lhcb-ftcalib 0.1.1

Library for calibrating flavour tagging algorithms at LHCb

lhcb_ftcalib

LHCb Flavour Tagging calibration software

At high-energy proton-proton collider experiments, the production flavour of neutral B mesons needs to be reconstructed from particle charges from hadronisation processes in the associated event, i.e. from additional hadronisations on the signal meson side, as well as hadronisation and decays of the partner B hadron. This is commonly done with ML techniques like (recurrent) neural networks or boosted decision trees. The mistag probability estimates of these models (probability that predicted production flavour is wrong) usually need to have the property of probabilities. This calibration tool optimizes a GLM function to predict the mistag probabilities and takes into account the fact that neutral mesons can undergo oscillation before they decay. In addition, it provides helper functions to measure the performance and correlations of these models.

Documentation: Read the Docs

Installation

pip install lhcb_ftcalib

Command Line Interface Examples

Run ftcalib --help for a list of all options or read the docs

1. Calibrating opposite side taggers in a sample and saving result

ftcalib file:tree -OS VtxCh Charm OSElectronLatest OSMuonLatest OSKaonLatest \
        -mode Bd -tau B_tau -id B_ID -op calibrate -out output

2. Calibrating both tagging sides, combining them inidividually, and calibrating+saving the results

ftcalib file:tree -OS VtxCh Charm OSElectronLatest OSMuonLatest OSKaonLatest \
        -SS SSPion SSProton \
        -mode Bd -tau B_tau -id B_ID -op calibrate combine calibrate -out output

Note: The command line interface is by design not feature complete. Use the API to fine tune the calibration settings.

Requirements

A ROOT installation needs to be present on the system, for example through a conda-forge environment. When uproot4 can write data to files, the dependencies root and root_pandas (and root_numpy) can be dropped.

• root
• uproot >= 4
• root_pandas
• iminuit >= 2.3.0
• pandas
• numpy
• scipy
• matplotlib
• uncertainties