state_value_fn_from_action_value_fn

Functions

get_state_value_fn(→ Callable)

It is assumed that every action is possible in every state.

Module Contents

state_value_fn_from_action_value_fn.get_state_value_fn(action_value_fn: Callable, policy: gridmind.policies.base_policy.BasePolicy, actions: List) Callable[source]

It is assumed that every action is possible in every state. If not policy should return 0 action probility for that action.