Efficient Bayesian analysis of kilonovae and gamma ray burst afterglows with FIESTA

Open Access

Table 1

Overview of FIESTA surrogates for different models.

Physical base model	Parameters	Symbol	Range	Surrogate architecture
	Inclination [rad]	ι	[0,π∕2]
	log isotropic kinetic energy [erg]	log₁₀(E₀)	[47,57]
	Jet core angle [rad]	θ_c	[0.01,π/5]
AFTERGLOWPY	Wing factor	α_w	[0.2, 3.5]	FLUXMODEL (MLP or cVAE)
(Gaussian jet)	log interstellar medium density [cm⁻³]	log₁₀( n_ism)	[−6,2]
	Electron power index	P	[2, 3]
	log electron energy fraction	log₁₀(ε_e)	[−4,0]
	log magnetic energy fraction	log₁₀(ε_B)	[−8,0]
PYBLASTAFTERGLOW	Initial Lorentz factor	Γ₀	[100,1000]	FLUXMODEL (MLP or cVAE)
(Gaussian jet)	Same as above for AFTERGLOWPY	"	"
	Inclination [rad]	ι	[0,π/2]
	log dynamical ejecta mass [M_⊙]	log₁₀(m_ej,dyn)	[−3,−1.3]
	Average dynamical ejecta velocity [c]	${\bar{v}}_{ej, dyn}$ $\bar{v}_{\text{ej, dyn}}$	[0.12,0.28]
POSSIS	Average dynamical ejecta electron fraction	${\bar{Y}}_{e, dyn}$ $\bar{Y}_{\text{e, dyn}}$	[0.15,0.35]	FLUXMODEL (MLP or cVAE) or LIGHTCURVEMODEL (MLP)
	log wind ejecta mass [M_⊙]	log₁₀(m_{ej, wind})	[−2, −0.9]
	Average wind ejecta velocity [c]	ῡ_{ej, wind}	[0.05,0.15]
	Wind ejecta electron fraction	Y_{e, wind}	[0.2,0.4]

Notes. We list the parameters, their ranges in the training data set, as well as the architectures employed for the ML surrogates. The models marked in bold are the best-performing surrogate architectures and are used in Sects. 4 and 5. The bar above the kilonova parameters indicates mass-averaged quantities.

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.