Parton-Level Cross Section Interpolation
Interpolate the relationship between a (parametrised) underlying distribution of particles and the probability of such an event occurring in a collider.
Hosted by
CERN
Posted
Data Set for Parton-Level Cross Section Interpolation
Hey,
Just wondered if anyone could offer any insight on understanding the data set for this challenge. I'm looking at the events.hdf5 file.
As I currently understand, we want to look at the 'events' and 'randomNumbers' elements of this file, and the challenge description says that randomNumbers contains the phase space points, and events the diff. cross section.
For each row I get len(events) = 10 and len(randomNumbers) = 5. I don't currently understand how this fits with the dimensionality n in the problem (either that or I've parsed the data wrong).
Any ideas would be helpful!
Also, for submission, it says that there should be a csv upload with event_id, cross_section, but as far as I can see the data does not give an event ID/or I'm missing something here as well?
Thanks! :)
Order by:
3
Posted by zeniheisser •
Pardon, we wrote the description before fully deciding whether to strip the files of irrelevant data or not.
For randomNumbers, all elements are relevant, but for events only the final element is needed; this final element is the differential cross section. For each process, the task is thus to interpolate all entries in randomNumbers to the final element of events.
Regarding the event ID --- I believe you are correct. For EventID, simply put the position of the event in the array (indexed at 0).
I'll update the overview to include this info!
2
Posted by imo_f •
Thank you, this is really helpful!