Estimating False Alarm Rates of Sub-Dominant Quasi-normal Modes in GW190521

Data release for "Estimating False Alarm Rates of Sub-Dominant Quasi-normal Modes in GW190521" [arXiv:2209.00640]. Included are data files and scripts needed to generate figures and values that calculate the probability of getting Bayes factors larger than X in noise.

To use, download the data_relase.tar.gz and extract (e.g., on the command line run `tar -xzvf data_release.tar.gz`). The tar ball contains the following files/folders:

collected_vals-control.hdf : HDF5 file containg key results about the injections that do not have an observable 330 mode in the ringdown. Data is organized in groups by the QNM modes used in the analysis.

collected_vals-signal.hdf : HDF5 file containing key results about the injections that have an observable 330 mode in the ringdown. Data is organized in groups by the QNM modes used in the analysis.

gw190521: folder containing full posteriors from GW190521 for QNM analyses that include the 220, 220+221, and 220+330 modes, for all start times that were analyzed. These results were originally published in doi:10.1103/PhysRevLett.131.221402. They are included here in order to reproduce the plots.

scripts: folder containing scripts needed to reproduce the cumulative fraction vs Bayes factor plots, and to calculate the probability of getting Bayes factors larger than X in noise. These use the data files in the other directories. To generate, run `create_bf_plots.sh`. This will run `plot_bf.py` with appropriate inputs to generate the culuative fraction vs. bayes factor plots in the text, as well as the histograms of the time at maximum BF.

Data for other plots in the text are not provided due to size limitations, but can be provided upon request.