(Date: 19.1.2015 – 23.1.2015)
I adapt my hit counting code from a few weeks ago a bit and manage to compile it on the supercomputer. I submit my first 100 jobs, totaling to 200 000 events. After a day most of them are ready (I only have to resubmit 2 %) and I can start looking at my hit count studies with a bit more statistics.
The benchmarking tool for the tracking algorithm comes of age, slowly. By the end of the week, I’m able to produce some plots concerning the algorithm’s reconstruction quality. Without further optimization, the algorithm is able to reconstruct 70 % of the tracks it possibly can. Not too good, but quite decent. It surely needs more tuning of the parameters and investigating, what goes wrong with the leftover 30 %. My adapted version of the benchmarking tool is over at the experiment’s SVN repo.
In two weeks, I’m going to present my work in front of the institute in Bochum. It’s kind of a more-detailed, albeit preliminary, less serious PhD defense talk. About 45 Minutes. I have to prepare the talk, produce plots; there are test talks to be held and such. So: Some work. I have pretty much the plots I want to show ready, but not as single files but still as ROOT canvases. The weekend is going to be used to make PDFs, and eventually the talk, from these. I hope the writing does not suffer too much.
Writing: During the week, I brought the QCD chapter to a status where I would call it ready-ish. Now I have to take care about bound quark states (mesons, in particular) and then the theory chapter is more or less done. Maybe that’s something I can do over the weekend, although it does depend on my progress concerning the talk – the preliminary version of this needs to be ready by Monday.
Also, during the week I prepare new renderings of the PANDA detector. I asked for new images from a colleague from Münster. They are now located in a neat little table in the PANDA wiki and also in my Github repo, to have some versioning (and means of hosting the binaries). Before I post the link to the experiment’s forum (and make it more known), I want to create labeled versions of the pictures, though.