Hi there! I’m Pablo de Castro, the AMVA4NewPhysics’ Early Stage Researcher, or ESR, at the INFN – Sezione di Padova. In this blog post, the first of many in the following three years,  I would like to tell you a bit about myself and the stuff I’ve done before joining this exciting project. I opted for a non-standard way to present myself, I hope you like it and/or give me some feedback!

From a crib to know-how

photo_pablo_interview
Me, on my previous job at IFCA, the same day of the AMVA4NP’s interview

I was born and raised in Cantabria, which is a region in the north of Spain. I was an avid reader and quite inquiring during my childhood, always wanting to know how things worked. Not surprisingly, that led me to math and science as I was growing.  I had a hard time choosing a higher education path, because everything seemed interesting from Biology to Computer Science, however I finally opted for doing a Bachelor’s in Physics at the University of Cantabria.

Three years later, one of them as an Erasmus student at London, I felt confident that I had the tools and the skills to face complex scientific problems. The only missing piece was a research topic, for testing the mentioned hypothesis and also writing a Bachelor’s thesis. At that time, I got an undergraduate research scholarship to analyze data from the CMS detector at the LHC, so the battlefield was set.

One does not simply …

It was Winter 2013 and the aim of the project was to accurately measure the W+W– cross section, avoiding vetoing jets in the final state if possible. So I had a new challenge, but before any serious approach I would have to learn the tools (and lingo) of  the trade of experimental particle physics. Several months passed, I was now pretty comfortable with ROOT and cluster queues, knew what fiducial meant and was getting used to speak of  systematic uncertainties all the time.

I kept making steady progress with the project and learning more along the way until Summer 2014. I had been accepted as a Summer Student at CERN, I was going to work on a lab using lasers to produce transient currents in silicon detectors in order to study the effect of radiation damage on them. I did work in the lab but also embarked there in my first serious software project, which is now called TRACS and is a simulation of transient currents and charge collection in semiconductor detectors based on Shockley–Ramo’s theorem.

tracs_potentials_screenshot
Screenshot of the TRACS GUI, the  colour maps are the weighting and electric potential of a microstrip detector, which are obtained solving Poisson’s equation with FEM. This can then be used to predict the movement and signal produced by any charge carrier (electrons and holes) distribution in the detector with some differential equation magic.

I had some fun, met new people and led a CERN Webfest team;  but summer was ending and I had to present my Bachelor’s thesis. So I came back, went to a high-energy physics school in the middle of the Pyrenees and put the finishing touches on the thesis project. I cannot disclosure plots or results because their are not approved, but worth saying that the alternative approach without vetoing jets in the final state performed quite well.

Doubts, chores and b-jets

Just graduated in Physics, the usual career doubts arose. Should I continue in this field? I decided to stay in high energy physics, but deadlines were past for most Master’s programs. The University of Cantabria has a Master’s in Physics program with an annoyingly long official name but with courses on particle physics, scientific computing, statistics and machine learning, so I signed in.

While I was completing my Master’s, I worked part time as a research project associate at IFCA, which is the research institution where I carried out my  Bachelor’s project.  Because I wanted to become a CMS member, I joined the b-tagging group and was assigned the task of release validation, which is basically checking and comparing some reference results every time the CMS software is updated. I was about to get involved in the early measurement of the top quark pair production cross section at 13TeV, when a new opportunity emerged.

b_tagging_roc
b-tagging, which doubtlessly deserves a post on its own in this blog, is the algorithmic identification of jets coming from the decay of b hadrons. Several jet properties are particular for those jets, which can even be combined in fancy MVA discriminators denoted with awful acronyms. This ROC curve, obtained from simulation and officially approved by CMS,  shows the powerfulness of the different algorithms to separate real b jets from light jets.

I obtained a scholarship for doing my Master’s thesis project at Brown University in Providence (RI, USA), which has also a group within the CMS Collaboration. I was able to get a research project consistent with my previous experiences, but which would require new data analysis skills. The point of this project was to measure the efficiency of b-tagging algorithms from data as a function of jet properties, through a maximum likelihood fit based on the tagged and untagged jets per event, simulation flavour composition and a crazy combinatorial model.

On coherence and fairness

I had managed to complete the mentioned data analysis and I was writing my Master’s thesis, when I saw the announcement of the ESR at the INFN Padova, which is combined with a PhD at the Department of Physics and Astronomy of the University of Padova. I thought that I was well suited for it, the project seemed challenging and formative, living in Padova was likely to be fun, the salary was engaging and so I applied. You will read the rest here soon.

Now that we are synchronized, I would like to make one thing clear. I have not talked about anyone other than myself in the previous paragraphs, as if those projects were isolatedly carried out  by  me. That is obviously not true, most of the stuff could not have been done without the help of mentors, teachers, supervisors, colleagues, fellow students, wiki contributors, technical bloggers, open source developers, quantum chromodynamics, friends and family.

Beyond Standard Model

In my spare time, I enjoy travelling, reading, watching good TV series and films and lifting weights in the gym several times a week. I also like to tinker with diverse software, hardware, concepts and data; as in NOSQL databases, pieces of wood, convolutional neural networks and IoT. It seems that I enjoy creative endeavours as writing this blog post or explaining things, which is appropriate given that I am expected to write of the order of 3×52/2≈78 blog posts here.

Do not expect however the other posts to be this personal and strange, I can be seriously scientific when I want to. Indeed, while I was thinking about this post I decided that I will create a blog of my own and there it is: pablodecm.com. The day I wrote this, its content was a null set, but I will dump there whatever I am thinking and I want to share, technical explanations and recipes with formulae and code and much more. As an intent to enlarge my online presence (or lack thereof), I can point you also to my Twitter and Github accounts.

Comment if you dare! Share if you are brave enough!

(Written by P. de Castro)