In large-scale graph dataset, it's slow to load the real_imag for all nodes. To speed up the loading process, I think concurrent.futures.ThreadPoolExecutor could make a difference here. Any suggestions for this? Could you please approve the PR? Hope for your reply.
In large-scale graph dataset, it's slow to load the
real_imag
for all nodes. To speed up the loading process, I thinkconcurrent.futures.ThreadPoolExecutor
could make a difference here. Any suggestions for this? Could you please approve the PR? Hope for your reply.Bests,