Re: [igraph] Faster way to detect duplicate graphs in a vector ptr?

[Timothy Rice]

Hi Gabor,

Thank you for responding.

> first, all this would be much-much simpler in R or Python, so you might
> consider using a high level language.

The thought crossed my mind but I think coding in C is good for the soul
:-)


> Second, you can define a hash function on the graph, and then calculate
> that for each graph. If the hash function is good, then you don't need to
> compare most pairs of graphs, because their hash will be different if they
> are different. If their hash is the same, then you need to compare them.
> This can be done by querying their edge lists, sorting them and then
> comparing them.
> 
> An additional trick would be to sort the graphs according to their hash,
> and then only compare neighboring elements in the list with the same hash.

This seems like something I could do. I haven't implemented hashing before
so it will be a good learning exercise :-)

Thanks for your help!


Cheers,


Tim
-- 
PhD Candidate
A: Room 210, Richard Berry (Maths & Stats), Melbourne University
E: address@hidden
W: http://www.ms.unimelb.edu.au/~trice
G: https://github.com/cryptarch

pgp3rXWN50uit.pgp
Description: PGP signature