This project uses 3 different search algorithms :

• The A* algorithm
• The Greedy Search algorithm
• The Uniform Cost algorithm

And 4 different heuristics :

• The manhanttan distance
• The Euclidian distance
• The diagonal distance
• The Hamming distance

These algorithms are very similar, using the same search function but different cost functions :

• A* uses h(x) + g(x)
• Greedy uses h(x)
• Uniform uses g(x)

With h(x) being an admissible heuristic function, and g(x) being the cost of the path from the start node.

Note that the program stops at the first valid solution and won't go for the shortest, as it is an NP-Hard problem.

Clone the repo, and simply run :

stack build
stack run

Make sure that you have installed The Haskell Tool Stack first.

The program takes a file as argument, being your puzzle, and optionally a custom search and custom heuristic, like so :

stack run path/to/your/file [optional custom search] [optional custom heuristic]

The input file should be as following :

3       # size of your puzzle
0 2 3   # the puzzle itself
1 4 5
8 7 6

You should get the following as output :

Solving grid using the A* algorithm and the Manhattan distance
0 2 3
1 4 5
8 7 6
-----
1 2 3
0 4 5
8 7 6
-----
1 2 3
8 4 5
0 7 6
-----
1 2 3
8 4 5
7 0 6
-----
1 2 3
8 4 5
7 6 0
-----
1 2 3
8 4 0
7 6 5
-----
1 2 3
8 0 4
7 6 5
-----
Solved with :
- Time complexity  : 6
- Space complexity : 3

• Add proper solvability check
• Add more heurstics
• Update data structures for better perfs
• Correct output to get the number of steps properly (current path is too large)
• Add number of moves needed from initial state to final state