Jaroslav Hajek skrev: > On Mon, Sep 13, 2010 at 11:26 AM, Paul Dreik <[email protected]> wrote: >> Hi! >> I needed topological sort for a project and wrote a quick and dirty >> implementation in octave, since I could not find such functionality >> among the existing code. >> >> I think topological sort would be a good thing to have in octave. >> >> *should I add the function to octave-forge? >> *if so, which package would fit best? >> >> I know the implementation is not efficient, but it is always a start. >> Nothing stops writing a better implementation using the same interface >> later. >> > > Even the interface may still be discussed. For instance, I don't think > that an adjacency list using a cell array is the best idea to store a > directed graph in Octave. I think a logical matrix is, either dense or > a sparse one. > > See also > http://www.mathworks.com/help/toolbox/bioinfo/ref/graphtopoorder.html > I think the cell array is easier to create and work with. A sparse matrix is probably more efficient, both with regards to performance and space. Either way, I guess the best thing is to keep matlab compatibility.
Maybe it is good to start a package "graph" for graph theory related functions? In Matlab, the bioinformatics toolbox contains some graph functions (see the link from Jaroslav). If noone objects, I will create such a package. Paul ------------------------------------------------------------------------------ Start uncovering the many advantages of virtual appliances and start using them to simplify application deployment and accelerate your shift to cloud computing. http://p.sf.net/sfu/novell-sfdev2dev _______________________________________________ Octave-dev mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/octave-dev
