Bayesian Analysis of Quasar Lightcurves with a Running Optimal Average: PyROA Fits to COSMOGRAIL Data



Available as .zip files are the data/results of using PyROA to fit to the COSMOGRAIL gravitationally lensed quasar data. In each .zip are the individual lightcurves for each image as a .dat file, a plot of the results, a corner plot of the sampled parameters and three pickle objects. These are samples.obj, samples_flat.obj and X_t.obj.

The first of these, samples.obj, contains all the MCMC samples in an array with shape (Nsamples, Nwalkers, Ndim).
The second, samples_flat.obj, contains the flattened samples, where the burn-in has been removed and is an array with shape (Nsamples_final, Ndim), where Nsamples_final is the number of samples with burn-in removed. This was used to generate the corner plot and obtain the best fit parameters.
The last, X_t.obj, contains the driving lightcurve, X(t), as described in Donnan et al. 2021. This is an array of the form [t, X, X_errs], where X is the value of the driving lightcurve at time, t, with errors, X_errs.

For objects with more than two images, folders where other images were the reference are included. The sigma parameter in the corner plot is the extra error added to the flux data which is labelled as si in the paper. The original data file downloaded from the COSMOGRAIL website is also included as a csv file.
Date made available2 Jul 2021

Cite this