Tamás Nepusz

igraph's core is written in C, and eventually all `read_graph()` and `write_graph()` functions end up in C, where we need to pass a `FILE*` object to the underlying C code....

> Is it possible that the virus is spreading through the workspace file? I find that highly unlikely.

> What if the workspace contains code with a bug that results in arbitrary code execution Any other code on your machine (including igraph, the R language, the GUI of...

It is not the number of nodes that primarily matters when plotting a graph but the number of edges. A bipartite projection can easily have more edges than the initial...

The Intel MKL libraries would have no effect here. > Is there a way to speed up the plot? There are two significant time-consuming parts of a plotting call: the...

Also note that the behaviour that you will see at times of over-allocation is heavily system-dependent. For instance, Linux systems may overcommit memory allocations (i.e. pretend that it successfully managed...

Thanks a lot, I'll look into this! It looks like `python-igraph` is accidentally holding on to a reference of a Python object that should otherwise be freed.

So I've spent a bit of time debugging the problem, but I could not reproduce it so far. My setup was: * python-igraph `master` branch (but no related change since...

Sorry, still no luck -- I have now tried on a graph with 200K vertices and 800K edges, and the memory usage seems to fluctuate around 180MB, sometimes a bit...

I'm kinda stuck on this as I still haven't found a clear way to reproduce this on my machine yet. Moving to the next milestone.