Re: transitive closure again

paul c wrote:
> Here's one for the theorists (further to the thread from a month or two
> ago where I said some nonsense about obtaining the closure without
> recursion and which has been proved wrong many times and although I find
> the proofs hard to follow).
>
> Let me turn it around and ask a converse question:
>
> If we have a relation that is the closure (including trivial
> relationships if necessary), can we obtain what Mikito (if I recall),
> calls the adjacency list, using the RM algebra but without recursion?

Seems easy: for all pairs of nodes making an edge (x,y ) in the TC graph, select only those that aren't represented as (x,z) & (z,y). Received on Sat Nov 05 2005 - 00:10:41 CET