Package kuimaze :: Module gym_wrapper :: Class EasyMazeEnv

[frames] | no frames]

Class EasyMazeEnv

gym.Env --+    
          |    
    MazeEnv --+
              |
             EasyMazeEnv

EasyMazeEnv is version of maze closer to graph search. It is possible to move agent from any state to different one already visited or neighbour state of current one. EasyMaze has all methods of HardMaze. Unlike the HardMaze, EasyMaze has additional method set_path - which can set different path than agent movement.

Instance Methods

__init__(self, informed, map_image_dir=None, grad=(0,0))
Class wrapping Maze into gym enviroment.

source code

step(self, action)

source code

set_path(self, path)
This method sets enviroment to visualize your found path.

source code

_is_available(self, new_state)
returns true if new state is available

source code

_easy_result(self, state_list)
Gives result of desired action in parameter

source code

_get_cost(self, curr, last)
returns cost of movement from last to curr

source code

expand(self, position)
returns tuple of positions with associated costs that can be visited from "position"

source code

Inherited from MazeEnv: close, get_all_states, render, reset, save_eps, save_path, seed, visualise

Inherited from MazeEnv (private): _get_action_space, _get_depth, _get_observation, _get_reward

Class Variables

Inherited from MazeEnv: MAP, metadata

Inherited from MazeEnv (private): _path, _visited

Method Details

init(self, informed, map_image_dir=None, grad=(0,0))
(Constructor)

Class wrapping Maze into gym enviroment.

Parameters:

informed - boolean
gym_compatible - boolean - T = HardMaze, F = EasyMaze
deter - boolean - T = deterministic maze, F = probabilistic maze
map_image_dir - string - path to image of map
grad - tuple - vector tuning the tilt of maze`

Overrides: MazeEnv.__init__

(inherited documentation)

step(self, action)

Overrides: MazeEnv.step

set_path(self, path)

This method sets enviroment to visualize your found path. Method render, must be called afterwards.

Parameters:

path - list of lists in format: [[x1, y1], [x2, y2], ... ]

Returns:

None

_is_available(self, new_state)

returns true if new state is available

Parameters:

new_state

Returns:

boolean

_easy_result(self, state_list)

Gives result of desired action in parameter

Parameters:

state_list - list or tuple of coordinates [x, y]

Returns:

state - new position of agent

_get_cost(self, curr, last)

returns cost of movement from last to curr

Parameters:

curr - new state
last - last state

Returns:

float

expand(self, position)

returns tuple of positions with associated costs that can be visited from "position"

Parameters:

position - position in the maze defined by coordinates (x,y)

Returns:

tuple of coordinates [x, y] with "cost" for movement to these positions: [[[x1, y1], cost1], [[x2, y2], cost2], ... ]