higgs_boson_visualized

This Jupyter notebook is now available (and executable), thanks to MyBinder. Or at least it should be, soon. I'm still figuring out the MyBinder thing. Try the link and see what happens. If it's nothing, look at the Jupyter Notebook in the repo.

Brief Description

This is my progress on what's now my favorite visualization. I've been working on doing a visualization of the Higgs-boson-discovery data, but with my own twist. I've done some nuclear/particle physics research (I got my MS Physics while doing such research.) More links/sources/etc. will be found further on in the README. If you want a shorter version of what this is about, with more links and attempts at humor, you can go there, first. You can think of that link a bit like a Wanna Skip Physics? button, a bit like the ones in the actual notebook.

We are going to see if I can do an analysis that leads to something similar to the discovery plot for the Higgs, which you can see right after this text.

I have rather enjoyed this visualization, when I wasn't cussing it out. That's how things seem to go when you're working with data, especially with learning new tools with a new dataset using some skills you haven't much used for ten years (so say I 20230321T175500-0600).

My-History-And-Sources

The first visualization I'll do here is from my academic background - MS Physics at the University of California at Riverside with research at the Reletavistic Heavy Ion Collider - RHIC (archived), specifically at a Pioneering High Energy Nuclear Interaction eXperiment - PHENIX (archived). I was able to be part of the research and discussion - much more discussion than research - concerning the discovery of the Higgs Boson, which happened at CERN's (archived) LHC - in France and Switzerland.

I'm following a guide (archived) from the Helsinki Institute of Physics (archived). The data comes from the Compact Muon Solenoid - CMS (archived) - experiment, which lives way under the ground of France on the Large Hadron Collider - https://www.hasthelargehadroncolliderdestroyedtheworldyet.com/ (archived), oops, the official link is for the LHC (archived), though I believe that the code from the former site is legit. Check it out.

<script type="text/javascript">
if (!(typeof worldHasEnded == "undefined")) {
document.write("YUP.");
} else {
document.write("NOPE.");
}
</script>

Anyway, I'll use the data from CMS to visualize something analagous to discovery of the Higgs through the 4-lepton channel. Look at the pictures - they'll be cool.

Some More Details

Roadblocks - Things that got in my way that may or may not be interesting

The big bug

Not the right number (nJxAgzeSAI0&t=877s)

The number you have called is being checked for trouble.

https://telephoneworld.org/wp-content/uploads/2020/12/4143270494.mp3

While analyzing the 4-lepton data from the Higgs search (2011 and 2012 runs), I planned not only to do the standard spectra on which one can see the Higgs's mass peak, but to use my particle physics background to do a deeper look. The discovery image (archived TODO) - I've made it easier to get straight to the image here (archived TODO)) is of the invariant masses of the reconstructed particles. You can learn more about invariant masses and reconstructed particles, i.e. parent particles, in the Jupyter Notebook (though I did create a Wanna Skip Physics? button) if you absolutely need it). I decided to also look at the invariant mass of one of the daughter particles, with the idea of finding the invariant mass of each of the four daughter leptons.

However, I was not getting reasonably close to the right answer. I wasn't sure what the error bars for invariant mass were, but I'm pretty sure they weren't on the order of 10 GeV/c^2. Don't worry if none of that made sense; it means I was misunderstanding something or just doing something wrong. I chased down the problem. I'd mismatched a letter. I plan on putting a link to a screenshot of the diff showing the problem that was fixed.

CSVs from online and My(new)Binder

I also had to overcome some other obstacles. One was a problem with pandas not reading in the CSVs correctly, or at least not doing so as a guide I was following suggested it should. I also had the normal experience of solving problems as I learned a new technology - MyBinder - which is awesome and which I highly suggest to anyone. The notebook I'm putting up is a compromise between trying to show how everything is done and a pure, talk-turkey presentation of the data. I think it does a good job of a showing the relationship between data, coding, and a person in data science/machine learning.

More questions that with which Physics StackExchange might help with

I must admit, I still have some questions. I might still post on Physics StackExchange . I hope some of the good people there might answer some questions I have about the radicand in the invariant mass formula, i.e. the $E^2 - \lvert \textbf{p} \rvert^2$ part of the whole $m_0 = \sqrt{E^2 - \lvert \textbf{p} \rvert^2}$. Part of my question is to ask if I'm handling things right when $\lvert \textbf{p} \rvert^2 > E^2$. Another thing I'd like to know is if my current computed observations of the invariant masses of daughter leptons are within expected error.

I also look at the invariant mass of the system of particles to double- check its invariant mass (and thus the invariant mass of the parent particle) as a sort of sanity check on the whole invariant mass thing. For the physics people, you might want to know that I've also included the final mass histogram, Monte Carlo events, analysis of Higgs in the mass regime, etc. in the notebook.

For the future

I can also cross-reference the ATLAS experiment (archived).

I'll also probably to a mass spectrum from PHENIX to relive the good ol' days. Their github (archived) will likely be useful.