Beauty in a plot

In the last few days, I finally managed to extract a solid, as-predicted, meaningful Z signal from the data collected by the ZBB trigger. It took us more than a year to decide that we cannot model well enough jets of low energy to be able to use them when we reconstruct the jet-jet mass, which is the variable where a Z boson signal shows up as a bump in an otherwise smooth distribution. So, since the Z does not produce jets of low energy, I decided in the end to just cut these events away, increasing the signal to noise ratio, and forgetting about the idea of keeping more background at low jet-jet mass to model its shape better (to understand what the hell I'm talking about, have a look at the plot from Run I here).

The Z signal is buried in a much more numerous background, so it is a bit like looking for the needle in the haystack. In a dataset of 21 million events, only about 20,000 of them are present (one in a thousand), so a careful selection is needed. But even after that, the signal looks ridiculously small... That is as good as one can do, and it also is why the analysis is interesting to me: it is challenging! (I remember when I first discussed the idea of extracting it from our Run I CDF data: some of my colleagues looked at me in disbelief, not taking me seriously. But then I did do it, and the looks changed).

So, is there beauty in a plot ? I think so. The plot is the final result of many analyses. If crafted skillfully, it speaks volumes by itself. It has to be pleasing to the eye, not overcrowded, esthetically composed. The data it shows has to look well modeled by whatever mean one uses to describe it - a fit function overimposed to it, for instance, has to pass through most of the data points - physicists are born with a eyes-interfaced fitting algorithm in a part of their brain, and they will be able to tell if a fit has a bad chisquared (the measure how how well the fit models the data) by just looking at it.

In the case of the plot that shows the Z boson peak, beauty is there for two other reasons. The first is that this is the mass of two b-quark-originated jets, and b quarks are also known as "beauty quarks" (others call them bottom quarks, but the usage is deprecated for victorianesque reasons by the scientific community). The second is that, when you find a small signal in a large background stack, you feel as if you were able to salvage these beautiful events from oblivion. That reminds me of another piece of beauty, which I used in the frontispiece of my PhD thesis:

"Full many a gem of purest ray serene

The dark unfathomed caves of ocean bear,

Full many a flower is born to blush unseen,

And waste its sweetness on the desert air."

(Thomas Gray, "Elegy written in a Country Churchyard", 1751, vv 53-56).

So, having found the gems, I have to do a plot today. I'm struggling with root, a program physicists use to do their final data analysis and to produce graphs and plots. I unfortunately cannot show here the results of my efforts, since they are not blessed (see here for what is a blessing of results in CDF) yet! However, here is how my computer screen looks today:

So, since I am not telling you anything about the plot on the screen, I feel I am not breaking our publishing rules... But if you look closely, you will see a teeny-tiny green histogram - that is the Z signal!

Comments

[Gordon Watts]

Yo! The more info you get out the better. But there is the competition to consider, and I'm sure they are watching this blog (they would be crazy not to after this post. :-)). I point this out because several of us proposed a joint workshop on single-top a while back and the upper management in both experiments was quite leary. Z->bb is one of the hot topics this time around (or should be -- imagine what we could do with the b-jet energy scale if we had enough!!) -- and thus I think will be subject to the same sort of... well, you know. :-)

[Tommaso Dorigo]

Such is my fear to break CDF rules these days, that I actually just went double-checking with the Spokespersons whether they agreed to the picture above being posted here. They did....
Gordon: we can indeed do something with the few thousand Z's in the plot. The graph on the upper right is in fact a fit to the chi2 vs jet energy scale, and it does fit nicely close to one.