All these settings can be changed in settings.py
-
To play the game manually change the variable AGENT to False
-
To use an LRTA* agent to play the game set the variable AGENT to True
-
The agent is stopped after 1000 runs where the agent gets to the goal
-
While simulating u can use:
- Key P to pause/unpause
- Key R to remove/put the cap of 60 actions/sec (Default is on)
- Key D to activate/remove drawing of the enviorment (Default is on)
-
The variable NOISE can be changed from 0 to 1 to simulate an human error where its value represents the probability of a human error and a random action is chosen
-
The variable ACTIONS represent the actions that the agent is capable of doing:
-
0 - NOP
-
1 - RIGHT
-
2 - LEFT
-
3 - JUMP
-
4 - JUMP + RIGHT
-
5 - JUMP + LEFT
By default its [1,3,0]
-
-
-The map can be changed in the variable MAP and new maps can also be created:
-
'1' - Floor
-
'P' - Player (2 squares height)
-
'F' - Flag (8 squares height)
-
'2' - Pipe (2 squares height)
-
'.' - used to simulate empty
Maps can have any height and width but the reccomended height is 13 squares