data for testing graph alogrithms

Question

Hi,

i am looking for a source of huge data sets to test some graph algrothm implemention. The files should be in an easy to read file format somthing like:

$Node1

Node23

Node322334

Node43432

$Node2:

Node232

...

Thanks,

Chris

Answer 1

Have you considered using Facebook's Graph API? It provides data in a JSON format, so it is very easy to read and should provide some large graphs depending on which data you query for.

Answer 2

A quick python hack:

def generateGraph(n=100, avgNeigbors=10):
    from random import randint
    for i in range(n):
        print "$"+str(i)
        for m in range(avgNeigbors-randint(-avgNeigbors/2,avgNeigbors/2)):
            print (randint(0,n))

Answer 3

I found this which may or may not contain what you need:

http://people.sc.fsu.edu/~jburkardt/datasets/graffiti/graffiti.html

http://people.sc.fsu.edu/~jburkardt/datasets/sgb/sgb.html

If you repost your question at http://math.stackexchange.com/ or at http://cstheory.stackexchange.com/ you may attract the attention of algorithmic graph theorists or computer scientists specialising in graph algorithms.

Do post a link here if you repost your question as I'm slightly interested in where to obtain such dataset. Thanks.

Answer 4

IMDB's dataset can be used for free (non-commercially!) which can be downloaded in flat text files. It's huge: 100's of megabytes of raw text you can build a graph of.