I am working on compressing graphs. The graph is represented as an adjacency list i.e each node in graph maintains a list of adjacent nodes. The lists contain Node_IDs(4 byte integers) and they are ordered by popularity of each Node_ID(score). So, I end with up unsorted list of Node_ID numbers and I have to compress them while maintaining their present order, to store efficiently.
+2
A:
Something simple would be to do a basic Huffman Encoding on the Node_IDs or even sequences of Node_IDs and convert the most common one to take up a few bits and so on.
Zugwalt
2010-07-22 01:17:12
A:
If you can get anytime the scores by knowing the ID-s then you can store them in any order.
ruslik
2010-07-22 01:17:54
No. The scores of ID's depend on the list, they are in. So, knowing just the ID can't get you the score.
rvijax
2010-07-23 05:35:19