[Nauty-list] python bindings, graphs for tile sets

lloyd konneker bootch at nc.rr.com
Sun Jul 26 01:37:33 EST 2009

I saw on the web that the old (by P. Dobc..) pynauty binding of python
to nauty may not be supported and not working.  It does seem to be
crashing for me.

I also saw that the SAGE project has another binding of python to nauty.

I also saw that the SAGE project has implemented the nauty algorithm in
pure python, open source code.

Does anyone have any experience with these bindings of python to nauty,
especially whether the old one still works with the latest nauty.  (A
few years ago it worked for me.)

And now for something completely different...

My application is tiling the plane.  I am generating tile sets (e.g.
Wang tiles, but not exactly, more like ghostdiagrams).  I convert the
tile sets to graphs to filter out isomorphic tile sets.  Does anyone
have a reference for isomorphism of tile sets?  

My plan is to link the sides of each tile in a directed cycle, and also
link sides to a vertex representing the symbol (or color, for Wang
tiles) of the side.  That is, all side (vertices) with the same symbol
would be linked to a common vertex.  Maybe I don't need the power of
graph isomorphism, but it seems like a natural solution.

More information about the Nauty mailing list