatNLO
>> launch
> ...
> WARNING: Fail to compile the Subprocesses
> Command "launch" interrupted with error:
> AttributeError : 'list' object has no attribute 'clear'
>
> looking inside the debug
>
> Traceback (most recent call last):
> File "/afs/cern.ch/work/a/alkaloge/Gridpacks/VLQ/genproductions/bin/MadGraph5_aMCatNLO/TestGrid/MG5_aMC_v2_9_13/bW_t_NLOv2913_QCD1_VLQ1_tp/bin/internal/extended_cmd.py", line 1541, in onecmd
> return self.onecmd_orig(line, **opt)
> File "/afs/cern.ch/work/a/alkaloge/Gridpacks/VLQ/genproductions/bin/MadGraph5_aMCatNLO/TestGrid/MG5_aMC_v2_9_13/bW_t_NLOv2913_QCD1_VLQ1_tp/bin/internal/extended_cmd.py", line 1490, in onecmd_orig
> return func(arg, **opt)
> File "/afs/cern.ch/work/a/alkaloge/Gridpacks/VLQ/genproductions/bin/MadGraph5_aMCatNLO/TestGrid/MG5_aMC_v2_9_13/bW_t_NLOv2913_QCD1_VLQ1_tp/bin/internal/amcatnlo_run_interface.py", line 1698, in do_launch
> self.compile(mode, options)
> File "/afs/cern.ch/work/a/alkaloge/Gridpacks/VLQ/genproductions/bin/MadGraph5_aMCatNLO/TestGrid/MG5_aMC_v2_9_13/bW_t_NLOv2913_QCD1_VLQ1_tp/bin/internal/amcatnlo_run_interface.py", line 5252, in compile
> compile_cluster.remove()
> File "/afs/cern.ch/work/a/alkaloge/Gridpacks/VLQ/genproductions/bin/MadGraph5_aMCatNLO/TestGrid/MG5_aMC_v2_9_13/bW_t_NLOv2913_QCD1_VLQ1_tp/bin/internal/cluster.py", line 755, in remove
> self.demons.clear()
> AttributeError: 'list' object has no attribute 'clear'
>
> any hints?
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.