4.

G REEDY A LGORITHMS II

‣ Dijkstra’s algorithm demo

‣ Dijkstra’s algorithm demo
(efficient implementation)

Lecture slides by Kevin Wayne

Copyright © 2005 Pearson-Addison Wesley
http://www.cs.princeton.edu/~wayne/kleinberg-tardos

Last updated on 11/15/17 10:10 AM

4. G REEDY A LGORITHMS II

‣ Dijkstra’s algorithm demo

‣ Dijkstra’s algorithm demo
(efficient implementation)
Dijkstra’s algorithm demo

Initialize S ← { s } and d[s] ← 0.

Repeatedly choose unexplored node v ∉ S which minimizes

(v) = min d[u] + e

e = (u,v) : u S the length of a shortest path from s
to some node u in explored part S,
add v to S; set d[v] ← π(v) and pred[v] ← argmin. followed by a single edge e = (u, v)

d[s] 0

S s 16 2

4 8 7 5 6

3 1

3
Dijkstra’s algorithm demo

Initialize S ← { s } and d[s] ← 0.

Repeatedly choose unexplored node v ∉ S which minimizes

(v) = min d[u] + e

e = (u,v) : u S the length of a shortest path from s
to some node u in explored part S,
add v to S; set d[v] ← π(v) and pred[v] ← argmin. followed by a single edge e = (u, v)

d[s] 0 0 + 16 = 16

S s 16

4 8

0+4=4 0+8=8
4
Dijkstra’s algorithm demo

Initialize S ← { s } and d[s] ← 0.

Repeatedly choose unexplored node v ∉ S which minimizes

(v) = min d[u] + e

e = (u,v) : u S the length of a shortest path from s
to some node u in explored part S,
add v to S; set d[v] ← π(v) and pred[v] ← argmin. followed by a single edge e = (u, v)

0 0 + 16 = 16

s 16

S 8
pred[v]

v 3

4 4+3=7
0+8=8 5
Dijkstra’s algorithm demo

Initialize S ← { s } and d[s] ← 0.

Repeatedly choose unexplored node v ∉ S which minimizes

(v) = min d[u] + e

e = (u,v) : u S the length of a shortest path from s
to some node u in explored part S,
add v to S; set d[v] ← π(v) and pred[v] ← argmin. followed by a single edge e = (u, v)

0 + 16 = 16
0 7 + 7 = 14

s 16

S 7

4 7 7+1=8
6
Dijkstra’s algorithm demo

Initialize S ← { s } and d[s] ← 0.

Repeatedly choose unexplored node v ∉ S which minimizes

(v) = min d[u] + e

e = (u,v) : u S the length of a shortest path from s
to some node u in explored part S,
add v to S; set d[v] ← π(v) and pred[v] ← argmin. followed by a single edge e = (u, v)

0 + 16 = 16
7 + 7 = 14
0 8 + 5 = 13 8 + 6 = 14

s 16

S 7 5 6

4 7 8
7
Dijkstra’s algorithm demo

Initialize S ← { s } and d[s] ← 0.

Repeatedly choose unexplored node v ∉ S which minimizes

(v) = min d[u] + e

e = (u,v) : u S the length of a shortest path from s
to some node u in explored part S,
add v to S; set d[v] ← π(v) and pred[v] ← argmin. followed by a single edge e = (u, v)

13 + 2 = 15
0 13 8 + 6 = 14

s 2

S 6

4 7 8
8
Dijkstra’s algorithm demo

Initialize S ← { s } and d[s] ← 0.

Repeatedly choose unexplored node v ∉ S which minimizes

(v) = min d[u] + e

e = (u,v) : u S

add v to S; set d[v] ← π(v) and pred[v] ← argmin.

0 13 14

4 7 8
9
Dijkstra’s algorithm demo

Initialize S ← { s } and d[s] ← 0.

Repeatedly choose unexplored node v ∉ S which minimizes

(v) = min d[u] + e

e = (u,v) : u S

add v to S; set d[v] ← π(v) and pred[v] ← argmin.

0 13 14

pred[v]

π[v] 4 7 8
10
4. G REEDY A LGORITHMS II

‣ Dijkstra’s algorithm demo

‣ Dijkstra’s algorithm demo
(efficient implementation)
Dijkstra’s algorithm demo (efficient implementation)

Initialization.
For all v ≠ s : π [v] ← ∞.
For all v ≠ s : pred [v] ← null.
S ← ∅ and π [s] ← 0.

π[s] 0

s 16 2

4 8 7 5 6

3 1

12
Dijkstra’s algorithm demo (efficient implementation)

Basic step. Choose unexplored node u ∉ S with minimum π [u].

Add u to S.
For each edge e = (u, v) leaving u, if π [v] > π [u] + e then:
π [v] ← π [u] + e

pred[v] ← e

0 0 + 16 = 16

s 16

4 8

0+4=4 0+8=8
13
Dijkstra’s algorithm demo (efficient implementation)

Basic step. Choose unexplored node u ∉ S with minimum π [u].

Add u to S.
For each edge e = (u, v) leaving u, if π [v] > π [u] + e then:
π [v] ← π [u] + e

pred[v] ← e

0 0 + 16 = 16

s 16

4 8

pred[v]

0+4=4 0+8=8
14
Dijkstra’s algorithm demo (efficient implementation)

Basic step. Choose unexplored node u ∉ S with minimum π [u].

Add u to S.
For each edge e = (u, v) leaving u, if π [v] > π [u] + e then:
π [v] ← π [u] + e

pred[v] ← e

0 16

4 8
4+3=7 15
Dijkstra’s algorithm demo (efficient implementation)

Basic step. Choose unexplored node u ∉ S with minimum π [u].

Add u to S.
For each edge e = (u, v) leaving u, if π [v] > π [u] + e then:
π [v] ← π [u] + e

pred[v] ← e

0 16

4 8
4+3=7 16
Dijkstra’s algorithm demo (efficient implementation)

Basic step. Choose unexplored node u ∉ S with minimum π [u].

Add u to S.
For each edge e = (u, v) leaving u, if π [v] > π [u] + e then:
π [v] ← π [u] + e

pred[v] ← e

7 + 7 = 14
0 16

4 7 7+1=8
17
Dijkstra’s algorithm demo (efficient implementation)

Basic step. Choose unexplored node u ∉ S with minimum π [u].

Add u to S.
For each edge e = (u, v) leaving u, if π [v] > π [u] + e then:
π [v] ← π [u] + e

pred[v] ← e

7 + 7 = 14
0 16

4 7 7+1=8
18
Dijkstra’s algorithm demo (efficient implementation)

Basic step. Choose unexplored node u ∉ S with minimum π [u].

Add u to S.
For each edge e = (u, v) leaving u, if π [v] > π [u] + e then:
π [v] ← π [u] + e

pred[v] ← e

8 + 5 = 13
0 14 8 + 6 = 14

5 6

4 7 8
19
Dijkstra’s algorithm demo (efficient implementation)

Basic step. Choose unexplored node u ∉ S with minimum π [u].

Add u to S.
For each edge e = (u, v) leaving u, if π [v] > π [u] + e then:
π [v] ← π [u] + e

pred[v] ← e

8 + 5 = 13
0 14 8 + 6 = 14

5 6

4 7 8
20
Dijkstra’s algorithm demo (efficient implementation)

Basic step. Choose unexplored node u ∉ S with minimum π [u].

Add u to S.
For each edge e = (u, v) leaving u, if π [v] > π [u] + e then:
π [v] ← π [u] + e

pred[v] ← e

13 + 2 = 15
0 13 14

s 2

4 7 8
21
Dijkstra’s algorithm demo (efficient implementation)

Basic step. Choose unexplored node u ∉ S with minimum π [u].

Add u to S.
For each edge e = (u, v) leaving u, if π [v] > π [u] + e then:
π [v] ← π [u] + e

pred[v] ← e

0 13 14

4 7 8
22
Dijkstra’s algorithm demo (efficient implementation)

Basic step. Choose unexplored node u ∉ S with minimum π [u].

Add u to S.
For each edge e = (u, v) leaving u, if π [v] > π [u] + e then:
π [v] ← π [u] + e

pred[v] ← e

0 13 14

4 7 8
23
Dijkstra’s algorithm demo (efficient implementation)

Termination.
π [v] = length of a shortest s↝v path.
pred[v] = last edge on a shortest s↝v path.

0 13 14

pred[v]

π[v] 4 7 8
24