# How to select of subset of diagram for LO computation?

- Keywords:
- diagram filtering

- Last updated by:
- Olivier Mattelaer

First a word of warning: Splitting diagrams into different set is something that you should not do lightly. They are two issue here.

- The first one is that you need to be sure that each set is independently gauge invariant.

Breaking gauge invariance when selecting diagram will lead to breaking lorentz invariance and will lead to completely unphysical number.

- The second is obviously that even if they are both gauge invariant, the two set typically have interference (and therefore by splitting them you are going to loose part of the computation (which can increase or decrease the cross-section depending of the relative phase betwenn the amplitude)

This being said.

Here are "simple" syntax for diagram filtering:

You can use

mu+ mu- > mu+ mu- h h $$ z

-> to fully remove all diagram where a Z is in S-channel

mu+ mu- > z > mu+ mu- h h

-> to require to have AT LEAST one Z in s-channel

mu+ mu- > z h h, z > mu+ mu-

-> to require to have a Z onshell (up to X times the width) that decays into two muons

If this is not enough, then you need to create your own python function to decide for each diagram if you want to keep it or not:

https:/

One good practise is to at least check numerically the lorentz/gauge invariance of your diagram selection.

You can use the check command for that:

check mu+ mu- > mu+ mu- h h $$ z