Pieter de Buck - Week 2 - Duke University

Hi everyone, it's Pieter and this is my second week at Duke.

I met my PI/professor and Dr. Nahrgang last week and we had a very long discussion about the efforts of the lab, the theory behind their experiments, and my role in all of this. So we started talking about the background information that I needed for my project and I felt that I was very prepared for it because of our efforts in class. However, to even remotely understand some of the more advanced topics in this field of science I would have to know at least college level math. And when I asked the postdoc Chris about some textbooks or other sources to strengthen my knowledge of the subject, he told me that we simply don't know enough about the topic of Quark-Gluon Plasma to write textbooks about it. I think that is actually pretty exciting, that this research is truly on the cutting edge of particle physics, we are investigating a phenomenon that has never been observed before, we just have enough trust in the already defined and documented parts of physics to apply them to this topic.

Last week I stepped up my project, and also moved from my laptop to the immensely powerful computer that Duke provided me. Now I am working with real data from the labs computer model, and this real data consists of 253 million lines of text, with each line containing information about a single particle, in this case a top quark (one of the six kinds of quarks). My program has been upgraded in the sense that it now has to take this data from many different files, and on its own find out which files contain relevant information. This has been challenging for me, but not out of my reach and it was really rewarding when it finally came together. I think this has been my last week working on this specific project, and I am now ready to move to the real deal. I will be creating a user-friendly, customisable and flexible program which my lab will actually use to evaluate the output of the computer model. The user needs to be able to specify certain conditions, and what kinds of graphs they want to see. The program needs to be able to deal with exceptions and irregularities.

A problem I am facing now is the fact that even this computer slows down to a crawl when the program is in operation, sometimes it even stops the program outright. This is of course to be expected when 250 million lines of text have to be read and interpreted, which I estimate to be at least 30 GB of data. I am looking for solutions to help this problem.

The world cup has been so amazing this year, I have not seen this many good games in a long time, and each appearance of my country, the Netherlands, has me nervous the whole day. They are now in the semis and my prediction of a Germany-Netherlands final seems very likely at this point. Good luck to the other countries too.

Pieter