def astar(maze):
# TODO: Write your code here
frontier = PriorityQueue()
frontier.put(maze.getStart(), 0)
# return path, num_states_explored
return [], 0
def bfs(maze):
"""
Runs BFS for part 1 of the assignment.
@param maze: The maze to execute the search on.
@return path: a list of tuples containing the coordinates of each state in the computed path
"""
# TODO: Write your code here
q = queue.Queue()
start = maze.getStart()
path =[]
obj = maze.getObjectives()
obj_num = len(obj)
visited = []
cur = mazenode(start,None)
visited.append(cur.position) # set visitied
while(not maze.isObjective(cur.row,cur.col)):
nei = maze.getNeighbors(cur.row,cur.col) # get neighbour
for i in nei:
if not (i in visited):
q.put(mazenode((i[0],i[1]),cur)) # add the next node into queue
visited.append((i[0],i[1]))
cur = q.get()
# print(visited)
# print(cur.position)
# get the path
while (not cur.lastnode is None): # extract the path
path.append(cur.position)
cur = cur.lastnode
path.append(cur.position)
return path[::-1]
def bfs(maze):
"""
Runs BFS for part 1 of the assignment.
@param maze: The maze to execute the search on.
@return path: a list of tuples containing the coordinates of each state in the computed path
"""
# TODO: Write your code here
start = maze.getStart()
objective = maze.getObjectives()
path = deque()
explored = deque()
frontier = deque()
frontier.append(start)
parent = {}
while frontier:
V = frontier.popleft()
if V in explored:
V = frontier.popleft()
if V == objective[0]: #objective in the form of [(5,1)]
explored.append(V)
break
else:
for i in set(maze.getNeighbors(V[0], V[1])):
if maze.isValidMove(
i[0], i[1]
) == True and i not in explored: #can lead to infinite loop
frontier.append(i)
parent[i] = V #key is son,value is parent
explored.append(V)
a = objective[0]
path.append(a)
while start not in path:
path.appendleft(parent[a])
a = parent[a]
return list(path)
def astar_corner(maze):
"""
Runs A star for part 2 of the assignment in the case where there are four corner objectives.
@param maze: The maze to execute the search on.
@return path: a list of tuples containing the coordinates of each state in the computed path
"""
# TODO: Write your code here
q = queue.PriorityQueue()
start = maze.getStart()
path = []
obj = maze.getObjectives()
obj_num = len(obj)
visited = []
cur = mazenode(start,None,0,0,obj)
# while(len(obj)>0):
d = cal_dis_mat(obj)
cur.cur_mst = findmst(d,0,len(obj))
visited.append((cur.position,cur.remain_obj))
while(len(cur.remain_obj)>0):
if(cur.position in cur.remain_obj):
# print(cur.position, cur.remain_obj)
if (len(cur.remain_obj)==1):break
cur.remain_obj.remove(cur.position)
d = cal_dis_mat(cur.remain_obj)
cur.cur_mst = findmst(d,0,len(cur.remain_obj))
# visited.append((cur.position,cur.remain_obj)) # set visitied
nei = maze.getNeighbors(cur.row,cur.col) # get neighbour
for i in nei:
if not_visited(i,cur.remain_obj,visited):
gx = cur.gx+1
hx = cur.cur_mst + find_min_dis(cur.remain_obj,i)[0]
# print(cur.position,'->',i)
# print('hx:',hx,' ','gx:',gx,' ','fx:',gx+hx)
q.put((gx+hx,mazenode((i[0],i[1]),cur,gx,hx,cur.remain_obj.copy(),cur.cur_mst))) # add the next node into queue
visited.append((i,cur.remain_obj))
cur = q.get()[1]
# visited = []
# print(cur.position)
# print(len(cur.remain_obj))
# obj.remove(cur.position)
while (not cur.lastnode is None): # extract the path
path.append(cur.position)
cur = cur.lastnode
path.append(cur.position)
# print(path)
return path[::-1]
def fast(maze):
"""
Runs suboptimal search algorithm for part 4.
@param maze: The maze to execute the search on.
@return path: a list of tuples containing the coordinates of each state in the computed path
"""
# TODO: Write your code here
start = maze.getStart()
path = []
obj = maze.getObjectives()
obj_num = len(obj)
visited = [False for i in obj]
cur = mazenode(start,None,0,0,obj)
min,idx = find_min_dis(obj,start)
gx,path = astar_dis_path(maze,start,obj[idx])
visited[idx] = True
visited_num = 1
while(1):
if visited_num == len(obj):
break
min = sys.maxsize
next_idx = idx
for i in range(len(obj)):
if not visited[i]:
dis = mah_d(obj[i],obj[idx])
if dis<min:
min = dis
next_idx = i
gx,minpath = astar_dis_path(maze,obj[idx],obj[next_idx])
idx = next_idx
visited[next_idx] = True
visited_num+=1
# print(visited_num)
# print(next_idx)
for i in minpath[1:]:
path.append(i)
return path
def astar(maze):
"""
Runs A star for part 1 of the assignment.
@param maze: The maze to execute the search on.
@return path: a list of tuples containing the coordinates of each state in the computed path
"""
# TODO: Write your code here
# a priority queue
q = queue.PriorityQueue()
start = maze.getStart()
path =[]
obj = maze.getObjectives()
obj_num = len(obj)
visited = []
cur = mazenode(start,None,mah_d(start,start),mah_d(start,obj[0]))
# visited.append(cur.position)
while(not maze.isObjective(cur.row,cur.col)):
visited.append(cur.position) # set visitied
nei = maze.getNeighbors(cur.row,cur.col) # get neighbour
for i in nei:
if not i in visited:
gx = 1+cur.gx
hx = mah_d(i,obj[0])
q.put((gx+hx,mazenode((i[0],i[1]),cur,gx,hx))) # add the next node into queue
# visited.append((i[0],i[1]))
cur = q.get()[1]
# get the path
while (not cur.lastnode is None): # extract the path
path.append(cur.position)
cur = cur.lastnode
path.append(cur.position)
return path[::-1]
def astar_multi(maze):
"""
Runs A star for part 3 of the assignment in the case where there are
multiple objectives.
@param maze: The maze to execute the search on.
@return path: a list of tuples containing the coordinates of each state in the computed path
"""
# TODO: Write your code here
q = queue.PriorityQueue()
start = maze.getStart()
path = []
obj = maze.getObjectives()
# print(obj)
obj_num = len(obj)
d = find_dis_goal_astar(maze,obj)
visited = []
cur = mazenode(start,None,0,0,obj)
# while(len(obj)>0):
cur.cur_mst = findmst(d,0,len(obj))
cur.v_g = [False for i in obj]
visited.append((cur.position,cur.remain_obj))
counter = 0
while(len(cur.remain_obj)>0):
if(cur.position in cur.remain_obj):
# print(cur.position, cur.remain_obj)
#case that breaks
if (len(cur.remain_obj)==1):break
# set and remove reached goal
cur.v_g[obj.index(cur.position)] = True
cur.remain_obj.remove(cur.position)
d1 = adjust_dis_max(d,cur.v_g,obj_num)
# print('len:',len(d1[0]),len(cur.remain_obj))
#calculate mst
cur.cur_mst = findmst(d1,0,len(cur.remain_obj))
# visited.append((cur.position,cur.remain_obj)) # set visitied
nei = maze.getNeighbors(cur.row,cur.col) # get neighbour
for i in nei:
if not_visited(i,cur.remain_obj,visited):
gx = cur.gx+1
hx = cur.cur_mst + find_min_dis(cur.remain_obj,i)[0]
if not find_same_in_queue(q,i,gx+hx,cur.remain_obj):
#print(cur.position,'->',i)
#print('hx:',hx,' ','gx:',gx,' ','fx:',gx+hx)
q.put((gx+hx,mazenode((i[0],i[1]),cur,gx,hx,cur.remain_obj.copy(),cur.cur_mst,cur.v_g.copy()))) # add the next node into queue
visited.append((i,cur.remain_obj))
cur = q.get()[1]
# visited = []
# print(cur.hx+cur.gx,' ',gx)
#print(len(cur.remain_obj))
counter+=1
# if counter==7703:break
# obj.remove(cur.position)
while (not cur.lastnode is None): # extract the path
path.append(cur.position)
cur = cur.lastnode
path.append(cur.position)
# print(path)
return path[::-1]
def astar(maze):
"""
Runs A star for part 1 of the assignment.
@param maze: The maze to execute the search on.
@return path: a list of tuples containing the coordinates of each state in the computed path
"""
# TODO: Write your code here
#following pseudocode as much as I understood it on :https://www.geeksforgeeks.org/a-search-algorithm/ and http://coecsl.ece.illinois.edu/ge423/lecturenotes/AstarHandOut.pdf
start = maze.getStart()
objective = maze.getObjectives()
path = deque()
explored = deque()
frontier = deque()
parent = {}
gdist = {}
hdist = {}
fdist = {}
frontier.append(start)
gdist[start] = 0
def heuristic(point1, point2):
return abs(point1[0] - point2[0]) + abs(point1[1] - point2[1])
hdist[start] = heuristic(start, objective[0])
fdist[start] = gdist[start] + hdist[start]
while frontier:
temp = 100000
for i in frontier:
if fdist[i] < temp:
temp = fdist[
i] #getting minmium fdict value in frontier elements
else:
temp = temp
for k in frontier:
if fdist[k] == temp:
V = k
break
frontier.remove(V)
if V in explored:
continue
explored.append(V)
#if the objective judgement is written in the loop of neighbor handlement, the states explored will increase around 100 in the bigMaze, found by trial-and-error
if V == objective[0]:
parent[V] = Last
break
for i in set(maze.getNeighbors(
V[0], V[1])): #get Neighbors already contained isValidMove
if i not in explored: # every i is a neighbor of V, a child of parent V
temphdist = heuristic(i, objective[0])
tempgdist = gdist[V] + heuristic(
i, V) #current cost to i(successor)
hdist[i] = temphdist
gdist[i] = tempgdist
if (
i in frontier and fdist[i] < tempgdist + temphdist
): #explored doesn't need to be contained here since admissible heruistic mentioned by instructor in Piazza
continue
else:
fdist[i] = hdist[i] + gdist[i]
parent[i] = V
frontier.append(i)
Last = V
else:
continue
a = objective[0]
path.append(a)
while start not in path:
path.appendleft(parent[a])
a = parent[a]
return list(path)
def astar_multi(maze):
"""
Runs A star for part 3 of the assignment in the case where there are
multiple objectives.
@param maze: The maze to execute the search on.
@return path: a list of tuples containing the coordinates of each state in the computed path
"""
# TODO: Write your code here
start_point = maze.getStart()
objective_lists = maze.getObjectives() #objective is a list
fullpath = deque()
counter = 1
def heuristic_objectives(
start, objectivelist
): #return an objective following the heuristic out of the whole list
temp = list()
for i in objectivelist:
temp.append(abs(start[0] - i[0]) + abs(start[1] - i[1]))
threshold = min(temp)
return threshold
#heruistic=heruistic_objectives(nearest goal's distance)+ MST distance of the rest of the objectivelists
def heuristic_mst(start_point, objective_lists):
temp = list()
for i in objective_lists:
temp.append(
abs(start_point[0] - i[0]) + abs(start_point[1] - i[1]))
for i in objective_lists:
if abs(start_point[0] - i[0]) + abs(start_point[1] -
i[1]) == min(temp):
objective_lists.remove(i)
break
#got rid of the nearest goal--objective_lists contain the remaining objectives now
edgedist = {}
tempobjective_lists = objective_lists.copy()
tempobjective_lists.append(start_point)
for i in objective_lists:
for j in tempobjective_lists:
if i != j:
edgedist[i, j] = abs(i[0] - j[0]) + abs(i[1] - j[1])
else:
continue
tempobjective_lists.remove(i)
# construct a combination between objectivelists(n length) nC2
# construct a dict edgelist[(1,2),(2,3):2]
explored = list()
# print('edgedist',edgedist)
explored.append(start_point)
T = 0
while len(
set(explored)) < len(objective_lists): #len(objective_lists=9)
temp_dist = list()
keychoice = list()
for single_key in edgedist.keys():
if (single_key[0] in explored or single_key[1]
in explored) and not (single_key[0] in explored
and single_key[1] in explored):
temp_dist.append(edgedist[single_key])
keychoice.append(single_key)
for key in keychoice:
if edgedist[key] == min(temp_dist):
minimum_key = key
break
# print('minimum_key=',minimum_key)
key_toadd = [k for k in minimum_key
if k not in explored] #[(2,3)] form
# print('key_toadd=',key_toadd)
T += edgedist[minimum_key]
explored.append(key_toadd[0])
# print('explored=',explored)
return T * 1.5
while objective_lists: # objectlist is not empty
path = deque()
explored = deque()
parent = {}
gdist = {}
hdist = {}
fdist = {}
gdist[start_point] = 0
hdist[start_point] = heuristic_objectives(
start_point, objective_lists) + heuristic_mst(
start_point, objective_lists)
fdist[start_point] = gdist[start_point] + hdist[start_point]
frontier = deque()
frontier.append(start_point)
# for a single path
while frontier:
V = frontier.popleft()
if V in explored:
V = frontier.popleft()
if V in objective_lists: #objective in the form of [(5,1)]
print('obectives_list before movement', objective_lists)
explored.append(V)
objective_lists.remove(V)
print('removed objective=', V, 'remaining objectives=',
objective_lists)
#last element in explored is the objective in this path
#maybe should put path command here 16:04-2/4
break
else:
for i in maze.getNeighbors(V[0], V[1]):
if maze.isValidMove(
i[0], i[1]
) == True and i not in explored: #can lead to infinite loop
temphdist = heuristic_objectives(i, objective_lists)
tempgdist = gdist[V] + 1 #current cost to i(successor)
if i in frontier and fdist[i] < tempgdist + temphdist:
continue
elif i in explored:
if fdist[i] < tempgdist + temphdist:
continue
else:
explored.remove(i)
frontier.append(i)
else:
fdist[i] = temphdist + tempgdist
hdist[i] = temphdist
gdist[i] = tempgdist
frontier.append(i)
parent[i] = V #key is son,value is parent
Last = V
explored.append(V)
print('start point=', start_point)
a = explored[-1]
path.appendleft(a)
while start_point not in path:
a = parent[a]
path.appendleft(a)
#print('a=',a)
print('path=', path)
start_point = explored[-1]
if counter == 1:
fullpath += path
else:
path.remove(a)
# print('path to +',path)
fullpath += path
counter += 1
# print('counter=',counter)
print('fullpath', fullpath)
return list(fullpath)
def astar_corner(maze):
"""
Runs A star for part 2 of the assignment in the case where there are four corner objectives.
@param maze: The maze to execute the search on.
@return path: a list of tuples containing the coordinates of each state in the computed path
"""
# TODO: Write your code here
start_point = maze.getStart()
objective_lists = maze.getObjectives() #objective is a list
fullpath = deque()
counter = 1
print('start=', start_point)
print('objectives are', objective_lists)
def heuristic_objectives(
start, objectivelist
): #return an objective following the heuristic out of the whole list
temp = list()
for i in objectivelist:
temp.append(abs(start[0] - i[0]) + abs(start[1] - i[1]))
threshold = min(temp)
return threshold
while objective_lists: # objectlist is not empty
path = deque()
explored = deque()
parent = {}
gdist = {}
hdist = {}
fdist = {}
gdist[start_point] = 0
hdist[start_point] = heuristic_objectives(start_point, objective_lists)
fdist[start_point] = gdist[start_point] + hdist[start_point]
frontier = deque()
frontier.append(start_point)
# for a single path
while frontier:
V = frontier.popleft()
if V in explored:
V = frontier.popleft()
if V in objective_lists: #objective in the form of [(5,1)]
explored.append(V)
objective_lists.remove(V)
#last element in explored is the objective in this path
break
else:
for i in maze.getNeighbors(V[0], V[1]):
if maze.isValidMove(
i[0], i[1]
) == True and i not in explored: #can lead to infinite loop
temphdist = heuristic_objectives(i, objective_lists)
tempgdist = gdist[V] + 1 #current cost to i(successor)
if i in frontier and fdist[i] < tempgdist + temphdist:
continue
else:
fdist[i] = temphdist + tempgdist
hdist[i] = temphdist
gdist[i] = tempgdist
frontier.append(i)
parent[i] = V #key is son,value is parent
Last = V
explored.append(V)
a = explored[-1]
path.appendleft(a)
while start_point not in path:
a = parent[a]
path.appendleft(a)
start_point = explored[-1]
if counter == 1:
fullpath += path
else:
path.remove(a)
fullpath += path
counter += 1
return list(fullpath)
