PyrootCK

Improving PyROOT for better productivity.

Collection of utilities are organized into subpackages::

• mathutils:

  • asymvar: class for variable with asymmetric-error, inspired from uncertainties.ufloat.
  • Eff, EffU, EffU_unguard: functions to compute efficiencies with Clopper-Pearson uncertainty.
  • weighted_average, weighted_harmonic_average: when a simple average is not enough.
  • combine_fully_correlated, combine_uncorrelated, combine_BLUE: for combining multiple observables with uncertainty into one, given choices of correlation. For BLUE (Best Linear Unbiased Estimator), see Valassi, 2013.

• iouils

  • import_tree to quickly load TTree from (multiple) TFile overloaded for different source types (local, ganga, eos, xrootd, ...).

• tmvautils

  • TMVA_Adapter to help setup TMVA.Reader variables, and return TTree of mva-response weights.

As well as miscellaneous monkey-patching on ROOT and uncertainties for more methods::

• ROOT:

  • Misc conversion to/from ROOT (TH,TGraph,RooWorkspace,RooFitResult,...) and pandas (Series, DataFrame).
  • TFile.slice_tree to extract TTree into smaller one.
  • TTree.drop to make index-unique TTree.
  • TH1.vlookup, TH2.vlookup: like in Microsoft Excel, to retrive value in a bin given point(s) on the axis.
  • TMultiGraph.brazillian: for the upper limits plot.

• uncertainties:

  • class var, based on ufloat but ready-made for statistical (Poisson) error.
  • Additional methods on ufloat: rerr, upperlim, low, high, interval, rounding_PDG
  • More methods involving error tag: tags, get_error, get_rerr

See the docstring from module index for more details.

Installation

It's available on pip: pip install pyrootck

Dependency: uncertainties, pandas, root_numpy, pyroot_zen, PythonCK

Disclaimer

This packacge was written and used during my PhD in 2013-2017 at EPFL (Lausanne) and LHCb collaboration (CERN), for the work in Z->tau tau cross-section measurement and H->mu tau searches at LHCb (8TeV).

I hope it can be of a good use for future analysis...