CS 225 Data Structures No Novem ember er 20 20 MS MSTs: - - PowerPoint PPT Presentation

▶
cs 225
SMART_READER_LITE
LIVE PREVIEW

CS 225 Data Structures No Novem ember er 20 20 MS MSTs: - - PowerPoint PPT Presentation

Sep 09, 2023 223 likes •375 views

CS 225 Data Structures No Novem ember er 20 20 MS MSTs: Kruskal + Prims Algorithm G G Carl Evans Kr Kruskals Algorithm (A, D) B 15 1 KruskalMST(G): 5 (E, H) 2 DisjointSets forest C 5 16 A (F, G) 3 foreach (Vertex

slide-1
SLIDE 1

CS 225

Data Structures

No Novem ember er 20 20 – MS MSTs: Kruskal + Prim’s Algorithm

G G Carl Evans

slide-2
SLIDE 2

Kr Kruskal’s Algorithm

A C D E B F G H 16 5 5 2 15 16 10 11 12 2 8 9 4 12 17 13

KruskalMST(G): DisjointSets forest foreach (Vertex v : G): forest.makeSet(v) PriorityQueue Q // min edge weight foreach (Edge e : G): Q.insert(e) Graph T = (V, {}) while |T.edges()| < n-1: Vertex (u, v) = Q.removeMin() if forest.find(u) != forest.find(v): T.addEdge(u, v) forest.union( forest.find(u), forest.find(v) ) return T 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

(A, D) (E, H) (F, G) (A, B) (B, D) (G, E) (G, H) (E, C) (C, H) (E, F) (F, C) (D, E) (B, C) (C, D) (A, F) (D, F) A C D E B F G H

slide-3
SLIDE 3

Kr Kruskal’s Algorithm

KruskalMST(G): DisjointSets forest foreach (Vertex v : G): forest.makeSet(v) PriorityQueue Q // min edge weight foreach (Edge e : G): Q.insert(e) Graph T = (V, {}) while |T.edges()| < n-1: Vertex (u, v) = Q.removeMin() if forest.find(u) != forest.find(v): T.addEdge(u, v) forest.union( forest.find(u), forest.find(v) ) return T 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

Priority Queue: Heap Sorted Array Building :6-8 Each removeMin :13

slide-4
SLIDE 4

Kr Kruskal’s Algorithm

KruskalMST(G): DisjointSets forest foreach (Vertex v : G): forest.makeSet(v) PriorityQueue Q // min edge weight foreach (Edge e : G): Q.insert(e) Graph T = (V, {}) while |T.edges()| < n-1: Vertex (u, v) = Q.removeMin() if forest.find(u) != forest.find(v): T.addEdge(u, v) forest.union( forest.find(u), forest.find(v) ) return T 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

Priority Queue: Total Running Time Heap Sorted Array

slide-5
SLIDE 5

Kr Kruskal’s Algorithm

Which Priority Queue Implementation is better for running Kruskal’s Algorithm?

  • Heap:
  • Sorted Array:
slide-6
SLIDE 6

Pa Partition Property

Consider an arbitrary partition of the vertices on G into two subsets U and V.

A C D E B F 8 4 2 7 1 2 3 9 5

U V

slide-7
SLIDE 7

Pa Partition Property

Consider an arbitrary partition of the vertices on G into two subsets U and V.

A C D E B F 8 4 2 7 1 2 3 9 5

U V

Let e be an edge of minimum weight across the partition. Then e is part of some minimum spanning tree.

e

slide-8
SLIDE 8

Pa Partition Property

The partition property suggests an algorithm:

A C D E B F G H 16 5 5 2 15 16 10 11 8 9 12 4 17 13 9

slide-9
SLIDE 9

Pr Prim’s Algorithm

A C D E B F 16 5 2 15 11 8 12 17 13 9

PrimMST(G, s): Input: G, Graph; s, vertex in G, starting vertex Output: T, a minimum spanning tree (MST) of G foreach (Vertex v : G): d[v] = +inf p[v] = NULL d[s] = 0 PriorityQueue Q // min distance, defined by d[v] Q.buildHeap(G.vertices()) Graph T // "labeled set" repeat n times: Vertex m = Q.removeMin() T.add(m) foreach (Vertex v : neighbors of m not in T): if cost(v, m) < d[v]: d[v] = cost(v, m) p[v] = m return T 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

slide-10
SLIDE 10

Pr Prim’s Algorithm

PrimMST(G, s): foreach (Vertex v : G): d[v] = +inf p[v] = NULL d[s] = 0 PriorityQueue Q // min distance, defined by d[v] Q.buildHeap(G.vertices()) Graph T // "labeled set" repeat n times: Vertex m = Q.removeMin() T.add(m) foreach (Vertex v : neighbors of m not in T): if cost(v, m) < d[v]: d[v] = cost(v, m) p[v] = m 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

  • Adj. Matrix
  • Adj. List

Heap Unsorted Array

slide-11
SLIDE 11

Pr Prim’s Algorithm

PrimMST(G, s): foreach (Vertex v : G): d[v] = +inf p[v] = NULL d[s] = 0 PriorityQueue Q // min distance, defined by d[v] Q.buildHeap(G.vertices()) Graph T // "labeled set" repeat n times: Vertex m = Q.removeMin() T.add(m) foreach (Vertex v : neighbors of m not in T): if cost(v, m) < d[v]: d[v] = cost(v, m) p[v] = m 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

  • Adj. Matrix
  • Adj. List

Heap

O(n2 + m lg(n)) O(n lg(n) + m lg(n))

Unsorted Array

O(n2) O(n2)

Sparse Graph: Dense Graph:

slide-12
SLIDE 12

MS MST Algori

  • rithm

m Ru Runtime me:

  • Kruskal’s Algorithm:

O(n + m lg(n))

  • What must be true about the connectivity of a graph

when running an MST algorithm?

  • How does n and m relate?
  • Prim’s Algorithm:

O(n lg(n) + m lg(n))

slide-13
SLIDE 13

MS MST Algori

  • rithm

m Ru Runtime me:

  • Kruskal’s Algorithm:

O(n + m lg(n))

  • Prim’s Algorithm:

O(n lg(n) + m lg(n))