[igraph] reading large, sparse graph with read.graph

[David Hunkins]

Hello, I am using igraph inside R, and reading a Pajek-formatted  
weighted, directed graph with read.graph. The graph I am trying to  
load is quite large (17 million vertices, and 28 million edges). It  
requires a long time simply to read the graph into memory. I would  
have thought that the time it would take read.graph to load into  
memory would be O(|V|+|E|). But it seems to be taking much longer,  
perhaps O(|V||E|). Why should it take so long? Is it significantly  
faster with any other format? (I don't use any attributes other than  
the weight.)

If it helps to answer, I am contemplating going outside R, and using  
the igraph library for Ruby (in alpha). Would graphs load  
significantly faster outside of R?

Thanks!

David Hunkins
address@hidden