Possibility of sending the incomplete adjacency matrix of a graph in test

lrjconan / GRAN

Efficient Graph Generation with Graph Recurrent Attention Networks, Deep Generative Model of Graphs, Graph Neural Networks, NeurIPS 2019

MIT License

469 stars 96 forks source link

Possibility of sending the incomplete adjacency matrix of a graph in test #5

Closed sbhttchryy closed 4 years ago

sbhttchryy commented 4 years ago

In gran_mixture_bernoulli.py A_pad = input_dict['adj'] if 'adj' in input_dict else None

My question is during testing, is it possible to send the the adjacency matrix of an incomplete graph as A_pad and see if it gets regenerated again? If yes, is there any thing that I should keep in mind?

lrjconan commented 4 years ago

Yes, it is doable to generate the graph conditioned on some existing partial adjacency matrix. I think you need to replace the all-zero pre-allocated the full adjacency matrix with your input adjacency matrix and slightly modify the sampling function.