Skip to main content
. 2013 Oct;20(10):738–754. doi: 10.1089/cmb.2013.0073

Procedure.

RandomOptimalReconciliation

1:    for each Inline graphic in a pre-order traversal of Gdo
2:     ifg = rt(G) then
3:      Let P denote the set Inline graphic.
4:      For any Inline graphic, assign the mapping Inline graphic to be s with probability Inline graphic.
5:   Let {g′, g″} = ChG(g).
6:   Let Q denote the set {i: bi = 1} in the context of the Equations 1, 3, or 4, used when computing N(g, s).
7:   Choose an ai, where Inline graphic, with probability Inline graphic.
8:   Assign an event type to g based on whether the chosen ai was derived from cΣ(·,·), cΔ(·,·), or cΘ(·,·).
9:   Consider the sets X′ and X″ for the chosen ai, as defined in Equation 2.
10:   Assign the mapping Inline graphic to be node s from X′ with probability Inline graphic.
11:   Assign the mapping Inline graphic to be node s from X″ with probability Inline graphic.