`import "github.com/yourbasic/graph"`

Package graph contains generic implementations of basic graph algorithms.

The algorithms in this library can be applied to any graph data structure implementing the two Iterator methods: Order, which returns the number of vertices, and Visit, which iterates over the neighbors of a vertex.

All algorithms operate on directed graphs with a fixed number of vertices, labeled from 0 to n-1, and edges with integer cost. An undirected edge {v, w} of cost c is represented by the two directed edges (v, w) and (w, v), both of cost c. A self-loop, an edge connecting a vertex to itself, is both directed and undirected.

The type Mutable represents a directed graph with a fixed number of vertices and weighted edges that can be added or removed. The implementation uses hash maps to associate each vertex in the graph with its adjacent vertices. This gives constant time performance for all basic operations.

The type Immutable is a compact representation of an immutable graph. The implementation uses lists to associate each vertex in the graph with its adjacent vertices. This makes for fast and predictable iteration: the Visit method produces its elements by reading from a fixed sorted precomputed list. This type supports multigraphs.

The subpackage graph/build offers a tool for building virtual graphs. In a virtual graph no vertices or edges are stored in memory, they are instead computed as needed. New virtual graphs are constructed by composing and filtering a set of standard graphs, or by writing functions that describe the edges of a graph.

The Basics example shows how to build a plain graph and how to efficiently use the Visit iterator, the key abstraction of this package.

The DFS example contains a full implementation of depth-first search.

Build a plain graph and visit all of its edges.

Code:

// Build a graph with four vertices and four undirected edges. // (Each of these edges are, in fact, represented by two directed // edges pointing in opposite directions.) g := graph.New(4) g.AddBoth(0, 1) // 0 -- 1 g.AddBoth(0, 2) // | | g.AddBoth(2, 3) // 2 -- 3 g.AddBoth(1, 3) // The vertices of all graphs in this package are numbered 0..n-1. // The edge iterator is a method called Visit; it calls a function // for each neighbor of a given vertex. Together with the Order // method — which returns the number of vertices in a graph — it // constitutes an Iterator. All algorithms in this package operate // on any graph implementing this interface. // Visit all edges of a graph. for v := 0; v < g.Order(); v++ { g.Visit(v, func(w int, c int64) (skip bool) { // Visiting edge (v, w) of cost c. return }) } // The immutable data structure created by Sort has an Iterator // that returns neighbors in increasing order. // Visit the edges in order. for v := 0; v < g.Order(); v++ { graph.Sort(g).Visit(v, func(w int, c int64) (skip bool) { // Visiting edge (v, w) of cost c. return }) } // The return value of an iterator function is used to break // out of the iteration. Visit, in turn, returns a boolean // indicating if it was aborted. // Skip the iteration at the first edge (v, w) with w equal to 3. for v := 0; v < g.Order(); v++ { aborted := graph.Sort(g).Visit(v, func(w int, c int64) (skip bool) { fmt.Println(v, w) if w == 3 { skip = true // Aborts the call to Visit. } return }) if aborted { break } }

Output:

0 1 0 2 1 0 1 3

Show how to use this package by implementing a complete depth-first search.

Code:

package main import ( "fmt" "github.com/yourbasic/graph" ) const ( White = iota Gray Black ) // The package doesn't support vertex labeling. However, // since vertices are always numbered 0..n-1, it's easy // to add this type of data on the side. This implementation // of depth-first search uses separate slices to keep track of // vertex colors, predecessors and discovery times. type DFSData struct { Time int Color []int Prev []int Discover []int Finish []int } func DFS(g graph.Iterator) DFSData { n := g.Order() // Order returns the number of vertices. d := DFSData{ Time: 0, Color: make([]int, n), Prev: make([]int, n), Discover: make([]int, n), Finish: make([]int, n), } for v := 0; v < n; v++ { d.Color[v] = White d.Prev[v] = -1 } for v := 0; v < n; v++ { if d.Color[v] == White { d.dfsVisit(g, v) } } return d } func (d *DFSData) dfsVisit(g graph.Iterator, v int) { d.Color[v] = Gray d.Time++ d.Discover[v] = d.Time // Visit calls a function for each neighbor w of v, // with c equal to the cost of the edge (v, w). // The iteration is aborted if the function returns true. g.Visit(v, func(w int, c int64) (skip bool) { if d.Color[w] == White { d.Prev[w] = v d.dfsVisit(g, w) } return }) d.Color[v] = Black d.Time++ d.Finish[v] = d.Time } // Show how to use this package by implementing a complete depth-first search. func main() { // Build a small directed graph: // // 0 ---> 1 <--> 2 3 // g := graph.New(4) g.Add(0, 1) g.AddBoth(1, 2) fmt.Println(g) fmt.Println(DFS(g)) }

- Constants
- func Acyclic(g Iterator) bool
- func BFS(g Iterator, v int, do func(v, w int, c int64))
- func Bipartition(g Iterator) (part []int, ok bool)
- func Components(g Iterator) [][]int
- func Connected(g Iterator) bool
- func Equal(g, h Iterator) bool
- func EulerDirected(g Iterator) (walk []int, ok bool)
- func EulerUndirected(g Iterator) (walk []int, ok bool)
- func MST(g Iterator) (parent []int)
- func ShortestPath(g Iterator, v, w int) (path []int, dist int64)
- func ShortestPaths(g Iterator, v int) (parent []int, dist []int64)
- func String(g Iterator) string
- func StrongComponents(g Iterator) [][]int
- func TopSort(g Iterator) (order []int, ok bool)
- type Immutable
- func Sort(g Iterator) *Immutable
- func Transpose(g Iterator) *Immutable
- func (g *Immutable) Degree(v int) int
- func (g *Immutable) Edge(v, w int) bool
- func (g *Immutable) Order() int
- func (g *Immutable) String() string
- func (g *Immutable) Visit(v int, do func(w int, c int64) bool) bool
- func (g *Immutable) VisitFrom(v int, a int, do func(w int, c int64) bool) bool
- type Iterator
- type Mutable
- func Copy(g Iterator) *Mutable
- func New(n int) *Mutable
- func (g *Mutable) Add(v, w int)
- func (g *Mutable) AddBoth(v, w int)
- func (g *Mutable) AddBothCost(v, w int, c int64)
- func (g *Mutable) AddCost(v, w int, c int64)
- func (g *Mutable) Cost(v, w int) int64
- func (g *Mutable) Degree(v int) int
- func (g *Mutable) Delete(v, w int)
- func (g *Mutable) DeleteBoth(v, w int)
- func (g *Mutable) Edge(v, w int) bool
- func (g *Mutable) Order() int
- func (g *Mutable) String() string
- func (g *Mutable) Visit(v int, do func(w int, c int64) bool) bool
- type Stats

- BFS
- Components
- EulerDirected
- EulerUndirected
- ShortestPath
- String
- StrongComponents
- package (Basics)
- package (DFS)

bfs.go bipart.go euler.go graph.go heap.go immutable.go maxflow.go mst.go mutable.go path.go strong.go top.go weak.go

The maximum and minum value of an edge cost.

Acyclic tells if g has no cycles.

BFS traverses g in breadth-first order starting at v. When the algorithm follows an edge (v, w) and finds a previously unvisited vertex w, it calls do(v, w, c) with c equal to the cost of the edge (v, w).

Find the shortest distances from a vertex in an unweighted graph.

Code:

gm := graph.New(6) gm.AddBoth(0, 1) // 0--1--2 gm.AddBoth(0, 3) // | | | gm.AddBoth(1, 2) // 3--4 5 gm.AddBoth(1, 4) gm.AddBoth(2, 5) gm.AddBoth(3, 4) g := graph.Sort(gm) dist := make([]int, g.Order()) graph.BFS(g, 0, func(v, w int, _ int64) { fmt.Println(v, "to", w) dist[w] = dist[v] + 1 }) fmt.Println("dist:", dist)

Output:

0 to 1 0 to 3 1 to 2 1 to 4 2 to 5 dist: [0 1 2 1 2 3]

Bipartition returns a subset U of g's vertices with the property that every edge of g connects a vertex in U to one outside of U. If g isn't bipartite, it returns an empty slice and sets ok to false.

Components produces a partition of g's vertices into its (weakly) connected components.

Find the weakly connected components in a directed graph.

Code:

g := graph.New(6) g.AddBoth(0, 1) // 0 <--> 1 ---> 2 g.Add(1, 2) // ^ g.Add(4, 3) // | g.AddBoth(5, 2) // 3 <--- 4 5 fmt.Println(graph.Components(g))

Output:

[[0 1 2 5] [3 4]]

Connected tells if g has exactly one (weakly) connected component.

Equal tells if g and h have the same number of vertices, and the same edges with the same costs.

EulerDirected returns an Euler walk in a directed graph. If no such walk exists, it returns an empty walk and sets ok to false.

Find an Euler walk in a directed graph.

Code:

```
// 0 <--> 1 --> 2 3
g := graph.New(4)
g.AddBoth(0, 1)
g.Add(1, 2)
fmt.Println(graph.EulerDirected(g))
```

Output:

[1 0 1 2] true

EulerUndirected returns an Euler walk following undirected edges in only one direction. If no such walk exists, it returns an empty walk and sets ok to false.

Find an Euler walk in an undirected graph.

Code:

```
// 0 1
// |
// 2--3---
// | |
// ----
g := graph.New(4)
g.AddBoth(1, 3)
g.AddBoth(2, 3)
g.AddBoth(3, 3)
fmt.Println(graph.EulerUndirected(g))
```

Output:

[1 3 3 2] true

MST computes a minimum spanning tree for each connected component of an undirected weighted graph. The forest of spanning trees is returned as a slice of parent pointers: parent[v] is either the parent of v in a tree, or -1 if v is the root of a tree.

The time complexity is O(|E|⋅log|V|), where |E| is the number of edges and |V| the number of vertices in the graph.

ShortestPath computes a shortest path from v to w. Only edges with non-negative costs are included. The number dist is the length of the path, or -1 if w cannot be reached.

The time complexity is O((|E| + |V|)⋅log|V|), where |E| is the number of edges and |V| the number of vertices in the graph.

Find a shortest path between two vertices in a graph.

Code:

g := graph.New(6) g.AddBothCost(0, 1, 8) // 0==1--2 g.AddBothCost(0, 3, 2) // | | | g.AddBothCost(1, 2, 2) // 3--4==5 g.AddBothCost(1, 4, 2) // g.AddBothCost(2, 5, 2) // -- cost 2 g.AddBothCost(3, 4, 2) // == cost 8 g.AddBothCost(4, 5, 8) path, dist := graph.ShortestPath(g, 0, 5) fmt.Println("path:", path, "length:", dist)

Output:

path: [0 3 4 1 2 5] length: 10

ShortestPaths computes the shortest paths from v to all other vertices. Only edges with non-negative costs are included. The number parent[w] is the predecessor of w on a shortest path from v to w, or -1 if none exists. The number dist[w] equals the length of a shortest path from v to w, or is -1 if w cannot be reached.

The time complexity is O((|E| + |V|)⋅log|V|), where |E| is the number of edges and |V| the number of vertices in the graph.

String returns a description of g with two elements: the number of vertices, followed by a sorted list of all edges.

Print a graph.

Code:

g0 := graph.New(0) fmt.Println(g0) g1 := graph.New(1) g1.Add(0, 0) fmt.Println(g1) g4 := graph.New(4) // 8 g4.AddBoth(0, 1) // 0 <--> 1 <--- 2 3 g4.AddCost(2, 1, 8) fmt.Println(g4)

Output:

0 [] 1 [(0 0)] 4 [{0 1} (2 1):8]

StrongComponents produces a partition of g's vertices into its strongly connected components.

A component is strongly connected if all its vertices are reachable from every other vertex in the component. Each vertex of the graph appears in exactly one of the strongly connected components, and any vertex that is not on a directed cycle forms a strongly connected component all by itself.

Find the strongly connected components in a directed graph.

Code:

g := graph.New(6) g.AddBoth(0, 1) // 0 <--> 1 <--> 2 g.AddBoth(1, 2) // ^ ^ ^ g.Add(3, 0) // | | | g.AddBoth(3, 4) // 3 <--> 4 ---> 5 g.Add(4, 1) g.Add(4, 5) g.Add(5, 2) fmt.Println(graph.StrongComponents(g))

Output:

[[2 1 0] [5] [4 3]]

TopSort returns a topological ordering of the vertices in a directed acyclic graph; if the graph is not acyclic, no such ordering exists and ok is set to false.

In a topological order v comes before w for every directed edge from v to w.

❖

```
type Immutable struct {
// contains filtered or unexported fields
}
```

Immutable is a compact representation of an immutable graph. The implementation uses lists to associate each vertex in the graph with its adjacent vertices. This makes for fast and predictable iteration: the Visit method produces its elements by reading from a fixed sorted precomputed list. This type supports multigraphs.

Sort returns an immutable copy of g with a Visit method that returns its neighbors in increasing numerical order.

Transpose returns the transpose graph of g. The transpose graph has the same set of vertices as g, but all of the edges are reversed compared to the orientation of the corresponding edges in g.

Degree returns the number of outward directed edges from v.

Edge tells if there is an edge from v to w.

Order returns the number of vertices in the graph.

String returns a string representation of the graph.

Visit calls the do function for each neighbor w of v, with c equal to the cost of the edge from v to w. The neighbors are visited in increasing numerical order. If do returns true, Visit returns immediately, skipping any remaining neighbors, and returns true.

VisitFrom calls the do function starting from the first neighbor w for which w ≥ a, with c equal to the cost of the edge from v to w. The neighbors are then visited in increasing numerical order. If do returns true, VisitFrom returns immediately, skipping any remaining neighbors, and returns true.

❖

type Iterator interface { // Order returns the number of vertices in a graph. Order() int // Visit calls the do function for each neighbor w of vertex v, // with c equal to the cost of the edge from v to w. // // • If do returns true, Visit returns immediately, skipping // any remaining neighbors, and returns true. // // • The calls to the do function may occur in any order, // and the order may vary. // Visit(v int, do func(w int, c int64) (skip bool)) (aborted bool) }

Iterator describes a weighted graph; an Iterator can be used to describe both ordinary graphs and multigraphs.

MaxFlow computes a maximum flow from s to t in a graph with nonnegative edge capacities. The time complexity is O(|E|²⋅|V|), where |E| is the number of edges and |V| the number of vertices in the graph.

❖

```
type Mutable struct {
// contains filtered or unexported fields
}
```

Mutable represents a directed graph with a fixed number of vertices and weighted edges that can be added or removed. The implementation uses hash maps to associate each vertex in the graph with its adjacent vertices. This gives constant time performance for all basic operations.

Copy returns a copy of g. If g is a multigraph, any duplicate edges in g will be lost.

New constructs a new graph with n vertices, numbered from 0 to n-1, and no edges.

Add inserts a directed edge from v to w with zero cost. It removes the previous cost if this edge already exists.

AddBoth inserts edges with zero cost between v and w. It removes the previous costs if these edges already exist.

AddBothCost inserts edges with cost c between v and w. It overwrites the previous costs if these edges already exist.

AddCost inserts a directed edge from v to w with cost c. It overwrites the previous cost if this edge already exists.

Cost returns the cost of an edge from v to w, or 0 if no such edge exists.

Degree returns the number of outward directed edges from v.

Delete removes an edge from v to w.

DeleteBoth removes all edges between v and w.

Edge tells if there is an edge from v to w.

Order returns the number of vertices in the graph.

String returns a string representation of the graph.

Visit calls the do function for each neighbor w of v, with c equal to the cost of the edge from v to w. If do returns true, Visit returns immediately, skipping any remaining neighbors, and returns true.

The iteration order is not specified and is not guaranteed to be the same every time. It is safe to delete, but not to add, edges adjacent to v during a call to this method.

❖

type Stats struct { Size int // Number of unique edges. Multi int // Number of duplicate edges. Weighted int // Number of edges with non-zero cost. Loops int // Number of self-loops. Isolated int // Number of vertices with outdegree zero. }

Stats holds basic data about an Iterator.

Check collects data about an Iterator.

Path | Synopsis |
---|---|

build | Package build offers a tool for building virtual graphs. |

Package graph imports 2 packages (graph) and is imported by 1 packages. Updated 2018-10-18. Refresh now. Tools for package owners.