On Feb 24, 1:33 am, Jeff Rose wrote:
>
> That would be great. Especially if the algorithms can operate over some
> kind of generalized interface, so the algorithms can run over different
> graph implementations. I'm writing a semi-structured storage library on
> top of the neo4j graph database,
I did make one important update. The nodes no longer need to be integers.
You can build a graph of nodes and adjacency lists without the pain of
mapping to indexes. Also, the "adjacency lists" can be a provided map, or
any function from node->neighbors.
(The fact that Clojure maps are functions
It would be easy to convert from your form to adjacency lists, so if you
want it write a converter :) I think we should keep the algorithms
*basically* efficient. I don't see that as premature optimization at all.
On Mon, Feb 23, 2009 at 9:37 PM, Eric wrote:
>
>
>
> On Feb 23, 6:38 pm, Jeffrey
On Feb 23, 6:38 pm, Jeffrey Straszheim
wrote:
> Well, right now I'm just handling directed graphs, and basically treating
> nodes as integer indexes, with a simple formula from index to adjacency list
> of nodes.
>
I would actually like to see an implementation that more closely
resembles the
Well, right now I'm just handling directed graphs, and basically treating
nodes as integer indexes, with a simple formula from index to adjacency list
of nodes.
I'm not opposed to a more elaborate implementation. Patch welcome :)
On Mon, Feb 23, 2009 at 7:33 PM, Jeff Rose wrote:
>
> Jeffrey St
Jeffrey Straszheim wrote:
> As part of my Datalog work I'm putting together some directed graph
> algorithms, mostly things like finding strongly connected components,
> and building dependency stratifications (think topological sort but with
> the results groups in tiers of non-interdependent
I don't remember that post, but my work also deals with cycles, and a
strongly connected component algorithm will be one of my first submissions.
On Mon, Feb 23, 2009 at 10:52 AM, cliffc wrote:
>
> As a compiler writer I do a lot of graph-work, but nearly all of it
> has cycles.
> There was a th
As a compiler writer I do a lot of graph-work, but nearly all of it
has cycles.
There was a thread earlier about defining cyclic graphs in Clojure.
Can someone point me to it?
Thanks,
Cliff
On Feb 22, 7:55 am, Rich Hickey wrote:
> On Feb 22, 2009, at 10:11 AM, Jeffrey Straszheim wrote:
>
> > J
I agree -- an important goal is to define a suitable common way to
represent graphs in terms of idiomatic Clojure data structures.
Francesco
On Feb 22, 4:11 pm, Jeffrey Straszheim
wrote:
> Just as a point of fact, I don't plan to make a complete *every algorithm
> you can think of* package, jus
On 22.02.2009, at 02:59, Jeffrey Straszheim wrote:
> As part of my Datalog work I'm putting together some directed graph
> algorithms, mostly things like finding strongly connected
> components, and building dependency stratifications (think
> topological sort but with the results groups in
Graphs has always inspired me and seeing implementation in Clojure
will be no less inspiring.
http://www.visualcomplexity.com/vc/
http://planarity.net/
Frantisek
On 22 Ún, 16:11, Jeffrey Straszheim
wrote:
> Just as a point of fact, I don't plan to make a complete *every algorithm
> you can thi
On Feb 22, 2009, at 10:11 AM, Jeffrey Straszheim wrote:
> Just as a point of fact, I don't plan to make a complete *every
> algorithm you can think of* package, just the ones I need. However,
> it would be easy to add others submitted by the community. I could
> be a gathering place of g
Just as a point of fact, I don't plan to make a complete *every algorithm
you can think of* package, just the ones I need. However, it would be easy
to add others submitted by the community. I could be a gathering place of
graph algorithms.
On Sun, Feb 22, 2009 at 7:31 AM, Francesco Bellomi <
fr
+1
Francesco
On Feb 22, 2:59 am, Jeffrey Straszheim
wrote:
> As part of my Datalog work I'm putting together some directed graph
> algorithms, mostly things like finding strongly connected components, and
> building dependency stratifications (think topological sort but with the
> results group
+1! That'd be really useful.
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
"Clojure" group.
To post to this group, send email to clojure@googlegroups.com
To unsubscribe from this group, send email to
clojure+unsub
+1 :)
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
"Clojure" group.
To post to this group, send email to clojure@googlegroups.com
To unsubscribe from this group, send email to
clojure+unsubscr...@googlegroups.com
+1 on that!
Graphs are common and many people are going to write similar
algorithms to yours. Also, as a library, it is going to have nice
interface (than if put together with Datalog) and will be much easier
to reuse.
Frantisek
PS: Graphs are my favorite data structure, so I am lobbying for
th
As part of my Datalog work I'm putting together some directed graph
algorithms, mostly things like finding strongly connected components, and
building dependency stratifications (think topological sort but with the
results groups in tiers of non-interdependent nodes). Anyhow, I'm thinking
this stu
18 matches
Mail list logo
