Distracted by research

I'm delighted to report that progress on graph isomorphism is no longer blocked.

The area that I've been having problems with is the construction of the isomorphism map once it has been established that the two graphs are isomorphic. I found where I was missing a bit of code, wrestled with it, and got it working (mostly) yesterday afternoon.

I know what I need to do next to finish this part of the algorithm, though I'm a bit surprised that it is necessary.

I started running the algorithm on some of the graphs in the (Amalfi graph database. It nailed the random graphs and ran very slowly on the 2D-mesh graphs. I'm not sure what the problem is, but I'll check it out and fix it.

The isomorphism-map construction bug's importance is that it has kept me from seeing whether the changes I made to the graph-isomorphism-determination algorithm a few months ago are effective. I'm delighted to report that this algorithm seems to be behaving correctly now for all graphs.

More bugs to fix, of course, but I can now justify the time needed to work on the theoretical side again.

Anyway, I made a lot of progress over the weekend.

Lagniappe

I'm delighted to introduce you to xkcd, a "webcomic of romance, sarcasm, math, and language".

Here is a movie of the stunning solar flare that is displayed above.

Copyright and Legal Notices

All Rights Reserved.