BITS F232

aft
Foundations of
Data Structures and Algorithms
Lect 27,28
Sameera Muhamed Salam
Dept of CSIS, BITS Pilani, Hyderabad Campus

Dr
Graph Traversals

BITS F232 Sameera M S 2

Depth First Search in an undirected graph
▶ Used for finding
▶ Path from one vertex to another
▶ Testing Connectivity
▶ To compute Spanning tree
▶ To compute cycle in G
▶ Searching starts from a Starting node s
▶ Mark it as visited
▶ select one arbitrary edge e incident on s
▶ let u be the other endpoint of e. Mark u as visited
▶ select an arbitrary edge e incident on u
▶ let v be the other endpoint of e
▶ If v is already visited, backtrack to u to find another unvisited vertex from v
▶ Repeat this process untill we reach a dead end (a vertex whose all neighbours are
visited)
▶ Then we backtrack along the visited nodes until we find a vertex with an unvisited
neighbour or we reach the start vertexSameera
BITS F232
s. M S 3
DFS: Pseudo code

BITS F232 Sameera M S 4

Recursive DFS

BITS F232 Sameera M S 5

DFS: Running Time
Theorem
Let G be an undirected graph on which a DFS Traversal starting from a vertex s has
been performed. Then the traversal visits all the vertices in the connected component
of s and the discovery edges forms a spanning tree of the connected component of s

▶ DFS is called exactly once on each vertex

▶ Each edge is visited exactly twice
▶ If ms is the number of edges in the connected component of s then DFS starting
at s takes O(ms ) time provided
▶ incidentEdges(v ) takes O(degree(v )) time
▶ hasNext() and nextEdge() returned by incidentEdges(v ) each takes O(1) time
▶ opposite(v , e) takes O(1) time
▶ Which implementation satisfies all these conditions?

BITS F232 Sameera M S 6

DFS: Running Time
Theorem
Let G be an undirected graph on which a DFS Traversal starting from a vertex s has
been performed. Then the traversal visits all the vertices in the connected component
of s and the discovery edges forms a spanning tree of the connected component of s

▶ DFS is called exactly once on each vertex

▶ Each edge is visited exactly twice
▶ If ms is the number of edges in the connected component of s then DFS starting
at s takes O(ms ) time provided
▶ incidentEdges(v ) takes O(degree(v )) time
▶ hasNext() and nextEdge() returned by incidentEdges(v ) each takes O(1) time
▶ opposite(v , e) takes O(1) time
▶ Which implementation satisfies all these conditions?
▶ Adjacency List
▶ Marking a vertex can be done in O(1) time.
BITS F232 Sameera M S 7
DFS: Running Time

DFS: Running Time

Let G be a graph with n vertices and m edges represented with adjacency list
structure. A DFS traversal of G can be performed in O(m + n) time. There exists a
O(m + n) time for the following problems
▶ Testing whether G is connected
▶ Computing Spanning forest of G
▶ Computing connected components of G
▶ Computing path between two vertices in G (if exists)
▶ Computing a cycle in G (if exists)

BITS F232 Sameera M S 8

Biconnected Components
▶ cut vertex or seperating vertex: A vertex whose removal disconnects the graph
▶ Bridge or seperating edge: An edge whose removal disconnects the graph
▶ Biconnected Component of G: satisfies any one of the following
▶ A subgraph of G which is biconnected and adding any additional vertices or edges
would force it to stop being biconnected
▶ A seperating edge and its end points

▶ Let G be a connected graph. The

following are equivalent
▶ G is biconnected
▶ For any 2 vertices of G, there is a
simple cycle containing them
▶ G does not have any seperation
edges or seperation vertices
Biconnected Components are important in Computer networks where vertices are
routers and edges represents connections.
BITS F232 Sameera M S 9
Breadth-First Search
▶ BFS is less adventurous than DFS
▶ Proceeds in rounds and subdivides vertices in to levels

BITS F232 Sameera M S 10

Discovery edges, Tree Edges and Back Edges
If you orient edges along the direction in which they are explored,edges in graph can be
divided as the following
▶ Discovery edges: Edges used to discover new vertices
▶ Tree edges: Edges in the DFS/BFS tree
▶ Back edges: current vertex to already visited vertex (DFS)
▶ Cross edges: current vertex to already visited vertex (BFS)

Theorem

BITS F232 Sameera M S 11

Running Time

BITS F232 Sameera M S 12

Compairing DFS and BFS

▶ BFS is better for finding shortest path

▶ BFS produces spanning tree such that all non tree edges are cross edges
▶ DFS produces spanning tree such that all non tree edges are back edges
▶ DFS is good to find disjoint paths between vertices

BITS F232 Sameera M S 13

Directed graphs and Reachability

→
−
we represent a directd graph or a digraph as G
→
−
▶ Given two vertices u and v in a digraph(Directed graph) G , we say u reaches v or
→
−
v is reachable from u if G has a directed path from u to v
→
−
▶ G is strongly connected if, for any two pair of vertices u and v
▶ u reaches v and v reaches u
→
− →
−
▶ G is acyclic if G does not contain directed cycles
−
→
▶ Transitive Closure (G ∗ ):
→
− −
→
▶ All vertices of G are vertices in G ∗
−
→∗ →
−
▶ uv is an edge in G whenever G has adirected path from u to v.

BITS F232 Sameera M S 14

Directed graph:Example

BITS F232 Sameera M S 15

Interesting Problems in Digraphs

Trvaersing a diagraph can be done using BFS and DFS. Directed DFS partitioned
edges into following
▶ Tree edges or discovery edges
▶ non tree edges
▶ Back edges: Connects a vertex to its ancestor in the tree
▶ Forward edges: Connects a vertex to its descendant in the tree
▶ Cross edges: Connects a vertex to vertex which is neither its ancestor nor its
descendant in the tree
▶ There will not be any forwarded edges after traversing in a BFS Manner.
BITS F232 Sameera M S 16
Testing for Strong Connectivity

▶ Can check for strong connectivity faster than O(n(n + m)) time using two DFS.

BITS F232 Sameera M S 17

Testing for Strong Connectivity

▶ Can check for strong connectivity faster than O(n(n + m)) time using two DFS.
→
−
▶ Do DFS from an arbitraty vertex s in G
→
− →
−
▶ If there is any vertex v in G that is not visited and is not reachable from s then G is
not strongly connected
▶ Reverse direction of all edges
▶ Do another DFS starting at s
→
−
▶ If every vertex is visited by second DFS also then G is strongly connected.
▶ Running time is O(n + m) time
BITS F232 Sameera M S 18
Dynamic Programming
▶ Like the divide-and-conquer method
▶ Solves problems by combining the solutions to subproblems
▶ Divide-and-conquer algorithms partition the problem into independent
subproblems, solve the subproblems recursively, and then combine their solutions
to solve the original problem.
▶ In contrast, dynamic programming is applicable when the subproblems are not
independent, that is, when subproblems share subsubproblems.
▶ A dynamic-programming algorithm solves every subsubproblem just once and then
saves its answer in a table, thereby avoiding the work of recomputing the answer
every time the subsubproblem is encountered.
▶ Dynamic programming is typically applied to optimization problems.
▶ In such problems there can be many possible solutions.
▶ We wish to find a solution with the optimal (minimum or maximum) value. We
call such a solution an optimal solution to the problem.
BITS F232 Sameera M S 19
Dynamic Programming

BITS F232 Sameera M S 20

Transitive closure using Dynamic Programming

▶ Constructing using series of rounds

→
−
▶ Number the vertices in G as v1 , v2 , . . . , vn
→
− →
−
▶ We start with G 0 = G
→
− →
−
▶ In each step k we start G k = G k−1
→
−
▶ Add edge (vi , vj ) if G k−1 contains (vi , vk ) and (vk , vj )

Lemma
→
− →
−
For i − 1, 2, . . . , n, disgraph G k has an edge (vi , vj ), iff G has a directed path from vi
to vj whose intermediate vertices are in the set {v1 , v2 , . . . , vk }. In particular,
→
− →
−
Gn = G∗

BITS F232 Sameera M S 21

Floyd-Warshall Algorithm

Running Time
▶ Adjacency Matrix: n DFS takes O(n3 ) time
▶ Adjacency List: O(n(n + m)). For dense graph O(n3 ) time
BITS F232 Sameera M S 22
Example

BITS F232 Sameera M S 23

Example

BITS F232 Sameera M S 24

Example

BITS F232 Sameera M S 25

References:

▶ Michael T. Goodrich and Roberto Tamassia, “ Algorithm Design Foundations,

Analysis, and Internet Examples”, Wiley Student Edition.
▶ Jon Kleinberg and Eva Tardos, “ Algorithm Design”, Pearson Publishers.
▶ Thomas H. Chorman, Charles E. Leiserson, Ronald L. Rivest and Clifford Stein, “
Introduction to algorithms”, MIT Press Publishers.
▶ Sanjoy Dasgupta, Umesh Vazirani, Christos Papadimitriou, “ Algorithms”,
McGraw-Hill Education Publishers.

BITS F232 Sameera M S 26

Discussion Time!

BITS F232 Sameera M S 27

BITS F232 Sameera M S 28