ghost
commented
7 years ago I have been able to separate momentum production and event generation. I have pretty much done my outline but will still be adding to it over the next week until our group meeting. I've sent you a copy regardless. Looking into how Pythia generates it's momentum now. Please let me know what you would like me to do this week.
Thomas has got a really nice way of making four momenta physical - just want to do a comparison between thomas's px,py,pz,E outputs and those from a real LHC 13 TeV data event, to see if any tweaking might be helpful.
Plan: Thomas to output px,py,pz,E into csv just for one output Lily to do the same for real "pseudo-"data px,py,pz,E Other group member using pythia px,py,pz,E
We change our particle production such that some of our 50000 events (1% ? 5% ?) are "special" and we add this information to the output file.
We then apply the same ML technique to the three inputs: we tell it there are two kinds of events and we ask it to decide which events go on which list. "Which events are special?".
We will have three results, one for each input format. Which input format results in the best discrimination?
Thomas to make a one page outline of report Thoughts: The generation of particle 4mo currently being done in Thomas's code could be done with Pythia to make a more realistic output. How does pythia do it differently to Thomas? http://home.thep.lu.se/~torbjorn/Pythia.html