Hi, do you think that this is a nice problems? For some set M, let there be partial orderings of M, <1, <2 of M.
We are looking for a "merge" < of the pos <1 <2 in he following sense: - < extends <1 - if there are x <2 y such that NOT x < y (that is, the merge has "rejected" that the relation x <2 y), then there is a "legitimate reason" for that, that is, a chain ( y = u0, u1, ..., ul = x) such that for all i: (ui <1 u_i+1) OR (ui <2 u_i+1) AND ui < u_i+1 (that is, the rejected relation is inconsistent with the transitive closure of <1 and that part of <2 that we already decided to throw into <.) Now what would be a good way to compute a merge of a sequence of partial orderings <1, <2, <3, ...? Is there something in o(n^2)? -- You received this message because you are subscribed to the Google Groups "Algorithm Geeks" group. To post to this group, send email to algogeeks@googlegroups.com. To unsubscribe from this group, send email to algogeeks+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/algogeeks?hl=en.
