Predicting stellar rotation periods using XGBoost | Astronomy & Astrophysics (A&A)

Open Access

Table 1

Names of the 170 predictors, as they appear in the original data set, and their description.

Variable	Description	Group	Observations	N
`teff`	T_eff, effective temperature of the star	astro		1
`teff_eup, teff_elo`	upper and lower errors for T_eff	astro		2
`logg`	log g, logarithm of the surface gravity	astro		1
`logg_eup, logg_elo`	upper and lower errors of log g	astro		2
`m`	M, the mass of the star, in solar masses	astro		1
`m_eup, m_elo`	upper and lower errors for M	astro		2
`f_07, f_7, f_20, f_50`	FliPer values for ν_C = 0.7, 7, 20 and 50 µHz	astro		4
`kepmag`	Kepler magnitude from the Kepler input catalogue	astro	parameter linked to the quality of the stellar target	1
`length`	length of the light curve, in days	ts	parameter linked to the quality of the acquired light curve	1
`n_bad_q`	number of bad quarters in the light curve	ts	parameter linked to the quality of the acquired light curve	1
`start_time, end_time`	starting and ending time of the light curve	ts	parameter linked to the quality of the acquired light curve	2
`sph_acf_xx, sph_acf_err_xx`	S_ph, photometric activity proxy, and its error, computed on the ACF for the xx-day filter	ts	mean of standard deviations over light curve segments of 5 × P_ACF	6
`sph, sph_e`	S _ph, photometric activity proxy	ts	it is a magnetic activity proxy	2
`g_acf_xx`	height of P_ACF for the xx-day filter	ts	P_ACF is the period of the highest peak in the ACF at a lag greater than zero; G_ACF is a control parameter	3
`h_acf_xx`	mean difference between height of P_ACF and the two local minima on both its sides for the xx-day filter	ts	H_ACF is a control parameter	3
`cs_noise_xx`	mean level of noise of the Gaussian functions fitted to CS for the xx-day filter	cs	this is the period uncertainty; the level of noise corresponds to the half width at half maximum (HWHM)	3
`cs_chiq_xx`	χ² of the fit of the Gaussian function on CS for the xx-day filter	cs		3
`cs_n_fit_xx`	number of Gaussian function fitted to each CS for the xx-day filter	cs		3
`cs_gauss_i_j_xx`	amplitude (i = 1), central period (i = 2), and standard deviation (i = 3) of the j^th (j = 1, 2, … 5) Gaussian fitted in the CS with the xx-day filter	cs	CS is the product between the normalised GWPS and the ACF	45
`h_cs_xx`	amplitude of P_CS for the xx-day filter	cs	P_CS is obtained as the period of the fitted Gaussian of highest amplitude; it is a control parameter	3
`sph_cs_xx, sph_cs_err_xx`	S_ph and its error computed on CS for the xx-day filter	cs	mean standard deviations over light curve segments of $5 \times P_{rot}^{CS}$ $5 \times P_{{\rm{rot }}}^{{\rm{CS}}}$	6
`gwps_noise_xx`	mean level of noise of the Gaussian functions fitted to GWPS for the xx-day filter	gwps	this is the period uncertainty; the level of noise is given by HWHM	3
`gwps_n_fit_xx`	number of Gaussian functions fitted to each GWPS for the xx-day filter	gwps		3
`gwps_gauss_i_j_xx`	amplitude (i = 1), central period (i = 2), and standard deviation (i = 3) of the j^th (j = 1, 2, … 6) Gaussian fitted in the GWPS with the xx-day filter	gwps		54
`sph_gwps_xx, sph_gwps_err_xx`	S_ph and its error computed on GWPS for the xx-day filter	gwps	mean standard deviation over light curve segments of $5 \times P_{rot}^{GWPS}$ $5 \times P_{{\rm{rot}}}^{{\rm{GWPS}}}$	6
`prot_acf_xx`	rotation period extracted from the ACF analysis for the xx-day filter	prot		3
`prot_cs_xx`	rotation period extracted from the CS analysis for the xx-day filter	prot		3
`prot_gwps_xx`	rotation period extracted from the GWPS analysis for the xx-day filter	prot		3

Notes. Indicated are the group to which the predictors belong, and the effective number of variables associated with each of them (N).

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.