seldonian.RL.Env_Description.Spaces.Continuous_Space

class Continuous_Space(bounds)

Bases: object

__init__(bounds)

Continuous space used for observations or actions

Parameters:

bounds (np.ndarray of shape (num_dims,2)) – Defines the min,max in each dimension of the space. For example, if the space is 3D with x in [-1.0,1.0], y in [5.1,5.4], and z in [-12.5,-6.8], then bounds would be: np.array([[-1.0,1.0],[5.1,5.4],[-12.5,-6.8]])

Variables:

num_dims – The number of dimensions in the space

__repr__()

Return repr(self).

Methods

check_bounds_valid(bounds)

Check that the bounds are in valid format and that in each dimension the bound maximum is larger than the bound minimum

Parameters:

bounds (np.ndarray of shape (num_dims,2)) – Defines the min,max in each dimension of the space. For example, if the space is 3D with x in [-1.0,1.0], y in [5.1,5.4], and z in [-12.5,-6.8], then bounds would be: np.array([[-1.0,1.0],[5.1,5.4],[-12.5,-6.8]])

get_num_dims()

Get the number of dimensions of the space