Rivet tutorials

How to contribute a routine

Contributing a Rivet routine of your (or any “missing”) analysis is an incredibly valuable contribution to the HEP community. And boosts citation rates and recognition, too! Thanks for even thinking about it… and if you need any help with the implementation, please see the various other tutorial pages, and contact the Rivet support team if you need more.

This page covers the process and requirements for getting your code added to the official analysis collection, in which case we will manage all code and (meta)data updates in future.

Making a merge request or issue

New analyses should all be contributed via a Gitlab merge request (MR), targetted at the (usually latest) release-* branch. Validation data should be included in the MR submission – see below for more detail on this.

This uniform process means that the MRs are a valuable future reference for how an analysis was checked and how it was behaving at the time of submission, in case issues arise later.

We are in the process of streamlining the routine contribution procedure using an MR template as well as a dedicated CI pipeline. In the meantime, feel free to make a merge request from your own repo fork (if you know where the relevant files should go) or open an issue on our issue tracker so we can discuss and guide you through it.

Analysis-submission checklist

Here is some advice and requirements, to help the MR process go smoothly:

  1. Make sure the reference data file is fully in sync with what’s on HEPData.

  2. Please provide all of the routine’s *.cc, *.plot and *.info files, and make sure the resulting plots are well formatted. These will go into the relevant analyses/pluginEXPERIMENTdirectory.

  3. Don’t override fonts, colours, etc. in your submitted plot file: keep that for your own papers, we need all the “managed/official” analyses to follow the default styling.

  4. Please provide the yoda file(s) used to produce the validation plots. Please also add key validation plot PNGs directly in the MR text, for review convenience. You can also attach the tarred/zipped rivet-plots directory if you wish, but we can reproduce this from the yoda file.

  5. Additionally, please identify a suitable sample in our suite of HepMC validation files, and specify its name in the info file e.g. like

    ReleaseTests:
 - $A LHC-13-Top-L

    and provide the output yoda file. Note that these files only have 1000 events and are mainly used for numerical regression tests: they are not expected to be good representations of the physics! Your validation should generally use much bigger samples. If there is not yet a suitable regression sample, please make a 1000-event version of your validation process and we will add it to the collection.

Thank you!!