Rebase newer branch

18 files changed, 2084 insertions, 0 deletions

diff --git a/python/GameState.py b/python/GameState.py
new file mode 100644
index 0000000..eed8f36
--- /dev/null
+++ b/python/GameState.py
@@ -0,0 +1,144 @@
+from random import choice
+
+# Based on https://github.com/DieterBuys/mcts-player/
+
+class GameState(object):
+
+    def __init__(self):
+        self.next_turn_player = 1
+        self.player = None
+
+    @property
+    def game_result(self):
+        return None
+
+    def get_moves(self):
+        return set()
+
+    def get_random_move(self):
+        moves = self.get_moves()
+        return choice(tuple(moves)) if moves != set() else None
+
+    def play_move(self, move):
+        pass
+
+
+class DotsAndBoxesState(GameState):
+    def __init__(self, nb_rows, nb_cols, player):
+        super(DotsAndBoxesState, self).__init__()
+
+        self.nb_rows = nb_rows
+        self.nb_cols = nb_cols
+        rows = []
+        for ri in range(nb_rows + 1):
+            columns = []
+            for ci in range(nb_cols + 1):
+                columns.append({"v": 0, "h": 0})
+            rows.append(columns)
+        self.board = rows
+
+        self.score = {1: 0, 2: 0}
+        self.player = player
+        print("Player: ", player)
+
+    @property
+    def game_result(self):
+        def game_decided(nb_cols, nb_rows, scoreP, scoreO):
+            # the game is decided if the winner is already known even before the game is ended
+            # you're guaranteed to win the game if you have more than halve of the total points that can be earned
+            total_points = nb_rows * nb_cols
+            if scoreP > total_points // 2 or scoreO > total_points // 2:
+                return True
+            else:
+                return False
+
+        # check if the board is full, then decide based on score
+        free_lines = self.get_moves()
+        player = self.player
+        opponent = self.player % 2 + 1
+
+        if not game_decided(self.nb_cols, self.nb_rows, self.score[player], self.score[opponent]) and len(free_lines) > 0:
+            return None
+        elif self.score[player] > self.score[opponent]:
+            return 1
+        elif self.score[player] < self.score[opponent]:
+            return 0
+        else:
+            return 0.5
+
+    def get_moves(self):
+        free_lines = []
+        for ri in range(len(self.board)):
+            row = self.board[ri]
+            for ci in range(len(row)):
+                cell = row[ci]
+                if ri < (len(self.board) - 1) and cell["v"] == 0:
+                    free_lines.append((ri, ci, "v"))
+                if ci < (len(row) - 1) and cell["h"] == 0:
+                    free_lines.append((ri, ci, "h"))
+        return set(free_lines)
+
+    def play_move(self, move):
+        r, c, o = move
+        assert move in self.get_moves()
+
+        # check if this move makes a box
+        makes_box = False
+        if o == "h":
+            if r - 1 >= 0:
+                # check above
+                if self.board[r-1][c]["h"] != 0 and self.board[r-1][c]["v"] != 0 and self.board[r-1][c+1]["v"] != 0:
+                    makes_box = True
+                    self.score[self.next_turn_player] += 1
+            if r + 1 <= self.nb_rows:
+                # check below
+                if self.board[r+1][c]["h"] != 0 and self.board[r][c]["v"] != 0 and self.board[r][c+1]["v"] != 0:
+                    makes_box = True
+                    self.score[self.next_turn_player] += 1
+
+        elif o == "v":
+            if c - 1 >= 0:
+                # check left
+                if self.board[r][c-1]["v"] != 0 and self.board[r][c-1]["h"] != 0 and self.board[r+1][c-1]["h"] != 0:
+                    makes_box = True
+                    self.score[self.next_turn_player] += 1
+
+            if c + 1 <= self.nb_cols:
+                # check right
+                if self.board[r][c+1]["v"] != 0 and self.board[r][c]["h"] != 0 and self.board[r+1][c]["h"] != 0:
+                    makes_box = True
+                    self.score[self.next_turn_player] += 1
+
+
+        # register move
+        self.board[r][c][o] = self.next_turn_player
+
+        if not makes_box:
+            # switch turns
+            self.next_turn_player = self.next_turn_player % 2 + 1
+
+    def __repr__(self):
+        str = ""
+        for r in range(self.nb_rows + 1):
+            for o in ["h", "v"]:
+                for c in range(self.nb_cols + 1):
+                    if o == "h":
+                        str += "."
+                        if c != self.nb_cols:
+                            if self.board[r][c][o] == 0:
+                                str += "  "
+                            else:
+                                str += "__"
+                        else:
+                            str += "\n"
+                    elif o == "v":
+                        if r != self.nb_rows:
+                            if self.board[r][c][o] == 0:
+                                str += " "
+                            else:
+                                str += "|"
+                        if c != self.nb_cols:
+                            str += "  "
+                        else:
+                            str += "\n"
+        return str
diff --git a/python/MCTS.py b/python/MCTS.py
new file mode 100644
index 0000000..a65e2d4
--- /dev/null
+++ b/python/MCTS.py
@@ -0,0 +1,151 @@
+import math
+from copy import deepcopy
+from time import clock
+from random import choice
+
+from GameState import GameState
+
+# Based on https://github.com/DieterBuys/mcts-player/
+
+class GameController(object):
+    def get_next_move(self, state):
+        # when you get a new move, it is assumed that the game is not ended yet
+        assert state.get_moves()
+
+
+class MCTSNode(object):
+    """Monte Carlo Tree Node.
+    Each node encapsulates a particular game state, the moves that
+    are possible from that state and the strategic information accumulated
+    by the tree search as it progressively samples the game space.
+    """
+
+    def __init__(self, state, parent=None, move=None):
+        self.parent = parent
+        self.move = move
+        self.state = state
+
+        self.plays = 0
+        self.score = 0
+
+        self.pending_moves = state.get_moves()
+        self.children = []
+
+    def select_child_ucb(self):
+        # Note that each node's plays count is equal
+        # to the sum of its children's plays
+        def ucb(child):
+            win_ratio = child.score / child.plays \
+                        + math.sqrt(2 * math.log(self.plays) / child.plays)
+            return win_ratio
+
+        return max(self.children, key=ucb)
+
+    def expand_move(self, move):
+        self.pending_moves.remove(move)  # raises KeyError
+
+        child_state = deepcopy(self.state)
+        child_state.play_move(move)
+
+        child = MCTSNode(state=child_state, parent=self, move=move)
+        self.children.append(child)
+        return child
+
+    def get_score(self, result):
+        # return result
+        if result == 0.5:
+            return result
+
+        if self.state.player == 2:
+            if self.state.next_turn_player == result:
+                return 0.0
+            else:
+                return 1.0
+        else:
+            if self.state.next_turn_player == result:
+                return 1.0
+            else:
+                return 0.0
+
+        if self.state.next_turn_player == result:
+            return 0.0
+        else:
+            return 1.0
+
+    def __repr__(self):
+        s = 'ROOT\n' if self.parent is None else ''
+
+        children_moves = [c.move for c in self.children]
+
+        s += """Score ratio: {score} / {plays}
+Pending moves: {pending_moves}
+Children's moves: {children_moves}
+State:
+{state}\n""".format(children_moves=children_moves, **self.__dict__)
+
+        return s
+
+
+class MCTSGameController(GameController):
+    """Game controller that uses MCTS to determine the next move.
+    This is the class which implements the Monte Carlo Tree Search algorithm.
+    It builds a game tree of MCTSNodes and samples the game space until a set
+    time has elapsed.
+    """
+
+    def select(self):
+        node = self.root_node
+
+        # Descend until we find a node that has pending moves, or is terminal
+        while node.pending_moves == set() and node.children != []:
+            node = node.select_child_ucb()
+
+        return node
+
+    def expand(self, node):
+        assert node.pending_moves != set()
+
+        move = choice(tuple(node.pending_moves))
+        return node.expand_move(move)
+
+    def simulate(self, state, max_iterations=1000):
+        state = deepcopy(state)
+
+        move = state.get_random_move()
+        while move is not None:
+            state.play_move(move)
+            move = state.get_random_move()
+
+            max_iterations -= 1
+            if max_iterations <= 0:
+                return 0.5  # raise exception? (game too deep to simulate)
+
+        return state.game_result
+
+    def update(self, node, result):
+        while node is not None:
+            node.plays += 1
+            node.score += node.get_score(result)
+            node = node.parent
+
+    def get_next_move(self, state, time_allowed=1.0):
+        super(MCTSGameController, self).get_next_move(state)
+
+        # Create new tree (TODO: Preserve some state for better performance?)
+        self.root_node = MCTSNode(state)
+        iterations = 0
+
+        start_time = clock()
+        while clock() < start_time + time_allowed:
+            node = self.select()
+
+            if node.pending_moves != set():
+                node = self.expand(node)
+
+            result = self.simulate(node.state)
+            self.update(node, result)
+
+            iterations += 1
+
+        # Return most visited node's move
+        return max(self.root_node.children, key=lambda n: n.plays).move
diff --git a/python/agent.py b/python/agent.py
new file mode 100644
index 0000000..49bc1cc
--- /dev/null
+++ b/python/agent.py
@@ -0,0 +1,55 @@
+import dotsandboxes.dotsandboxesagent as dba
+
+import sys
+import argparse
+import logging
+from GameState import GameState, DotsAndBoxesState
+from MCTS import MCTSNode, MCTSGameController
+
+
+logger = logging.getLogger(__name__)
+games = {}
+agentclass = None
+
+
+class Agent(dba.DotsAndBoxesAgent):
+    def __init__(self, player, nb_rows, nb_cols, timelimit):
+        super(Agent, self).__init__(player, nb_rows, nb_cols, timelimit)
+        self.GameStateClass = DotsAndBoxesState
+        self.game_state = self.GameStateClass(nb_rows, nb_cols, player)
+        self.controller = MCTSGameController()
+
+    def register_action(self, row, column, orientation, player):
+        super(Agent, self).register_action(row, column, orientation, player)
+        # adjust agent specific board representation
+        move = (row, column, orientation)
+        self.game_state.play_move(move)
+
+    def next_action(self):
+        r, c, o = self.controller.get_next_move(self.game_state, time_allowed=self.timelimit)
+        return r, c, o
+
+    def end_game(self):
+        super(Agent, self).end_game()
+
+
+# Adapted from provided code
+def main(argv=None):
+    global agentclass
+    parser = argparse.ArgumentParser(description='Start agent to play Dots and Boxes')
+    parser.add_argument('--verbose', '-v', action='count', default=0, help='Verbose output')
+    parser.add_argument('--quiet', '-q', action='count', default=0, help='Quiet output')
+    parser.add_argument('port', metavar='PORT', type=int, help='Port to use for server')
+    args = parser.parse_args(argv)
+
+    logger.setLevel(max(logging.INFO - 10 * (args.verbose - args.quiet), logging.DEBUG))
+    logger.addHandler(logging.StreamHandler(sys.stdout))
+
+    agentclass = Agent
+    dba.agentclass = Agent
+    dba.start_server(args.port)
+
+
+if __name__ == "__main__":
+    sys.exit(main())
+
diff --git a/python/alphaBeta.py b/python/alphaBeta.py
new file mode 100644
index 0000000..01f82cf
--- /dev/null
+++ b/python/alphaBeta.py
@@ -0,0 +1,30 @@
+
+def alpha_beta(node, alpha, beta):
+	
+	# Based on https://en.wikipedia.org/wiki/Alpha%E2%80%93beta_pruning#Pseudocode
+	# node needs to support three operations: isTerminal(), value(), getChildren(), maximizingPlayer()
+	
+	if node.isTerminal():
+		return node.value()
+	
+	if node.maximizingPlayer():
+		
+		v = float("-inf")
+		for child in node.getChildren():
+			
+			v = max(v, alpha_beta(child, alpha, beta))
+			alpha = max(alpha, v)
+			if beta <= alpha:
+				break
+		
+	else:
+		
+		v = float("inf")
+		for child in node.getChildren():
+			
+			v = min(v, alpha_beta(child, alpha, beta))
+			beta = min(beta, v)
+			if beta <= alpha:
+				break
+	
+	return v
diff --git a/python/ann.py b/python/ann.py
new file mode 100644
index 0000000..05ae647
--- /dev/null
+++ b/python/ann.py
@@ -0,0 +1,170 @@
+from numpy import *
+from math import sqrt
+from copy import deepcopy
+from time import time
+
+class ANN:
+	
+	"""ANN with one hidden layer, one output and full connections in between consecutive layers.
+	Initial weights are chosen from a normal distribution.
+	Activation function is tanh."""
+	
+	INIT_SIGMA = 0.02
+	REL_STOP_MARGIN = 0.01
+	MAX_ITERATIONS = 1000000
+	ACTIVATION = tanh
+	D_ACTIVATION = lambda x: 1 - tanh(x)**2 # Derivative of tanh
+	VEC_ACTIVATION = vectorize(ACTIVATION)
+	VEC_D_ACTIVATION = vectorize(D_ACTIVATION)
+	STEP_SIZE = 0.1
+	
+	def __init__(self, input_size, hidden_size):
+		
+		#self.input_size = input_size
+		#self.hidden_size = hidden_size
+		self.hidden_weights = random.normal(0, ANN.INIT_SIGMA, (hidden_size, input_size))
+		self.output_weights = random.normal(0, ANN.INIT_SIGMA, hidden_size)
+	
+	def get_weights(self):
+		return self.hidden_weights, self.output_weights
+	
+	def predict(self, input_vector):
+		
+		# Predicts the output for this input vector
+		# input_vector will be normalized
+		
+		input_vector = input_vector/linalg.norm(input_vector)
+		return ANN.ACTIVATION(dot(self.output_weights, ANN.VEC_ACTIVATION(dot(self.hidden_weights, input_vector))))
+	
+	@staticmethod
+	def frob_norm(a, b):
+		
+		# Calculates the total Frobenius norm of both matrices A and B
+		return sqrt(linalg.norm(a)**2 + linalg.norm(b)**2)
+	
+	def train(self, examples):
+		
+		#print("Training")
+		start = time()
+		
+		# examples is a list of (input, output)-tuples
+		# input will be normalized
+		# We stop when the weights have converged within some relative margin
+		
+		for example in examples:
+			example[0] = example[0]/linalg.norm(example[0])
+		
+		iteration = 0
+		while True:
+			
+			
+			# Store old weights to check for convergence later
+			prev_hidden_weights = deepcopy(self.hidden_weights)
+			prev_output_weights = deepcopy(self.output_weights)
+			
+			for k in range(len(examples)):
+				
+				input_vector, output = examples[k]
+				
+				# Calculate outputs
+				hidden_input = dot(self.hidden_weights, input_vector)
+				hidden_output = ANN.VEC_ACTIVATION(hidden_input)
+				final_input = dot(self.output_weights, hidden_output)
+				predicted_output = ANN.ACTIVATION(final_input)
+				
+				#print("Output:", output)
+				#print("Predicted output:", predicted_output)
+				
+				# Used in calculations
+				prediction_error = output - predicted_output
+				output_derivative = ANN.D_ACTIVATION(final_input)
+				
+				# Adjust output weights and calculate requested hidden change
+				requested_hidden_change = prediction_error*output_derivative*self.output_weights
+				self.output_weights = self.output_weights + ANN.STEP_SIZE*prediction_error*hidden_output
+				
+				#print("After adjusting output weights:", ANN.ACTIVATION(dot(self.output_weights, hidden_output)))
+				
+				# Backpropagate requested hidden change to adjust hidden weights
+				self.hidden_weights = self.hidden_weights + ANN.STEP_SIZE*outer(requested_hidden_change*(ANN.VEC_D_ACTIVATION(hidden_input)), input_vector)
+				
+				#print("After adjusting hidden weights:", ANN.ACTIVATION(dot(self.output_weights, ANN.VEC_ACTIVATION(dot(self.hidden_weights, input_vector)))))
+				
+				# Check stop criteria
+				iteration += 1
+				if iteration >= ANN.MAX_ITERATIONS:
+					break
+			
+			# Check stop criteria
+			if iteration >= ANN.MAX_ITERATIONS:
+				break
+			diff = ANN.frob_norm(self.hidden_weights - prev_hidden_weights, self.output_weights - prev_output_weights)
+			base = ANN.frob_norm(self.hidden_weights, self.output_weights)
+			#if base > 0 and diff/base < ANN.REL_STOP_MARGIN:
+			#	break
+		
+		print(time() - start)
+		print("Stopped training after %s iterations."%iteration)
+
+# TESTING
+
+def print_difference(ann1, ann2):
+	
+	# Prints the differences in weights in between two ANN's with identical topology
+	
+	hidden_weights1, output_weights1 = ann1.get_weights()
+	hidden_weights2, output_weights2 = ann2.get_weights()
+	hidden_diff = hidden_weights1 - hidden_weights2
+	output_diff = output_weights1 - output_weights2
+	
+	print(hidden_diff)
+	print(output_diff)
+	print("Frobenius norms:")
+	print("Hidden weights difference:", linalg.norm(hidden_diff))
+	print("Output weights difference:", linalg.norm(output_diff))
+	print("Both:", ANN.frob_norm(hidden_diff, output_diff))
+
+def RMSE(ann, examples):
+	
+	total = 0
+	for input_vector, output in examples:
+		total += (output - ann.predict(input_vector))**2
+	return sqrt(total/len(examples))
+
+def generate_examples(amount, input_size, evaluate):
+	# evaluate is a function mapping an input vector onto a numerical value
+	examples = []
+	inputs = random.normal(0, 100, (amount, input_size))
+	for i in range(amount):
+		input_vector = inputs[i]
+		examples.append([input_vector, evaluate(input_vector)])
+	return examples
+
+def test():
+	
+	# Test the ANN by having it model another ANN with identical topology but unknown weights
+	
+	input_size = 5
+	hidden_size = 3
+	real = ANN(input_size, hidden_size)
+	model = ANN(input_size, hidden_size)
+	
+	# Generate training data
+	training_data = generate_examples(10000, input_size, real.predict)
+	validation_data = generate_examples(10000, input_size, real.predict)
+	
+	# Print initial difference, train, then print new difference
+	print("Initial difference:")
+	print_difference(real, model)
+	print("Initial RMSE (on training data):", RMSE(model, training_data))
+	print("Initial RMSE (on validation data):", RMSE(model, validation_data))
+	model.train(training_data)
+	print("After training:")
+	print_difference(real, model)
+	print("After training RMSE (on training data):", RMSE(model, training_data))
+	print("After training RMSE (on validation data):", RMSE(model, validation_data))
+
+if __name__ == "__main__":
+	test()
+
+
diff --git a/python/board.py b/python/board.py
new file mode 100644
index 0000000..36cfe8c
--- /dev/null
+++ b/python/board.py
@@ -0,0 +1,110 @@
+from GameState import GameState
+
+class Board(GameState):
+	
+	# Interface methods
+	
+	def __init__(self, nb_rows, nb_cols, player):
+		
+		# Basic initialization
+		self.nb_rows = nb_rows
+		self.nb_cols = nb_cols
+		self.player = player # 1 or 2
+		self.scores = [0, 0, 0]
+		self.max_score = nb_rows*nb_cols
+		
+		# Construct edges and nodes matrix
+		self.edges = [] # Boolean, true means uncut
+		self.nodes = [] # Represents valence of nodes from strings-and-coins
+		for x in range(self.nb_cols):
+			self.edges.append([True]*self.nb_rows)
+			self.edges.append([True]*(self.nb_rows + 1))
+			self.nodes.append([4]*self.nb_rows)
+		self.edges.append([True]*self.nb_rows)
+		
+		# Construct all possible moves
+		self.moves_left = set() # Moves are represented as ints
+		for x in range(len(self.edges)):
+			for y in range(len(self.edges[x])):
+				self.moves_left.add(self.coords_to_edge(x, y))
+		
+		# TODO: chain updating
+		# Initialize chains
+		# Chains are represented as lists of nodes
+		self.closed_chains = [] # Chains which start and end in nodes of valence
+
+	@property
+	def game_result(self):
+		
+		own_score = self.scores[self.player]
+		opponent_score = self.scores[self.player % 2 + 1]
+		if len(self.moves_left) == 0:
+			diff = own_score - opponent_score
+			if diff > 0:
+				return 1
+			elif diff < 0:
+				return 0
+			else:
+				return 0.5
+		else:
+			# Check if one player already has at least half of all points
+			if own_score > self.max_score//2:
+				return 1
+			elif opponent_score > self.max_score//2:
+				return 0
+			else:
+				return None
+
+	def get_moves(self):
+		return self.moves_left
+
+	def get_random_move(self):
+		moves = self.get_moves()
+		return choice(tuple(moves)) if moves != set() else None
+
+	def play_move(self, move):
+		x, y = self.edge_to_coords(move)
+		self.edges[x][y] = False
+		self.moves_left.remove(move)
+		
+		# Update valence
+		if x%2 == 0:
+			# Horizontal edge, decrease valence of left and right nodes
+			for node_x in x//2 - 1, x//2:
+				if node_x >= 0 and node_x < nb_cols:
+					self.nodes[node_x][y] -= 1
+		else:
+			# Vertical edge, decrease valence of top and bottom nodes
+			for node_y in y - 1, y:
+				if node_y >= 0 and node_y < nb_rows:
+					self.nodes[x//2][node_y] -= 1
+		
+		# TODO: chain updating
+	
+	# Own methods
+	
+	def undo_move(self, move):
+		x, y = self.edge_to_coords(move)
+		self.edges[x][y] = True
+		self.moves_left.add(move)
+		
+		# Update valence
+		if x%2 == 0:
+			# Horizontal edge, decrease valence of left and right nodes
+			for node_x in x//2 - 1, x//2:
+				if node_x >= 0 and node_x < nb_cols:
+					self.nodes[node_x][y] += 1
+		else:
+			# Vertical edge, decrease valence of top and bottom nodes
+			for node_y in y - 1, y:
+				if node_y >= 0 and node_y < nb_rows:
+					self.nodes[x//2][node_y] += 1
+		
+		# TODO: chain updating
+	
+	def coords_to_edge(self, x, y):
+		return x*(self.nb_cols + 1) + y
+	
+	def edge_to_coords(self, move):
+		return move//(self.nb_cols + 1), move%(self.nb_cols + 1)
+	
diff --git a/python/dataStructures.py b/python/dataStructures.py
new file mode 100644
index 0000000..1e972fc
--- /dev/null
+++ b/python/dataStructures.py
@@ -0,0 +1,32 @@
+class DisjointSet:
+    def __init__(self, nbElements):
+        self.parent = []
+        self.rank = []
+        for i in range(0, nbElements):
+            self.parent[i] = i
+            self.rank[i] = 0
+
+    def find(self, x):
+        if self.parent[x] != x:
+            self.parent[x] = self.find(self.parent[x])
+        return self.parent[x]
+
+    def union(self, x, y):
+        xRoot = self.find(x)
+        yRoot = self.find(y)
+
+        # x and y already belong to the same set
+        if xRoot == yRoot:
+            return
+
+        # merge the set of x and y
+        if self.rank[xRoot] < self.rank[yRoot]:
+            self.parent[xRoot] = yRoot
+        elif self.rank[xRoot] > self.rank[yRoot]:
+            self.parent[yRoot] = xRoot
+        else:
+            self.parent[xRoot] = yRoot
+            self.rank[yRoot] += 1
+
+    def inSameSet(self, x, y):
+        return self.find(x) == self.find(y)
\ No newline at end of file
diff --git a/python/dotsandboxes/README.md b/python/dotsandboxes/README.md
new file mode 100644
index 0000000..e3f844c
--- /dev/null
+++ b/python/dotsandboxes/README.md
@@ -0,0 +1,134 @@
+Dots and Boxes application
+==========================
+
+Live demo: https://people.cs.kuleuven.be/wannes.meert/dotsandboxes/play
+
+![Screenshot of Dots and Boxes](https://people.cs.kuleuven.be/wannes.meert/dotsandboxes/screenshot.png?v=2)
+
+This setup is part of the course "Machine Learning: Project" (KU Leuven,
+Faculty of engineering, Department of Computer Science,
+[DTAI research group](https://dtai.cs.kuleuven.be)).
+
+
+Installation
+------------
+
+The example agent is designed for Python 3.6 and requires the
+[websockets](https://websockets.readthedocs.io) package. Dependencies can be
+installed using pip:
+
+    $ pip install -r requirements.txt
+
+
+Start the game GUI
+------------------
+
+This program shows a web-based GUI to play the Dots and Boxes
+game. This supports human-human, agent-human and agent-agent combinations.
+It is a simple Javascript based application that runs entirely in the browser.
+You can start it by opening the file `static/dotsandboxes.html` in a browser.
+Or alternatively, you can start the app using the included simple server:
+
+    $ ./dotsandboxesserver.py 8080
+
+The game can then be played by directing your browser to http://127.0.0.1:8080.
+
+
+Start the agent client
+----------------------
+
+This is the program that runs a game-playing agent. This application listens
+to [websocket](https://developer.mozilla.org/en-US/docs/Web/API/WebSockets_API)
+requests that communicate game information and sends back the next action it
+wants to play.
+
+Starting the agent client is done using the following command:
+
+    $ ./dotsandboxesagent <port>
+
+This starts a websocket on the given port that can receveive JSON messages.
+
+The JSON messages given below should be handled by your agent.
+Take into account the maximal time allowed to reply.
+
+### Initiate the game
+
+Both players get a message that a new game has started:
+
+    {
+        "type": "start",
+        "player": 1,
+        "timelimit", 0.5,
+        "grid": [5, 5],
+        "game": "123456"
+    }
+
+where `player` is the number assigned to this agent, `timelimit` is the
+time in seconds in which you need to send your action back to the server,
+and `grid` is the grid size in rows and columns.
+
+If you are player 1, reply with the first action you want to perform:
+
+    {
+        "type": "action",
+        "location": [1, 1],
+        "orientation": "v"
+    }
+
+The field `location` is expressed as row and column (zero-based numbering) and
+`orientation` is either "v" (vertical) or "h" (horizontal).
+
+
+### Action in the game
+
+When an action is played, the message sent to both players is:
+
+    {
+        "type": "action",
+        "game": "123456",
+        "player": 1,
+        "nextplayer": 2,
+        "score": [0, 0],
+        "location": [1, 1],
+        "orientation": "v"
+    }
+
+
+If it is your turn you should answer with a message that states your next
+move:
+
+    {
+        "type": "action",
+        "location": [1, 1],
+        "orientation": "v"
+    }
+
+
+### Game end
+
+When the game ends after an action, the message is slightly altered:
+
+    {
+        "type": "end",
+        "game": "123456",
+        "player": 1,
+        "nextplayer": 0,
+        "score": [3, 1],
+        "location": [1, 1],
+        "orientation": "v",
+        "winner": 1
+    }
+
+The `type` field becomes `end` and a new field `winner` is set to the player
+that has won the game.
+
+
+Contact information
+-------------------
+
+- Wannes Meert,        https://people.cs.kuleuven.be/wannes.meert
+- Hendrik Blockeel,    https://people.cs.kuleuven.be/hendrik.blockeel
+- Arne De Brabandere,  https://people.cs.kuleuven.be/arne.debrabandere
+- Sebastijan Dumančić, https://people.cs.kuleuven.be/sebastijan.dumancic
+- Pieter Robberechts,  https://people.cs.kuleuven.be/pieter.robberechts
+
diff --git a/python/dotsandboxes/dotsandboxesagent b/python/dotsandboxes/dotsandboxesagent
new file mode 100755
index 0000000..eecf719
--- /dev/null
+++ b/python/dotsandboxes/dotsandboxesagent
@@ -0,0 +1,5 @@
+#!/bin/bash
+# It is not necessary to use a shell script for this. Dropping the .py
+# extension and including the correct shebang is also correct.
+python3 $(dirname "$0")/dotsandboxesagent.py $@
+
diff --git a/python/dotsandboxes/dotsandboxesagent.py b/python/dotsandboxes/dotsandboxesagent.py
new file mode 100644
index 0000000..c8bc05e
--- /dev/null
+++ b/python/dotsandboxes/dotsandboxesagent.py
@@ -0,0 +1,213 @@
+#!/usr/bin/env python3
+# encoding: utf-8
+"""
+dotsandboxesagent.py
+
+Template for the Machine Learning Project course at KU Leuven (2017-2018)
+of Hendrik Blockeel and Wannes Meert.
+
+Copyright (c) 2018 KU Leuven. All rights reserved.
+"""
+import sys
+import argparse
+import logging
+import asyncio
+import websockets
+import json
+from collections import defaultdict
+import random
+
+
+logger = logging.getLogger(__name__)
+games = {}
+agentclass = None
+
+
+class DotsAndBoxesAgent:
+    """Example Dots and Boxes agent implementation base class.
+    It returns a random next move.
+
+    A DotsAndBoxesAgent object should implement the following methods:
+    - __init__
+    - add_player
+    - register_action
+    - next_action
+    - end_game
+
+    This class does not necessarily use the best data structures for the
+    approach you want to use.
+    """
+    def __init__(self, player, nb_rows, nb_cols, timelimit):
+        """Create Dots and Boxes agent.
+
+        :param player: Player number, 1 or 2
+        :param nb_rows: Rows in grid
+        :param nb_cols: Columns in grid
+        :param timelimit: Maximum time allowed to send a next action.
+        """
+        self.player = {player}
+        self.timelimit = timelimit
+        self.ended = False
+        self.nb_rows = nb_rows
+        self.nb_cols = nb_cols
+        rows = []
+        for ri in range(nb_rows + 1):
+            columns = []
+            for ci in range(nb_cols + 1):
+                columns.append({"v": 0, "h": 0})
+            rows.append(columns)
+        self.cells = rows
+
+    def add_player(self, player):
+        """Use the same agent for multiple players."""
+        self.player.add(player)
+
+    def register_action(self, row, column, orientation, player):
+        """Register action played in game.
+
+        :param row:
+        :param columns:
+        :param orientation: "v" or "h"
+        :param player: 1 or 2
+        """
+        self.cells[row][column][orientation] = player
+
+    def next_action(self):
+        """Return the next action this agent wants to perform.
+
+        In this example, the function implements a random move. Replace this
+        function with your own approach.
+
+        :return: (row, column, orientation)
+        """
+        logger.info("Computing next move (grid={}x{}, player={})"\
+                .format(self.nb_rows, self.nb_cols, self.player))
+        # Random move
+        free_lines = []
+        for ri in range(len(self.cells)):
+            row = self.cells[ri]
+            for ci in range(len(row)):
+                cell = row[ci]
+                if ri < (len(self.cells) - 1) and cell["v"] == 0:
+                    free_lines.append((ri, ci, "v"))
+                if ci < (len(row) - 1) and cell["h"] == 0:
+                    free_lines.append((ri, ci, "h"))
+        if len(free_lines) == 0:
+            # Board full
+            return None
+        movei = random.randint(0, len(free_lines) - 1)
+        r, c, o = free_lines[movei]
+        return r, c, o
+
+    def end_game(self):
+        self.ended = True
+
+
+## MAIN EVENT LOOP
+
+async def handler(websocket, path):
+    logger.info("Start listening")
+    game = None
+    # msg = await websocket.recv()
+    try:
+        async for msg in websocket:
+            logger.info("< {}".format(msg))
+            try:
+                msg = json.loads(msg)
+            except json.decoder.JSONDecodeError as err:
+                logger.error(err)
+                return False
+            game = msg["game"]
+            answer = None
+            if msg["type"] == "start":
+                # Initialize game
+                if msg["game"] in games:
+                    games[msg["game"]].add_player(msg["player"])
+                else:
+                    nb_rows, nb_cols = msg["grid"]
+                    games[msg["game"]] = agentclass(msg["player"],
+                                                    nb_rows,
+                                                    nb_cols,
+                                                    msg["timelimit"])
+                if msg["player"] == 1:
+                    # Start the game
+                    nm = games[game].next_action()
+                    print('nm = {}'.format(nm))
+                    if nm is None:
+                        # Game over
+                        logger.info("Game over")
+                        continue
+                    r, c, o = nm
+                    answer = {
+                        'type': 'action',
+                        'location': [r, c],
+                        'orientation': o
+                    }
+                else:
+                    # Wait for the opponent
+                    answer = None
+
+            elif msg["type"] == "action":
+                # An action has been played
+                r, c = msg["location"]
+                o = msg["orientation"]
+                games[game].register_action(r, c, o, msg["player"])
+                if msg["nextplayer"] in games[game].player:
+                    # Compute your move
+                    nm = games[game].next_action()
+                    if nm is None:
+                        # Game over
+                        logger.info("Game over")
+                        continue
+                    nr, nc, no = nm
+                    answer = {
+                        'type': 'action',
+                        'location': [nr, nc],
+                        'orientation': no
+                    }
+                else:
+                    answer = None
+
+            elif msg["type"] == "end":
+                # End the game
+                games[msg["game"]].end_game()
+                answer = None
+            else:
+                logger.error("Unknown message type:\n{}".format(msg))
+
+            if answer is not None:
+                print(answer)
+                await websocket.send(json.dumps(answer))
+                logger.info("> {}".format(answer))
+    except websockets.exceptions.ConnectionClosed as err:
+        logger.info("Connection closed")
+    logger.info("Exit handler")
+
+
+def start_server(port):
+    server = websockets.serve(handler, 'localhost', port)
+    print("Running on ws://127.0.0.1:{}".format(port))
+    asyncio.get_event_loop().run_until_complete(server)
+    asyncio.get_event_loop().run_forever()
+
+
+## COMMAND LINE INTERFACE
+
+def main(argv=None):
+    global agentclass
+    parser = argparse.ArgumentParser(description='Start agent to play Dots and Boxes')
+    parser.add_argument('--verbose', '-v', action='count', default=0, help='Verbose output')
+    parser.add_argument('--quiet', '-q', action='count', default=0, help='Quiet output')
+    parser.add_argument('port', metavar='PORT', type=int, help='Port to use for server')
+    args = parser.parse_args(argv)
+
+    logger.setLevel(max(logging.INFO - 10 * (args.verbose - args.quiet), logging.DEBUG))
+    logger.addHandler(logging.StreamHandler(sys.stdout))
+
+    agentclass = DotsAndBoxesAgent
+    start_server(args.port)
+
+
+if __name__ == "__main__":
+    sys.exit(main())
+
diff --git a/python/dotsandboxes/dotsandboxescompete.py b/python/dotsandboxes/dotsandboxescompete.py
new file mode 100644
index 0000000..ee2aee8
--- /dev/null
+++ b/python/dotsandboxes/dotsandboxescompete.py
@@ -0,0 +1,212 @@
+#!/usr/bin/env python3
+# encoding: utf-8
+"""
+dotsandboxescompete.py
+
+Template for the Machine Learning Project course at KU Leuven (2017-2018)
+of Hendrik Blockeel and Wannes Meert.
+
+Copyright (c) 2018 KU Leuven. All rights reserved.
+"""
+
+import sys
+import argparse
+import logging
+import asyncio
+import websockets
+import json
+from collections import defaultdict
+import random
+import uuid
+import time
+
+logger = logging.getLogger(__name__)
+
+
+def start_competition(address1, address2, nb_rows, nb_cols, timelimit):
+   asyncio.get_event_loop().run_until_complete(connect_agent(address1, address2, nb_rows, nb_cols, timelimit))
+
+
+async def connect_agent(uri1, uri2, nb_rows, nb_cols, timelimit):
+    cur_game = str(uuid.uuid4())
+    winner = None
+    cells = []
+    cur_player = 1
+    points = [0, 0, 0]
+    timings = [None, [], []]
+
+    for ri in range(nb_rows + 1):
+        columns = []
+        for ci in range(nb_cols + 1):
+            columns.append({"v":0, "h":0, "p":0})
+        cells.append(columns)
+
+    logger.info("Connecting to {}".format(uri1))
+    async with websockets.connect(uri1) as websocket1:
+        logger.info("Connecting to {}".format(uri2))
+        async with websockets.connect(uri2) as websocket2:
+            logger.info("Connected")
+
+            # Start game
+            msg = {
+              "type": "start",
+              "player": 1,
+              "timelimit": timelimit,
+              "game": cur_game,
+              "grid": [nb_rows, nb_cols]
+            }
+            await websocket1.send(json.dumps(msg))
+            msg["player"] = 2
+            await websocket2.send(json.dumps(msg))
+
+            # Run game
+            while winner is None:
+                ask_time = time.time()
+                logger.info("Waiting for player {}".format(cur_player))
+                if cur_player == 1:
+                    msg = await websocket1.recv()
+                else:
+                    msg = await websocket2.recv()
+                recv_time = time.time()
+                diff_time = recv_time - ask_time
+                timings[cur_player].append(diff_time)
+                logger.info("Message received after (s): {}".format(diff_time))
+                try:
+                    msg = json.loads(msg)
+                except json.decoder.JSONDecodeError as err:
+                    logger.debug(err)
+                    continue
+                if msg["type"] != "action":
+                    logger.error("Unknown message: {}".format(msg))
+                    continue
+                r, c = msg["location"]
+                o = msg["orientation"]
+                next_player = user_action(r, c, o, cur_player,
+                                          cells, points,
+                                          nb_rows, nb_cols)
+                if points[1] + points[2] == nb_cols * nb_rows:
+                    # Game over
+                    winner = 1
+                    if points[2] == points[1]:
+                        winner = 0
+                    if points[2] > points[1]:
+                        winner = 2
+                else:
+                    msg = {
+                        "type": "action",
+                        "game": cur_game,
+                        "player": cur_player,
+                        "nextplayer": next_player,
+                        "score": [points[1], points[2]],
+                        "location": [r, c],
+                        "orientation": o
+                    }
+                    await websocket1.send(json.dumps(msg))
+                    await websocket2.send(json.dumps(msg))
+
+                cur_player = next_player
+
+            # End game
+            logger.info("Game ended: points1={} - points2={} - winner={}".format(points[1], points[2], winner))
+            msg = {
+                "type": "end",
+                "game": cur_game,
+                "player": cur_player,
+                "nextplayer": 0,
+                "score": [points[1], points[2]],
+                "location": [r, c],
+                "orientation": o,
+                "winner": winner
+            }
+            await websocket1.send(json.dumps(msg))
+            await websocket2.send(json.dumps(msg))
+
+    # Timings
+    for i in [1, 2]:
+        logger.info("Timings: player={} - avg={} - min={} - max={}"\
+            .format(i,
+                    sum(timings[i])/len(timings[i]),
+                    min(timings[i]),
+                    max(timings[i])))
+
+    logger.info("Closed connections")
+
+
+def user_action(r, c, o, cur_player, cells, points, nb_rows, nb_cols):
+    logger.info("User action: player={} - r={} - c={} - o={}".format(cur_player, r, c, o))
+    next_player = cur_player
+    won_cell = False
+    cell = cells[r][c]
+    if o == "h":
+        if cell["h"] != 0:
+            return cur_player
+        cell["h"] = cur_player
+        # Above
+        if r > 0:
+            if cells[r - 1][c]["v"] != 0 \
+                and cells[r - 1][c + 1]["v"] != 0 \
+                and cells[r - 1][c]["h"] != 0 \
+                and cells[r][c]["h"] != 0:
+                won_cell = True
+                points[cur_player] += 1
+                cells[r - 1][c]["p"] = cur_player
+        # Below
+        if r < nb_rows:
+            if cells[r][c]["v"] != 0 \
+                and cells[r][c + 1]["v"] != 0 \
+                and cells[r][c]["h"] != 0 \
+                and cells[r + 1][c]["h"] != 0:
+                won_cell = True
+                points[cur_player] += 1
+                cells[r][c]["p"] = cur_player
+
+    if o == "v":
+        if cell["v"] != 0:
+            return cur_player
+        cell["v"] = cur_player;
+        # Left
+        if c > 0:
+            if cells[r][c - 1]["v"] != 0 \
+                and cells[r][c]["v"] != 0 \
+                and cells[r][c - 1]["h"] != 0 \
+                and cells[r + 1][c - 1]["h"] != 0:
+                won_cell = True
+                points[cur_player] += 1
+                cells[r][c - 1]["p"] = cur_player
+        # Right
+        if c < nb_cols:
+            if cells[r][c]["v"] != 0 \
+                and cells[r][c + 1]["v"] != 0 \
+                and cells[r][c]["h"] != 0 \
+                and cells[r + 1][c]["h"] != 0:
+                won_cell = True
+                points[cur_player] += 1
+                cells[r][c]["p"] = cur_player
+
+    if not won_cell:
+        next_player = 3 - cur_player
+    else:
+        next_player = cur_player
+        print("Update points: player1={} - player2={}".format(points[1], points[2]))
+    return next_player
+
+
+def main(argv=None):
+    parser = argparse.ArgumentParser(description='Start agent to play Dots and Boxes')
+    parser.add_argument('--verbose', '-v', action='count', default=0, help='Verbose output')
+    parser.add_argument('--quiet', '-q', action='count', default=0, help='Quiet output')
+    parser.add_argument('--cols', '-c', type=int, default=2, help='Number of columns')
+    parser.add_argument('--rows', '-r', type=int, default=2, help='Number of rows')
+    parser.add_argument('--timelimit', '-t', type=float, default=0.5, help='Time limit per request in seconds')
+    parser.add_argument('agents', nargs=2, metavar='AGENT', help='Websockets addresses for agents')
+    args = parser.parse_args(argv)
+
+    logger.setLevel(max(logging.INFO - 10 * (args.verbose - args.quiet), logging.DEBUG))
+    logger.addHandler(logging.StreamHandler(sys.stdout))
+
+    start_competition(args.agents[0], args.agents[1], args.rows, args.cols, args.timelimit)
+
+
+if __name__ == "__main__":
+    sys.exit(main())
+
diff --git a/python/dotsandboxes/dotsandboxesserver.py b/python/dotsandboxes/dotsandboxesserver.py
new file mode 100644
index 0000000..1b66372
--- /dev/null
+++ b/python/dotsandboxes/dotsandboxesserver.py
@@ -0,0 +1,60 @@
+#!/usr/bin/env python3
+# encoding: utf-8
+"""
+dotsandboxesserver.py
+
+Template for the Machine Learning Project course at KU Leuven (2017-2018)
+of Hendrik Blockeel and Wannes Meert.
+
+Copyright (c) 2018 KU Leuven. All rights reserved.
+"""
+
+import sys
+import argparse
+import logging
+import http.server
+import socketserver
+import json
+
+logger = logging.getLogger(__name__)
+
+
+class RequestHandler(http.server.SimpleHTTPRequestHandler):
+    def do_GET(self):
+        if self.path == "/":
+            self.send_response(302)
+            self.send_header("Location", "static/dotsandboxes.html")
+            self.end_headers()
+        return super().do_GET()
+
+    def do_PUT(self):
+        response = {
+            'result': 'ok'
+        }
+        self.send_response(200)
+        self.send_header('Content-type', 'application/json')
+        self.end_headers()
+        self.wfile.write(json.dumps(response).encode())
+
+
+def start_server(port):
+    with socketserver.TCPServer(("", port), RequestHandler) as httpd:
+        print("Running on http://127.0.0.1:{}".format(port))
+        httpd.serve_forever()
+
+
+def main(argv=None):
+    parser = argparse.ArgumentParser(description='Start server to play Dots and Boxes')
+    parser.add_argument('--verbose', '-v', action='count', default=0, help='Verbose output')
+    parser.add_argument('--quiet', '-q', action='count', default=0, help='Quiet output')
+    parser.add_argument('port', metavar='PORT', type=int, help='Port to use for server')
+    args = parser.parse_args(argv)
+
+    logger.setLevel(max(logging.INFO - 10 * (args.verbose - args.quiet), logging.DEBUG))
+    logger.addHandler(logging.StreamHandler(sys.stdout))
+
+    start_server(args.port)
+
+
+if __name__ == "__main__":
+    sys.exit(main())
diff --git a/python/dotsandboxes/requirements.txt b/python/dotsandboxes/requirements.txt
new file mode 100644
index 0000000..14774b4
--- /dev/null
+++ b/python/dotsandboxes/requirements.txt
@@ -0,0 +1 @@
+websockets
diff --git a/python/dotsandboxes/static/dotsandboxes.css b/python/dotsandboxes/static/dotsandboxes.css
new file mode 100644
index 0000000..71b1d3b
--- /dev/null
+++ b/python/dotsandboxes/static/dotsandboxes.css
@@ -0,0 +1,10 @@
+
+.footer {
+  color: #B3B3B3;
+  margin-bottom: 1ex;
+}
+
+.footer a {
+  color: #87A0B3;
+}
+
diff --git a/python/dotsandboxes/static/dotsandboxes.html b/python/dotsandboxes/static/dotsandboxes.html
new file mode 100644
index 0000000..ecbcbb4
--- /dev/null
+++ b/python/dotsandboxes/static/dotsandboxes.html
@@ -0,0 +1,56 @@
+<!DOCTYPE html>
+<html>
+<html lang="en">
+<meta charset="utf-8">
+<meta name="author" content="Wannes Meert">
+<meta name="description" content="Dots-and-Boxes game. Part of the Machine Learning: Project course at KU Leuven (Hendrik Blockeel, Wannes Meert).">
+<meta name="keywords" content="artificial intelligence,AI,machine learning,dots and boxes,KU Leuven">
+<meta name="viewport" content="width=device-width, initial-scale=1, shrink-to-fit=no">
+<title>Dots and Boxes</title>
+<link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/bootstrap/4.0.0/css/bootstrap.min.css" integrity="sha384-Gn5384xqQ1aoWXA+058RXPxPg6fy4IWvTNh0E263XmFcJlSAwiGgFAW/dAiS6JXm" crossorigin="anonymous">
+<link rel="stylesheet" href="dotsandboxes.css">
+</head>
+<body>
+	<div class="container">
+		<h1>Dots and Boxes</h1>
+		<div class="row">
+			<div class="col-md">
+				<div id="playing-area"></div>
+			</div>
+			<div class="col-md">
+				<div class="form-group">
+				<p>Size of game:</p>
+				<div class="input-group">
+				<div class="input-group-prepend">
+				<span class="input-group-text">Rows and Columns</span>
+				</div>
+				<input type="number" class="form-control" id="nb-rows" value=6>
+				<input type="number" class="form-control" id="nb-cols" value=6>
+				</div>
+				</div>
+				<div class="form-group">
+				<p>Players:</p>
+				<div class="input-group mb-3">
+				<div class="input-group-prepend"><span class="input-group-text" id="basic-addon3">Agent 1</span></div>
+				<input type="text" class="form-control" id="agent1" aria-describedby="basic-addon3">
+				</div>
+				<div class="input-group mb-3">
+				<div class="input-group-prepend"><span class="input-group-text" id="basic-addon3">Agent 2</span></div>
+				<input type="text" class="form-control" id="agent2" aria-describedby="basic-addon3">
+				</div>
+				<p>Fill in the address where an agent can be reached using WebSockets (e.g. ws://127.0.0.1:8089).
+				If a field is empty a human player is assumed.
+				</p>
+				<button type="button" class="btn btn-secondary" id="restart-btn">Restart game</button>
+				</div>
+			</div>
+		</div>
+		<div class="footer">
+		<small>&copy; <a href="https://dtai.cs.kuleuven.be">DTAI Research Group</a>, KU Leuven &mdash; <a href="https://github.com/wannesm/dotsandboxes">Source</a></small>
+		</div>
+	</div>
+	<script src="https://d3js.org/d3.v4.min.js"></script>
+	<script src="dotsandboxes.js"></script>
+</body>
+</html>
+
diff --git a/python/dotsandboxes/static/dotsandboxes.js b/python/dotsandboxes/static/dotsandboxes.js
new file mode 100644
index 0000000..11e9447
--- /dev/null
+++ b/python/dotsandboxes/static/dotsandboxes.js
@@ -0,0 +1,454 @@
+/**
+ * dotsandboxes.js
+ *
+ * Template for the Machine Learning Project course at KU Leuven (2017-2018)
+ * of Hendrik Blockeel and Wannes Meert.
+ *
+ * Copyright (c) 2018 KU Leuven. All rights reserved.
+ **/
+
+function generateGuid() {
+  var result, i, j;
+  result = '';
+  for(j=0; j<32; j++) {
+    if( j == 8 || j == 12|| j == 16|| j == 20) 
+      result = result + '-';
+    i = Math.floor(Math.random()*16).toString(16).toUpperCase();
+    result = result + i;
+  }
+  return result;
+}
+
+// GAME LOGIC
+
+var cur_game = generateGuid();
+var cur_player = 1;
+var cur_ended = false;
+var points = [0, 0, 0];
+var timelimit = 0.5;
+var nb_cols = 6;
+var nb_rows = 6;
+var data = new Array(0);
+
+function restart_game() {
+  //console.log("Restarting game");
+  cur_game = generateGuid();
+  nb_cols = parseInt(document.getElementById('nb-cols').value);
+  if (nb_cols == "" || isNaN(nb_cols)) {
+    nb_cols = 6;
+  }
+  nb_rows = parseInt(document.getElementById('nb-rows').value);
+  if (nb_rows == "" || isNaN(nb_rows)) {
+    nb_rows = 6;
+  }
+  cur_ended = false;
+  console.log("Starting game", cur_game);
+  points = [0, 0, 0];
+  cur_player = 1;
+  var old_length = 0;
+  for (var ri=0; ri<nb_rows + 1; ri++) {
+    if (ri >= data.length) {
+      data.push(new Array(0));
+    }
+    var row = data[ri];
+    for (var ci=0; ci<nb_cols + 1; ci++) {
+      if (ci >= row.length) {
+        row.push({l:0, t:0, p:0, r:0, c:0});
+      }
+      var l = 0;
+      var t = 0;
+      var p = 0;
+      if (ri == nb_rows) {
+        l = undefined;
+        p = undefined;
+      }
+      if (ci == nb_cols) {
+        t = undefined;
+        p = undefined
+      }
+      var cell = row[ci];
+      cell.l = l;
+      cell.t = t;
+      cell.p = p;
+      cell.r = ri;
+      cell.c = ci;
+    }
+    old_length = row.length;
+    for (var ci=nb_cols + 1; ci<old_length; ci++) {
+      row.pop();
+    }
+  }
+  old_length = data.length;
+  for (var ri=nb_rows + 1; ri<old_length; ri++) {
+    data.pop();
+  }
+}
+
+function user_click(cell, o) {
+  if (cur_ended) {
+    //console.log('Game ended, ignoring click');
+    return;
+  }
+  console.log('User click', cell, o);
+  var won_cell = false;
+  var c = cell.c;
+  var r = cell.r;
+  var msg = {
+    type: "action",
+    game: cur_game,
+    player: cur_player,
+    nextplayer: cur_player,
+    score: [points[1], points[2]],
+    location: [r, c],
+    orientation: o
+  };
+  if (o == "h") {
+    if (cell.t != 0) {
+      return;
+    }
+    cell.t = cur_player;
+    // Above
+    if (r > 0) {
+      if (data[r - 1][c].l != 0
+        && data[r - 1][c + 1].l != 0
+        && data[r - 1][c].t != 0
+        && data[r][c].t != 0) {
+        won_cell = true;
+        points[cur_player] += 1;
+        data[r - 1][c].p = cur_player;
+      }
+    }
+    // Below
+    if (r < nb_rows) {
+      if (data[r][c].l != 0
+        && data[r][c + 1].l != 0
+        && data[r][c].t != 0
+        && data[r + 1][c].t != 0) {
+        won_cell = true;
+        points[cur_player] += 1;
+        data[r][c].p = cur_player;
+      }
+    }
+  }
+
+  if (o == "v") {
+    if (cell.l != 0) {
+      return;
+    }
+    cell.l = cur_player;
+    // Left
+    if (c > 0) {
+      if (data[r][c - 1].l != 0
+        && data[r][c].l != 0
+        && data[r][c - 1].t != 0
+        && data[r + 1][c - 1].t != 0) {
+        won_cell = true;
+        points[cur_player] += 1;
+        data[r][c - 1].p = cur_player;
+      }
+    }
+    // Right
+    if (c < nb_cols) {
+      if (data[r][c].l != 0
+        && data[r][c + 1].l != 0
+        && data[r][c].t != 0
+        && data[r + 1][c].t != 0) {
+        won_cell = true;
+        points[cur_player] += 1;
+        data[r][c].p = cur_player;
+      }
+    }
+  }
+
+  msg["score"] = [points[1], points[2]];
+
+  if (!won_cell) {
+    cur_player = 3 - cur_player;
+    msg.nextplayer = cur_player;
+  }
+  update_board();
+  if (points[1] + points[2] == nb_cols * nb_rows) {
+    // Game over
+    var winner = 1
+    if (points[2] == points[1]) {
+      winner = 0;
+    }
+    if (points[2] > points[1]) {
+      winner = 2;
+    }
+    cur_ended = true;
+    msg.type = "end";
+    msg.nextplayer = 0;
+    msg.winner = winner;
+  }
+  send_to_agents(msg);
+}
+
+var field_margin = 10;
+var cell_width = 40;
+var cell_margin = 4;
+var player_height = 40;
+var width = 400;
+var height = 600;
+var line_width = 5;
+
+var player_color = [
+  "#E6E6E6",
+  "#FC6666",
+  "#0F80FF"
+];
+
+var svg = d3.select("#playing-area").append("svg")
+  .attr("width", width)
+  .attr("height", height)
+  .append("g")
+  .attr("transform", "translate("+field_margin+","+field_margin+")");
+
+var player = svg.append("g")
+  .attr("class", "player")
+  .attr("transform", "translate(0,10)");
+
+var field = svg.append("g")
+  .attr("class", "field")
+  .attr("transform", "translate(0,"+player_height+")");
+
+
+function update_board() {
+  // PLAYERS - enter & update
+  var player_text = player.selectAll("text")
+    .data([cur_player, cur_player]);
+
+  player_text = player_text.enter().append("text")
+    .attr("x", function(c, i) { return i * 100;})
+    .merge(player_text)
+      .text(function(c, i) {return "Player " + (i + 1) + ": "+points[i + 1];})
+      .attr("fill", function(c, i) {
+        if (c == i + 1) {
+          return player_color[c];
+        } else {
+          return player_color[0];
+        }
+      });
+
+  // ROWS - enter & update
+  var rows = field.selectAll(".row")
+    .data(data)
+      .attr("fill", function() {return null;});
+
+  rows.exit().remove();
+
+  rows = rows.enter().append("g")
+      .attr("class", "row")
+      .attr("transform", function(row, i) {return "translate(0," + cell_width * i + ")";})
+    .merge(rows);
+
+  // COLS - enter & update
+  var cols = rows.selectAll(".col")
+    .data(function(col) {return col;});
+
+  cols.exit().remove();
+
+  var cols_enter = cols.enter().append("g")
+      .attr("class", "col")
+      .attr("transform", function(col, ri) {return "translate("+cell_width * ri+",0)";});
+
+  // CELL - enter
+  cols_enter.append("rect")
+    .attr("class", "cell")
+    .attr("rx", cell_margin)
+    .attr("ry", cell_margin)
+    .attr("opacity", 0.25)
+    .attr("x", cell_margin)
+    .attr("y", cell_margin)
+    .attr("width", cell_width - 2*cell_margin)
+    .attr("height", cell_width - 2*cell_margin);
+
+  // HLINE - enter
+  cols_enter.append("line")
+    .attr("class", "hline")
+    .attr("x1", function(cell, ci) {return cell_margin;})
+    .attr("x2", function(cell, ci) {return cell_width - cell_margin;})
+    .attr("y1", 0)
+    .attr("y2", 0)
+    .attr("stroke-linecap", "round")
+    .attr("stroke", function(cell) {return player_color[cell.t];});
+
+  cols_enter.append("path")
+    .attr("d", "M"+cell_margin+",0"+
+               "L"+(cell_width/2)+",-"+(cell_width/3)+
+               "L"+(cell_width-cell_margin)+",0"+
+               "L"+(cell_width/2)+","+(cell_width/3)+"Z")
+    .attr("stroke", "black")
+    .attr("stroke-width", 2)
+    .attr("opacity", "0")
+    .on("click", function(cell) {
+      if (agents[cur_player].active == true) {
+        console.log("Ignoring click, automated agent")
+      } else {
+        user_click(cell, "h");
+      }
+    });
+
+  // VLINE - enter
+  cols_enter.append("line")
+    .attr("class", "vline")
+    .attr("y1", function(cell, ci) {return cell_margin;})
+    .attr("y2", function(cell, ci) {return cell_width - cell_margin;})
+    .attr("x1", 0)
+    .attr("x2", 0)
+    .attr("stroke-linecap", "round")
+    .attr("stroke", function(cell) {return player_color[cell.l];});
+
+  cols_enter.append("path")
+    .attr("d", "M0,"+cell_margin+
+               "L-"+(cell_width/3)+","+(cell_width/2)+
+               "L0,"+(cell_width-cell_margin)+
+               "L"+(cell_width/3)+","+(cell_width/2)+"Z")
+    .attr("stroke", "black")
+    .attr("stroke-width", 2)
+    .attr("opacity", "0")
+    .on("click", function(cell) {
+      if (agents[cur_player].active == true) {
+        console.log("Ignoring click, automated agent");
+      } else {
+        user_click(cell, "v");
+      }
+    });
+
+  cols = cols_enter
+    .merge(cols);
+
+  // HLINE - update
+  cols.selectAll(".hline")
+    .attr("stroke-width", function(cell) {
+      if (typeof(cell.t) == "undefined") {
+        return 0;
+      }
+      return line_width;
+    })
+    .attr("stroke", function(cell) {return player_color[cell.t];});
+
+  // VLINE - update
+  cols.selectAll(".vline")
+    .attr("stroke-width", function(cell, ci) {
+      if (typeof(cell.l) == "undefined") {
+        return 0;
+      }
+      return line_width;
+    })
+    .attr("stroke", function(cell) {return player_color[cell.l];});
+
+  // CELL - update
+  cols.selectAll(".cell")
+    .attr("fill", function(cell) {
+      if (cell.p == undefined) {
+        return "white";
+      }
+      return player_color[cell.p];
+    });
+}
+
+
+// AGENT CONNECTIONS
+
+var agents = [
+  {},
+  {address: undefined, active: false, socket: undefined},
+  {address: undefined, active: false, socket: undefined}
+];
+
+var msg_queue = [];
+
+
+function start_connections() {
+  for (var i=1; i<3; i++) {
+    agents[i] = {address:undefined, active: false, socket: undefined};
+    var address = document.getElementById('agent'+i).value;
+    if (address != "") {
+      //console.log("Starting websocket for agent "+i+" on address "+address);
+      var agent = agents[i];
+      agent.address = address;
+      agent.socket = new WebSocket(address);
+      agent.socket.onopen = (function (ii, iagent) { return function(event) {
+        console.log("Agent "+ii+" connected")
+        iagent.active = true;
+        iagent.socket.onmessage = function(event) {
+          var msg = JSON.parse(event.data);
+          //console.log("Get msg from agent "+ii, msg);
+          if (msg.type == "action") {
+            if (cur_player == ii) {
+              console.log("Received action from ACTIVE player "+ii, msg);
+              user_click(data[msg.location[0]][msg.location[1]], msg.orientation);
+            } else {
+              console.log("Received action from NON-ACTIVE player "+ii, msg);
+            }
+          }
+          return false;
+        };
+        iagent.socket.onclose = function(event) {
+          console.log("Closing connection to agent "+ii);
+        };
+        iagent.socket.onerror = function(event) {
+          console.log("Error on connection to agent "+ii, event);
+        };
+        msg = {
+          "type": "start",
+          "player": ii,
+          "timelimit": timelimit,
+          "game": cur_game,
+          "grid": [nb_rows, nb_cols]
+        };
+        iagent.socket.send(JSON.stringify(msg));
+      };}(i, agent));
+    }
+  }
+}
+
+
+function send_to_agents(msg) {
+  msg_queue.push(JSON.stringify(msg));
+  try_sending_to_agents();
+}
+
+
+function try_sending_to_agents() {
+  var all_connected = true;
+  for (var i=1; i<3; i++) {
+    if (agents[i].address !== undefined && agents[i].active == false) {
+      all_connected = false;
+      break;
+    }
+  }
+  if (!all_connected) {
+    // Wait until all are connected
+    setTimeout(try_sending_to_agents, 100);
+  } else {
+    if (msg_queue.length == 0 ) {
+      return;
+    }
+    var msg = msg_queue.shift();
+    console.log("Send msg to agents", msg);
+    for (var i=1; i<3; i++) {
+      if (agents[i].active == true) {
+        agents[i].socket.send(msg);
+      }
+    }
+  }
+}
+
+
+// STARTUP
+
+function restart() {
+  restart_game();
+  update_board();
+  start_connections();
+}
+
+var restartbtn = document.getElementById("restart-btn");
+restartbtn.onclick = function() {
+  console.log("Restart game");
+  restart();
+};
+
+restart();
diff --git a/python/evaluate.py b/python/evaluate.py
new file mode 100644
index 0000000..fb60211
--- /dev/null
+++ b/python/evaluate.py
@@ -0,0 +1,240 @@
+#!/usr/bin/env python3
+# encoding: utf-8
+"""
+dotsandboxescompete.py
+
+Template for the Machine Learning Project course at KU Leuven (2017-2018)
+of Hendrik Blockeel and Wannes Meert.
+
+Copyright (c) 2018 KU Leuven. All rights reserved.
+"""
+
+import sys
+import argparse
+import logging
+import asyncio
+import websockets
+import json
+from collections import defaultdict
+import random
+import uuid
+import time
+import csv
+
+logger = logging.getLogger(__name__)
+OUTPUTWRITER = None
+
+
+def start_competition(address1, address2, nb_rows, nb_cols, timelimit):
+   asyncio.get_event_loop().run_until_complete(connect_agent(address1, address2, nb_rows, nb_cols, timelimit))
+
+
+async def connect_agent(uri1, uri2, nb_rows, nb_cols, timelimit):
+    cur_game = str(uuid.uuid4())
+    winner = None
+    cells = []
+    cur_player = 1
+    points = [0, 0, 0]
+    timings = [None, [], []]
+
+    for ri in range(nb_rows + 1):
+        columns = []
+        for ci in range(nb_cols + 1):
+            columns.append({"v":0, "h":0, "p":0})
+        cells.append(columns)
+
+    logger.info("Connecting to {}".format(uri1))
+    async with websockets.connect(uri1) as websocket1:
+        logger.info("Connecting to {}".format(uri2))
+        async with websockets.connect(uri2) as websocket2:
+            logger.info("Connected")
+
+            # Start game
+            msg = {
+              "type": "start",
+              "player": 1,
+              "timelimit": timelimit,
+              "game": cur_game,
+              "grid": [nb_rows, nb_cols]
+            }
+            await websocket1.send(json.dumps(msg))
+            msg["player"] = 2
+            await websocket2.send(json.dumps(msg))
+
+            # Run game
+            while winner is None:
+                ask_time = time.time()
+                logger.info("Waiting for player {}".format(cur_player))
+                if cur_player == 1:
+                    msg = await websocket1.recv()
+                else:
+                    msg = await websocket2.recv()
+                recv_time = time.time()
+                diff_time = recv_time - ask_time
+                timings[cur_player].append(diff_time)
+                logger.info("Message received after (s): {}".format(diff_time))
+                try:
+                    msg = json.loads(msg)
+                except json.decoder.JSONDecodeError as err:
+                    logger.debug(err)
+                    continue
+                if msg["type"] != "action":
+                    logger.error("Unknown message: {}".format(msg))
+                    continue
+                r, c = msg["location"]
+                o = msg["orientation"]
+                next_player = user_action(r, c, o, cur_player,
+                                          cells, points,
+                                          nb_rows, nb_cols)
+                if points[1] + points[2] == nb_cols * nb_rows:
+                    # Game over
+                    winner = 1
+                    if points[2] == points[1]:
+                        winner = 0
+                    if points[2] > points[1]:
+                        winner = 2
+                else:
+                    msg = {
+                        "type": "action",
+                        "game": cur_game,
+                        "player": cur_player,
+                        "nextplayer": next_player,
+                        "score": [points[1], points[2]],
+                        "location": [r, c],
+                        "orientation": o
+                    }
+                    await websocket1.send(json.dumps(msg))
+                    await websocket2.send(json.dumps(msg))
+
+                cur_player = next_player
+
+            # End game
+            logger.info("Game ended: points1={} - points2={} - winner={}".format(points[1], points[2], winner))
+            msg = {
+                "type": "end",
+                "game": cur_game,
+                "player": cur_player,
+                "nextplayer": 0,
+                "score": [points[1], points[2]],
+                "location": [r, c],
+                "orientation": o,
+                "winner": winner
+            }
+            await websocket1.send(json.dumps(msg))
+            await websocket2.send(json.dumps(msg))
+
+    # Timings
+    for i in [1, 2]:
+        logger.info("Timings: player={} - avg={} - min={} - max={}"\
+            .format(i,
+                    sum(timings[i])/len(timings[i]),
+                    min(timings[i]),
+                    max(timings[i])))
+
+    logger.info("Closed connections")
+    OUTPUTWRITER.writeln({'score1': points[1], 'score2': points[2], 'winner': winner})
+
+
+def user_action(r, c, o, cur_player, cells, points, nb_rows, nb_cols):
+    logger.info("User action: player={} - r={} - c={} - o={}".format(cur_player, r, c, o))
+    next_player = cur_player
+    won_cell = False
+    cell = cells[r][c]
+    if o == "h":
+        if cell["h"] != 0:
+            return cur_player
+        cell["h"] = cur_player
+        # Above
+        if r > 0:
+            if cells[r - 1][c]["v"] != 0 \
+                and cells[r - 1][c + 1]["v"] != 0 \
+                and cells[r - 1][c]["h"] != 0 \
+                and cells[r][c]["h"] != 0:
+                won_cell = True
+                points[cur_player] += 1
+                cells[r - 1][c]["p"] = cur_player
+        # Below
+        if r < nb_rows:
+            if cells[r][c]["v"] != 0 \
+                and cells[r][c + 1]["v"] != 0 \
+                and cells[r][c]["h"] != 0 \
+                and cells[r + 1][c]["h"] != 0:
+                won_cell = True
+                points[cur_player] += 1
+                cells[r][c]["p"] = cur_player
+
+    if o == "v":
+        if cell["v"] != 0:
+            return cur_player
+        cell["v"] = cur_player;
+        # Left
+        if c > 0:
+            if cells[r][c - 1]["v"] != 0 \
+                and cells[r][c]["v"] != 0 \
+                and cells[r][c - 1]["h"] != 0 \
+                and cells[r + 1][c - 1]["h"] != 0:
+                won_cell = True
+                points[cur_player] += 1
+                cells[r][c - 1]["p"] = cur_player
+        # Right
+        if c < nb_cols:
+            if cells[r][c]["v"] != 0 \
+                and cells[r][c + 1]["v"] != 0 \
+                and cells[r][c]["h"] != 0 \
+                and cells[r + 1][c]["h"] != 0:
+                won_cell = True
+                points[cur_player] += 1
+                cells[r][c]["p"] = cur_player
+
+    if not won_cell:
+        next_player = 3 - cur_player
+    else:
+        next_player = cur_player
+        print("Update points: player1={} - player2={}".format(points[1], points[2]))
+    return next_player
+
+
+def main(argv=None):
+    parser = argparse.ArgumentParser(description='Start agent to play Dots and Boxes')
+    parser.add_argument('--verbose', '-v', action='count', default=0, help='Verbose output')
+    parser.add_argument('--quiet', '-q', action='count', default=0, help='Quiet output')
+    parser.add_argument('--cols', '-c', type=int, default=2, help='Number of columns')
+    parser.add_argument('--rows', '-r', type=int, default=2, help='Number of rows')
+    parser.add_argument('--timelimit', '-t', type=float, default=0.5, help='Time limit per request in seconds')
+    parser.add_argument('--number', '-n', type=int, default=1, help='Number of games that will be played for the evaluation')
+    parser.add_argument('--output', '-o', default="output.csv", help='File where game results will be written to')
+    parser.add_argument('agents', nargs=2, metavar='AGENT', help='Websockets addresses for agents')
+    args = parser.parse_args(argv)
+
+    logger.setLevel(max(logging.INFO - 10 * (args.verbose - args.quiet), logging.DEBUG))
+    logger.addHandler(logging.StreamHandler(sys.stdout))
+
+    global OUTPUTWRITER
+    OUTPUTWRITER = OutputWriter(args.output)
+
+    for i in range(args.number):
+        start_competition(args.agents[0], args.agents[1], args.rows, args.cols, args.timelimit)
+
+    OUTPUTWRITER.close()
+
+
+class OutputWriter:
+    def __init__(self, outputfile):
+        self.csvfile = open(outputfile, 'w', newline='')
+        try:
+            fieldnames = ['score1', 'score2', 'winner']
+            self.writer = csv.DictWriter(self.csvfile, fieldnames=fieldnames)
+            self.writer.writeheader()
+        except IOError:
+            self.csvfile.close()
+
+    def writeln(self, csvdict):
+        self.writer.writerow(csvdict)
+
+    def close(self):
+        self.csvfile.close()
+
+
+if __name__ == "__main__":
+    sys.exit(main())
+
diff --git a/python/start.sh b/python/start.sh
new file mode 100755
index 0000000..6be69a2
--- /dev/null
+++ b/python/start.sh
@@ -0,0 +1,7 @@
+(cd dotsandboxes; python3 dotsandboxesserver.py 8080) &
+python3 agent.py 10001 &
+python3 agent.py 10002 &
+read -p "Press enter to close all programs."
+
+trap "exit" INT TERM
+trap "kill 0" EXIT