Difference between revisions of "Computer Science/61b"

From lensowiki
Jump to: navigation, search
(Gameboard & chips ADT: -bb – fix parameters for chip methods)
(Methods: -bb – updating in light of recent coding)
Line 47: Line 47:
 
  /* Andrew? */
 
  /* Andrew? */
 
  // generates an array of Gameboards of all the possible moves from the current Gameboard g for a given player. returns an array of Gameboards
 
  // generates an array of Gameboards of all the possible moves from the current Gameboard g for a given player. returns an array of Gameboards
  Gameboard[] generateMoves(Gameboard g, int color)
+
  Object[][] generateMoves(Gameboard g, int color)
 
----
 
----
  /* Paul? */
+
  /* Paul */
  // evaluates an array of gameboards using the minimax algorithm and alpha-beta pruning.
+
  // evaluates all possible moves to be generated from the '''this''' board using the minimax algorithm and alpha-beta pruning.
  // if necessary, generates additional move levels (i.e. if none of the gameboards in this array result in a win)
+
  // continues searching until it reaches '''depth''' or a win, whichever comes first
  Move evalTree(Gameboard[] arr)
+
  Move evalTree(int color, int depth)
 
----
 
----
  /* Jordan? */
+
  /* Jordan */
 
  // evaluates the given network for a winning color
 
  // evaluates the given network for a winning color
  // returns the winning color (as Chip.color), -10 if there is no winner
+
  // returns a probability of winning if there is no win
  // '''NOTE''': we need to figure out how to rate a board that doesn't win and we're already at the lowest level – some sort of probability technique, how many connections there are with the current chips, etc....
+
  // returns 999 if WHITE wins, -999 if BLACK wins
 
  int winner(Gameboard g)
 
  int winner(Gameboard g)

Revision as of 02:36, 29 October 2006

Project 2 modules

Members: Paul, -bb; Andrew, -fe; Jordan, -er

Gameboard & chips ADT

/* Paul can do this - it's a essentially a copy from the Oceans project, so it's quick */
// Gameboard abstract data type
// board field stores a board using an array of Chips (not that important – no one has access to this field anyway)
// methods are required to retrieve/store chips
Gameboard class
fields: private Chip[] board
methods:
// inserts a chip of the given color in the specified place as determined by an x, y coord pair
// returns true if the insertion succeeds; false otherwise
boolean insertChip(int color, int x, int y)
// move a given chip to a specified place as determined by an x, y coord pair
// returns true if the move succeeds; false otherwise
boolean moveChip(Chip c, int x, int y)
// performs a move and returns true/false depending on whether it succeeded
boolean performMove(int color, Move m)
// returns the Chip at a given location as determined by an x, y coord pair
Chip retrieveChip(int x, int y)
// returns true if the given move is valid; false otherwise
boolean validMove(Move m)
   // returns true if the chip cluster (if any) is small enough to permit a chip insertion at the given location as determined by an x, y coord pair; false otherwise
 ->could be inside or another method: boolean checkClusterSize(int x, int y)

// abstract data type to hold information about chips
// color field stores the chip's color as a static final int; x and y fields store the chip's x and y coords, respectively
Chip class
fields (all private): int color, int x, int y
	public static final WHITE/BLACK = 1/0
methods:
// returns the chip's color
int getColor()
// returns the x coord of the chip
int getX()
// returns the y coord of the chip
int getY()

Changes to MachinePlayer class

New fields

int color
int opponent
int depth
Gameboard board

Methods

/* Andrew? */
// generates an array of Gameboards of all the possible moves from the current Gameboard g for a given player. returns an array of Gameboards
Object[][] generateMoves(Gameboard g, int color)

/* Paul */
// evaluates all possible moves to be generated from the this board using the minimax algorithm and alpha-beta pruning.
// continues searching until it reaches depth or a win, whichever comes first
Move evalTree(int color, int depth)

/* Jordan */
// evaluates the given network for a winning color
// returns a probability of winning if there is no win
// returns 999 if WHITE wins, -999 if BLACK wins
int winner(Gameboard g)