I have an undirected graph of locations in an indoor map.
When given a set of vertices , i want to find the shortest path which will cover all those vertices.
Graph contains 52 vertices and 150 - 250 Edges.
What is the best algorithm i can use to find the shortest path.
Please don't get confused that this is a Travelling Salesman Problem. It doesn't have to cover all the nodes.Only the given set of nodes.
As i've commented, this is a hard problem, so don't expect a polynomial time algorithm.
But if you're looking for an algorithm that you may be able to compute in acceptable time for the problem instances you mentioned, this might work:
Let G(V,E) be the original graph, let N be the set of nodes that must be visited.
1. Compute the shortest-path matrix M for the entire graph (|V|x|V| matrix that contains
the length of the shortest path between each two nodes).
2. Generate a new graph G`, containing N alone, with the distances between each
two nodes taken from the shortest-path matrix M.
3. Solve the Minimum Weight Hamiltonian Path Problem on G`.
Note that the "hardest" part here is the third part, which takes exponential time. but if the group N is not too big, you'll be able to solve it:
Bruteforce algorithm will let you solve problems where N contains about 11 nodes within seconds ( O(|N|!) complexity )
Dynamic Programming will let you solve problems where N contains about 20 nodes within seconds ( O(2^|N|*|N|^2) complexity.
You can basically apply any algorithm that solves the Minimum Weight Hamiltonian Path Problem to the third part, these algorithms are usually equivalent to TSP algorithms (the only difference between these problems is that in the TSP you return to the source node after visiting all the other nodes).
Related
I'm taking the Algorithms: Design and Analysis II class, one of the questions asks:
Assume that P ≠ NP. Consider undirected graphs with nonnegative edge
lengths. Which of the following problems can be solved in polynomial
time?
Hint: The Hamiltonian path problem is: given an undirected graph with
n vertices, decide whether or not there is a (cycle-free) path with n
- 1 edges that visits every vertex exactly once. You can use the fact that the Hamiltonian path problem is NP-complete. There are relatively
simple reductions from the Hamiltonian path problem to 3 of the 4
problems below.
For a given source s and destination t, compute the length of a shortest s-t path that has exactly n - 1 edges (or +∞, if no such path
exists). The path is allowed to contain cycles.
Amongst all spanning trees of the graph, compute one with the smallest-possible number of leaves.
Amongst all spanning trees of the graph, compute one with the minimum-possible maximum degree. (Recall the degree of a vertex is the
number of incident edges.)
For a given source s and destination t, compute the length of a shortest s-t path that has exactly n - 1 edges (or +∞, if no such path
exists). The path is not allowed to contain cycles.
Notice that a Hamiltonian path is a spanning tree of a graph and only has two leaf nodes, and that any spanning tree of a graph with exactly two leaf nodes must be a Hamiltonian path. That means that the NP-Complete problem of determining whether a Hamiltonian path exists in a graph can be solved by finding the minimum-leaf spanning tree of the graph: the path exists if and only if the minimum-leaf spanning tree has exactly two leaves. Thus, problem 2 is NP-Complete.
Problem 3 is NP-Hard; here is a paper that proves that.
That means, between 1 and 4, one is NP-Complete, another is in P. It seems like problem 4 reduces trivially to the the Hamiltonian path problem, but I'm not able to understand how having a cycle makes it solvable? Or is it the other way?
For the first one you can use Dijkstra to get shortest even and odd distances possible. To this end for every vertex you need to store not a single minimum number, but two of them. One is minimum weight of an odd path, another one is for minimum weight of an even path. After you have these two lengths you can easily increase path length by even number of edges if cycles are allowed. So, the first problem is from P. Step-be-step algorithm would be:
Find shortest even and odd length paths.
Increase length of one of these paths which has the same parity as n-1 to n-1 by adding cycle of length 2 required number of times.
Suppose I have a weighted, undirected graph. Each edge has a positive weight. I would like to find a simple path (no vertices appear in the path twice) from a given source node (s) to a target node (t) which has the total sum of weights close to a given value (P).
Even though it sounds like a well-studied problem, I couldn't find a satisfying solution. Many graph algorithms are aiming to find the shortest path (in a sense of steps or cost), but not to find the "matched" cost path.
A naive solution would be finding all paths from s to t, compute sum of weights for each path and select the one that is close to P. However, finding all paths between two nodes in a graph is known to be #P-hard.
A possible solution could be modified the A* algorithm, so that for each node in the frontier we get the cost from the root to that node (g), and estimate the cost from that node to the goal (h). Then instead of choosing a node with the smallest g+h, we choose a node with the smallest |P - (g+h)|. However, I am not sure if this is the best solution.
Another thought is inspired from the linear programming since the objective function of this problem is sum(weights of a path from s to t) - P = 0. I know the shortest path problem can be formed as a linear programming task but not sure how to formulate this problem as a one.
Please help, thanks in advance!
This problem is NP-hard via a reduction from the Hamiltonian path problem. In that problem, you are given a graph and a pair of nodes s and t and are asked whether there's a simple path from s to t that passes through all the nodes in the graph. You can solve the Hamiltonian path problem via your problem as follows:
Assign each edge in the graph weight 1.
Find the s-t simple path whose weight is as close to n-1 as possible, where n is the number of nodes in the graph.
Returns whether this path has cost exactly n-1.
If the graph has a Hamiltonian path, then that path will have cost n-1. Otherwise, it doesn't, and the best path found will have a cost that's lower than n-1.
You are given a weighted connected graph (20 nodes) with all edges having positive weight. We have a robot that starts at point A and it must pass at points B, D and E for example. The idea is to find the shortest path that connects all these 4 points. The robot also has a limited battery, but it can be recharged in some points.
After researching on the internet I have two algorithms in mind: Dijkstra's and TSP. Dijkstra's will find the shortest path between a node and every other node and TSP will find the shortest path that connects all points. Is there any variant of the TSP that only finds the shortest path between a set of nodes? After all, in the TSP all nodes are labeled "must-pass". I'm still not taking in account the battery constraint.
Thanks in advance!
You can reduce your graph to a TSP and then invoke a TSP algorithm on it:
Use Floyd-Warshall algorithm to find the distance u,v for ALL pairs of vertices u and v.
Create a new graph, containing only the "desired" vertices, and set the weight between two such vertices u and v as the distance found by Floyd-Warshall.
Run TSP Solver on the modified graph to get the path in the modified graph, and switch each edge in the modified graph with a shortest path from the original graph.
The above is optimal, because assume there is a shorter path.
D0=u->...D1->...->D2->...->Dk->...->t=D{k+1}
Di->...->D{i+1} has at least the weight of FloydWarshall(Di,D{i+1}) (correctness of Floyd-Warshall), and thus the edges (D0,D1),(D1,D2),...,(Dk,D{k+1) exist in the modified graph with a weight smaller/equal the weight in the given path.
Thus, from correctness of your TSP-Solver, by using D0->D1->...->Dk->D{k+1}, you get a path that is at least as good as the candidate optimal path.
You might also want to look into the generalized traveling salesman problem (GTSP): The nodes are partitioned into subsets, and the problem is to find the minimum-length route that visits exactly one node in each subset. The model is allowed to choose whichever node it wants from each subset. If there are nodes that must be visited, you can put them in a subset all by themselves.
I am trying to do c++ program.I am trying to do problem in which i have numbers of points. Now i need to find the path that goes through all the points. This is not actually TSP because as per my knowledge in TSP it is possible to travel from all points to every other points. But in my case the path network between the points is fixed and i just need to find the suitable path that goes through all the points provided that all points may not have connection to every other point..so what algorithm am i supposed to follow.
It seems you are looking for a way to traverse a graph? If so have you tried Breadth first search http://en.wikipedia.org/wiki/Breadth-first_search or Depth first search http://en.wikipedia.org/wiki/Depth-first_search to traverse your graph.
You want to find a Hamiltonian path for a graph.
In the mathematical field of graph theory, a Hamiltonian path (or
traceable path) is a path in an undirected graph that visits each
vertex exactly once. A Hamiltonian cycle (or Hamiltonian circuit) is a
Hamiltonian path that is a cycle. Determining whether such paths and
cycles exist in graphs is the Hamiltonian path problem, which is
NP-complete.
Some techniques that exist :
There are n! different sequences of vertices that might be Hamiltonian
paths in a given n-vertex graph (and are, in a complete graph), so a
brute force search algorithm that tests all possible sequences would
be very slow. There are several faster approaches. A search procedure
by Frank Rubin divides the edges of the graph into three classes:
those that must be in the path, those that cannot be in the path, and
undecided. As the search proceeds, a set of decision rules classifies
the undecided edges, and determines whether to halt or continue the
search. The algorithm divides the graph into components that can be
solved separately. Also, a dynamic programming algorithm of Bellman,
Held, and Karp can be used to solve the problem in time O(n2 2n). In
this method, one determines, for each set S of vertices and each
vertex v in S, whether there is a path that covers exactly the
vertices in S and ends at v. For each choice of S and v, a path exists
for (S,v) if and only if v has a neighbor w such that a path exists
for (S − v,w), which can be looked up from already-computed
information in the dynamic program.
Andreas Björklund provided an alternative approach using the
inclusion–exclusion principle to reduce the problem of counting the
number of Hamiltonian cycles to a simpler counting problem, of
counting cycle covers, which can be solved by computing certain matrix
determinants. Using this method, he showed how to solve the
Hamiltonian cycle problem in arbitrary n-vertex graphs by a Monte
Carlo algorithm in time O(1.657n); for bipartite graphs this algorithm
can be further improved to time O(1.414n).
For graphs of maximum degree three, a careful backtracking search can
find a Hamiltonian cycle (if one exists) in time O(1.251n).
Is there an established algorithm for finding a path from point A to point B in a directed-weighted graph which visits exactly N nodes, but not necessarily any nodes in particular?
This problem is known to be NP-hard via a reduction from Hamiltonian path. In particular, you can solve Hamiltonian path with a polynomial-time reduction to this problem as follows: for each possible pair of nodes (s, t) in a graph with n nodes, ask if there is a path from s to t that passes through exactly n nodes. This makes only polynomial calls to your solver, so any polynomial-time solution to your problem would result in a polynomial-time solution to the Hamiltonian path problem.
So in short, you shouldn't expect a polynomial-time algorithm for this problem unless P = NP.
I'm going to assume you're trying to find the shortest/longest weight with N nodes. This probably isn't optimal, but from your original graph, you could generate a state graph with 'N*(#Nodes)' nodes representing consisting of the original node and number of steps taken and run it thorough some shortest path algorithm like http://en.wikipedia.org/wiki/Dijkstra's_algorithm.
i.e.,
A->B->C
\___/
becomes
(A,0)->(B,1)->(C,2)
\>(C,1)
Your target node would then be node (B,N) - B with N steps. This approach would allow for loops in the original graph if it's not a DAG ( (X,0)->(Y,1)->(X,2) )
I am not sure if I undersand it correctly, can you do a depth first search (up to N-1 layers away from the source)?
If you can visit your destination in that layer. you can get a path down there.