On Thu, Jul 12, 2012 at 10:46 AM, Tamás Nepusz
<address@hidden> wrote:
> because the problem seems related to the way igraph processes the data internally, i'm not sure about what to do.
> it seems reasonable to think I should reduce the size of the network, but how much?
Let's clarify the issue first. Are you using read.graph(filename, format="edgelist") or something else? How does your file look like? Is it only two columns with some numbers? Are your vertex indices (i.e. the numbers) continuous from 0 to N-1, where N is the number of vertices in the graph? If not, have you tried using read.graph(filename, format="ncol") instead? ncol is basically a symbolic edge list where vertices are assigned IDs from 0 to N-1 in the order of appearance and the numbers in the file (that are used to identify the vertices) are stored in the "name" vertex attribute.
Best,
T.