# Module: tf_agents.bandits.environments.bernoulli_action_mask_tf_environment

<br />

|---|
| [![](https://www.tensorflow.org/images/GitHub-Mark-32px.png) View source on GitHub](https://github.com/tensorflow/agents/blob/v0.19.0/tf_agents/bandits/environments/bernoulli_action_mask_tf_environment.py) |

Environment wrapper that adds action masks to a bandit environment.

This environment wrapper takes a `BanditTFEnvironment` as input, and generates
a new environment where the observations are joined with boolean action
masks. These masks describe which actions are allowed in a given time step. If a
disallowed action is chosen in a time step, the environment will raise an
error. The masks are drawn independently from Bernoulli-distributed random
variables with parameter `action_probability`.

The observations from the original environment and the mask are joined by the
given `join_fn` function, and the result of the join function will be the
observation in the new environment.

#### Usage:

'''
env = MyFavoriteBanditEnvironment(...)
def join_fn(context, mask):
return (context, mask)
masked_env = BernoulliActionMaskTFEnvironment(env, join_fn, 0.5)
'''

## Classes

[`class BernoulliActionMaskTFEnvironment`](https://www.tensorflow.org/agents/api_docs/python/tf_agents/bandits/environments/bernoulli_action_mask_tf_environment/BernoulliActionMaskTFEnvironment): An environment wrapper that adds action masks to observations.

<br />

<br />

<br />

| ## Other Members ||
|---|---|
| absolute_import | Instance of `__future__._Feature` |
| division | Instance of `__future__._Feature` |
| print_function | Instance of `__future__._Feature` |

<br />