2->4 process with dim=5 operators

Please, send me your model and batch file.
May e-mail address: <email address hidden>
Best
    Alexander Pukhov

On 05/21/2016 07:02 AM, dimitris wrote:
> Question #293870 on CalcHEP changed:
> https://answers.launchpad.net/calchep/+question/293870
>
> Description changed to:
> Hello!
>
> I'm trying to calculate a cross section of a 2->4 process using calchep_batch. Calchep runs for about an hour nicely and then
> ram usage blows up to 100% (8Gb) and takes another 5Gb of swap. I left it to run for ~9 hours but nothing happend. I'm
> suspecting some kind of memory leak.
>
> In my model, in the Feynman gauge, there are only two diagrams for this
> process. There are dipole operators (two fermions with the Z boson) and
> a 4point interaction with G^0 (two fermions with the neutral goldstone
> boson), which I believe make tis calculation even more difficult than a
> similar calculation in a UV-complete model .
>
> Also, there are other 2->4 processes which are similar to this one and
> calchep calculates them relatively quickly (~3hrs).
>
> Is there another way to do the symbolic calculation?
> Any ideas of whats going on?
>
> Thanks in advance.
>

Thanks for the answer,
I've sent you the files!

Hello again!

I think I've found a workaround.

I've written a python script to simplify the vertices (under some assumptions that I have in my model) and now the process u,u->u,u,~n1,~n1 runs in under 13 mins and calchep uses a lot less memory.

There are some other processes like u,d->u,d,~n1,~n1, that need ~11 hrs.
But before I simplify the vertices, calchep would crush in the symbolic calculation.

So, the situation has improved a lot.

Thanks!

I'll do that. I already have the lanhep files, so I only need to
use my assumptions.
If I encounter a problem, I'll return here.

Thanks!

Thanks Alexander Pukhov, that solved my question.