Analysis of the observed and intrinsic durations of Swift/BAT gamma-ray bursts

•Duration distribution of Swift/BAT gamma-ray bursts (GRBs) is investigated.•For GRBs with known z, the analysis is performed in the observer and rest frames.•Mixtures of two and three log-normal distributions are fitted.•Maximum log-likelihood, Akaike and Bayesian information criterion are employed to choose the best fit.•It is found that the data is better followed by a two-Gaussian.

The duration distribution of 947 GRBs observed by Swift/BAT, as well as its subsample of 347 events with measured redshift, allowing to examine the durations in both the observer and rest frames, are examined. Using a maximum log-likelihood method, mixtures of two and three standard Gaussians are fitted to each sample, and the adequate model is chosen based on the value of the difference in the log-likelihoods, Akaike information criterion and Bayesian information criterion. It is found that a two-Gaussian is a better description than a three-Gaussian, and that the presumed intermediate-duration class is unlikely to be present in the Swift duration data.