Modified from those two repos as utils in Time2Graph.
Original README:
weighted random walk implementation for deepwalk (in python)
As deepwalk [1] can only generate sequence in a uniform distribution, weighted edge is not accepted.
In order to faciliate the usage, I implemented one alternative version with weighted edge. (just add one more function)
-
For most users, you can install original deepwalk package in your computer frist by:
pip install deepwalk
-
Then, download this package, and put the 'main.py' and 'weighted_random_walk.py' into your python lib depository directory for original deepwalk.
-
Now, you can run your deepwalk with:
deepwalk --input input.file --format weighted_edgelist --output output.file
[1] deepwalk: https://github.com/phanein/deepwalk
[2] weighed_deepwalk: https://github.com/shun1024/weighted-deepwalk