Error: column index out of bounds

[jfredfan]

Hi - I installed ProNE (python 2.7) and was able to train embeddings for the PPI example with the command: python proNE.py -graph data/PPI.ungraph -emb1 emb/PPI_sparse.emb -emb2 emb/PPI_spectral.emb -dimension 128 -step 10 -theta 0.5 -mu 0.2

However, when applied on my own data it ran into the following exception. Appreciate any advice what went wrong. The input is about 50MB so I don't know how to share it for debugging. Thanks much in advance.

Traceback (most recent call last): File "proNE.py", line 178, in main() File "proNE.py", line 157, in main model = ProNE(args.graph, args.emb1, args.emb2, args.dimension) File "proNE.py", line 31, in init matrix0[e[0], e[1]] = 1 File "/home/m1234567/miniconda2/lib/python2.7/site-packages/scipy/sparse/lil.py", line 332, in setitem self.rows, self.data, i, j, x) File "_csparsetools.pyx", line 63, in scipy.sparse._csparsetools.lil_insert File "_csparsetools.pyx", line 89, in scipy.sparse._csparsetools.lil_insert IndexError: column index (17654) out of bounds

[jfredfan]

I think I figured out. When preparing the input node pairs, there is no need to perform dedup or remove self-link. Those preprocesses seem to cause problem instead.

[shuiyuejihua]

Node index is continuous without interruption.

[XquaKe]

Did you solve this problem?