Tamás Nepusz

The Trio loop was contributed to `urwid` by me; I guess that the `urwid` devs don't use Trio regularly. I have sent a PR months ago but it was not...

@SpotlightKid Did you have a chance to look at this PR? It would be great to have a new set of wheels on PyPI as Python 3.10 has been released;...

Thanks! In the meanwhile I tried the PR In my own fork and it seems like it still works nicely and generates correct wheels for Win32, macOS and Linux; see...

I repeated the microbenchmark proposed by @jessexknight with and without `igraph_options(return.vs.es=F)` to get a baseline comparison; here are the results: ``` > microbenchmark(times=100, + {k1 = unlist(adjacent_vertices(G,s))}, + {k2 =...

I've spent a bit more time debugging the vs/es creation process and commenting bits and pieces in and out to see how they would affect performance. It turns out that...

Also, do the expected vertex and edge counts match? I'm asking because it looks like your input file might have gaps in the IDs. igraph requires the numeric vertex IDs...

> I rebuild the python-igraph development version using igraph-4d2a03c30d303f6e6b9a64c7bf0c334b856c3c8c. This igraph version is work? Yes, that should work. I'm thinking about two possibilities but I don't know yet whether any...

That's true, but I cannot think of any other reason; that's why it would be good to see a stack trace.

Is the conversion possible now via `igraph` -> `tidygraph` -> `d3`?

_From @brooksambrose on May 24, 2015 22:46_ Ok, just a documentation issue, the argument is actually `vertex.label.cex` not `label.cex`. From the `igraph.plotting` help page: `Vertex parameters first, note that the...