/// <summary>
/// The main IA algorithm to get the best move to play. Inspired by the code in the AI course HE-Arc 2018-2019.
/// </summary>
/// <param name="node">The current node to search for</param>
/// <param name="depth">The max depth to search in</param>
/// <param name="minOrMax">Either -1 or 1 so we use the same code for both</param>
/// <param name="parentValue">The best value found at the moment</param>
/// <returns>2 values containing the move chosen and the value that it generates</returns>
private static Tuple <int, Move> Alphabeta2(MiniMaxTreeNode node, int depth, int minOrMax, int parentValue)
{
if (depth == 0 || node.IsLeaf())
{
return(new Tuple <int, Move>(node.Evaluate() + DEPTH_BONUS * depth, null));
}
int optVal = minOrMax == 1 ? int.MinValue : int.MaxValue;
Move optMove = null;
foreach (Move move in node.GetMoves())
{
MiniMaxTreeNode newNode = node.ApplyMove(move);
Tuple <int, Move> valMove = Alphabeta2(newNode, depth - 1, -minOrMax, optVal);
if (valMove.Item1 * minOrMax > optVal * minOrMax)
{
optVal = valMove.Item1;
optMove = move;
if (optVal * minOrMax > parentValue * minOrMax)
{
break;
}
}
}
return(new Tuple <int, Move>(optVal, optMove));
}
/// <summary>
/// Method called by anyone to calculate the best move to play.
/// </summary>
/// <param name="node">The current node to search for</param>
/// <param name="depth">The max depth to search in</param>
/// <param name="whitePlayer">Starting with the white player (true) or black(false)</param>
/// <returns></returns>
public static Move GetMove(int[,] board, int depth, bool whitePlayer)
{
MiniMaxTreeNode rootNode = new MiniMaxTreeNode(board, whitePlayer);
Tuple <int, Move> moveSelected = Alphabeta2(rootNode, depth, 1, int.MaxValue);
return(moveSelected.Item2);
}
