CS311

Algorithm Design and

Analysis
LECTURE 11 : GRAPH ALGORITHMS (3)
Contents
▪ Shortest-path problem
▪ Single-Source Shortest-Path
▪ Relaxation
▪ Dijkstra’s algorithm
▪ All-Pairs Shortest Path Problem
▪ Floyd-Warshall algorithm

ALGORITHM DESIGN AND ANALYSIS 2

Shortest-Path Problem
▪ Shortest path = a path with the minimum weight.
▪ Applications
➢Static/Dynamic network routing
➢Robot motion planning
➢Map/Route generation in traffic
▪ The input to a shortest-paths problem is a weighted, directed graph G = (V, E),
with a weight function W: E →R (assigning real values to edges).
▪ Weight of path p = v1 →v2 →… →vk is
k −1
w( p) =  w(vi , vi +1 )
i =1

ALGORITHM DESIGN AND ANALYSIS 3

Single-Source Shortest-Path problems
▪ Given a graph G = (V, E) , find a shortest path from a given source vertex s ϵ V to every
vertex v ϵ V .
▪ The algorithm for the single-source problem can solve many other problems, including
the following variants:

➢Single-destination shortest-paths problem: Find a shortest path

to a given destination vertex t from each vertex v .
➢Single-pair shortest-paths problem : Find a shortest path from u
to v for given vertices u and v.
➢All-pairs shortest-paths problem: Find a shortest path from u to v
for every pair of vertices u and v.

ALGORITHM DESIGN AND ANALYSIS 4

Shortest-path Estimate
For each vertex v ϵ V , the single-source shortest paths algorithms maintain an
attribute v.d, represents the weight of a shortest path from source s to v .

ALGORITHM DESIGN AND ANALYSIS 5

Relaxation

ALGORITHM DESIGN AND ANALYSIS 6

Dijkstra’s algorithm
▪ Dijkstra’s algorithm solves the single-source shortest-paths problem on a weighted,
directed graph, but it requires nonnegative weights on all edges.
▪ The running time of Dijkstra’s algorithm is lower than that of the Bellman-Ford
algorithm.
▪ Greedy, similar to Prim's algorithm for MST.
▪ Uses Q, a priority queue ADT keyed by v.d, which is re-organized whenever some d
decreases.
▪ Basic idea
✓Maintain a set S of solved vertices whose SP from s has been determined.
✓Repeatedly: Selects u ϵ V–S with the minimum SP estimate (greedy choice).
✓Then, adds u into S , and relaxes all edges leaving u .

ALGORITHM DESIGN AND ANALYSIS 7

Dijkstra’s algorithm

ALGORITHM DESIGN AND ANALYSIS 8

Dijkstra’s algorithm

V d π
S 0 NIL
Q = {s , t , y , x , z} T  NIL

S= Y  NIL
X  NIL
z  NIL

ALGORITHM DESIGN AND ANALYSIS 9

Dijkstra’s algorithm

V d π
S 0 NIL
Extract-MIN(Q)  s T 
10 S
NIL
Q = {s, t , y , x , z} Y 
5 S
NIL
S = {s} X  NIL
z  NIL

ALGORITHM DESIGN AND ANALYSIS 10

Dijkstra’s algorithm

V d π
Extract-MIN(Q)  y S 0 NIL
Q = { y, t , x, z } T 8
10 Y
S
S = {s, y} Y 5 S
X 
14 Y
NIL
Zz 
7 Y
NIL

ALGORITHM DESIGN AND ANALYSIS 11

Dijkstra’s algorithm

V d π
Extract-MIN(Q)  z S 0 NIL
Q = { z, t , x } T 8 Yy
S = {s, y, z} Y 5 S
X 13
14 YZ
z 7 Yy

ALGORITHM DESIGN AND ANALYSIS 12

Dijkstra’s algorithm

V d π
Extract-MIN(Q)  t S 0 NIL
Q = { t , x} T 8 y
S = {s, y, z, t} Y 5 S
X 9
13 ZT
z 7 y

ALGORITHM DESIGN AND ANALYSIS 13

Dijkstra’s algorithm

V d π
Extract-MIN(Q)  x S 0 NIL
Q={x} T 8 y
S = {s, y, z, t, x} Y 5 S
X 9 T
z 7 y

ALGORITHM DESIGN AND ANALYSIS 14

 Dijkstra’s algorithm Analysis
O(V)
Q is a Min-heap

O(V lg V)
The total running time is
O(V lg V)
O(E lg V)
V
in the typical case that
E O(E lg V)
E = Ω (V)
O(lg V)

ALGORITHM DESIGN AND ANALYSIS 15

All-Pairs Shortest Path Problem
▪ The input is a weighted, directed graph G =(V , E) with a weight function W that
maps edges to real-valued weights.

▪ The goal is to find, for every pair of vertices u,v ϵ V , a shortest (least-weight)
path from u to v.

▪ We can solve it by running a single-source shortest-paths algorithm V times,

once with each vertex as the source.

ALGORITHM DESIGN AND ANALYSIS 16

All-Pairs Shortest Path Problem
▪ If all edge weights are nonnegative, we can use Dijkstra’s algorithm:

➢ The priority queue implemented as Min-heap, the running time is

O(VE lg V).

▪Is there an algorithm that guarantee a much better asymptotic running time??

ALGORITHM DESIGN AND ANALYSIS 17

All-Pairs Shortest Path Problem
▪ Represent the graph by an adjacency matrix.
▪ The input is an n X n matrix W = (wij) representing the edge weights of an n-vertex, where

Example :

ALGORITHM DESIGN AND ANALYSIS 18

Floyd-Warshall algorithm
▪ The Floyd-Warshall algorithm presents a dynamic-programming solution to the all-
pairs shortest path problem runs in Θ(V3) time.
▪ Let dij(k) be the weight of a shortest path from vertex i to vertex j for which all
intermediate vertices belong to the set {1, 2, …, k}.
▪ When k = 0 , a path from vertex i to vertex j with no intermediate vertex. The path has
at most one edge.

ALGORITHM DESIGN AND ANALYSIS 19

Floyd-Warshall algorithm
The matrix dij(k) is defined recursively by:

Because for any path, all intermediate vertices belong to the set {1, 2, … , n) ,
the matrix D(n) = dij(n) gives the final answer.
Example:
d3,2(1) is the min (d3,2 (0) , d3,1(0) + d1,2(0) )
∞ 12

ALGORITHM DESIGN AND ANALYSIS 20

Floyd-Warshall algorithm
D(1):

d2,3 (1) is the min (d2,3 (0) , d2,1(0) + d1,3(0) )

d2,4 (1) is the min (d2,4 (0) , d2,1(0) + d1,4(0) )
d3,2 (1) is the min (d3,2 (0) , d3,1(0) + d1,2(0) )

ALGORITHM DESIGN AND ANALYSIS 21

Floyd-Warshall algorithm
D(2):

D(1):

ALGORITHM DESIGN AND ANALYSIS 22

Floyd-Warshall algorithm
D(3):

D(4):
D(3)

ALGORITHM DESIGN AND ANALYSIS 23

Floyd-Warshall algorithm

Final
Answer

ALGORITHM DESIGN AND ANALYSIS 24

Floyd-Warshall algorithm

O(1)

▪ The running time of the Floyd-Warshall algorithm is determined by the triply nested
for loops.
▪ The algorithm runs in Θ(n3) time.

ALGORITHM DESIGN AND ANALYSIS 25

Questions

ALGORITHM DESIGN AND ANALYSIS 26