Please see Gagné et al. (2018) for the details of the BANYAN Sigma classifier and Gagné et al. (2026) for the updated BANYAN Sigma models, and please acknowledge both papers when you use results from this tool.

The stored BANYAN Sigma results in MOCAdb include an Obs column describing the observables used for the stored results of this star: pm is the minimum set of observables, equatorial coordinates and proper motion; rv is the radial velocity; and plx is the parallax or photometric distance. The YA(%) result is the total membership probability for all possible young associations combined, in percent. Most of the time, this is concentrated in a single association, but for some ambiguous members it can be distributed across more than one. When this happens, all_prob_yas contains more than one moca_aid key with its relative share of the total association probability, in percent. For example, all_prob_yas: BPMG(60); ABDMG(40), with YA(%): 90%, means the total association probability is divided 60/40 between BPMG and ABDMG; the absolute BPMG probability is 54%, and the ABDMG probability is 36%.

The priors of BANYAN Sigma are adjusted such that a selection threshold of 90% in total association probability will recover 90% of the true members across associations when using input measurements with a precision similar to Gaia; see Section 6 of Gagné et al. 2018. The contamination rate even at a 90% probability threshold varies widely between associations and is usually worse for the nearest associations or those with larger UVW models. When choosing a threshold other than 90%, the recovery rate also becomes association-dependent.