To run the ABC experiment:
$ bash abc.sh

The printed output is a list of Python dictionaries, where each dictionary is a result for a particular horizon H, sample complexity N, algorithm, and random seed. For example:
{'H': 5, 'N': 500, 'Algorithm': 'VGB', 'avg_steps': 25.15, 'kl': 0.11721284974703168}

To visualize the results, the following Python code can be used (where "data" is the list of dictionaries):
>>> import vis_abc
>>> vis_abc.plot(data, 'kl', "Error (KL-divergence)")
>>> vis_abc.plot(data, 'avg_steps', "Average Steps")

