Multi-wavelength GRB Classifier

Multi-wavelength GRB Classifier

This Multi-wavelength GRB Classifier physically classify Gamma Ray Bursts (GRBs) as Type II (massive star core collapse) and Type I GRB (compact star merger) based on multi-wavelength criteria, with Naive Bayes method ultilized. Both the prompt emission information and host galaxy information are taken into consideration. This method is based on Li, Zhang & Yuan, 2020, The Astrophysical Journal, Volume 897, page 154, also available at arXiv:2005.13663.

Input parameter
(keep the unknown parameters blank)
name (not required): log T_90 (s):
log E_iso (erg): alpha:
log E_p (keV): log (f_eff-1):
log sSFR (Gyr^-1): log M (M_sun):
[X/H]: log R_50 (kpc):
log r_off (=R_off/R_50): F_light:

Input parameters:

  • log T_90: The duration of GRB prompt emission, in unit of seconds.
  • log E_iso: The bolometric isotropic energy of GRBs, in 1-10000 keV, in unit of erg.
  • alpha: The low energy index of the GRB gamma-ray spectrum. It is usually a negative value.
  • log E_p: The peak energy of the GRB gamma-ray spectrum, in unit of keV.
  • log (f_eff-1): The ratio between the peak flux and the flux to make the duration to be 2 seconds. See Lü, H.-J. et al. 2014 for more information.
  • log sSFR: specific star formation rate of the host galaxy, in unit of Gyr^-1.
  • log M: stellar mass of the host galaxy, in unit of M_sun.
  • [X/H]: metallicity relative to the Sun. The Sun has a metallicity [X/H]=0.0 here.
  • log R_50: Half light radius of the host galaxy, in unit of kpc.
  • log r_off: offset from the GRB to the center of the host galaxy, in unit of R_50.
  • F_light: Cumulative light fraction of position of the transient relative to other position of the host galaxy. See Fruchter et al. 2016 for more information.

Output:

  • pL: The probability to be Type II (massive star core collapse).
  • pS: The probability to be Type I (compact star merger).
  • logOdds: The logarithmic ratio of the prabability to be Type II to Type I.
  • logOdds_prompt: The logOdds for prompt emission properties.
  • logOdds_host: The logOdds for host galaxy properties.
  • O=logOdds+0.7: correlation corrected logOdds. Type II GRBs have O > 0 and Type I GRBs have O < 0.
  • Type: The preferred type of GRB, estimated the multi-wavelength properties.

Program Version History

V1.0: 2020 May. Original program version corresponding to submitted version of Li, Zhang and Yuan (2020).

Download

Download the python code. Run python grb_cls.py to see the results in data_out.csv.
Contact Ye Li (liye_kiaa@pku.edu.cn/yeli@pmo.ac.cn) or Bing Zhang (zhang@physics.unlv.edu) if any problem is encountered.