concepts.benchmark.gridworld.minigrid.gym_minigrid.minigrid.Grid#
- class Grid[source]#
Bases:
object
Represent a grid and operations on it
Methods
copy
()decode
(array)Decode an array grid encoding back into a grid
encode
([vis_mask])Produce a compact numpy encoding of the grid
get
(i, j)horz_wall
(x, y[, length, obj_type])process_vis
(agent_pos)render
(tile_size[, agent_pos, agent_dir, ...])Render this grid at a given scale :param r: target renderer object :param tile_size: tile size in pixels
render_tile
(obj[, agent_dir, highlight, ...])Render a tile and cache the result
Rotate the grid to the left (counter-clockwise)
set
(i, j, v)slice
(topX, topY, width, height)Get a subset of the grid
vert_wall
(x, y[, length, obj_type])wall_rect
(x, y, w, h)Attributes
- __new__(**kwargs)#
- render(tile_size, agent_pos=None, agent_dir=None, highlight_mask=None)[source]#
Render this grid at a given scale :param r: target renderer object :param tile_size: tile size in pixels
- classmethod render_tile(obj, agent_dir=None, highlight=False, tile_size=TILE_PIXELS, subdivs=3)[source]#
Render a tile and cache the result
- tile_cache = {}#