Run with command:
python test.py
In file test.py uncomment the start_seq and goal_seq variables that correspond with the 15-puzzle.
Then comment out start_seq and goal_seq that correspond with 8-puzzle.
# Uncomment these variables for the 15 puzzle
# Also change ROWS and COLS values in file environment.py
# start_seq = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15]
# goal_seq = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15]
# 8 puzzle
start_seq = [0, 1, 2, 3, 4, 5, 6, 7, 8]
goal_seq = [0, 1, 2, 3, 4, 5, 6, 7, 8]
# Uncomment these variables for the 15 puzzle
# Also change ROWS and COLS values in file environment.py
start_seq = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15]
goal_seq = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15]
# 8 puzzle
# start_seq = [0, 1, 2, 3, 4, 5, 6, 7, 8]
# goal_seq = [0, 1, 2, 3, 4, 5, 6, 7, 8]
In the file environment.py change the values of COLS and ROWS to 4.
# Change ROWS and COLS values for N dimensions
COLS = 3
ROWS = 3
# Change ROWS and COLS values for N dimensions
COLS = 4
ROWS = 4
This program implements the A* Search, and Greedy-Best-First Search algorithms on an N-Puzzle (Only tested on 8 and 15 puzzles).
By running the file test.py you will be presented with the starting sequence, solution sequence, number of steps taken, and elapsed time of these algorithms being ran with 3 different heuristics:
h1 = Manhattan Distance
h2 = Number of displaced tiles
h3 = Combined heuristic, that returns minimum of h1, h2, and a heuristic that takes the average number of misplaced tiles per row.