Re: Graph Theory problem

Tom Renfro wrote:
>
> I think what you are wanting is a topological sort.

I think you're probably right. I was thrown off by the definition of "topological order" which requires a directed acyclic graph whereas, if I now understand correctly, "topological sort" does not. Have I got that right?

> my $G = new Graph::Undirected;

>...

Yes, very clever. I think I accomplished the same thing by using a directed graph instead of an undirected one and creating bi-directional links between the tables in a pair:

sub order_joins {

    my $joins = shift;
    my $G = new Graph::Directed;
    foreach $pair (_at_$joins) {

        my ($first,$second) = _at_$pair;
        $G->add_path( $first, $second );
        $G->add_path( $second, $first );

My way seems like it is closer to what is actually being modeled, but should it produce the same result? (in my test cases, it seems to)

Thanks much Tom!

-- 
Jeff