Outline and Reading Campus Tour

Overview of the assignment Review

Adjacency matrix structure (6.2.3) Kruskals MST algorithm (7.3.1)

Partition ADT and implementation (4.2.2) The decorator pattern (6.5.1) The traveling salesperson problem

Definition Approximation algorithm (13.4.3)

6/8/2002 2:07 PM

Campus Tour

6/8/2002 2:07 PM

Campus Tour

Graph Assignment
Goals

Adjacency Matrix Structure

Edge list structure Augmented vertex objects

a u

b w

Learn and implement the adjacency matrix structure an Kruskals minimum spanning tree algorithm Understand and use the decorator pattern and various JDSL classes and interfaces Implement the adjacency matrix structure for representing a graph Implement Kruskals MST algorithm Computation and visualization of an approximate traveling salesperson tour
Campus Tour 3

Integer key (index) associated with vertex 0 u 1 0 0 1 a 2 b v 1 2 2 w

Your task

2D-array adjacency array

Frontend

Reference to edge object for adjacent vertices Null for non nonadjacent vertices

6/8/2002 2:07 PM

6/8/2002 2:07 PM

Campus Tour

Kruskals Algorithm
The vertices are partitioned into clouds

Example
B 1 A 5 7 10 8 9 C 11 E 3 D H G 4 6 2 A F 1 B 5 7 10 8 9 C 11 E 3 D H G 4 6 2 F

We start with one cloud per vertex Clouds are merged during the execution of the algorithm makeSet(o): create set {o} and return a locator for object o find(l): return the set of the object with locator l union(A,B): merge sets A and B

Algorithm KruskalMSF(G) Input weighted graph G Output labeling of the edges of a minimum spanning forest of G Q new heap-based priority queue for all v G.vertices() do l makeSet(v) { elementary cloud } setLocator(v,l) for all e G.edges() do Q.insert(weight(e), e) while Q.isEmpty() e Q.removeMin() [u,v] G.endVertices(e) A find(getLocator(u)) B find(getLocator(v)) if A B setMSFedge(e) { merge clouds } union(A, B)
5

Partition ADT:

B 1 A 5 7

8 9 C 11 10 E 3 D

G 4 6 H 2 A
Campus Tour

B F 1 5 7

8 9 C 11 10 E 3 D

G 4 6 H 2 F

6/8/2002 2:07 PM

Campus Tour

6/8/2002 2:07 PM

Example (contd.)
B 1 A 5 7 10 8 9 C 11 E 3 D H G 4 6 2 A F 1 B 5 7 10 8 9 C 11 E 3 D H G 4 6 2 F

Partition Implementation
Partition implementation

st e

ps

four steps
B 1 A 5 7 10 8 9 C 11 E 3 D H G 4 6 2 F

B 1 A 5 7

8 9 C 11 10 E 3 D

4 6 H 2 F

tw o

A set is represented the sequence of its elements A position stores a reference back to the sequence itself (for operation find) The position of an element in the sequence serves as locator for the element in the set In operation union, we move the elements of the smaller sequence into to the larger sequence makeSet, find: O(1) union: O(min(nA, nB))
Campus Tour

Amortized analysis

Worst-case running times

Consider a series of k Partiton ADT operations that includes n makeSet operations Each time we move an element into a new sequence, the size of its set at least doubles An element is moved at most log2 n times Moving an element takes O(1) time The total time for the series of operations is O(k + n log n)
8

6/8/2002 2:07 PM

Campus Tour

6/8/2002 2:07 PM

Analysis of Kruskals Algorithm

Graph operations

Decorator Pattern
Labels are commonly used in graph algorithms

Methods vertices and edges are called once Method endVertices is called m times We perform m insert operations and m removeMin operations We perform n makeSet operations, 2m find operations and no more than n 1 union operations

Priority queue operations

Auxiliary data Output DFS: unexplored/visited label for vertices and unexplored/ forward/back labels for edges Dijkstra and Prim-Jarnik: distance, locator, and parent labels for vertices Kruskal: locator label for vertices and MSF label for edges
Campus Tour

The decorator pattern extends the methods of the Position ADT to support the handling of attributes (labels)

Examples

Partition operations

Label operations

We set vertex labels n times and get them 2m times

Kruskals algorithm runs in time O((n + m) log n) time provided the graph has no parallel edges and is represented by the adjacency list structure
6/8/2002 2:07 PM Campus Tour 9

has(a): tests whether the position has attribute a get(a): returns the value of attribute a set(a, x): sets to x the value of attribute a destroy(a): removes attribute a and its associated value (for cleanup purposes)

The decorator pattern can be implemented by storing a dictionary of (attribute, value) items at each position
10

6/8/2002 2:07 PM

Traveling Salesperson Problem

A tour of a graph is a spanning cycle (e.g., a cycle that goes through all the vertices) A traveling salesperson tour of a weighted graph is a tour that is simple (i.e., no repeated vertices or edges) and has has minimum weight No polynomial-time algorithms are known for computing traveling salesperson tours The traveling salesperson problem (TSP) is a major open problem in computer science

TSP Approximation
4

B 2 A 5 6 C

7 2 8 1

D F E 3

We can approximate a TSP tour with a tour of at most twice the weight for the case of Euclidean graphs

Find a polynomial-time algorithm computing a traveling salesperson tour or prove that none exists
Campus Tour

Example of traveling salesperson tour (with weight 17)

Vertices are points in the plane Every pair of vertices is connected by an edge The weight of an edge is the length of the segment joining the points Compute a minimum spanning tree Form an Eulerian circuit around the MST Transform the circuit into a tour
Campus Tour 12

Approximation algorithm

6/8/2002 2:07 PM

11

6/8/2002 2:07 PM