Open garyhsu29 opened 1 year ago
I think the symmetrical problem could be solved by
editing https://github.com/VHRanger/CSRGraph/blob/db3caaa58fc84de36e4645e37b268e8f656f090a/csrgraph/methods.py#L33 to
sparse.csr_matrix((weights, dst, new_src), shape = (new_src.shape[0] - 1, new_src.shape[0] - 1))
Thanks for the great work on nodevectors & CSR graph! I encounter an issue while using nodevectors, but it turns out it's a problem on sparse matrix generating process. Error message: