Generating Networks

Code

After learning the parameters, you can generate a dynamic network by running generate_dynamic_graph.py directly on the command line:

python generate_dynamic_graph.py 'path/to/dataset_directory' 'num. timesteps'

or from your own Python code:

  from generate_dynamic_graph import dymond_generate

  dymond_generate(dataset_dir, num_timesteps)

Note: num_timesteps and 'num. timesteps' are the number of timesteps you want to generate

Output files

Generated graph will be saved to path/to/dataset_directory/learned_parameters/generated_graph/generated_graph.pklz