最小生成树
Kruskal算法
Kruskal算法是基于贪心的MST算法。
Step1:把图里的边按照权重从小到大排序
Step2:从小到大遍历每条边,用并查集来确认这条边是否有效(连接两个非联通区域)
Step3:直到所有节点都联通
class Solution:
def minCostConnectPoints(self, points: List[List[int]]) -> int:
def find(idx):
if parent[idx] != idx:
parent[idx] = find(parent[idx])
return parent[idx]
def union(idx1,idx2):
p1,p2 = find(idx1),find(idx2)
if p1 < p2:
parent[p2] = p1
else:
parent[p1] = p2
n = len(points)
parent = [i for i in range(n)]
dist = dict()
for i in range(n):
for j in range(i+1,n):
dist[(i,j)] = abs(points[i][0]-points[j][0])+abs(points[i][1]-points[j][1])
distance = sorted(dist.items(),key = lambda x:x[1])
res = 0
for pair,d in distance:
i,j = pair
if sum(parent) == 0:
return res
if find(i) != find(j):
union(i,j)
res += d
return res
Prim算法
定义一个数组 D i s t Dist Dist 为选中区域到所有节点的距离,如果已选中则置为 − 1 -1 −1(或者新建一个visited数组来存放选中节点
Step1:随机选择一个节点作为起点,遍历它的所有边,更新
D
i
s
t
Dist
Dist中的值
Step2:从
D
i
s
t
Dist
Dist中选取最小正数对应的节点,选中该节点,更新
D
i
s
t
Dist
Dist值
Step3:循环Step2直到所有节点都被选中
Trick:不断找距离最近的下一个节点的过程可以用heap来实现加速。
class Solution:
def minCostConnectPoints(self, points: List[List[int]]) -> int:
p1 = points.pop(0)
visited = [p1]
dist = []
res = 0
while points:
for p2 in points:
heapq.heappush(dist,(abs(p1[0]-p2[0])+abs(p1[1]-p2[1]),p2))
d,p2 = heapq.heappop(dist)
while p2 in visited:
d,p2 = heapq.heappop(dist)
points.remove(p2)
visited.append(p2)
p1 = p2
res += d
return res
Dijkstra算法
注意注意,DJ算法跟MST是不一样的,DJ算法的目标是找出某个节点到其他各节点的最短距离。是最短路算法。而MST是在图里找出联通的最小生成树。
具体算法可以参考https://www.zhihu.com/question/20630094/answer/758191548
今天先不写了,累了,ヾ(•ω•`)o拜拜