Tau ID - No track consideration?

[Daschm]

Dear Developers,

to understand how the efficiencies in the Delphes tau-ID are actually defined, I had a look at the corresponding module code. I realized that for the tagging, the number of associated tracks to the jet candidate is not taken into account, but I remember that this used to be the case in Delphes 2.

I am wondering if there is a reason why this has changed for Delphes 3. Especially, for 1-track taus it makes a large impact if the corresponding track has been reconstructed or not and (at least for ATLAS) one of the mandatory prerequisites for a tau identification is the reconstruction of at least one associated track [dR < 0.2].

Sincerly,
Daniel

[Michele Selvaggi]

Hi Daniel,

In Delphes 2 we were in fact following the PGS approach and the tau-tagging algorithm did not properly reproduce experimental performances. Therefore we chose to adopt a naive procedure based on MC truth, similar to the one we use for b-tagging.

If you need to have a more physics driven tau-tagging module based on the jet constituents, you can easily implement it in Delphes 3.
We can provide help if needed. In any case, you can find the documentation on how to write a new module here:

https://cp3.irmp.ucl.ac.be/projects/delphes/wiki/WorkBook/ModuleSystem

Once you module is written, don't forget to update the ModulesLinkDef.h file
then type in Delphes main directory:

./configure
make

Let us know if you need any help, and if you decide to write it and you are willing to share it with the community, we will be happy to include it in one of the following releases.

Michele