add neighborhood

doc/build.jl

@@ -354,7 +354,9 @@ Any graph traversal  will traverse an edge only if it is present in the graph. W
 ## Connectivity / Bipartiteness
 `Graph connectivity` functions are defined on both undirected and directed graphs:
 
-{{is_connected, is_strongly_connected, is_weakly_connected, connected_components, strongly_connected_components, weakly_connected_components, has_self_loop, attracting_components, is_bipartite, condensation, period}}
+{{is_connected, is_strongly_connected, is_weakly_connected, connected_components,
+ strongly_connected_components, weakly_connected_components, has_self_loop,
+ attracting_components, is_bipartite, condensation, period, neighborhood}}
 
 ## Cycle Detection
 In graph theory, a cycle is defined to be a path that starts from some vertex

doc/pathing.md

@@ -131,6 +131,13 @@ period(g::LightGraphs.DiGraph)
 ```
 Computes the (common) period for all nodes in a strongly connected graph.
 
+### neighborhood
+```julia
+neighborhood(g, v::Int, d::Int; dir=:out)
+```
+
+Returns a vector of the vertices in `g` at distance less or equal to `d` from `v`. If `g` is a `DiGraph` the `dir` optional argument specifies the edge direction with respect to `v` (i.e. `:in` or `:out`) to be considered.
+
 ## Cycle Detection
 In graph theory, a cycle is defined to be a path that starts from some vertex
 `v` and ends up at `v`.

src/LightGraphs.jl

@@ -62,7 +62,7 @@ randomwalk, saw, non_backtracking_randomwalk,
 # connectivity
 connected_components, strongly_connected_components, weakly_connected_components,
 is_connected, is_strongly_connected, is_weakly_connected, period,
-condensation, attracting_components,
+condensation, attracting_components, neighborhood,
 
 # cliques
 maximal_cliques,
@@ -100,7 +100,7 @@ maximum_weight_maximal_matching, MatchingResult,
 # randgraphs
 erdos_renyi, watts_strogatz, random_regular_graph, random_regular_digraph, random_configuration_model,
 StochasticBlockModel, make_edgestream, nearbipartiteSBM, blockcounts, blockfractions,
-stochastic_block_model, 
+stochastic_block_model,
 
 #community
 modularity, community_detection_nback, core_periphery_deg,

src/connectivity.jl

@@ -227,3 +227,29 @@ function attracting_components(g::DiGraph)
     end
     return scc[attracting]
 end
+
+"""
+    neighborhood(g, v::Int, d::Int; dir=:out)
+
+Returns a vector of the vertices in `g` at distance less or equal to `d`
+from `v`. If `g` is a `DiGraph` the `dir` optional argument specifies the edge direction
+with respect to `v` (i.e. `:in` or `:out`) to be considered.
+"""
+function neighborhood(g::SimpleGraph, v::Int, d::Int; dir=:out)
+    neig = Set{Int}(v)
+    ∂neig = copy(neig)
+    fneig = dir == :out ? out_neighbors : in_neighbors
+    for l=1:d
+        newneigs = Set{Int}()
+        for i in ∂neig
+            for j in fneig(g, i)
+                if j ∉ neig
+                    push!(newneigs, j)
+                    push!(neig, j)
+                end
+            end
+        end
+        ∂neig = newneigs
+    end
+    return collect(neig)
+end

test/connectivity.jl

@@ -132,3 +132,22 @@ fig8 = DiGraph(fig8)
 
 @test attracting_components(fig1) == Vector[[2,5]]
 @test attracting_components(fig3) == Vector[[3,4],[8]]
+
+g10 = StarGraph(10)
+@test neighborhood(g10, 1 , 0) == [1]
+@test length(neighborhood(g10, 1, 1)) == 10
+@test length(neighborhood(g10, 2, 1)) == 2
+@test length(neighborhood(g10, 1, 2)) == 10
+@test length(neighborhood(g10, 2, 2)) == 10
+
+g10 = StarDiGraph(10)
+@test neighborhood(g10, 1 , 0, dir=:out) == [1]
+@test length(neighborhood(g10, 1, 1, dir=:out)) == 10
+@test length(neighborhood(g10, 2, 1, dir=:out)) == 1
+@test length(neighborhood(g10, 1, 2, dir=:out)) == 10
+@test length(neighborhood(g10, 2, 2, dir=:out)) == 1
+@test neighborhood(g10, 1 , 0, dir=:in) == [1]
+@test length(neighborhood(g10, 1, 1, dir=:in)) == 1
+@test length(neighborhood(g10, 2, 1, dir=:in)) == 2
+@test length(neighborhood(g10, 1, 2, dir=:in)) == 1
+@test length(neighborhood(g10, 2, 2, dir=:in)) == 2