Long gamma-ray burst light curves as the result of a common stochastic pulse–avalanche process

Home

All issues

Volume 689 (September 2024)

A&A, 689 (2024) A266

Full HTML

Open Access

Table 2.

Results of the GA optimisation on the BATSE and Swift datasets.

Parameter	SS96	BATSE	Swift/BAT
μ	1.20	$1 . 10_{- 0.02}^{+ 0.03}$ $1.10^{+0.03}_{-0.02}$	$1 . 34_{- 0.02}^{+ 0.03}$ $1.34^{+0.03}_{-0.02}$
μ₀	1.00	$0 . 91_{- 0.07}^{+ 0.06}$ $0.91^{+0.06}_{-0.07}$	$1 . 16_{- 0.10}^{+ 0.18}$ $1.16^{+0.18}_{-0.10}$
α	4.00	$2 . 57_{- 0.52}^{+ 0.07}$ $2.57^{+0.07}_{-0.52}$	$2 . 53_{- 0.01}^{+ 0.25}$ $2.53^{+0.25}_{-0.01}$
δ₁	−0.50	$- 1 . 28_{- 0.05}^{+ 0.16}$ $-1.28^{+0.16}_{-0.05}$	$- 0 . 75_{- 0.29}^{+ 0.11}$ $-0.75^{+0.11}_{-0.29}$
δ₂	0	$0 . 28_{- 0.03}^{+ 0.01}$ $0.28^{+0.01}_{-0.03}$	$0 . 27_{- 0.02}^{+ 0.01}$ $0.27^{+0.01}_{-0.02}$
τ_min	0.02 s	$0 . 02_{- 0.01}^{+ 0.02} s$ $0.02^{+0.02}_{-0.01}\,\mathrm{s}$	$0 . 03_{- 0.02}^{+ 0.02} s$ $0.03^{+0.02}_{-0.02}\,\mathrm{s}$
τ_max	26.0 s	$40 . 2_{- 1.2}^{+ 0.9} s$ $40.2^{+0.9}_{-1.2}\,\mathrm{s}$	$56 . 8_{- 1.3}^{+ 0.4} s$ $56.8^{+0.4}_{-1.3}\,\mathrm{s}$

Loss (Train best)	–	0.72	0.38
Loss (Train avg.)	–	0.98	0.66

Loss (Test)	1.47	0.88	0.56

Loss (Test: ⟨F/F_p⟩)	1.01	0.67	0.46
Loss (Test: ⟨(F/F_p)³⟩)	0.40	0.20	0.20
Loss (Test: ⟨ACF⟩)	2.24	0.64	0.49
Loss (Test: T_20%)	2.22	2.04	1.08

Notes. Column 2 presents the parameters given by SS96 (for the BATSE dataset), while Cols. 3 and 4 show the optimised parameters obtained after 30 generations of the GA for BATSE and Swift/BAT, respectively. From the distribution of the seven parameters in the last generation we estimated their best-fitting values as the median, and their corresponding errors as the 16th and 84th percentiles. ‘Train best’ is the loss of the best generation, while ‘Train avg.’ is the average loss in the last generation. The test set is a newly produced set of 5000 simulated LCs; the last four rows show all the single contributions to the ‘Test’ loss.

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.