zhangrui111/selfdefinedenvs

Some selfdefined environments for RL algorithms

Some selfdefined environments for RL algorithms.

1. gui_HellOrTreasure/* & gui_maze/*

These two environments are very similar in terms of implementation, i.e., most codes are exactly same.

I. About the map (json) file

An example of gui_HellOrTreasure/*

{
  "height": 6,
  "width": 6,
  "exit": [[1, 5]],
  "player": [[3, 0]],
  "hells": [[1, 2], [2, 2], [3, 2], [3, 5], [4, 5]]
}

• "height" & "width" must be int type.
• "exit", "player" & "hells"/"obstacles" must be list in list, even though only comprise one element.

II. Action space

Legal actions are of int type in [0, 3]

• 0 -- "move up"
• 1 -- "move down"
• 2 -- "move left"
• 3 -- "move right"

III. Its coordination style and two types of observations

1) Coordination style

Coordination provided in the map/json file is a matrix transpose of the real coordination in the maze. For example,

The real full observation corresponding to the above map/json file is as follows:

[[ 0.  0.  0.  1.  0.  0.]
 [ 0.  0.  0.  0.  0.  0.]
 [ 0. -1. -1. -1.  0.  0.]
 [ 0.  0.  0.  0.  0.  0.]
 [ 0.  0.  0.  0.  0.  0.]
 [ 0.  2.  0. -1. -1.  0.]]

2) gui_HellOrTreasure/* & gui_maze/* all support two types of observations.

• full observation -- The whole maze in the form of a matrix, e.g.,

[[ 0.  0.  0.  0.]
 [ 0.  1. -1.  0.]
 [ 0. -1.  2.  0.]
 [ 0.  0.  0.  0.]]

• real coordination -- The real coordination of a object in the form of a four-elements array. Now, this type of observation only support "the player" object. e.g.,

[45.0, 5.0, 75.0, 35.0]