Random Graph Models Using Variational Autoencoders
Switch branches/tags
Clone or download
Bidisha Samanta
Bidisha Samanta Adding README.md
Latest commit b3bbb5d Jun 6, 2018
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
graphold local debugging commit Nov 13, 2017
README.md Adding README.md Jun 6, 2018
cell.py Clean code Feb 11, 2018
cell.pyc Clean code Feb 11, 2018
config.py Version 0 Nov 1, 2017
drawgraph.py Clean code Feb 11, 2018
generate_erdos_renyi.py Clean code Feb 11, 2018
layer.py Clean code Feb 11, 2018
main.py Clean code Feb 11, 2018
model.py Clean code Feb 11, 2018
sample.py Adding new file Feb 11, 2018
utils.py Clean code Feb 11, 2018

README.md

Designing Random Graph Models Using Variational Autoencoders With Applications to Chemical Design

Paper : https://arxiv.org/pdf/1802.05283.pdf

Required packages

-tensorflow 1.4.1

-rdkit >= 2016.03.4

-networkx 2.0

Command to learn model

python main.py --num_epochs 10 --learning_rate 0.0001 --log_every 5 --z_dim <z> --random_walk <k> --edges <e> --nodes <n> --graph_file <graph> --z_dir <zspace> --sample_file <sampledir> --out_dir <outputdir> > log.out

Command to sample graph

python sample.py --num_epochs 10 --learning_rate 0.0001 --log_every 5 --z_dim <z> --random_walk <k> --edges <e> --nodes <n> --graph_file <graph> --z_dir <zspace> --sample_file <sample> --out_dir <outputfile> > log.out

For real data please checkout the branch node_label