This is a small python package to analyse time series data for Quasi-periodic Oscillations (QPOs) using periodograms and Gaussian processes (GPs).
I implemented GP models using celerite and george, and use bilby for Bayesian inference.
My collaborators and I published the results on periodograms here. We also have a forthcoming publication on GPs.
Not all simulated data is available on the repository since it would take up too much space. I can provide the data upon request.
Use bilby version 1.20 or higher to run this. If 1.20 has not been released yet, install from the feature branch on https://git.ligo.org/lscsoft/bilby/-/merge_requests/1086.
The intended usage is shown in analyse.py and analyse_periodogram.py. inject.py shows how to create simulated data.
Most of the keywords should be fairly self-explanatory.
Most other scripts are for creating plots or post-processing of results.
I also provide a jupyter notebook in the scripts folder, which mostly follows analyse.py.