Chapter 5: Greedy Algorithms
Activity 11: The Interval Scheduling Problem
In this activity, we will find the optimal scheduling of tasks to maximize the number of tasks that can be completed. Follow these steps to complete the activity:
- Add the required header files and define the Task struct as follows:
#include <list>
#include <algorithm>
#include <iostream>
#include <random>
// Every task is represented as a pair <start_time, end_time>
struct Task
{
unsigned ID;
unsigned start_time;
unsigned end_time;
};
- The following function can be used to generate a list of N tasks with random data:
auto initialize_tasks(size_t num_tasks)
{
std::random_device rd;
std::mt19937 rand(rd());
std::uniform_int_distribution<std::mt19937::result_type>
uniform_dist(1, num_tasks);
// Create and initialize a set of tasks
std::list<Task> tasks;
for (unsigned i = 1; i <= num_tasks; i++)
{
auto start_time = uniform_dist(rand);
auto duration = uniform_dist(rand);
tasks.push_back({i, start_time, start_time + duration });
}
return tasks;
}
- Implement the scheduling algorithm as follows:
auto schedule(std::list<Task> tasks)
{
// Sort the list of tasks by their end times
tasks.sort([](const auto& lhs, const auto& rhs)
{ return lhs.end_time < rhs.end_time; });
// Remove the tasks that interfere with one another
for (auto curr_task = tasks.begin(); curr_task != tasks.end(); curr_task++)
{
// Point to the next task
auto next_task = std::next(curr_task, 1);
// While subsequent tasks interfere with the current task in iter
while (next_task != tasks.end() &&
next_task->start_time < curr_task->end_time)
{
next_task = tasks.erase(next_task);
}
}
return tasks;
}
- The following utility functions are used to print the list of tasks, test our implementation, and include the driver code for the program:
void print(std::list<Task>& tasks)
{
std::cout << "Task ID \t Starting Time \t End time" << std::endl;
for (auto t : tasks)
std::cout << t.ID << "\t\t" << t.start_time << "\t\t" << t.end_time << std::endl;
}
void test_interval_scheduling(unsigned num_tasks)
{
auto tasks = initialize_tasks(num_tasks);
std::cout << "Original list of tasks: " << std::endl;
print(tasks);
std::cout << "Scheduled tasks: " << std::endl;
auto scheduled_tasks = schedule(tasks);
print(scheduled_tasks);
}
int main()
{
test_interval_scheduling(20);
return 0;
}
Activity 12: The Welsh-Powell Algorithm
We will implement the Welsh-Powell algorithm on the graph in this activity. A reference implementation is given here:
- Add the required header files and declare the graph that will be implemented later:
#include <unordered_map>
#include <set>
#include <map>
#include <string>
#include <vector>
#include <algorithm>
#include <iostream>
template <typename T> class Graph;
- Implement the struct, representing edges like so:
template<typename T>
struct Edge
{
size_t src;
size_t dest;
T weight;
// To compare edges, only compare their weights,
// and not the source/destination vertices
inline bool operator< (const Edge<T>& e) const
{
return this->weight < e.weight;
}
inline bool operator> (const Edge<T>& e) const
{
return this->weight > e.weight;
}
};
- The following function allows us to serialize and print graphs by overloading the << operator for the graph datatype:
template <typename T>
std::ostream& operator<<(std::ostream& os, const Graph<T>& G)
{
for (auto i = 1; i < G.vertices(); i++)
{
os << i << ":\t";
auto edges = G.outgoing_edges(i);
for (auto& e : edges)
os << "{" << e.dest << ": " << e.weight << "}, ";
os << std::endl;
}
return os;
}
- Implement the graph with the edge list representation, as shown here:
template<typename T>
class Graph
{
public:
// Initialize the graph with N vertices
Graph(size_t N) : V(N)
{}
// Return number of vertices in the graph
auto vertices() const
{
return V;
}
// Return all edges in the graph
auto& edges() const
{
return edge_list;
}
void add_edge(Edge<T>&& e)
{
// Check if the source and destination vertices are within range
if (e.src >= 1 && e.src <= V &&
e.dest >= 1 && e.dest <= V)
edge_list.emplace_back(e);
else
std::cerr << "Vertex out of bounds" << std::endl;
}
// Returns all outgoing edges from vertex v
auto outgoing_edges(size_t v) const
{
std::vector<Edge<T>> edges_from_v;
for (auto& e : edge_list)
{
if (e.src == v)
edges_from_v.emplace_back(e);
}
return edges_from_v;
}
// Overloads the << operator so a graph be written directly to a stream
// Can be used as std::cout << obj << std::endl;
template <typename T>
friend std::ostream& operator<< <>(std::ostream& os, const Graph<T>& G);
private:
size_t V; // Stores number of vertices in graph
std::vector<Edge<T>> edge_list;
};
- Initialize the set of colors that we will use in our implementation of the Welsh-Powell algorithm. Let this number of colors be 6, as implemented in the following unordered_map:
// Initialize the colors that will be used to color the vertices
std::unordered_map<size_t, std::string> color_map = {
{1, "Red"},
{2, "Blue"},
{3, "Green"},
{4, "Yellow"},
{5, "Black"},
{6, "White"}
};
- Implement the Welsh-Powell graph coloring algorithm like so:
template<typename T>
auto welsh_powell_coloring(const Graph<T>& G)
{
auto size = G.vertices();
std::vector<std::pair<size_t, size_t>> degrees;
// Collect the degrees of vertices as <vertex_ID, degree> pairs
for (auto i = 1; i < size; i++)
degrees.push_back(std::make_pair(i, G.outgoing_edges(i).size()));
// Sort the vertices in decreasing order of degree
std::sort(degrees.begin(),
degrees.end(),
[](const auto& a, const auto& b)
{ return a.second > b.second; });
std::cout << "The vertices will be colored in the following order: " << std::endl;
std::cout << "Vertex ID \t Degree" << std::endl;
for (auto const i : degrees)
std::cout << i.first << "\t\t" << i.second << std::endl;
std::vector<size_t> assigned_colors(size);
auto color_to_be_assigned = 1;
while (true)
{
for (auto const i : degrees)
{
if (assigned_colors[i.first] != 0)
continue;
auto outgoing_edges = G.outgoing_edges(i.first);
std::set<size_t> neighbour_colors;
// We assume that the graph is bidirectional
for (auto e : outgoing_edges)
{
auto dest_color = assigned_colors[e.dest];
neighbour_colors.insert(dest_color);
}
if (neighbour_colors.find(color_to_be_assigned) == neighbour_colors.end())
assigned_colors[i.first] = color_to_be_assigned;
}
color_to_be_assigned++;
// If there are no uncolored vertices left, exit
if (std::find(assigned_colors.begin() + 1, assigned_colors.end(), 0) ==
assigned_colors.end())
break;
}
return assigned_colors;
}
- The following function outputs the vector of colors:
void print_colors(std::vector<size_t>& colors)
{
for (auto i = 1; i < colors.size(); i++)
{
std::cout << i << ": " << color_map[colors[i]] << std::endl;
}
}
- Finally, the following driver code creates the required graph, runs the vertex coloring algorithm, and outputs the results:
int main()
{
using T = unsigned;
Graph<T> G(9);
std::map<unsigned, std::vector<std::pair<size_t, T>>> edges;
edges[1] = { {2, 2}, {5, 3} };
edges[2] = { {1, 2}, {5, 5}, {4, 1} };
edges[3] = { {4, 2}, {7, 3} };
edges[4] = { {2, 1}, {3, 2}, {5, 2}, {6, 4}, {8, 5} };
edges[5] = { {1, 3}, {2, 5}, {4, 2}, {8, 3} };
edges[6] = { {4, 4}, {7, 4}, {8, 1} };
edges[7] = { {3, 3}, {6, 4} };
edges[8] = { {4, 5}, {5, 3}, {6, 1} };
for (auto& i : edges)
for (auto& j : i.second)
G.add_edge(Edge<T>{ i.first, j.first, j.second });
std::cout << "Original Graph" << std::endl;
std::cout << G;
auto colors = welsh_powell_coloring<T>(G);
std::cout << "Vertex Colors: " << std::endl;
print_colors(colors);
return 0;
}
