Anton Tsitsulin

The results are based on a random approximation, the default value of niters=100 gives reasonable default results for many datasets (as used in the paper). If you want more stable...

Do you mean that the metric has multi-mode distribution? If so, this is not expected, and we haven't observed this effect. If you can publish the arrays in question, I...

Can you link the original implementation you are mentioning explicitly? I have reimplemented the official Python code that used the Gensim library underneath. The subsampling actually comes from word2vec implementation...

Yeah, I tried to keep it close to Python implementation (to be honest, there is not much in Python code per se, I was just reimplementing Gensim). I believe this...

You will need to deal with multiprocessing slightly better than I do in the training loop. One option would be to just run the random walk generation and write to...

There is definetely a couple of weird things here: number of nodes is larger than the number of edges, and the graph seems almost disconnected. Are you sure you have...

If most of the nodes lie in small disconnected subgraphs, node2vec will be much faster (with unknown quality). From the output I would assume that process is ran correctly, and...

Interesting. Could you run the default couple of times and see if it crashes anytime?

If you post the graph file, I will try to look at the issue some time in the future. If you find a bug yourself, please notify/submit a PR.

Hi! It may be that there are more second-order degrees than the system can allocate. Can you make a quick check that deepwalk-c (from the other repository) runs on that...