// Copyright ©2015 The Gonum Authors. All rights reserved. // Use of this source code is governed by a BSD-style // license that can be found in the LICENSE file. package path import ( "gonum.org/v1/gonum/graph" "gonum.org/v1/gonum/graph/internal/linear" "gonum.org/v1/gonum/graph/traverse" ) // BellmanFordFrom returns a shortest-path tree for a shortest path from u to all nodes in // the graph g, or false indicating that a negative cycle exists in the graph. If the graph // does not implement Weighted, UniformCost is used. // // If g is a graph.Graph, all nodes of the graph will be stored in the shortest-path // tree, otherwise only nodes reachable from u will be stored. // // The time complexity of BellmanFordFrom is O(|V|.|E|). func BellmanFordFrom(u graph.Node, g traverse.Graph) (path Shortest, ok bool) { if h, ok := g.(graph.Graph); ok { if h.Node(u.ID()) == nil { return Shortest{from: u}, true } path = newShortestFrom(u, graph.NodesOf(h.Nodes())) } else { if g.From(u.ID()) == graph.Empty { return Shortest{from: u}, true } path = newShortestFrom(u, []graph.Node{u}) } path.dist[path.indexOf[u.ID()]] = 0 path.negCosts = make(map[negEdge]float64) var weight Weighting if wg, ok := g.(Weighted); ok { weight = wg.Weight } else { weight = UniformCost(g) } // Queue to keep track which nodes need to be relaxed. // Only nodes whose vertex distance changed in the previous iterations // need to be relaxed again. queue := newBellmanFordQueue(path.indexOf) queue.enqueue(u) // TODO(kortschak): Consider adding further optimisations // from http://arxiv.org/abs/1111.5414. var loops int64 for queue.len() != 0 { u := queue.dequeue() uid := u.ID() j := path.indexOf[uid] to := g.From(uid) for to.Next() { v := to.Node() vid := v.ID() k, ok := path.indexOf[vid] if !ok { k = path.add(v) } w, ok := weight(uid, vid) if !ok { panic("bellman-ford: unexpected invalid weight") } joint := path.dist[j] + w if joint < path.dist[k] { path.set(k, joint, j) if !queue.has(vid) { queue.enqueue(v) } } } // The maximum number of edges in the relaxed subgraph is |V_r| * (|V_r|-1). // If the queue-loop has more iterations than the maximum number of edges // it indicates that we have a negative cycle. maxEdges := int64(len(path.nodes)) * int64(len(path.nodes)-1) if loops > maxEdges { path.hasNegativeCycle = true return path, false } loops++ } return path, true } // BellmanFordAllFrom returns a shortest-path tree for shortest paths from u to all nodes in // the graph g, or false indicating that a negative cycle exists in the graph. If the graph // does not implement Weighted, UniformCost is used. // // If g is a graph.Graph, all nodes of the graph will be stored in the shortest-path // tree, otherwise only nodes reachable from u will be stored. // // The time complexity of BellmanFordAllFrom is O(|V|.|E|). func BellmanFordAllFrom(u graph.Node, g traverse.Graph) (path ShortestAlts, ok bool) { if h, ok := g.(graph.Graph); ok { if h.Node(u.ID()) == nil { return ShortestAlts{from: u}, true } path = newShortestAltsFrom(u, graph.NodesOf(h.Nodes())) } else { if g.From(u.ID()) == graph.Empty { return ShortestAlts{from: u}, true } path = newShortestAltsFrom(u, []graph.Node{u}) } path.dist[path.indexOf[u.ID()]] = 0 path.negCosts = make(map[negEdge]float64) var weight Weighting if wg, ok := g.(Weighted); ok { weight = wg.Weight } else { weight = UniformCost(g) } // Queue to keep track which nodes need to be relaxed. // Only nodes whose vertex distance changed in the previous iterations // need to be relaxed again. queue := newBellmanFordQueue(path.indexOf) queue.enqueue(u) // TODO(kortschak): Consider adding further optimisations // from http://arxiv.org/abs/1111.5414. var loops int64 for queue.len() != 0 { u := queue.dequeue() uid := u.ID() j := path.indexOf[uid] for _, v := range graph.NodesOf(g.From(uid)) { vid := v.ID() k, ok := path.indexOf[vid] if !ok { k = path.add(v) } w, ok := weight(uid, vid) if !ok { panic("bellman-ford: unexpected invalid weight") } joint := path.dist[j] + w if joint < path.dist[k] { path.set(k, joint, j) if !queue.has(vid) { queue.enqueue(v) } } else if joint == path.dist[k] { path.addPath(k, j) } } // The maximum number of edges in the relaxed subgraph is |V_r| * (|V_r|-1). // If the queue-loop has more iterations than the maximum number of edges // it indicates that we have a negative cycle. maxEdges := int64(len(path.nodes)) * int64(len(path.nodes)-1) if loops > maxEdges { path.hasNegativeCycle = true return path, false } loops++ } return path, true } // bellmanFordQueue is a queue for the Queue-based Bellman-Ford algorithm. type bellmanFordQueue struct { // queue holds the nodes which need to be relaxed. queue linear.NodeQueue // onQueue keeps track whether a node is on the queue or not. onQueue []bool // indexOf contains a mapping holding the id of a node with its index in the onQueue array. indexOf map[int64]int } // enqueue adds a node to the bellmanFordQueue. func (q *bellmanFordQueue) enqueue(n graph.Node) { i, ok := q.indexOf[n.ID()] switch { case !ok: panic("bellman-ford: unknown node") case i < len(q.onQueue): if q.onQueue[i] { panic("bellman-ford: already queued") } case i == len(q.onQueue): q.onQueue = append(q.onQueue, false) case i < cap(q.onQueue): q.onQueue = q.onQueue[:i+1] default: q.onQueue = append(q.onQueue, make([]bool, i-len(q.onQueue)+1)...) } q.onQueue[i] = true q.queue.Enqueue(n) } // dequeue returns the first value of the bellmanFordQueue. func (q *bellmanFordQueue) dequeue() graph.Node { n := q.queue.Dequeue() q.onQueue[q.indexOf[n.ID()]] = false return n } // len returns the number of nodes in the bellmanFordQueue. func (q *bellmanFordQueue) len() int { return q.queue.Len() } // has returns whether a node with the given id is in the queue. func (q bellmanFordQueue) has(id int64) bool { idx, ok := q.indexOf[id] if !ok || idx >= len(q.onQueue) { return false } return q.onQueue[idx] } // newBellmanFordQueue creates a new bellmanFordQueue. func newBellmanFordQueue(indexOf map[int64]int) bellmanFordQueue { return bellmanFordQueue{ onQueue: make([]bool, len(indexOf)), indexOf: indexOf, } }