Tamás Nepusz
Tamás Nepusz
All of these seem okay to me, no reason not to remove them.
Please go ahead if you have time to do so, the proposal sounds good to me. But I personally won't have time to do this myself in the near future.
The sampling of the upper triangle plus the diagonal for a matrix of size N x N can be done easily: pick a random integer _n_ between 0 and N...
I think it's a reasonable idea.
Can we move this back to 0.9.0?
The `DEGSEQ_VL` method is based on the original implementation of Viger and Latapy, and I presume that it follows [their paper](https://arxiv.org/pdf/cs/0502085.pdf). From the paper, it seems like they claim that...
> The issue is that a reasonable fix to these problems would involve an API change I am not particularly concerned about API changes at this stage; I don't know...
Yup, with the usual deprecation path (i.e. keep the original name as an alias).
`igraph_order_t` was added in lieu of a boolean in 0.10.0. It's also used that way in `igraph_sort_vertex_ids_by_degree()`. If we really want to push this consistency thing to the extreme, then...
Yes, that's correct.