[BioSQL-l] ontology for transitive closure table

[Hilmar Lapp]

On Thursday, March 27, 2003, at 12:38  PM, Aaron J Mackey wrote:

> The changes I've made to Graph::Directed are to simply store the set of
> vertices that were traversed (i.e. the actual "path") to generate a new
> edge in the transitive closure

I'm wondering whether the APSP graph wouldn't work as transitive 
closure too? The APSP_Floyd_Warshall() returns a graph with length and 
path recorded for all edges. Floyd-Warshall is greedy, no? So, you 
won't have all possible paths, would you? Am I missing something?

	-hilmar
-- 
-------------------------------------------------------------
Hilmar Lapp                            email: lapp at gnf.org
GNF, San Diego, Ca. 92121              phone: +1-858-812-1757
-------------------------------------------------------------