You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Some algorithms may be different, which can make direct comparisons difficult to do well (for example, stopping criteria for e.g. PageRank could be different).
The text was updated successfully, but these errors were encountered:
for example, stopping criteria for e.g. PageRank could be different
Currently, igraph does not use power iteration at all for PageRank. Older versions did, but that code was removed, as it was slow, and did not support all features (such as personalization).
Mathematically, PageRank can be formulated either as a linear equation (assuming a damping factor less than 1), or as an eigenvector problem. At the moment, igraph has two methods: it can use the PRPACK library, which uses the linear equation picture, or it can use ARPACK to compute PageRank as an eigenvector.
We would like to be able to compare performance to
igraph
: https://github.com/igraph/python-igraphSome algorithms may be different, which can make direct comparisons difficult to do well (for example, stopping criteria for e.g. PageRank could be different).
The text was updated successfully, but these errors were encountered: