`mkmom infinite factor` error in batch mode

You haven't given us enough information. What model? What process? What cuts? In particular, did you remove any perturbative singularities?

Sorry I thought it was about computer here is the details, it may occur due to the offshell higgs now I'm trying to do the process by excluding higgs and I will write the result

Model: Standard Model
Model changed: False
Gauge: Feynman

Process: p,p->le,le,n,n
Remove: Z,A,G,h
Composite: p=u,U,d,D,s,S,c,C,b,B,G
Composite: le=e,E,m,M
Composite: n=ne,Ne,nm,Nm

pdf1: cteq6l (proton)
pdf2: cteq6l (proton)

p1: 7000
p2: 7000

Parameter: Mh=126

Cut parameter: J(le,le)
Cut invert: False
Cut min: 0.5
Cut max:

Cut parameter: T(le)
Cut invert: False
Cut min: 15
Cut max:

Cut parameter: N(le)
Cut invert: False
Cut min: -2.5
Cut max: 2.5
Cut parameter: J(le,le)
Cut invert: False
Cut min: 0.5
Cut max:

Cut parameter: T(le)
Cut invert: False
Cut min: 15
Cut max:

Cut parameter: N(le)
Cut invert: False
Cut min: -2.5
Cut max: 2.5

Dist parameter: N(e)
Dist min: -5
Dist max: 5
Dist n bins: 300
Dist title: p,p->le,le,n,n
Dist x-title: N(e)

Dist parameter: T(e)
Dist min: 0
Dist max: 600
Dist n bins: 100
Dist title: p,p->le,le,n,n
Dist x-title: T(e) (GeV)

Dist parameter: T(m)
Dist min: 0
Dist max: 600
Dist n bins: 300
Dist title: p,p->le,le,n,n
Dist x-title: T(m) (GeV)

Dist parameter: J(e,m)
Dist min: 0
Dist max: 7
Dist n bins: 300
Dist title: p,p->le,le,n,n
Dist x-title: J(e,m)

Number of events (per run step): 1000
Filename: test
NTuple: False
Cleanup: False

other informations are basicaly default. Thanks...

sorry should be no higgs in the Remove part if its not the problem.

same problem again higgs is not the problem this time it gave NAN at u,C->e,Ne,M,nm

One remark and one further question to clarify.

1. Remark -- you have repeated the same cuts for N,J,T twice -- this is redundant

2. Question what is your operating system? Is it 32 or 64 bit one?

Sasha

And one more question:
what is the CalcHEP version are you running?

Sasha

1. Its due to the copy my mouse is broken the second cut is not written in batch file

2.Ubuntu 14.04, 64 bit Server

Linux throne 3.13.0-24-generic #47-Ubuntu SMP Fri May 2 23:30:00 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux

No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 14.04 LTS
Release: 14.04
Codename: trusty

last version of calchep 3.4.6

In my own computer I didnt face with such error but mine couldn't finished the events (Its ubuntu 12.04 32 bit, calchep 3.4) it couldnt even finished a single diagram for 5 hours

There are two points
1. You did not specify number of Vegas calls, default is 10K which is too low for 2->4 process,
so there were not enough points in the phase space which led to a zero cross section and which gave NaN
I suggest number of calls 100K
2. Another problem is that the accuracy of the integration is also bad because of the presence of two W-boson resonances.
The accuracy of the integration is displayed by the the calchep_batch as well as can be seen in details in the prt_1 file
by clicking it.

The solution is to
a) introduce proper kinematic
b) introduce regularisation

as it is shown in the batch file I attached below.

-------------------------------
Model: Standard Model
Model changed: False
Gauge: Feynman

Process: p,p->le,Le,N,n
Remove: Z,A,G,h
Composite: p=u,U,d,D,s,S,c,C,b,B,G

Composite: le=e,m
Composite: Le=E,M
Composite: n=ne,nm
Composite: N=Ne,Nm

pdf1: cteq6l (proton)
pdf2: cteq6l (proton)

p1: 7000
p2: 7000

Parameter: Mh=126

Cut parameter: J(le,le)
Cut invert: False
Cut min: 0.5
Cut max:

Cut parameter: T(le)
Cut invert: False
Cut min: 15
Cut max:

Cut parameter: N(le)
Cut invert: False
Cut min: -2.5
Cut max: 2.5
Cut parameter: J(le,le)
Cut invert: False
Cut min: 0.5
Cut max:

Kinematics : 12 -> 35, 46
Kinematics : 35 -> 3 , 5
Kinematics : 46 -> 4 , 6

Regularisation momentum: 35
Regularization mass: MW
Regularization width: wW
Regularization power: 2

Regularisation momentum: 46
Regularization mass: MW
Regularization width: wW
Regularization power: 2

Dist parameter: N(e)
Dist min: -5
Dist max: 5
Dist n bins: 300
Dist title: p,p->le,le,n,n
Dist x-title: N(e)

Dist parameter: T(e)
Dist min: 0
Dist max: 600
Dist n bins: 100
Dist title: p,p->le,le,n,n
Dist x-title: T(e) (GeV)

Dist parameter: T(m)
Dist min: 0
Dist max: 600
Dist n bins: 300
Dist title: p,p->le,le,n,n
Dist x-title: T(m) (GeV)

Dist parameter: J(e,m)
Dist min: 0
Dist max: 7
Dist n bins: 300
Dist title: p,p->le,le,n,n
Dist x-title: J(e,m)

Number of events (per run step): 1000
Filename: test
NTuple: False
Cleanup: False
Max number of cpus: 8

nSess_1: 5
nCalls_1: 100000
nSess_2: 5
nCalls_2: 100000

Thanks Alexander Belyaev, that solved my question.

Sorry for the same question. This time process is p,p -> e,E again my computer (calchep 3.4) does not give any NAN but didnt generate events. Server (calchep 3.4.6) directly gave me NAN in crossection calculation i tried to exclude Z or A but same problem repeated. Tried kinematics, regulatization but didn't helped. here is my batch file:

Model: Standard Model
Model changed: False
Gauge: Feynman

Process: p,p->e,E

Composite: p=u,U,d,D,s,S,c,C,b,B,G

pdf1: cteq6l (proton)
pdf2: cteq6l (proton)

p1: 7000
p2: 7000

Parameter: Mh=126

Kinematics : 12 -> 3, 4

Dist parameter: E(e,E)
Dist min: 0
Dist max: 500
Dist n bins: 300
Dist title: p,p->e,E
Dist x-title: E(e,E) (GeV)

Number of events (per run step): 1000
Filename: test
NTuple: False
Cleanup: False

Parallelization method: local
Max number of cpus: 3
sleep time: 3
nice level : 19

nSess_1: 5
nCalls_1: 100000
nSess_2: 5
nCalls_2: 100000

I don't understand why I got this much error in such an easy process. Hope any one can help Thanks.

Do you have a perturbative singularity? Forget the batch until you understand your process. Open the gui and run each process and try different cuts and generate distributions until you understand what is going on.

I tried it also. in sub processes it gave randomly NAN for instance in U,u -> eE first time it calculated (but the plot were meaningless) after I tried other processes (which gave NAN) I tried U,u again and get a NAN. tried to put cuts, breit wigner but couldn't solve it. Tried to exclude A and Z respectively and tried cuts, breit wigner, again NAN. I couldn't see where the problem is.

What cuts would be required to see final state particles in a detector in general? The detectors do not cover the full sphere, so some angle must be cut. The detectors do not see the interaction point, so the particle must have some energy to hit the detectors and they must deposit enough energy to be seen. So, some sort of energy cut is needed. If two particles are exactly collinear, then the detectors (forget magnetic fields for the moment) will not see the separate particles.

Note that I am not directly answering the question of how to remove perturbative singularities. And, furthermore, not all of these cuts are required to remove perturbative singularities for every process. But, I am giving you something to think about. Beyond this, you need to look up some articles on collider phenomenology to learn more about these basic questions. Best of luck.

Dear jackaraz
(sorry I do not have your name since you did not introduce yourself, so I am using your ID),

adding words to what Neil said, I just would like to notice that your question is a bout different problem,
so you have opened the new ticket.

About the problem -- this is not about CalcHEP -- this is about physics.
The pp->e,E process has a singularity in the zero e+e- invariant mass region of the parameter space
and you are integrating this region. Moreover PDF functions also have singularity in this region since they are diverge when x_{1,2}->0. Including this region into calculation does not make much sense since there are physical problems involving this region where divergences mention above takes place.

There are indications of a divergence in this case or similar cases which you can see:
1) very large cross section
2) very unstable result -- see accuracy in the numerical session and in the prt_1 files

So avoid such divergences you need to apply proper kinematical cuts.
In this particular case, for example, you can apply M(e,E) cut, for example M(e,E)>50 GeV etc
which should solve the problem, i.e.

Cut parameter: M(e,E)
Cut invert: False
Cut min: 50
Cut max:

In addition I suggest also to use regularisation indicating that there is a singularity in the zero mass region

Regularization momentum: 34
Regularization mass: 0
Regularization width: 0
Regularization power: 2

which will make your result converge faster

Regards,
     Sasha Belyaev

It looks like I owe you an apology. It looks like there is a bug and we are looking into it. I am sorry.
Best wishes,
Neil

Actually I owe both of you a gratitude, since yesterday I read a tons of paper and learned so much but still have problems in some cuts which I took from experimental data of both pp->m,M and pp->e,E. I will check them again after the bug been fixed. Gratitude again for your patience.

Thanks Dr. Christensen and Dr. Belyaev

Bug is fixed, see 3.4.7 version on the Web,
cross section will be the same but for very low M(e,E) cut it is much more stable and does not require regularisation anymore
to avoid NANs

Sasha