Reading and writing in different formats

Heppy histograms can be written to and read from the native text-based format using heppy.basehistogram.to_file() and heppy.from_file().

ROOT

Reading from a ROOT file:

heppy.readroot(rootfile, histpath, variation_paths={}, ignore_missing_variations=False, **kwargs)

Reads a histogram (possibly with systematic variations) from a ROOT file.

Parameters:

• rootfile (str) – path to ROOT file
• histopath (str) – path of the histogram inside the file
• variation_paths (dict of str : str) – optional dictionary of variation names (keys) and paths to the variation histograms inside the same ROOT file (values). NOTE: we can easily add the ability to read variation histograms also from other ROOT files than the nominal, get in touch if you want that.
• **kwargs –

  get passed on to histogram constructor. An important one is areas=False if retrieving ratios (e.g. efficiencies).

Converting to ROOT histograms and writing to ROOT files can be done using heppy.histogram1d.to_root() and heppy.histogram1d.to_root_file() (for 1D histograms) and heppy.histogram2d.to_root() and heppy.histogram2d.to_root_file() (for 2D histograms).

Rivet (YODA)

One-dimensional histograms can be written to the Rivet YODA format using heppy.histogram1d.to_yoda() or its synonym heppy.histogram1d.to_rivet().

TRExFitter (YAML)

Reading TRExFitter YAML format:

heppy.readtrex(trexfile, unfolded_format=False)

Reads Heppy histogram from TRExFitter YAML format.

Requires PyYAML.

IMPORTANT: the uncertainties are treated as correlated uncertainties, since their actual correlations cannot be inferred from the histogram. So treat them with care: they’re fine for plotting and looking at individual bins, but do not rebin the histogram or do statistical testing involving its shape and assume that you will find correct results.

Parameters:

• trexfile (str) – path to input file.
• unfolded_format (bool) – set this to True if reading the YAML format that TRExFitter uses for unfolded cross sections.