Data release for Lattice investigations of the chimera baryon spectrum in the Sp(4) gauge theory

This repository contains the code used to prepare the plots and results included in Lattice investigations of the chimera baryon spectrum in the Sp(4) gauge theory[2311.14663]. It uses data from the corresponding data release on Zenodo.

Requirements

• Python 3.11
• GNU Make
• LaTeX

Setup

Environment

All required Python packages will be automatically installed into a Python virtual environment by make. This requires an internet connection. If you plan to run the workflow on a cluster node without internet access, first, prepare the environment by running

make venv

Data

To run the full analysis end to end, two sets of data are required.

• Download the file raw_data.zip from the data release, and unzip it into the raw_data directory.
• Download the file metadata.zip from the data release, and unzip it into the metadata directory.

For convenience, we also provide intermediary files to skip the most computational parts of the analysis.

• To avoid the step of collating the raw data into HDF5 format, download the file data.h5 from the data release and place it into the tmp_data directory. In this case, raw_data.zip is not needed.

• To skip the Akaike Information Criterion-based analysis (which takes around 300 core hours on an x86-64 cluster), download the file FIT_mass.csv from the data release and place it into the CSVs directory. To ensure that make does not try to recompute it, use touch to set its modification date to be in the future; for example,

  touch -A 120000 CSVs/fit_mass.csv
  

Running the analysis

From the root directory of the repository, run

make

Plots and tables will be output in the figs and tabs directories respectively.

Viewing the 3D plot

As 3D plots can be significantly easier to interpret if they can be interacted with, we provide a method to do this. After running the full analysis described above, the commands

source venv/bin/activate
python generate/plot_figs.py show_3d

will regenerate all plots, and additionally display the 3D spectrum plot interactively, so that it may be panned, zoomed, and rotated.

Reproducibility and reusability

Some aspects of this analysis make use of scipy.optimize.curve_fit and .minimize; as such, results are not bitwise reproducible between different CPU architectures. Additionally, some aspects (most notably analysis/analysis_AIC.py) are, for reasons we do not fully understand, slightly non-deterministic, even on the same hardware, giving results that differ in the 10th significant figure or beyond. This is well beyond the precision we consider in our analysis, and does not affect our conclusions.

The code included in this repository was written with the analysis of the specific dataset above in mind; as such, it may not readily generalise to other data without additional work.