Data from: "Deep Generative Modeling of Periodic Variable Stars Using Physical Parameters"

This dataset was used for the training of a conditioned Variational Autoencoder that generates physically informed light curves of periodic variable stars. The light curves correspond to data obtained from The Optical Gravitational Lensing Experiment (OGLE), while ancillary information was obtained from the Gaia Data Release 2 (GAIA DR2). This repository contains the preprocessed OGLE light curves and the GAIA measurements corresponding to each cross-matched source. We also provided a subsample of cross-matched sources that were carefully validated following several steps described in the companion article (paper reference).

This dataset is realized in tandem with the corresponding GitHub and article.