ParticleNet

[IvanMM27]

ParticleNeT model for Jet Tagging with particle cloud like data used in KM3NeT implemented succesfully in GraphNeT. It uses all utilities from GraphNeT, mainly DynEdgeConv blocks to build its architecture. References:

• https://arxiv.org/pdf/1902.08570
• https://ecap.nat.fau.de/wp-content/uploads/2023/06/2023-06_LukasHennig_MSc_ORCA_TauID_GNNs.pdf

[RasmusOrsoe]

@IvanMM27 thank you for adding the ParticleNeT model!

In order to merge this pull request, we need the automatic checks to pass. As you might notice, quite a few of the checks are currently failing. In addition, this PR appears to contain not only changes related to the ParticleNeT model itself, but also code that looks like KM3NeT integration code.

I would suggest that this PR is split up into at least two smaller PR's to reduce the complexity. I.e. One pull request that contains the ParticleNeT model, and then a second pull request that contains the KM3NeT integration.

@IvanMM27 would that be alright with you?

Tagging @giogiopg for completeness

[IvanMM27]

@IvanMM27 thank you for adding the ParticleNeT model!

In order to merge this pull request, we need the automatic checks to pass. As you might notice, quite a few of the checks are currently failing. In addition, this PR appears to contain not only changes related to the ParticleNeT model itself, but also code that looks like KM3NeT integration code.

I would suggest that this PR is split up into at least two smaller PR's to reduce the complexity. I.e. One pull request that contains the ParticleNeT model, and then a second pull request that contains the KM3NeT integration.

@IvanMM27 would that be alright with you?

Tagging @giogiopg for completeness

@RasmusOrsoe I think it is better to re-do this pull request again as we want to push only the ParticleNeT model and not any of KM3NeT utilities. Would it be okay if I delete this one and we do another one? Apologies for the misunderstanding.

@giogiopg

[RasmusOrsoe]

@IvanMM27 thank you for adding the ParticleNeT model!
In order to merge this pull request, we need the automatic checks to pass. As you might notice, quite a few of the checks are currently failing. In addition, this PR appears to contain not only changes related to the ParticleNeT model itself, but also code that looks like KM3NeT integration code.
I would suggest that this PR is split up into at least two smaller PR's to reduce the complexity. I.e. One pull request that contains the ParticleNeT model, and then a second pull request that contains the KM3NeT integration.
@IvanMM27 would that be alright with you?
Tagging @giogiopg for completeness

@RasmusOrsoe I think it is better to re-do this pull request again as we want to push only the ParticleNeT model and not any of KM3NeT utilities. Would it be okay if I delete this one and we do another one? Apologies for the misunderstanding.

@giogiopg

Yes! That is totally fine!