Choose timezone
Your profile timezone:
Powered by Indico
v3.3.4
Check Run2015A data content. There is a typo in "CastorTowersP4", it very likely is "CastorTowerp4".
you may either use ROOT directly to read the files and check, or even simpler: use the printTTree.py on LXPLUS to dump the content.
Another topic: please count jet-multiplicity only after eta-cut.
And of course, send my your analysis code, so that I can have a look.
this is remarkable progress. Very good, and I am impressed. Here are comments:
- when you send plots, I can not read the screenshots (too small). It would be much better, if you paste plots in good quality (save as pdf) into a powerpoint and send me this. Really try to improve image quality.
- Yellow line, yellow as color in plots: don't use yellew: it is almost impossible to see and on a projector in sunlight you won't see it at all. Only dark colors. Also light green is particularly bad.
- in you "threshold values" First: please name your thrsholds so that I can see where they belong to, e.g. "em_bin3" or "had_bin1" or so. And second: how can it be that all values are negative? This makes not so much sense. RMS is a strictly positive number. Please send me formulas/code that you used to calculate that. But also, the yellow lines in your plots are all at positive values. Why the yellow lines are not exactly at the values of this list?
- and to your final (main) question: ZeroBias (ZB) are always real collision data form Run2015A; they are not MC and not "gen"-level. Only the "EPOS" and "QGSJetII" samples are obviously MC and also "gen"-level.
- thus, your plot "ratio": for the left and center plots (em and had dE/dEta): those plots are currently made from calo towers. This is very good. Please make two further plots exactly like this, name them by adding "_gen", thus e.g. "EnergyEmEta_gen", and fill them only for MC samples with generator-level particle energies (in the same way as you would fill the current plots with calo towers).
- When you have that, we should look for EPOS and QGSJetII at the "correction factors" from measured calo energy to gen-level energy by dividing EnergyEmEta_gen by EnergyEmEta. This basically tells you: what is the fraction of em energy I measure with the calorimeters with a specific noise-threshold-cut, compared to the true total generator level particle energy. This is obviously a very important quantity. We need it for both, e.m. and had.
- In your code, you can protect these plots with the "if self.isMC:" trick that you introduced already. Otherwise, this WILL crash for zero-bias data since this has no generator-level data. You have experienced this before.