def hotPotato(namelist, num):
simqueue = Queue()
for name in namelist:
simqueue.enqueue(name)
while simqueue.size() > 1:
for i in range(num):
simqueue.enqueue(simqueue.dequeue())
simqueue.dequeue()
return simqueue.dequeue()
def shortest_path(
walls, start, end
): # can access walls from agents state.data.walls gives numpy 2D array, true = walls, false = space
start = [start[0], start[1]]
end = [end[0], end[1]]
if start == end:
return [start]
neighbours = Queue() # queue storing the next positions to explore
neighbours.enqueue(start)
counts = np.zeros(
(walls.width, walls.height), dtype=int
) # 2D array to store the distance from the start to all visted points
predecessors = np.zeros(
(counts.shape[0], counts.shape[1], 2), dtype=int
) # 2D array storing the predecessors (past points allowing path to be retraced)
counts[start[0], start[1]] = 1
# loop until the end position is found
while not neighbours.isEmpty():
n = neighbours.dequeue()
print n
if n == end:
print "path found!", n, counts[n[0]][n[1]]
break # path found
# add all the valid neighbours to the list and remember from where they came from
for neighbour in [[n[0] - 1, n[1]], [n[0] + 1, n[1]], [n[0], n[1] - 1],
[n[0], n[1] + 1]]:
if not walls[neighbour[0]][neighbour[1]] and counts[
neighbour[0], neighbour[1]] == 0:
neighbours.enqueue(neighbour)
predecessors[neighbour[0], neighbour[1]] = n
counts[neighbour[0], neighbour[1]] = counts[n[0], n[1]] + 1
print 'here 0'
if counts[end[0], end[1]] == 0:
return [] # path not found
print 'here 1'
path = []
n = end
print end, counts[end[0], end[1]]
# reconstruct the path
print counts
while n != start:
#print n
if n == start:
break
path.append(n)
n = predecessors[n[0], n[1]].tolist()
path.append(start)
print 'here 2'
print counts
return path
