« A powerful new algorithm | Main | In Rome and Sweating »

June 27, 2005


manuel visaya

the microeconomics of particle physics just gets cheesier the more elusive a particle that is caught. you catch a neutrino, next day, the lowest temperature attainable is lowered, like chasing down a vein. some things just don't change. the lesson being to always follow through but still there is multitasking and the quasi-issue of trust. the duck is ubiquitous and yet easily ignored at this point. hard is done (BRNT).

#10 LaBs

Gordon T Watts

This is just a way of getting a scale factor as a function of n-dimensions when you don't have the MC data and don't know the functional form to fit the n-dimensional scale factor to, yes? For example we have to scale from MC to data b-tagging rates we use a scale factor that has n-parameters. Since n is small, we can define and easily fit a function. Am I getting what you are trying to do here?

As far as minimizing -- why not feed it to MINUIT? Your speed will depending only on how finely you "bin" the MC (use a gaussian or similar to find the center of each bin, width for error). Or are you going for something different here?

Tommaso Dorigo

Hi Gordon,

well, sort of. You can think of hyperballs as bins of variable size, yes. The algorithm is more subtle, since the bin center is by definition the point where you are evaluating the field, and a weighted mean is made depending on a generalized distance in the n dimensions. But the generalities are the same.
I did try the algorithm to estimate a b-tagging efficiency (not a scale factor), but now I am using it to correct the measured b-jet energy.

riqie arneberg

i find your explainations very easy to understand, and i will use your blog to educate myself. i have read every popular work in physics that the local library carries, guess u might say im a fan]

we played bridge earlier, i am frprncss7


Hi Riqie, thank you for visiting my site and for the comment. Hope we can play bridge again soon!

The comments to this entry are closed.