`evorl.ec.evox_algorithm.openes`¶

Module Contents¶

Classes¶

OpenAI ES.

Functions¶

`compute_centered_ranks`
`compute_ranks`	Returns ranks in [0, len(x)-1].

API¶

class evorl.ec.evox_algorithm.openes.OpenES(center_init, pop_size, learning_rate, noise_std, optimizer=None, mirror_sampling=True)¶

Bases: evox.Algorithm

OpenAI ES.

Paper: Evolution Strategies as a Scalable Alternative to Reinforcement Learning.

ask(state)¶

setup(key)¶

tell(state, fitness)¶

evorl.ec.evox_algorithm.openes.compute_centered_ranks(x)¶

evorl.ec.evox_algorithm.openes.compute_ranks(x)¶

Returns ranks in [0, len(x)-1].

Note: This is different from scipy.stats.rankdata, which returns ranks in [1, len(x)].