NU-Lec 19 Graphs Search

Data Structures
Graphs – BFS & DFS

Graph Searching
Given: a graph G = (V, E), directed or undirected

Goal: methodically explore every vertex and every edge
Two main approaches:

Breadth-First Search
Depth-First Search
Some Applications of BFS and DFS
• BFS
• Find the shortest path from a vertex s to a vertex v
• Find the length of such a path
• Find if a strongly connected directed graph contains cycles
• Construct a BSF tree/forest from a graph
• DFS
• Find a path from a vertex s to a vertex v.
• Find the length of such a path.
• Construct a DSF tree/forest from a graph.
• Visit all vertices and edges
3
Breadth First Search
When search is limited to essentially two operations:
• Visit and inspect a node of a graph
• Gain access to visit the nodes that neighbor the

currently visited node
4
The algorithm uses a queue data structure to store

intermediate results as it traverses the graph, as follows:
1. Enqueue the root node
2. Dequeue a node and examine it
1.If the element required is found in this node, quit the
search and return a result.
2. Otherwise enqueue any successors (the direct child
nodes) that have not yet been discovered.
3. If the queue is empty, every node on the graph has been
examined – quit the search and return "not found".
4. If the queue is not empty, repeat from Step 2.
5
2 4 8
s 5 7
3 6 9
6
2 4 8
0 s 5 7
3 6 9
Undiscovered
Discovered Queue: s
Top of queue
Finished
7
2 4 8
0 s 5 7
3 6 9
Undiscovered
Discovered Queue: s 2
Top of queue
Finished
8
2 4 8
0 s 5 7
3 6 9
Undiscovered
Discovered Queue: s 2 3
Top of queue
Finished
9
2 4 8
0 s 5 7
3 6 9
Undiscovered
Discovered Queue: 2 3 5
Top of queue
Finished
10
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
11
2 4 8
5 already discovered:
0 s 5 7
don't enqueue
3 6 9
Undiscovered
Discovered Queue: 2 3 5 4
Top of queue
Finished
12
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
13
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
14
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
15
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
16
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
17
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
18
2 4 8
0 s 5 7
3 6 9
Undiscovered
Discovered Queue: 4 6
Top of queue
Finished
19
3
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
20
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
21
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
22
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
23
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
24
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
25
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
26
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
27
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
28
2 4 8
0 s 5 7
3 6 9
Undiscovered
Top of queue
Finished
29
2 4 8
0 s 5 7
3 6 9
Undiscovered
Discovered Queue: 9
Top of queue
Finished
30
2 4 8
0 s 5 7
3 6 9
Undiscovered
Discovered Queue: 9
Top of queue
Finished
31
2 4 8
0 s 5 7
3 6 9
Undiscovered
Discovered Queue: 9
Top of queue
Finished
32
2 4 8
0 s 5 7
3 6 9
Undiscovered
Discovered Queue:
Top of queue
Finished
33
2 4 8
0 s 5 7
3 6 9
Level Graph
34
Depth First Search
Depth-first search: Strategy
Go as deep as can visiting un-visited nodes
Choose any un-visited vertex when you have a choice
If stuck at a dead-end, backtrack as little as Possible
Back up to where you could go to another unvisited vertex

Then continue to go on from that point
Eventually you’ll return to where you started
Depth-first searching
A depth-first search (DFS)

A explores a path all the way
to a leaf before
B C
backtracking and exploring
another path
For example, after searching
D E F G
A, then B, then D, the
search backtracks and tries
H I J K another path from B
Node are explored in the order
L M N O P Q ABDEHLMNIOPCF
GJKQ
N will be found before J
36
DFS Algorithm
create a stack S
mark v as visited and push v onto S
while S is non-empty
pop at the top u of S
if u has an (unvisited)neighbour w,
mark w and push it onto S
else
pop S
Undirected Depth First Search
A
H I
B C G
Adjacency Lists
D E
A: F C B G
B: A
C: A
F D: F E
E: G F D
F: A E D:
G: E A:
H: I:
I: H:
38
A
H I
B C G
D E
F newly
discovered F
Undiscovered
Marked visit(A)
Active (A, F) (A, C) (A, B) (A, G)
Finished
Stack 39
A
H I
A already B C G
marked
D E
visit(F)
Undiscovered (F, A) (F, E) (F, D)
Marked visit(A)
Finished
Stack 40
A
H I
B C G
D E E newly
discovered
visit(F)
Marked visit(A)
Finished
Stack 41
A
H I
B C G G newly
discovered
D E
visit(E)
F (E, G) (E, F) (E, D)
visit(F)
Marked visit(A)
Finished
Stack 42
A
H I
E already B C G
marked
visit(G)
D E (G, E) (G, A)
visit(E)
F (E, G) (E, F) (E, D)
visit(F)
Marked visit(A)
Finished
Stack 43
A
H I
A already B C G
marked
visit(G)
D E (G, E) (G, A)
visit(E)
F (E, G) (E, F) (E, D)
visit(F)
Marked visit(A)
Finished
Stack 44
A
H I
Finished G
B C G
visit(G)
D E (G, E) (G, A)
visit(E)
F (E, G) (E, F) (E, D)
visit(F)
Marked visit(A)
Finished
Stack 45
A
H I
F already B C G
marked
D E
visit(E)
F (E, G) (E, F) (E, D)
visit(F)
Marked visit(A)
Finished
Stack 46
A
H I
B C G
D newly
D E
discovered
visit(E)
F (E, G) (E, F) (E, D)
visit(F)
Marked visit(A)
Finished
Stack 47
A
H I
F already B C G
marked
visit(D)
D E (D, F) (D, E)
visit(E)
F (E, G) (E, F) (E, D)
visit(F)
Marked visit(A)
Finished
Stack 48
A
H I
E already B C G
marked
visit(D)
D E (D, F) (D, E)
visit(E)
F (E, G) (E, F) (E, D)
visit(F)
Marked visit(A)
Finished
Stack 49
A
H I
Finished D
B C G
visit(D)
D E (D, F) (D, E)
visit(E)
F (E, G) (E, F) (E, D)
visit(F)
Marked visit(A)
Finished
Stack 50
A
H I
B C G
Finished E
D E
visit(E)
F (E, G) (E, F) (E, D)
visit(F)
Marked visit(A)
Finished
Stack 51
A
H I
D already B C G
marked
D E
visit(F)
Marked visit(A)
Finished
Stack 52
A
H I
B C G
D E
Finished F
F
visit(F)
Marked visit(A)
Finished
Stack 53
A
H I
C newly
B C G
discovered
D E
Undiscovered
Marked visit(A)
Finished
Stack 54
A
H I
A already B C G
marked
D E
visit(C)
Undiscovered (C, A)
Marked visit(A)
Finished
Stack 55
A
H I
B C G
D E
Finished C
F
visit(C)
Undiscovered (C, A)
Marked visit(A)
Finished
Stack 56
A
H I
B newly
B C G
discovered
D E
Undiscovered
Marked visit(A)
Finished
Stack 57
A
H I
A already B C G
marked
D E
visit(B)
Undiscovered (B, A)
Marked visit(A)
Finished
Stack 58
A
H I
B C G
D E
Finished B
F
visit(B)
Undiscovered (B, A)
Marked visit(A)
Finished
Stack 59
A
H I
G already B C G
finished
D E
Undiscovered
Marked visit(A)
Finished
Stack 60
A
H I
B C G
D E
Finished A
Undiscovered
Marked visit(A)
Finished
Stack 61
A
H I
B C G
D E
Undiscovered
Marked
Active
Finished
Stack 62
A
H I I newly
discovered
B C G
D E
Undiscovered
Marked visit(H)
Active (H, I)
Finished
Stack 63
A
H I
H already B C G
marked
D E
visit(I)
Undiscovered (I, H)
Marked visit(H)
Active (H, I)
Finished
Stack 64
A
H I
B C G
D E
Finished I
visit(I)
Undiscovered (I, H)
Marked visit(H)
Active (H, I)
Finished
Stack 65
A
H I
B C G
D E
F
FinishedH
Undiscovered
Marked visit(H)
Active (H, I)
Finished
Stack 66
A
H I
B C G
D E
Undiscovered
Marked
Active
Finished
Stack 67
Differences between DFS and BFS
Depending on the data and what you are looking

for, either DFS or BFS could be advantageous.
Consider your family tree
If you are searching for some of your siblings

cousins then it would be safe to assume that person
would be on the bottom of the tree.
Which approach is better in this case?
68
Consider your family tree
If you are searching for some of your ancestors

then it would be safe to assume that person would
be closer to the top of the tree.
Which approach is better in this case?
In general, both approaches have the same time

complexity. In worst case, they need to visit all the
nodes.
69
In terms of memory – DFS is better than BFS.
DFS needs to store only the path from the root to

the leaf node, beside the siblings of each node on
the path, remember that BFS needs to store all
the explored nodes in memory.
What can BFS do and DFS can’t (efficently)?

Finding shortest paths (in unweighted graphs)
70
References
Nell Dale -- Chapter 9

http://hci.iwr.uni-heidelberg.de/MIP/Teaching/alda/handouts-20-mst.pdf
http://www.cs.uah.edu/~rcoleman/Common/CodeVault/Code/Code301_Graphs.html
http://en.wikipedia.org/wiki/Breadth-first_search

NU-Lec 19 Graphs Search

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

NU-Lec 19 Graphs Search

Uploaded by

Copyright:

Available Formats

Data Structures

Graphs – BFS & DFS

Given: a graph G = (V, E), directed or undirected

Two main approaches:

• Visit and inspect a node of a graph

• Gain access to visit the nodes that neighbor the

The algorithm uses a queue data structure to store

Back up to where you could go to another unvisited vertex

A depth-first search (DFS)

Depending on the data and what you are looking

Consider your family tree

If you are searching for some of your siblings

Which approach is better in this case?

Consider your family tree

If you are searching for some of your ancestors

Which approach is better in this case?

In general, both approaches have the same time

In terms of memory – DFS is better than BFS.

DFS needs to store only the path from the root to

What can BFS do and DFS can’t (efficently)?

Nell Dale -- Chapter 9

You might also like