PRISM-Astro/h2f_delight_train

Description

This is the dataset from DELIGHT [1], a model that locates the host galaxy of a transient from multi-resolution images. The difference with the original is that the images are not the cutouts used in the paper: they were regenerated from hips2fits (PanSTARRS DR1 HiPS, served by CDS).

It contains 7,350 objects, meant for training and validation. The fold column carries a ready-made 5-fold cross-validation split, so the same file covers both roles. Held-out evaluation is kept in a separate dataset, h2f_delight_test, which shares no host galaxy with this one.

Every cutout is centered on the real supernova position, and the host_pos label is the offset from there to the host galaxy.

This split includes Sérsic metadata extracted from CasJobs (PanSTARRS DR1). The rSer* columns are populated for every object, which makes it possible to sample synthetic transient positions from the galaxy light profile.

Image format

image has shape (5, 5, 30, 30), which maps to (resolution levels, photometric bands, height, width)

Columns

Column	Type	Description
Image and label
image	(5,5,30,30) float32	Multi-resolution PanSTARRS cutout, centered on the supernova.
host_pos	float32 × 2	Offset [dx, dy] from the supernova to the host galaxy, in level-0 pixels.
Identification
oid	string	Transient identifier (ZTF, or IAU name for historical supernovae).
host_name	string	Catalog name of the host galaxy.
Host galaxy
host_redshift	float64	Redshift of the host; NaN when unknown.
hostsize	float64	Angular size of the host galaxy from the catalog.
host_ra	float64	Right ascension of the host, in degrees.
host_dec	float64	Declination of the host, in degrees.
Transient position
meanra	float64	Mean right ascension of the transient, in degrees. This is the center of the cutout.
meandec	float64	Mean declination of the transient, in degrees. This is the center of the cutout.
sn_ra	float64	Right ascension of the supernova, in degrees.
sn_dec	float64	Declination of the supernova, in degrees.
Sérsic fit — r band, from CasJobs
rSerRadius	float64	Effective radius Re of the r-band Sérsic fit, in arcsec.
rSerMag	float64	Magnitude of the r-band Sérsic fit.
rSerAb	float64	Axis ratio b/a of the fitted ellipse, between 0 and 1.
rSerPhi	float64	Position angle of the fitted ellipse, in degrees.
rSerRa	float64	Right ascension of the Sérsic fit center, in degrees.
rSerDec	float64	Declination of the Sérsic fit center, in degrees.
Cross-validation
fold	int64	Assigned cross-validation fold, from 1 to 5.

References

[1] Förster, F., Muñoz Arancibia, A. M., Reyes-Jainaga, I., et al. 2022, DELIGHT: Deep Learning Identification of Galaxy Hosts of Transients using Multiresolution Images, The Astronomical Journal, 164, 195. doi:10.3847/1538-3881/ac912a

@article{Förster_2022,
  doi = {10.3847/1538-3881/ac912a},
  url = {https://doi.org/10.3847/1538-3881/ac912a},
  year = {2022},
  month = {oct},
  publisher = {The American Astronomical Society},
  volume = {164},
  number = {5},
  pages = {195},
  author = {Förster, Francisco and Muñoz Arancibia, Alejandra M. and Reyes-Jainaga, Ignacio and Gagliano, Alexander and Britt, Dylan and Cuellar-Carrillo, Sara and Figueroa-Tapia, Felipe and Polzin, Ava and Yousef, Yara and Arredondo, Javier and Rodríguez-Mancini, Diego and Correa-Orellana, Javier and Bayo, Amelia and Bauer, Franz E. and Catelan, Márcio and Cabrera-Vives, Guillermo and Dastidar, Raya and Estévez, Pablo A. and Pignata, Giuliano and Hernández-García, Lorena and Huijse, Pablo and Reyes, Esteban and Sánchez-Sáez, Paula and Ramírez, Mauricio and Grandón, Daniela and Pineda-García, Jonathan and Chabour-Barra, Francisca and Silva-Farfán, Javier},
  title = {DELIGHT: Deep Learning Identification of Galaxy Hosts of Transients using Multiresolution Images},
  journal = {The Astronomical Journal},
}