Mad Andrew's Journal

Of Peanuts and Plots
Ahem. If I were the sort of person that cracked open champagne to celebrate, yesterday I would have.
Why?
A little background. In my thesis, I will cite some limits on the extent to which the data from our detector agrees with the Standard Model. This involve forming a Bayesian likelihood, which tells you how much your data is like the model, or how much it is like something else. If the data's most likely point is near the model, then you set limits on how far the parameters can stray and still have some probability of producing this data. This results in a nice elipse in parameter space (a confidence level contour for those who are statistically inclined. You know who you are.).
UNLESS- The data doesn't agree with the model at all. Worst case would be that you form a 'boomerang' that excludes the Standard Model entirely, or you have some other weird shape like a peanut, away from the point.
Yesterday I made the first likelihood fit for the data. I was nervous, because I didn't know what the answer would be. And then I get this beautiful elipse that looks exactly how I thought it should. Woot. It also means that I got to this point in a DAY unlike some post-docs I could mention where it took months.

Of course, I proudly send this off to my boss, who then turns up his nose. Couldn't care less. Wants to see a completely different plot. My world is anger, and bloodthirsty vengeance.

Advertisement