The routine may be called by the names h03adf or nagf_mip_shortestpath.
3Description
h03adf attempts to find the shortest path through a directed or undirectedacyclic network, which consists of a set of points called vertices and a set of curves called arcs that connect certain pairs of distinct vertices. An acyclic network is one in which there are no paths connecting a vertex to itself. An arc whose origin vertex is and whose destination vertex is can be written as . In an undirected network the arcs and are equivalent (i.e., ), whereas in a directed network they are different. Note that the shortest path may not be unique and in some cases may not even exist (e.g., if the network is disconnected).
The network is assumed to consist of vertices which are labelled by the integers . The lengths of the arcs between the vertices are defined by the
distance matrix
, in which the element gives the length of the arc ; if there is no arc connecting vertices and (as is the case for an acyclic network when ). Thus the matrix is usually sparse. For example, if and the network is directed, then
If the network is undirected, is symmetric since (i.e., the length of the arc the length of the arc ).
The method used by h03adf is described in detail in Section 9.
4References
Dijkstra E W (1959) A note on two problems in connection with graphs Numer. Math.1 269–271
5Arguments
1: – IntegerInput
On entry: , the number of vertices.
Constraint:
.
2: – IntegerInput
3: – IntegerInput
On entry: and , the labels of the first and last vertices, respectively, between which the shortest path is sought.
Constraints:
;
;
.
4: – LogicalInput
On entry: indicates whether the network is directed or undirected.
The network is directed.
The network is undirected.
5: – IntegerInput
On entry: the number of nonzero elements in the distance matrix .
Constraints:
if , ;
if , .
6: – Real (Kind=nag_wp) arrayInput
On entry: the nonzero elements of the distance matrix , ordered by increasing row index and increasing column index within each row. More precisely, must contain the value of the nonzero element with indices (); this is the length of the arc from the vertex with label to the vertex with label . Elements with the same row and column indices are not allowed. If , then only those nonzero elements in the strict upper triangle of need be supplied since . (f11zaf may be used to sort the elements of an arbitrarily ordered matrix into the required form.)
Constraint:
, for .
7: – Integer arrayInput
8: – Integer arrayInput
On entry: and must contain the row and column indices, respectively, for the nonzero element stored in .
Constraints:
irow and icol must satisfy the following constraints (which may be imposed by a call to f11zaf):
;
and , for .
In addition, if , , and ;
if , .
9: – Real (Kind=nag_wp)Output
On exit: the length of the shortest path between the specified vertices and .
10: – Integer arrayOutput
On exit: contains details of the shortest path between the specified vertices and . More precisely, for some . The remaining elements are set to zero.
11: – Integer arrayWorkspace
12: – Real (Kind=nag_wp) arrayWorkspace
13: – IntegerInput/Output
On entry: ifail must be set to , or to set behaviour on detection of an error; these values have no effect when no error is detected.
A value of causes the printing of an error message and program execution will be halted; otherwise program execution continues. A value of means that an error message is printed while a value of means that it is not.
If halting is not appropriate, the value or is recommended. If message printing is undesirable, then the value is recommended. Otherwise, the value is recommended. When the value or is used it is essential to test the value of ifail on exit.
On exit: unless the routine detects an error or a warning has been flagged (see Section 6).
6Error Indicators and Warnings
If on entry or , explanatory error messages are output on the current error message unit (as defined by x04aaf).
Errors or warnings detected by the routine:
On entry, .
Constraint: .
On entry, and .
Constraint: .
On entry, and .
Constraint: .
On entry, and .
Constraint: .
On entry, and .
Constraint: if , .
On entry, and .
Constraint: if , .
On entry, , , and .
Constraint: , ; when .
On entry, , , and .
Constraint: when .
On entry, , .
Constraint: .
On entry, , , , , .
Constraints: or and .
On entry, , , .
Constraint: or .
On entry, and . No connected network exists between vertices ns and ne.
An unexpected error has been triggered by this routine. Please
contact NAG.
See Section 7 in the Introduction to the NAG Library FL Interface for further information.
Your licence key may have expired or may not have been installed correctly.
See Section 8 in the Introduction to the NAG Library FL Interface for further information.
Dynamic memory allocation failed.
See Section 9 in the Introduction to the NAG Library FL Interface for further information.
7Accuracy
The results are exact, except for the obvious rounding errors in summing the distances in the length of the shortest path.
8Parallelism and Performance
Background information to multithreading can be found in the Multithreading documentation.
h03adf is not threaded in any implementation.
9Further Comments
h03adf is based upon Dijkstra's algorithm (see Dijkstra (1959)), which attempts to find a path between two specified vertices and of shortest length .
The algorithm proceeds by assigning labels to each vertex, which may be temporary or permanent. A temporary label can be changed, whereas a permanent one cannot. For example, if vertex has a permanent label , then is the distance and is the previous vertex on a shortest length path. If the label is temporary, then it has the same meaning but it refers only to the shortest path found so far. A shorter one may be found later, in which case the label may become permanent.
The algorithm consists of the following steps.
1.Assign the permanent label to vertex and temporary labels to every other vertex. Set and go to 2.
2.Consider each vertex adjacent to vertex with a temporary label in turn. Let the label at be and at . If , then a new temporary label is assigned to vertex ; otherwise no change is made in the label of . When all vertices with temporary labels adjacent to have been considered, go to 3.
3.From the set of temporary labels, select the one with the smallest second component and declare that label to be permanent. The vertex it is attached to becomes the new vertex . If go to 4. Otherwise go to 2 unless no new vertex can be found (e.g., when the set of temporary labels is ‘empty’ but , in which case no connected network exists between vertices and ).
4.To find the shortest path, let denote the label of vertex . The column label () gives while the row label () then links back to the previous vertex on a shortest length path. Go to vertex . Suppose that the (permanent) label of vertex is , then the next previous vertex is on a shortest length path. This process continues until vertex is reached. Hence the shortest path is
which has length .
10Example
This example finds the shortest path between vertices and for the undirected network