def init_dfs_stack(g, eqs):
# heap key: eq, value: nzeros
heap = None
connected = is_connected(g)
if connected:
heap = PriorityQueue()
heap.populate((eq, len(g[eq])) for eq in sorted(eqs))
root_node = Node(0, [], g, heap, connected)
return [root_node]
def init_dfs_stack(g, eqs, indent):
# heap key: eq, value: nzeros
heap = None
connected = is_connected(g)
if connected:
heap = PriorityQueue()
heap.populate((eq, len(g[eq])) for eq in sorted(eqs))
remaining_eqs = set(eqs)
asm_key = to_asm_key(remaining_eqs)
root_node = Node(0, [ ], g, heap, connected, remaining_eqs, None, asm_key,
None, None, get_id())
register('init', indent)
return [root_node]
def dijkstra(graph: Graph, source: object) -> tuple:
"""
calcule les plus courts chemins entre les sommets d'un graphe à partir d'une origine source
:param graph: le graphe
:param source: le sommet d'origine
:return: une tuple avec les distance stockés dans la variable dist
et le tableau de routage
"""
dist: dict = {source: 0} # initialization
routage: dict = {} # routing table
Q: PriorityQueue = PriorityQueue() # create vertex priority queue Q
for v in graph.get_vertices():
if v != source:
dist[v] = math.inf # unknown distance from source to v
routage[v] = None # predecessor of v
Q.add(v, priority=dist[v])
while not Q.is_empty(): # the main loop
vertex: tuple = Q.pop() # remove and return best vertex
u: object = vertex[1]
for v in graph.neighbors(u): # only v that are still in Q
alt = dist[u] + graph.get_weight(u, v)
if alt < dist[v]:
dist[v] = alt
routage[v] = u
Q.add(v, priority=alt)
return dist, routage
def __init__(self, cfg, width, height):
self.cfg = cfg
# Set up the state of the game board using the LevelState
# object. This is the main class that holds things like the
# amount of fuel that the player has left.
initialFuel = 30 # A default value of fuel to use if one is
# not specified in config.json
try:
initialFuel = self.cfg["initialfuel"]
except:
pass
self.state = LevelState(initialFuel)
# The actual board, represented as an width-by-height matrix
# (array of arrays)
self.board = [[0 for x in range(height)] for y in range(width)]
# A "dirty" matrix. This is used for efficiency: don't redraw
# tiles that don't need to be redrawn.
self.dirty = [[True for x in range(height)] for y in range(width)]
# Set the width/height
self.width = width
self.height = height
# Build the board as a matrix of priority queues
for x in range(width):
for y in range(height):
self.board[x][y] = PriorityQueue()
# The set of things that want to listen to clock ticks.
self.clockTickListeners = []
self.ferrets = []
self.healthpacks = []
self.stones = []
def TCPSimulator():
"""
This function implements the test program for the assignment.
It begins by creating the client and server objects and then
executes an event loop driven by a priority queue for which
the priority value is time.
"""
eventQueue = PriorityQueue()
client = TCPClient()
server = TCPServer()
client.server = server
server.client = client
client.queueRequestMessage(eventQueue, 0)
while not eventQueue.isEmpty():
e, t = eventQueue.dequeueWithPriority()
if EVENT_TRACE:
print(str(e) + " at time " + str(t))
e.dispatch(eventQueue, t)
def applyDijkstra(g, start, finish):
"""
Applies Dijkstra's algorithm to the graph g, updating the
distance from start to each node in g.
"""
#MAKE SURE THAT THE ALGORITHM IS APPLIED ONLY TO THE START/FINISH NODE
#USE node.predecessor, node.distance
#initializeSingleSource(g, start)
finalized = set()
frontier = PriorityQueue()
finishedendnode = False
#explored = set()
start.distance = 0
start.predecessor = None
frontier.enqueue(start)
while not frontier.isEmpty() and finishedendnode == False:
node = frontier.dequeue()
#print(type(node))
finalized.add(node)
if node == finish:
finishedendnode = True
break
for arc in node.getArcsFrom():
n1 = arc.getStart()
n2 = arc.getFinish()
#print("n1: "+ str(n1))
#print("n2: "+ str(n2))
#print("node: "+ str(node))
if n2 not in finalized:
#print("n2: "+ str(n2))
if n2 not in frontier:
initIndivNode(n1, n2, arc)
frontier.enqueue(n2, n2.distance)
#init single source
else:
oldDistance = n2.distance
relax(n1, n2, arc.getCost())
if n2.distance < oldDistance:
frontier.raisePriority(n2, n2.distance)
def initial_solution(g_orig, eqs):
# Duplication of heap_md.min_degree with none forbidden
g = fast_copy(g_orig)
eq_nzeros = PriorityQueue()
eq_nzeros.populate((eq, len(g[eq])) for eq in sorted(eqs))
rowp, n_tears = [ ], len(g) - len(eqs)
while eq_nzeros:
nzeros, eq = eq_nzeros.popitem()
if nzeros:
n_tears -= 1
rowp.append(eq)
vrs = sorted(g[eq])
eqs_update = set(chain.from_iterable(g[v] for v in vrs))
eqs_update.discard(eq)
g.remove_node(eq)
g.remove_nodes_from(vrs)
for e in sorted(eqs_update):
eq_nzeros[e] = len(g[e])
assert len(rowp) == len(eqs)
return n_tears, rowp
def applyDijkstra(g, start, finish=None):
"""
Applies Dijkstra's algorithm to the graph g, updating the
distance from start to each node in g.
"""
initializeSingleSource(g, start)
finalized = set()
pq = PriorityQueue()
pq.enqueue(start)
while not pq.isEmpty():
node = pq.dequeue()
finalized.add(start)
if node == finish: #if we are at the finish node then we are all done!
break
for arc in node.getArcsFrom():
n1 = arc.getStart()
n2 = arc.getFinish()
if n2 not in finalized:
if n2.distance is math.inf: #check if the node is one that we have not visited before, in which case we will add it to the queue
pq.enqueue(n2)
oldDistance = n2.distance
relax(n1, n2, arc.getCost())
if n2.distance < oldDistance:
pq.raisePriority(n2, n2.distance)
def eliminate(g, eqs, eq):
rowp = [eq]
remaining_eqs = sorted(n for n in g if n in eqs and n != eq)
heap = PriorityQueue()
heap.populate((r, len(g[r])) for r in remaining_eqs)
elimination_step(g, heap, eq)
while heap:
eq, nzeros = heap.peekitem()
if nzeros > 1:
break
heap.popitem()
rowp.append(eq)
elimination_step(g, heap, eq)
return rowp, heap
def __init__(self, cfg, width, height):
self.cfg = cfg
fuel = 20
try:
fuel = self.cfg["initialfuel"]
except:
pass
self.state = LevelState(fuel)
self.board = [[0 for x in range(width)] for y in range(height)]
self.dirty = [[True for x in range(width)] for y in range(height)]
self.width = width
self.height = height
for x in range(width):
for y in range(height):
self.board[x][y] = PriorityQueue()
def eliminate(g, remaining_eqs, eq):
rowp = [ eq ]
remaining_eqs.remove(eq)
heap = PriorityQueue()
heap.populate((r, len(g[r])) for r in sorted(remaining_eqs))
elimination_step(g, heap, eq)
while heap:
eq, nzeros = heap.peekitem()
if nzeros > 1:
break
remaining_eqs.remove(eq)
heap.popitem()
rowp.append(eq)
elimination_step(g, heap, eq)
return rowp, heap
def applyDijkstraOrig(g, start):
"""
Applies Dijkstra's algorithm to the graph g, updating the
distance from start to each node in g.
"""
initializeSingleSource(g, start)
finalized = set()
pq = PriorityQueue()
for node in g.getNodes():
pq.enqueue(node, node.distance)
while not pq.isEmpty():
node = pq.dequeue()
finalized.add(node)
for arc in node.getArcsFrom():
n1 = arc.getStart()
n2 = arc.getFinish()
if n2 not in finalized:
oldDistance = n2.distance
relax(n1, n2, arc.getCost())
if n2.distance < oldDistance:
pq.raisePriorityOrig(n2, n2.distance)
def astar(graph, start, goal):
# Fringe. Nodes not visited yet
openList = PriorityQueue()
# Visited Nodes. Each one will store it's parent position
closedList = {}
node = Node(start)
openList.push(node)
while openList:
node, _ = openList.pop()
position, cost = node.position, node.cost
if position in closedList:
# Oops. Already expanded.
continue
# Save the node in the closed list, and keep track of its parent
closedList[position] = node.parent
if position == goal:
break
for childPosition, actionCost in graph.getChildren(position):
# Only add to the open list if it's not expanded yet
if not childPosition in closedList:
childNode = Node(childPosition, cost + actionCost, position)
openList.push(
childNode,
childNode.cost + graph.getHCost(childPosition, goal))
path = []
if position == goal:
# Ensure a path has been found
path.insert(0, position)
while position and position != start:
position = closedList[position]
path.insert(0, position)
return path
class WaitingQueue:
def __init__(self):
self.queue = PriorityQueue()
self.log = []
def push(self, current_time, p):
self.queue.push([p.healthy, current_time], p)
self.log.append((Log.ENTER, p.healthy, current_time))
def empty(self):
return self.queue.empty()
def pop(self, current_time):
p = self.queue.pop()[-1]
self.log.append((Log.EXIT, p.healthy, current_time))
return p
def remove(self, current_time, p):
self.queue.remove(p)
self.log.append((Log.EXIT, p.healthy, current_time))
def size(self):
return self.queue.size()
def astar(graph, start, goal):
# Fringe. Nodes not visited yet
openList = PriorityQueue()
# Visited Nodes. Each one will store it's parent position
closedList = {}
node = Node(start)
openList.push(node)
while openList:
node, _ = openList.pop()
position, cost = node.position, node.cost
if position in closedList:
# Oops. Already expanded.
continue
# Save the node in the closed list, and keep track of its parent
closedList[position] = node.parent
if position == goal:
break
for childPosition, actionCost in graph.getChildren(position):
# Only add to the open list if it's not expanded yet
if not childPosition in closedList:
childNode = Node(childPosition, cost + actionCost, position)
openList.push(childNode, childNode.cost + graph.getHCost(childPosition, goal))
path = []
if position == goal:
# Ensure a path has been found
path.insert(0, position)
while position and position != start:
position = closedList[position]
path.insert(0, position)
return path
def solve(mode):
if mode == 1:
# test code for grid class
print("sup")
a = Grid([0, 1, 2, 3, 4, 5, 6, 7, 8])
b = Grid([1, 2, 3, 8, 0, 4, 7, 6, 5])
c = Grid([0, 1, 2, 3, 4, 5, 6, 7, 8])
print(a)
print(b)
print(c)
print("a = b? %s" % a.equals(b))
print("a = c? %s" % a.equals(c))
print("a = c? %s" % Grid.equal(a, c))
print("is a soln? %s" % a.is_solution())
print("is b soln? %s" % b.is_solution())
if mode == 2:
# test code for node class
a = Node(Grid([0, 1, 2, 3, 4, 5, 6, 7, 8]), 0, 0)
b = Node(a.data, a, 0)
a.print_tree()
if mode == 3:
# test code for get_moves()
soln = Grid([1, 2, 3, 8, 0, 4, 7, 6, 5])
a = Node(generate_problem(soln), 0, 0)
a.print_tree()
for move in a.data.get_moves():
print(move)
Node(Grid(a.data.move(move)), a, 0)
a.print_tree()
if mode == 4:
# A* search algorithm
pq = PriorityQueue()
soln = Grid([1, 2, 3, 8, 0, 4, 7, 6, 5])
root = Node(generate_problem(soln), 0, 0)
def expand_node(node):
# expands node to generate all possible moves
if len(node.children) == 0:
for move in node.data.get_moves():
m = Grid(node.data.move(move))
if not root.find_item(m):
# makes sure there are no duplicates
Node(m, node, move)
if root.data.is_solution():
return
pq.push(root, 0)
gui = EightPuzzleGUI()
best = B(root)
gui.draw_grid(best.data.data, gui.mainFrame, 50)
def increment(button, button1):
gui.mainFrame.update()
nxt = pq.pop() # gets next thing in PQ
expand_node(nxt)
if len(pq) > 0:
gui.draw_grid(pq.data.data[0][0].data, gui.rightFrame1, 12)
if len(pq) > 1:
gui.draw_grid(pq.data.data[1][0].data, gui.rightFrame2, 12)
if len(pq) > 2:
gui.draw_grid(pq.data.data[2][0].data, gui.rightFrame3, 12)
if len(pq) > 3:
gui.draw_grid(pq.data.data[3][0].data, gui.rightFrame4, 12)
if len(pq) > 4:
gui.draw_grid(pq.data.data[4][0].data, gui.rightFrame5, 12)
for child in nxt.children:
# adds children to PQ with h(n) + g(n)
# gui.draw_grid(child.data, gui.mainFrame, 50)
if child.data.h_score() < best.data.data.h_score():
best.data = child
gui.draw_grid(best.data.data, gui.mainFrame, 50)
pq.push(child, child.data.h_score() + 0.5 * child.level)
if child.data.is_solution():
button.destroy()
button1.destroy()
# print(child.moves)
gui.draw_moves(child.moves)
return True
# pq.print()
return False
def loop_to_end(button, button1):
# loops until solution is found
while True:
if increment(button, button1):
return
start = tk.Button(gui.root,
text="Increment",
command=lambda: increment(start, loop),
width=5,
height=5)
start.grid(column=0, row=4, sticky=tk.N + tk.S + tk.E + tk.W)
loop = tk.Button(gui.root,
text="Loop",
command=lambda: loop_to_end(start, loop),
width=5,
height=5)
loop.grid(column=1, row=4, sticky=tk.N + tk.S + tk.E + tk.W)
gui.root.mainloop()
def __init__(self):
self.queue = PriorityQueue()
self.log = []
return True
return False
def is_bounded(x):
return x[0] >= 0 and x[0] <= 100 and x[1] >= 0 and x[1] <= 100
path = []
start = (0,50)
while in_circle(start):
start = (0,np.random.randint(101))
frontier = PriorityQueue()
frontier.insert(start,0)
visited = set()
while frontier.heap:
dist,node = frontier.pop()
path.append(node)
if node[0] == 100:
break
visited.add(node)
up = (node[0], node[1] + d)
up_right = (node[0] + d, node[1] + d)
right = (node[0] + d + 1, node[1])
def get_ready_queue(job_list, pattern, time_quantum):
# obtain job key function as key for priority queue comparison
job_key = job_keys[pattern]
if job_key == round_robin:
job_list.sort(key=first_come_first_serve)
i = 0
while i < len(job_list):
job = job_list[i]
job.instance_id = i
if job.duration > time_quantum:
split_job = Job(job.arrival + time_quantum,
job.duration - time_quantum, job.priority,
job.id, True)
job.terminate = False
job.duration = time_quantum
job_list.append(split_job)
i += 1
global last_index
last_index.clear()
# prepare job queue
job_queue = PriorityQueue(job_list, key=job_key)
# prepare the ready queue
ready_queue = []
global time_elapsed
time_elapsed = 0
global idd
idd = 0
def schedule(job):
# schedule a job to the ready queue
if ready_queue and ready_queue[-1].id == job.id:
# merge with the previous job because same ID
# this happens when pre-emptive
ready_queue[-1].duration += job.duration
ready_queue[-1].terminate = job.terminate
else:
# append new job
ready_queue.append(job)
# update time elapsed after scheduling
global time_elapsed
time_elapsed = ready_queue[-1].end()
global last_index, idd
last_index[job.id] = idd
idd += 1
# simulate scheduling algorithm
while job_queue:
# get the next job
job = job_queue.pop()
global last_index
if job.idd != (last_index[job.id] if job.id in last_index else -1):
job.idd = (last_index[job.id] if job.id in last_index else -1)
job_queue.push(job)
elif job.arrival < time_elapsed: # this job will arrive later, repush to the queue
job.arrival = time_elapsed
#job.instance_id = Job.instance_id
#Job.instance_id += 1
job_queue.push(job)
elif job_key == round_robin and job.duration > time_quantum:
# split job into two for round robin
part_two = Job(arrival=job.arrival + time_quantum,
duration=job.duration - time_quantum,
priority=job.priority,
job_id=job.id,
terminate=job.terminate)
# part_two.instance_id = job.instance_id + len(job_list)
job_queue.push(part_two) # repush part 2
# process part 1
job.duration = time_quantum
job.terminate = False
job_queue.push(job)
# schedule(job)
elif not pre_emptive(job_key):
# schedule immediately
schedule(job)
else: # pre_emptive
# special case: find possible jobs that can interrupt
put_back = []
interrupted = False
while job_queue and job_queue.top().arrival < job.end():
# check if the next job is better than having the other end of the current job
interrupt = job_queue.top()
split_job = Job(arrival=interrupt.arrival,
duration=job.end() - interrupt.arrival,
priority=job.priority,
job_id=job.id,
terminate=job.terminate)
# print interrupt, 'wants to interrupt', split_job
# print job_key(interrupt), job_key(split_job)
if job_key(interrupt) < job_key(split_job):
# job has been interrupted! split then repush into the queue
job.duration -= split_job.duration
job.terminate = False
# push new parts of the job
job_queue.push(job)
job_queue.push(split_job)
interrupted = True
break
put_back.append(job_queue.pop())
# return tested jobs back into queue
for tested_job in put_back:
job_queue.push(tested_job)
if not interrupted:
# finalize job schedule
schedule(job)
return ready_queue
def limited_infection(g, source=None, limit=5, threshold=0):
'''
Runs limited infection using adjusted BFS
params
------
g: input graph
source: node to start infection from
limit: max number of nodes
threshold: min value of edge weights to traverse on and accept
returns
------
iterations for animation
'''
if not source:
source = pick_source(g)
# create dictionary of nodes to store state for animation
infected = OrderedDict()
for node in g.nodes():
infected[node] = False
# store initial state
iterations = [list(infected.items())]
# run altered bfs algorithm
visited = set()
queue = PriorityQueue()
queue.add(source, 1)
count=0
while not queue.is_empty():
# exact if limit reached
if count >= limit:
break
# pop node from priority queue of nodes
p, n = queue.pop_smallest()
print(p, n)
user = g.node[n]['data']
if n not in visited:
# infect current node
user.version = 'new'
infected[n] = True
iterations.append(list(infected.items()))
visited.add(n)
# increment count of infected nodes
count+=1
# children of current node not yet infected
suc = [adj for adj in g.successors(n) if adj not in visited]
# parents of current node not yet infected
pre = [adj for adj in g.predecessors(n) if adj not in visited]
# if there are enough children to be under limit
# add them with priority 1
if suc and len(suc) <= (limit - count - queue.size(1)):
for adj in suc:
if g[n][adj]['weight'] > threshold:
queue.add(adj, 1)
# else add them with priority 2
else:
for adj in suc:
if g[n][adj]['weight'] > threshold:
queue.add(adj, 2)
# add parents with priority 3
for adj in pre:
if g[adj][n]['weight'] > threshold:
queue.add(adj, 3)
# add all adjacent nodes if nothing in queue
if queue.is_empty():
for adj in suc:
if g[n][adj]['weight'] > threshold:
queue.add(adj, 4)
for adj in pre:
if g[adj][n]['weight'] > threshold:
queue.add(adj, 4)
# flush output for testing
sys.stdout.flush()
return iterations
def get_ready_queue(job_list, pattern, time_quantum): # obtain job key function as key for priority queue comparison job_key = job_keys[pattern] if job_key == round_robin: job_list.sort(key=first_come_first_serve) i = 0 while i < len(job_list): job = job_list[i] job.instance_id = i if job.duration > time_quantum: split_job = Job(job.arrival + time_quantum, job.duration - time_quantum, job.priority, job.id, True) job.terminate = False job.duration = time_quantum job_list.append(split_job) i += 1 global last_index last_index.clear() # prepare job queue job_queue = PriorityQueue(job_list, key=job_key) # prepare the ready queue ready_queue = [] global time_elapsed time_elapsed = 0 global idd idd = 0 def schedule(job): # schedule a job to the ready queue if ready_queue and ready_queue[-1].id == job.id: # merge with the previous job because same ID # this happens when pre-emptive ready_queue[-1].duration += job.duration ready_queue[-1].terminate = job.terminate else: # append new job ready_queue.append(job) # update time elapsed after scheduling global time_elapsed time_elapsed = ready_queue[-1].end() global last_index, idd last_index[job.id] = idd idd += 1 # simulate scheduling algorithm while job_queue: # get the next job job = job_queue.pop() global last_index if job.idd != (last_index[job.id] if job.id in last_index else -1): job.idd = (last_index[job.id] if job.id in last_index else -1) job_queue.push(job) elif job.arrival < time_elapsed: # this job will arrive later, repush to the queue job.arrival = time_elapsed #job.instance_id = Job.instance_id #Job.instance_id += 1 job_queue.push(job) elif job_key == round_robin and job.duration > time_quantum: # split job into two for round robin part_two = Job( arrival = job.arrival + time_quantum, duration = job.duration - time_quantum, priority = job.priority, job_id = job.id, terminate = job.terminate ) # part_two.instance_id = job.instance_id + len(job_list) job_queue.push(part_two) # repush part 2 # process part 1 job.duration = time_quantum job.terminate = False job_queue.push(job) # schedule(job) elif not pre_emptive(job_key): # schedule immediately schedule(job) else: # pre_emptive # special case: find possible jobs that can interrupt put_back = [] interrupted = False while job_queue and job_queue.top().arrival < job.end(): # check if the next job is better than having the other end of the current job interrupt = job_queue.top() split_job = Job( arrival = interrupt.arrival, duration = job.end() - interrupt.arrival, priority = job.priority, job_id = job.id, terminate = job.terminate ) # print interrupt, 'wants to interrupt', split_job # print job_key(interrupt), job_key(split_job) if job_key(interrupt) < job_key(split_job): # job has been interrupted! split then repush into the queue job.duration -= split_job.duration job.terminate = False # push new parts of the job job_queue.push(job) job_queue.push(split_job) interrupted = True break put_back.append(job_queue.pop()) # return tested jobs back into queue for tested_job in put_back: job_queue.push(tested_job) if not interrupted: # finalize job schedule schedule(job) return ready_queue
def astarItemsMultiPath(graph, start, goal, itemsAvailable, initialRaftState,
illegalEdges):
#itemsAvailable = intialItems.copy()
#print("initial state") print(itemsAvailable)
# Fringe. Nodes not visited yet
openList = PriorityQueue()
# Visited Nodes. Each one will store it's parent position
closedList = {}
nodeTable = {}
# assign some start and goal points
# and push start point as a node to openList.
start_tuple = tuple(start)
goal_tuple = tuple(goal)
node = PathNode(start_tuple, itemsAvailable.copy())
node.setRaftState(initialRaftState)
openList.push(node)
while openList:
# if openList does not contain any node
# we will break.
curr_node, _ = openList.pop()
if not curr_node:
break
position, cost = curr_node.position, curr_node.cost
if position in closedList:
# Oops. Already expanded.
continue
# Save the node in the closed list, and keep track of its parent
if curr_node.parent != None:
closedList[position] = curr_node.parent.position
else:
closedList[position] = curr_node.position
#track of its items remainings
nodeTable[position] = curr_node
# arrived to goal so save it as goal node.
if position == goal_tuple:
goalNode = curr_node
break
# copy some parent properties
curr_available = curr_node.getItemAvailable()
raftState = curr_node.getRaftState()
parentStonePlace = curr_node.getStonePlaced().copy()
# for each legal child for current parent,
for childPosition, actionCost, itemRequired, element in graph.getChildren(
position, curr_available, raftState, parentStonePlace):
copyStonePlaced = parentStonePlace.copy()
# if illegal edges were given
# and current one is illegal, ignore current child.
pos1 = list(position).copy()
pos2 = childPosition
if illegalEdges and ([pos1, pos2] in illegalEdges
or [pos2, pos1] in illegalEdges):
continue
# Only add to the open list if it's not expanded yet
if not tuple(childPosition) in closedList:
# copyAvailableItem
copy_available_item = curr_available.copy()
# deduct from item list if it was required
if itemRequired:
copy_available_item = deductItem(copy_available_item,
itemRequired)
childRaftSate = raftState
# updating child raftstate based on where current child is
# and the item required to get to the child.
if itemRequired == 'r':
childRaftSate = 1
elif childRaftSate == 1 and element != '~':
childRaftSate = 0
# updating the stone coordinate that was used for child.
# append to the parent ones.
stoneCoord = []
if itemRequired == 'o':
stoneCoord = [childPosition]
#print(stoneCoord)
copyStonePlaced += stoneCoord
# creating new child node and push to the openlist
childNode = PathNode(tuple(childPosition), copy_available_item,
cost + actionCost, curr_node,
childRaftSate)
childNode.addStonePlaced(copyStonePlaced)
openList.push(
childNode,
childNode.cost + graph.getHCost(childPosition, goal))
# intialize the returning variables
path = []
totalItemUsed = []
totalNewItemCollected = []
itemUsedState = False
finalItemList = itemsAvailable
finalRaftState = 1
finalStonePlace = []
# if we arrive goal, we are updating those variables.
if position == goal_tuple:
finalItemList = goalNode.getItemAvailable()
itemUsedState = confirmItemUsed(finalItemList, itemsAvailable)
finalRaftState = goalNode.getRaftState()
finalStonePlace = goalNode.getStonePlaced()
# Ensure a path has been found
position_list = list(position)
path.insert(0, position_list)
while position and position != start_tuple:
position = closedList[position]
position_list = list(position)
path.insert(0, position_list)
return [
path, itemUsedState, finalItemList, finalRaftState, finalStonePlace,
nodeTable
] #finalItemList, ]
def solve_cube():
def expand_node(node):
# expands node to generate all possible moves
if len(node.children) == 0:
for i in range(6):
m = deepcopy(node.data)
m.rotate_face(list(Facing)[i])
if not root.find_item(m):
# makes sure there are no duplicates
Node(m, node, list(Facing)[i])
def increment(button, button1):
gui.mainFrame.update()
nxt = pq.pop() # gets next thing in PQ
expand_node(nxt)
# print(nxt.data)
if len(pq) > 0:
gui.draw_cube(pq.data.data[0][0].data, gui.rightFrame1, 10)
if len(pq) > 1:
gui.draw_cube(pq.data.data[1][0].data, gui.rightFrame2, 10)
if len(pq) > 2:
gui.draw_cube(pq.data.data[2][0].data, gui.rightFrame3, 10)
if len(pq) > 3:
gui.draw_cube(pq.data.data[3][0].data, gui.rightFrame4, 10)
if len(pq) > 4:
gui.draw_cube(pq.data.data[4][0].data, gui.rightFrame5, 10)
for child in nxt.children:
# adds children to PQ with h(n) + g(n)
# gui.draw_grid(child.data, gui.mainFrame, 40)
if child.data.h_score() < best.data.data.h_score():
best.data = child
gui.draw_grid(best.data.data, gui.mainFrame, 40)
pq.push(child, child.data.h_score() + 4 * child.level)
if child.data.h_score() == 0:
button.destroy()
button1.destroy()
# print(child.data)
# print(child.moves)
# print("done idiot")
gui.draw_moves(child.moves)
return True
# pq.print()
return False
def loop_to_end(button, button1):
for i in range(999):
if increment(button, button1):
return
pq = PriorityQueue()
root = Node(Cube(), 0, -1)
root.data.scramble(6)
# print(root.data)
gui = EightPuzzleGUI()
pq.push(root, 0)
best = B(root)
start = tk.Button(gui.root,
text="Increment",
command=lambda: increment(start, loop),
width=5,
height=5)
start.grid(column=0, row=4, sticky=tk.N + tk.S + tk.E + tk.W)
loop = tk.Button(gui.root,
text="Loop",
command=lambda: loop_to_end(start, loop),
width=5,
height=5)
loop.grid(column=1, row=4, sticky=tk.N + tk.S + tk.E + tk.W)
gui.root.mainloop()
