gA chiral-continuum extrapolation scripts

This repository contains Python scripts that are released with the publication of An accurate calculation of the nucleon axial charge with lattice QCD. They have been tested extensively with Python 2.7.

We release the scripts used to perform the chiral-continuum extrapolation along with bootstrapped correlator results necessary for the extrapolation in hdf5 format.

• The analysis is performed by running python c51_gA_analysis.py where user selectable options may be viewed in the help section by assigning the flag --h.
• User defined constants for the extrapolation, and plotting parameters for matplotlib are assigned in data_params.py.
• Various fit ansatzes and their corresponding chi^2 statistic is defined in ga_fit_funcs.py.
• Plotting scripts for the extrapolation curves as a function of mpi, a, and mL are defined in utils.py.
• c51_gA_analysis.py may be used to bootstrap the extrapolation results. If this is done, the bootstrapped results will be written to c51_ga.sqlite in sqlite3 format. This file may be read by running sqlite3 c51_ga.sqlite from the terminal.
• sqlite_store.py contains the script which creates and writes to the c51_ga.sqlite database.
• c51_gA_histogram.py generates histograms from bootstrapped extrapolation results. Options for this scripts are accessible under the function fit_list() inside the file. The user may choose the set of models considered in the AIC weighted average. The script will plot histograms of the individual extrapolations, as well as the AIC averaged histogram.

Dependencies

We rely on the numpy, scipy, matplotlib, iminuit, sqlite3, tqdm, and tables Python modules. They may be installed via the Python package utility pip via

pip install numpy scipy matplotlib iminuit tqdm tables

The tables module relies on HDF5, which is commonly installed in many supercomputing environments and can easily be installed on local machines.