Benchmark Datasets Used in "Improved Classification of Variable Stars with Phase-Invariant Neural Networks"

To aid the reproducibility of the results in the paper “Improved Classification of Variable Stars with Phase-Invariant Neural Networks,” we make our aggregation of the following data available. Code used to load the data and generate the results can be found at https://github.com/kmzzhang/periodicnetwork. These datasets have been constructed from publicly available data sources. If you use these datasets, please cite the original papers [1, 2, 3], in addition to ours [TBD]. Others might find this data useful for testing time-series inference techniques.

[1] Jayasinghe, T. et al. The ASAS-SN catalogue of variable stars I: The Serendipitous Survey. Monthly Notices of the Royal Astronomical Society 477, 3145–3163 (2018). URL https://academic.oup.com/mnras/article/477/3/3145/4961151.
[2] Alcock, C. et al. The MACHO Project LMC Variable Star Inventory.II.LMC RR Lyrae Stars- Pulsational Characteristics and Indications of a Global Youth of the LMC. The Astronomical Journal 111, 1146 (1996). URL http://adsabs.harvard.edu/abs/1996AJ....111.1146A.
[3] Udalski, A. The Optical Gravitational Lensing Experiment. Real Time Data Analysis Systems in the OGLE-III Survey. Acta Astronomica 53, 291–305 (2003). URL http: //adsabs.harvard.edu/abs/2003AcA....53..291U.