Single Source Shortest Path

Dijkstra’s Algorithm

Instructor: Ashok Singh Sairam

Lecture Plan
• Dijkstra’s Algorithm
▪ Algorithm
▪ Example
▪ Complexity
• Linear Programming
▪ System of difference constraints

MA512: Data Structures and Algorithms

2
Dijkstra’s Algorithm
• Single-source shortest path problem:
▪ No negative-weight edges: w(u, v) > 0,  (u, v)  E

• Each edge is relaxed only once! Lower running time

• Maintains two sets of vertices:
V

S V-S
d[v]=δ (s, v) d[v]>δ (s, v)
MA512: Data Structures and Algorithms
3
Dijkstra’s Algorithm: Data structure
• Vertices in V – S reside in a min-priority queue
▪ Keys in Q are estimates of shortest-path weights d[u]

• Repeatedly select a vertex u  V – S, with the

minimum shortest-path estimate d[u]
• Relax all edges leaving u

MA512: Data Structures and Algorithms

4
Dijkstra’s Algorithm

9 Update Q (DECREASE_KEY)

MA512: Data Structures and Algorithms

5
Example (1)
S=<> Q=<s,t,x,z,y> S=<s> Q=<y,t,x,z>
t 1 x t x
1
  
10 
10 9
10 9
2 3 4 6
s 0 2 3 4 6
s 0
5 7
5 7
   
2 5 2
y z y z

MA512: Data Structures and Algorithms

6
Example (2)
t 1 x t 1 x
8
10 
14 8 13
14
10 9 10 9
2 3 4 6 2 3 4 6
s 0 s 0
5 7 5 7
5 
7 5 7
2 2
y z y z

S=<s,y> Q=<z,t,x> S=<s,y,z> Q=<t,x>

MA512: Data Structures and Algorithms

7
Example (3)

S=<s,y,z,t> Q=<x> S=<s,y,z,t,x> Q=<>

t x t 1 x
1
8 13
9 8 9
10 9 10 9

2 4 2 3 4 6
s 0 3 6 s 0

7 5 7
5
5 7 5 7
2 2
y z y z

MA512: Data Structures and Algorithms

8
Dijkstra’s Algorithm: complexity

(V)

(V) O(VlgV)
O(lgV)
O(E) times
(total) O(ElgV)
9 Update Q (DECREASE_KEY) O(lgV)

Running time: O(VlgV + ElgV) = O(ElgV)

MA512: Data Structures and Algorithms
9 9
Exercise

MA512: Data Structures and Algorithms

10
Linear Programming
• Suppose you are given
▪ A matrix A with m rows and n columns
▪ A vector 𝑏 of length m
▪ A vector 𝑐Ԧ of length n
• Find a length-n vector 𝑥Ԧ such that
T
is as large (or small) as possible

such that

MA512: Data Structures and Algorithms

11
LP: Pictorial view
• The matrix inequality: 𝐴𝑥Ԧ ≤ 𝑏

• Each row of A gives coefficients of a linear expression:

• Each row of A along with an entry of b specifies a

linear inequality:
MA512: Data Structures and Algorithms
12
Ex:
• Given directed graph G = (V, E) with capacities ce and cost qe
on edge e. Sending 𝛼 units of flow on edge e costs 𝛼qe.
Find a flow that gets r units of flow from s to t, and minimizes
the cost

MA512: Data Structures and Algorithms

13
Feasibility Problem
• We wish to find any feasible solution
▪ That is any path from s to t that gets r flows. Not
concerned with the cost
• That is we wish to solve
𝐴𝑥Ԧ ≤ 𝑏

MA512: Data Structures and Algorithms

14
System of difference constraints
• Linear programming where each row of A contains
exactly one 1, one –1, and the rest 0’s.
▪ Example

MA512: Data Structures and Algorithms

15
Constraint graph
• Given 𝐴𝑥Ԧ ≤ 𝑏, the corresponding constraint graph is
(transpose of an incidence matrix) a weighted
directed graph G=(V,E) where
▪ Each vertex corresponds to one of the n unknown variables
▪ Each edge corresponds to the m inequalities

MA512: Data Structures and Algorithms

16
Ex: Constraint graph

One feasible solution: x = (-5, -3, 0, -1, -4)

MA512: Data Structures and Algorithms
17
Unsatisfiable constraints
• Theorem. If the constraint graph contains a negative-
weight cycle, then the system of differences is
unsatisfiable.

MA512: Data Structures and Algorithms

18
Satisfying the constraints

MA512: Data Structures and Algorithms

19
MA512: Data Structures and Algorithms
20
Acknowledgement
• Erik D. Demaine and Charles E. Leiserson

MA512: Data Structures and Algorithms

21