Scribd
Upload
0 views

Good Graph Questions

The document outlines exercises related to graph theory, specifically focusing on topological sorting, shortest and longest paths in directed acyclic graphs (DAGs), and semi-connected graphs. It provides algorithms for determining unique topological orders, performing topological sorting in O(n + m) time, and finding shortest and longest paths using topological ordering. Additionally, it discusses a practical application involving task dependencies in construction projects to determine completion time and critical tasks.

Uploaded by

mozhanhomavandi
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
0 views

Good Graph Questions

The document outlines exercises related to graph theory, specifically focusing on topological sorting, shortest and longest paths in directed acyclic graphs (DAGs), and semi-connected graphs. It provides algorithms for determining unique topological orders, performing topological sorting in O(n + m) time, and finding shortest and longest paths using topological ordering. Additionally, it discusses a practical application involving task dependencies in construction projects to determine completion time and critical tasks.

Uploaded by

mozhanhomavandi
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

ADS 1 — problem sheet 4

Exercise 1: How to determine if given graph has precisely one topological


order?
Solution: From lecture we know that the graph G has a topological order if
and only if it is a directed acyclic graph.
Let ≺ be a topological order. We say that two vertices a ≺ b are consecutive
if and only if there is no vertex c such that a ≺ c ≺ b. If there are two consecutive
vertices a and b not connected by an edge then ≺ is not unique, because we can
swap them and resulting order is topological again.
Conversely if every pair of consecutive vertices is connected by an edge, then
≺ is unique: one can check that there is only one source in G and that is the
unique vertex that can be first in a topological order. After removing it we have
again only one source, which is the unique vertex that can be second and so on.
Exercise 2: Design O(n + m) algorithm for topological sorting that is based
on the idea of repeated removal of sources.
Solution: Given graph G = (V, E) to the following
1. For every vertex v calculate i(v) to be the number of vertices w such that
(w, v) ∈ E (that the in-degree of v)
2. Create queue Q and put there all vertices such that i(v) = 0.
3. Create empty queue O.
4. Until Q is non-empty repeat:

(a) Dequeue vertex v from Q.


(b) Enqueue vertex v to O.
(c) For every w such that (v, w) ∈ E decrease i(w) and if i(w) is 0 then
add w to Q

This algorithm can be easily checked to run in O(n + m). If the input graph
G is a DAG the queue O, once algorithm terminates, will contain a topological
order: this can be verified by observing that Q is first initialised by all sources
of G and then decreasing i(w) corresponds to “removing” vertex form G and
updating the in-degrees.
Exercise 3: Describe algorithm to find a shortest path in a edge-labelled DAG
which makes use of the topological ordering of vertices. Edge-labelled DAG an
acyclic oriented graph G = (V, E) where every edge e ∈ E has associated length
`(e). The length of a given path is then sum of lengths of all edges in it. On next
class we will study an algorithm to answer this question on graphs in general.
Solution: This is similar to the problem of counting number of paths discussed
on the class. Given graph G = (V, E) and vertex v0 first put h(v0 ) = 0. Then
process vertices in the topological order. All vertices before v0 are not reachable
from v0 and thus we can put h(v0 ) = ∞ for other vertices we can put

h(v) = min h(w) + `(w, v).


w,(w,v)∈E

1
At the end of the algorithm h(v) will contain distances form v0 . As usual we
can also compute predecessors array which will identify a shortest path.
Exercise 4: Do the same to find the longest path in an edge-labelled DAG (this
is a problem we do now know how to solve effectively on general graphs).
Solution: This is the same algorithm replacing min with max
Exercise 5: We say that connected graph G = (V, E) is semi-connected if for
every pair of vertices u, v ∈ E there exists an oriented path from u to v or from
v to u (possibly both). Design linear-time algorithm to decide if given graph G
is semi-connected.
Solution: Observe that if G is DAG then it is semi-connected if and only if it
has unique topological order. We know how to solve this problem. To solve the
problem for general graphs, first construct component graph using the algorithm
from the class and then apply Exercise 1.

Homework 2: Construction of a house consists of multiple tasks where some


needs to be finished before others can start. We describe this as a graph: vertices
are tasks and each vertex has associated duration. Oriented edges represents
dependencies. Write algorithm to determine shortest time in which the house can
be constructed (assuming there is unlimited number of workers doing individual
tasks) and find critical ones (that is, tasks whose delays would result in longer
overall construction times).

You might also like