For more background on Trajectum please see its first two articles 2010.15130 and 2010.15134. The Trajectum code itself can be found at https://sites.google.com/view/govertnijs/trajectum 

The structure of the Trajectum collide routine is such that it takes a parameter file and produces several output files, including lists with all generated particles.

Secondly, an analyze routine has to be run that computes relevant user-specified observables, including experimental cuts. This produces summary files, with one line per event containing all necessary information for those observables.

Finally, a collect routine takes the summary files (which usually is much smaller than the full particle lists) and combines events into centrality classes. This is finally put into a Trajectum output file that is most conveniently processed using Mathematica.

At https://www.dropbox.com/sh/6rev1kpynwlq8vv/AACISEV8XFH-B9zRu-QzJ58Oa?dl=0 we provide a some_examples.nb notebook that shows some ways to analyse a set of Trajectum output files (make sure to download the complete directory, 800MB, to have access to all observables shown). It also includes the Python routines to do a full global analysis (basically a copy of jbernhard), but this part is as yet undocumented. Thanks a lot to Jonah Bernhard and Scott Moreland for their great code!

These are the Trajectum output files and Mathematica routines used to  make the figures as shown in 2112.xxxxx. In the future more notebooks will appear, including more detailed commenting. In the meantime we invite  all to submit question on Trajectum input and output to the authors by e-mail.

The Trajectum output and plotting routines can be found here: dropbox.