PU weights

Pile-Up weights are computed via a bin-by-bin division of the npu distributions in MonteCarlo and Data.

Data distributions are provided centrally and are stored in /weights/PUreweight/UL_PU_distributions/: we use the ones computed using 69.2 mb pp cross section (machine physicists like 80 mb). The ones computed using 66 mb and 72.4 mb are used to evaluated systematic uncertainties.

MC distributions can be produced using any samples: we typically use TT since they are the largest ones.

Use the macro /weights/PUreweight/ratioPU.C (https://github.com/LLRCMS/KLUBAnalysis/blob/VBF_UL/weights/PUreweight/ratioPU.C). In this line (https://github.com/LLRCMS/KLUBAnalysis/blob/3aa88f23ccb17fa287cf8bf22be1b10c0c79e1e6/weights/PUreweight/ratioPU.C#L3) you should change the maximum number of events (depends on the sample), the location of the file list and the root output file name.

root -l -b -q ratioPU.C

Use /weights/PUreweight/makeRatio.C to get the bin-by-bin values: they are printed on the terminal and should be copied to a txt file. Change https://github.com/LLRCMS/KLUBAnalysis/blob/3aa88f23ccb17fa287cf8bf22be1b10c0c79e1e6/weights/PUreweight/makeRatio.C#L1 with the name you choose in the previous script and https://github.com/LLRCMS/KLUBAnalysis/blob/3aa88f23ccb17fa287cf8bf22be1b10c0c79e1e6/weights/PUreweight/makeRatio.C#L11 with the data histogram name.

root -l makeRatio.C