[FRIAM] Graph/Network discursion.
Steven A Smith
sasmyth at swcp.com
Fri Jun 9 15:37:39 EDT 2017
Glen -
At the risk of boring the rest of the crowd silly, I'd be interested in
hearing more about the kinds of graphs you would like to talk about. I
agree that Partially Ordered Sets are a (relatively) special case.
My interest is in the structure/function duality, more in topological
than geometric structure. I've read D'Arcy Thompson (no relation Nick?)
but not studied him closely, and I defer for my intuition to Christopher
Alexander (A Pattern Language) for high-dimensional graph-relations in a
real-world (human-built environments) context I can relate to.
Perhaps I'm more interested in Networks, though I'd like to ask the
naive question of how folks here distinguish the two... I tend to think
of Networks as Graphs with flows along edges. I think in terms of
multi-graphs (allowing multiple edges between nodes) or more precisely,
edges with multiple strengths/lengths/flows or more precisely yet I
think, with vector properties on edges...
- Steve
On 6/9/17 9:05 AM, ┣glen┫ wrote:
> Because, as Steve rightly pointed out with that Joslyn paper, the point is the extent to which the system submits to ordering. A strict hierarchy (levels, like I think EricS drew) submits to a total order, whereas a brranching hierarchy (still levels) submits to a partial order. Graphs work, but not as analogy, per se ... more like exact representations. The kinds of graphs I'd like to talk about don't (necessarily) submit to ordering, even partial ordering. (no levels) It would be more complete to say that any "ordering" would be more complicated than simple relations like ≥ or ≤.
>
> On 06/08/2017 09:48 PM, Marcus Daniels wrote:
>> Why does there need to be any spatial property? Why not a graph?
>
More information about the Friam
mailing list