General Graph Drawing
General Graph Drawing
GraphPlot and GraphPlot3D calculate and plot a visually appealing 2D/3D layout of a graph. The functions are designed to work with very large graphs and handle both connected and disconnected graphs.
| GraphPlot[{vi1->vj1,vi2->vj2,…}] | generate a plot of the graph in which vertex vik is connected to vertex vjk |
| GraphPlot[{{vi1->vj1,lbl1},…}] | associate labels lblk with edges in the graph |
| GraphPlot[m] | generate a plot of the graph represented by the adjacency matrix m |
| GraphPlot3D[{vi1->vj1,vi2->vj2,…}] | generate a 3D plot of the graph in which vertex vik is connected to vertex vjk |
| GraphPlot3D[{{vi1->vj1,lbl1},…}] | associate labels lblk with edges in the graph |
| GraphPlot3D[m] | generate a 3D plot of the graph represented by the adjacency matrix m |
GraphPlot[{"d" -> "c", "e" -> "b", "e" -> "c", "e" -> "d", "f" -> "a", "f" -> "b", "f" -> "d", "f" -> "e", "g" -> "a", "g" -> "b", "g" -> "c", "g" -> "e"}, VertexLabels -> Automatic]For disconnected graphs, individual components are laid out in a visually appealing way and assembled:
GraphPlot[Table[i -> Mod[i ^ 2, 129], {i, 0, 128}]]This is a larger graph defined by a sparse adjacency matrix from a structural engineering application. The matrix comes from the Harwell–Boeing Collection:
a = Import["LinearAlgebraExamples/Data/dwt_1005.psa", "HarwellBoeing"];GraphPlot3D[a, VertexShapeFunction -> None]GraphPlot may produce slightly different output on different platforms, due to floating-point differences.
The following options are accepted for GraphPlot and GraphPlot3D; in addition, options for Graphics and Graphics3D are accepted.
option name | default value | |
| DirectedEdges | False | whether to show edges as directed arrows |
| EdgeLabels | Automatic | whether to include labels given for edges |
| EdgeShapeFunction | Automatic | function to give explicit graphics for edges |
| GraphLayout | Automatic | the method used to lay out the graph |
| MultiedgeStyle | Automatic | how to draw multiple edges between vertices |
| PlotRangePadding | Automatic | how much padding to put around the plot |
| "PackingLayout" | Automatic | method to use for packing components |
| PlotStyle | Automatic | style in which objects are drawn |
| SelfLoopStyle | Automatic | how to draw edges linking a vertex to itself |
| VertexCoordinates | Automatic | rules for explicit vertex coordinates |
| VertexLabels | Automatic | whether to show vertex names as labels |
| VertexShapeFunction | Automatic | function to give explicit graphics for vertices |
DirectedEdges
The option DirectedEdges specifies whether to draw edges as arrows. Possible values for this option are True or False. The default value for this option is False.
GraphPlot[{1 -> 6, 2 -> 6, 3 -> 6, 4 -> 6, 5 -> 6}, DirectedEdges -> True, VertexLabels -> Automatic]EdgeLabels
The option EdgeLabels specifies whether and how to display labels given for the edges. Possible values for this option are All, None, or Automatic. The default value for this option is Automatic, which displays the supplied edge labels on the graph.
By default, GraphPlot displays the supplied label for the edge between vertices 3 and 6:
GraphPlot[{1 -> 5, 1 -> 6, 2 -> 4, 2 -> 6, 3 -> 4, 3 -> 5, {3 -> 6, "edge 3->6"}, 4 -> 6, 5 -> 6}, VertexLabels -> Automatic]Alternatively, use Tooltip[vi->vj,lbl] to specify a tooltip for an edge. Place the cursor over the edge between vertices 3 and 6, as well as over the edge label on the edge between vertices 3 and 5, to see the tooltips:
GraphPlot[{1 -> 5, 1 -> 6, 2 -> 4, 2 -> 6, 3 -> 4, {3 -> 5, Tooltip[ "edge 3->5", "3->5"]}, Tooltip[3 -> 6, "3->6"], 4 -> 6, 5 -> 6}, VertexLabels -> Automatic]EdgeShapeFunction
The option EdgeShapeFunction specifies graphical representation of the graph edges. Possible values for this option are Automatic, None, or a function that gives a proper combination of graphics primitives and directives. With the default setting of Automatic, a line is drawn for each edge. With EdgeShapeFunction->None, edges are not drawn.
GraphPlot[Table[1, {30}, {30}], EdgeShapeFunction -> None]With EdgeShapeFunction->g, each edge is rendered with the graphics primitives and directives given by the function g that can take three or more arguments, in the form g[{ri,…,rj},{vi,vj},lblij,…], where ri, rj are the coordinates of the beginning and ending points of the edge, vi, vj are the beginning and ending vertices, and lblij is any label specified for the edge or None. Explicit settings for EdgeShapeFunction->g override settings for EdgeLabels and DirectedEdges.
This plots the edges as gray arrows with ends set back from vertices by a distance 0.1 (in the graph's coordinate system):
GraphPlot[{1 -> 2, 2 -> 3, 3 -> 4, 4 -> 5, 5 -> 6, 6 -> 7, 7 -> 1, 11 -> 12, 12 -> 13, 13 -> 14, 14 -> 15, 15 -> 16, 16 -> 17, 17 -> 11, 1 -> 11, 2 -> 12, 3 -> 13, 4 -> 14, 5 -> 15, 6 -> 16, 7 -> 17}, EdgeShapeFunction -> ({GrayLevel[0.5], Arrow[#1, 0.1]}&)]Table[
GraphPlot[{1 -> 6, 2 -> 6, 3 -> 6, 4 -> 6, 5 -> 6}, ImageSize -> 150, EdgeShapeFunction -> {{"Arrow", "ArrowSize" -> asize}}], {asize, {0.05, 0.1, 0.15}}]GraphPlot[{1 -> 5, 1 -> 6, 2 -> 4, 2 -> 6, 3 -> 4, 3 -> 5, 3 -> 6, 4 -> 6, 5 -> 6}, EdgeShapeFunction -> ({Line[#1], Text[#2, Total[#1] / 2., Background -> White]}&), VertexLabels -> Automatic]GraphPlot3D[{1 -> 2, 2 -> 3, 3 -> 4, 5 -> 1, 5 -> 2, 5 -> 3, 5 -> 4, 1 -> 4, 3 -> 5, 3 -> 3}, EdgeShapeFunction -> (If [Length[#1] > 2, {Red, Line[#1]}, Line[#1]]&)]GraphPlot3D[{2 -> 1, 5 -> 1, 6 -> 1, 3 -> 2, 11 -> 2, 4 -> 3, 16 -> 3, 5 -> 4, 21 -> 4, 26 -> 5, 7 -> 6, 10 -> 6, 8 -> 7, 30 -> 7, 9 -> 8, 42 -> 8, 10 -> 9, 38 -> 9, 12 -> 10, 12 -> 11, 15 -> 11, 13 -> 12, 14 -> 13, 37 -> 13, 15 -> 14, 33 -> 14, 17 -> 15, 17 -> 16, 20 -> 16, 18 -> 17, 19 -> 18, 32 -> 18, 20 -> 19, 53 -> 19, 22 -> 20, 22 -> 21, 25 -> 21, 23 -> 22, 24 -> 23, 52 -> 23, 25 -> 24, 48 -> 24, 27 -> 25, 27 -> 26, 30 -> 26, 28 -> 27, 29 -> 28, 47 -> 28, 30 -> 29, 43 -> 29, 32 -> 31, 35 -> 31, 54 -> 31, 33 -> 32, 34 -> 33, 35 -> 34, 36 -> 34, 56 -> 35, 37 -> 36, 40 -> 36, 38 -> 37, 39 -> 38, 40 -> 39, 41 -> 39, 57 -> 40, 42 -> 41, 45 -> 41, 43 -> 42, 44 -> 43, 45 -> 44, 46 -> 44, 58 -> 45, 47 -> 46, 50 -> 46, 48 -> 47, 49 -> 48, 50 -> 49, 51 -> 49, 59 -> 50, 52 -> 51, 55 -> 51, 53 -> 52, 54 -> 53, 55 -> 54, 60 -> 55, 57 -> 56, 60 -> 56, 58 -> 57, 59 -> 58, 60 -> 59}, EdgeShapeFunction -> ({Cylinder[#1, .05]}&), VertexShapeFunction -> ({Sphere[#, .15]}&)]Spring[{x0_, y0_}, n_ : 10] := Module[{x = x0 + (y0 - x0) * 0.05, y = y0 + (x0 - y0) * 0.1, theta}, theta = If[(y0 - x0)[[1]] == 0., Pi / 2., ArcTan[(y0 - x0)[[2]] / ((y0 - x0)[[1]])]];
Line[Join[{x0}, Table[x + (y - x) * t + 0.05 {Cos[ 2 Pi n t + theta], Sin[2 Pi n t + theta], 0}, {t, 0, 1, .005}], {y0}]]];
GraphPlot3D[{1 -> 2, 2 -> 3, 3 -> 1, 1 -> 4, 2 -> 4, 3 -> 4}, EdgeShapeFunction -> ({Blue, Spring[#1]}&), VertexShapeFunction -> (Sphere[#1, 0.1]&)]normal[{x_, y_}] := 0.03 * {-y, x} / Norm[{x, y}];
GraphPlot[{1 -> 2, 2 -> 3, 3 -> 4, 4 -> 5, 5 -> 6, 6 -> 1, 1 -> 2, 3 -> 4, 5 -> 6, 1 -> 7, 2 -> 8, 3 -> 9, 4 -> 10, 5 -> 11, 6 -> 12}, VertexShapeFunction -> (Text[Style[If[#2 ≤ 6, "C", "H"], Bold], #1, Background -> White]&), EdgeShapeFunction -> (If[Length[#1] > 2, norm = normal[First[#1] - Last[#1]];{Line[{First[#1] + norm, Last[#1] + norm}], Line[{First[#1] - norm, Last[#1] - norm}]}, Line[#1]]&)]GraphPlot[{1 -> 2, 2 -> 3, 3 -> 4, 4 -> 5, 5 -> 6, 6 -> 1}, EdgeShapeFunction -> ({Arrow[#], Inset["Mathematica", Mean[#1], Automatic, Automatic, #[[1]] - #[[2]], Background -> White]}&)]arrow = [image];GraphPlot[{1 -> 5, 1 -> 6, 2 -> 4, 2 -> 6, 3 -> 4, 3 -> 5, 3 -> 6, 4 -> 6, 5 -> 6}, EdgeShapeFunction -> ({Line[#], Inset[arrow, Mean[#1], Automatic, Automatic, #[[1]] - #[[2]], Background -> White]}&)]GraphLayout
Algorithms to be used in GraphPlot and GraphPlot3D can be specified using the GraphLayout option, either as GraphLayout->"name" or GraphLayout->{"name",opt1->val1,…}, where opti-> vali are method-specific options, described in separate sections. GraphLayout->Automatic uses the "RadialEmbedding" method for trees and "SpringElectricalEmbedding" otherwise.
| Automatic | a method suitable for the problem is chosen automatically |
| "CircularEmbedding" | lay out the vertices in a circle |
| "HighDimensionalEmbedding" | invoke the high-dimensional embedding method, in which the graph is first laid out in a high-dimensional space based on the graph distances of the vertices to k centers; this layout is then projected to 2D or 3D space by linear combination of the high-dimensional coordinates using principal component analysis |
| "RadialEmbedding" | invoke the radial embedding method, which is most suitable for tree or tree-like graphs; if the graph is not a tree, a spanning tree is first constructed, and a radial embedding of the spanning tree is used to derive the drawing for the graph |
| "RandomEmbedding" | lay out vertices randomly |
| "SpiralEmbedding" |
lay out the vertices in a spiral; in 3D, this distributes vertices uniformly on a sphere
|
| "SpringElectricalEmbedding" | invoke the spring-electrical embedding method, in which neighboring vertices are subject to an attractive spring force that is proportional to their physical distance, and all vertices are subject to a repulsive electrical force that is inversely proportional to their distance; the overall energy is minimized |
| "SpringEmbedding" | invoke the spring embedding method, in which a vertex is subject to either attractive or repulsive force from another vertex, as though they are connected by a spring; the spring has an ideal length equal to the graph distance between the vertices; the total spring energy is minimized |
Valid values of the GraphLayout option.
GraphPlot[{1 -> 3, 1 -> 4, 2 -> 4, 2 -> 5, 3 -> 5, 6 -> 7, 7 -> 8, 8 -> 9, 9 -> 10, 6 -> 10, 1 -> 6, 2 -> 7, 3 -> 8, 4 -> 9, 5 -> 10}, VertexLabels -> Automatic]This draws the graph using the "SpringEmbedding" algorithm:
GraphPlot[{1 -> 3, 1 -> 4, 2 -> 4, 2 -> 5, 3 -> 5, 6 -> 7, 7 -> 8, 8 -> 9, 9 -> 10, 6 -> 10, 1 -> 6, 2 -> 7, 3 -> 8, 4 -> 9, 5 -> 10}, GraphLayout -> "SpringEmbedding", VertexLabels -> Automatic]This draws the graph using the "HighDimensionalEmbedding" algorithm:
GraphPlot[{1 -> 3, 1 -> 4, 2 -> 4, 2 -> 5, 3 -> 5, 6 -> 7, 7 -> 8, 8 -> 9, 9 -> 10, 6 -> 10, 1 -> 6, 2 -> 7, 3 -> 8, 4 -> 9, 5 -> 10}, GraphLayout -> "HighDimensionalEmbedding", VertexLabels -> Automatic]This draws the complete graph of 30 vertices using the "CircularEmbedding" method:
GraphPlot[Table[1, {30}, {30}], GraphLayout -> "CircularEmbedding", VertexLabels -> Automatic]This draws the complete graph of 30 vertices using the "SpiralEmbedding" method in 3D:
GraphPlot3D[Table[1, {30}, {30}], GraphLayout -> "SpiralEmbedding", VertexLabels -> Automatic]This draws the complete graph using the "RandomEmbedding" method:
GraphPlot[Table[1, {30}, {30}], GraphLayout -> "RandomEmbedding", VertexLabels -> Automatic]MultiedgeStyle
The option MultiedgeStyle specifies whether to draw multiple edges between two vertices. Possible values for MultiedgeStyle are Automatic (the default), True, False, or a positive real number. With the default setting MultiedgeStyle->Automatic, multiple edges are shown for a graph specified by a list of rules, but not shown if specified by an adjacency matrix. With MultiedgeStyle->δ, the multiedges are spread out to a scaled distance of δ.
GraphPlot[{1 -> 2, 2 -> 1, 1 -> 2, 1 -> 2, 2 -> 3, 3 -> 4, 4 -> 5, 5 -> 6, 6 -> 1}, DirectedEdges -> True, VertexLabels -> Automatic]But multiple edges are not shown for graphs specified by an adjacency matrix:
GraphPlot[(| | | | | | |
| :- | :- | :- | :- | :- | :- |
| 0 | 3 | 0 | 0 | 0 | 0 |
| 1 | 0 | 1 | 0 | 0 | 0 |
| 0 | 0 | 0 | 1 | 0 | 0 |
| 0 | 0 | 0 | 0 | 1 | 0 |
| 0 | 0 | 0 | 0 | 0 | 1 |
| 1 | 0 | 0 | 0 | 0 | 0 |), VertexLabels -> Automatic]This spreads multiple edges by the specified amount:
GraphPlot[{1 -> 2, 2 -> 1, 1 -> 2, 1 -> 2, 2 -> 3, 3 -> 4, 4 -> 5, 5 -> 6, 6 -> 1}, MultiedgeStyle -> 1, DirectedEdges -> True, VertexLabels -> Automatic]PackingLayout
The option "PackingLayout" specifies the method used for packing disconnected components. Possible values for the option are Automatic (the default), "ClosestPacking", "ClosestPackingCenter", "Layered", "LayeredLeft", "LayeredTop", and "NestedGrid". With "PackingLayout"->"ClosestPacking", components are packed as close together as possible using a polyomino method [6], starting from the top left. With "PackingLayout"->"ClosestPackingCenter", components are packed starting from the center. With "PackingLayout"->"Layered", components are packed in layers starting from top left. With "PackingLayout"->"LayeredLeft" or "PackingLayout"->"LayeredTop", components are packed in layers starting from the top/left respectively. With "PackingLayout"->"NestedGrid", components are arranged in a nested grid. The typical effective default setting is "PackingLayout"->"Layered", and the packing starts with the components of the largest bounding box area.
g = Flatten[Table[x = RandomReal[1, {50 + 1}];Table[x[[i + 1]] -> x[[Mod[i ^ 2, 50] + 1]], {i, 0, 50}], {10}]];GraphPlot[g]g = Flatten[Table[x = RandomReal[1, {50 + 1}];Table[x[[i + 1]] -> x[[Mod[i ^ 2, 50] + 1]], {i, 0, 50}], {10}]];GraphPlot[g, GraphLayout -> {"PackingLayout" -> "ClosestPackingCenter"}]Users can adjust the packing by suboptions of "PackingLayout". The suboption "Padding" specifies the amount of space to allow between components; possible values are Automatic (the default), or a non-negative number. The suboption "PaddingFunction", which overrides "Padding", also specifies the amount of space to allow between components. It takes a list of the form {{w1, h1},…}, which are the width and height of the bounding box of the components, and returns a non-negative number. Options "PackingLayout"->"ClosestPacking" and "PackingLayout"->"ClosestPackingCenter" also accept a "PolyominoNumber" suboption, which specifies the average number of polyominoes used to approximate each disconnected component. Possible values for the "PolyominoNumber" suboption are Automatic (the default, which usually sets "PolyominoNumber" to 100), or a positive integer. A smaller "PolyominoNumber" typically has the effect of not allowing smaller components to embed in between large components.
g = Flatten[Table[x = RandomReal[1, {50 + 1}];Table[x[[i + 1]] -> x[[Mod[i ^ 2, 50] + 1]], {i, 0, 50}], {10}]];GraphPlot[g, GraphLayout -> {"PackingLayout" -> {"ClosestPackingCenter", "Padding" -> 1}}]g = Flatten[Table[x = RandomReal[1, {50 + 1}];Table[x[[i + 1]] -> x[[Mod[i ^ 2, 50] + 1]], {i, 0, 50}], {10}]];GraphPlot[g, GraphLayout -> {"PackingLayout" -> {"ClosestPackingCenter", "PolyominoNumber" -> 5, "Padding" -> 1}}]PlotRangePadding
PlotStyle
PlotStyle is a common option for graphics functions inherited by GraphPlot and GraphPlot3D. The option PlotStyle specifies the style in which objects are drawn.
GraphPlot[{1 -> 2, 2 -> 3, 3 -> 4, 4 -> 1, 6 -> 5, 7 -> 5, 5 -> 4, 9 -> 8, 10 -> 8, 8 -> 3, 12 -> 11, 13 -> 11, 11 -> 1, 15 -> 14, 16 -> 14, 14 -> 2}, VertexLabels -> Automatic, PlotStyle -> {Red, Thickness[0.02]}]SelfLoopStyle
The option SelfLoopStyle specifies whether and how to draw loops for vertices that are linked to themselves. Possible values of the option are Automatic (the default), True, False, or a positive real number. With SelfLoopStyle->Automatic, self-loops are shown if the graph is specified by a list of rules, but not by an adjacency matrix. With SelfLoopStyle->δ, the self-loops are drawn with a diameter of δ (relative to the average edge length).
GraphPlot[{3 -> 2, 4 -> 1, 4 -> 3, 5 -> 1, 5 -> 2, 6 -> 1, 6 -> 2, 6 -> 3, 6 -> 4, 6 -> 5, 1 -> 1, 1 -> 1}, VertexLabels -> Automatic]Self-loops are not shown if the graph is specified by an adjacency matrix:
GraphPlot[(| | | | | | |
| :- | :- | :- | :- | :- | :- |
| 0 | 1 | 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 | 0 |
| 1 | 0 | 0 | 1 | 0 | 0 |
| 0 | 0 | 0 | 2 | 0 | 0 |
| 0 | 1 | 0 | 1 | 0 | 0 |
| 1 | 1 | 1 | 1 | 1 | 0 |), VertexLabels -> Automatic]GraphPlot[{3 -> 2, 4 -> 1, 4 -> 3, 5 -> 1, 5 -> 2, 6 -> 1, 6 -> 2, 6 -> 3, 6 -> 4, 6 -> 5, 1 -> 1, 1 -> 1}, VertexLabels -> Automatic, SelfLoopStyle -> 1]VertexCoordinates
The option VertexCoordinates specifies the coordinates of the vertices. Possible values are None, a list of coordinates, or a list of rules specifying the coordinates of selected or all vertices.
GraphPlot[{1 -> 3, 1 -> 4, 2 -> 4, 2 -> 5, 3 -> 5, 6 -> 7, 7 -> 8, 8 -> 9, 9 -> 10, 6 -> 10, 1 -> 6, 2 -> 7, 3 -> 8, 4 -> 9, 5 -> 10}, VertexCoordinates -> {{0.309, 0.951}, {-0.809, -0.587}, {0.309, -0.951}, {-0.809, 0.587}, {1., 0}, {0.618, 1.902}, {-1.618, 1.175}, {-1.618, -1.175}, {0.618, -1.902}, {2., 0}}]GraphPlot[{1 -> 3, 1 -> 4, 2 -> 4, 2 -> 5, 3 -> 5, 6 -> 7, 7 -> 8, 8 -> 9, 9 -> 10, 6 -> 10, 1 -> 6, 2 -> 7, 3 -> 8, 4 -> 9, 5 -> 10}, GraphLayout -> "SpringEmbedding"]GraphPlot[{1 -> 2, 2 -> 3, 3 -> 4, 4 -> 1}, VertexCoordinates -> {1 -> {1, 1}, 2 -> {2, 3}}, Frame -> True, FrameTicks -> True, VertexLabels -> Automatic]
GraphPlot[{1 -> 4, 2 -> 3, 2 -> 5, 1 -> 3}, VertexCoordinates -> {1 -> {Automatic, 1}, 2 -> {Automatic, 1}, 3 -> {Automatic, 2}, 4 -> {Automatic, 2}, 5 -> {Automatic, 2}}, Frame -> True, FrameTicks -> True, VertexLabels -> Automatic]This draws a bipartite graph by fixing 
coordinates. "Anchors" are added to connect disconnected components:
coordinates. "Anchors" are added to connect disconnected components:bipartite = {1 -> b, 2 -> a, 3 -> a, 3 -> d, 4 -> c, 4 -> a, 4 -> b};
g = Join[bipartite, Map[(Left -> #[[1]])&, bipartite], Map[(Right -> #[[2]])&, bipartite]];GraphPlot[g, VertexCoordinates -> {left -> {-2, 0}, right -> {2, 0}, Sequence@@Flatten[Map[({#[[1]] -> {-1, Automatic}, #[[2]] -> {1, Automatic}})&, bipartite]]}, VertexShapeFunction -> (If[#2 =!= Left && #2 =!= Right, Text[#2, #1, Background -> White], {}]&), EdgeShapeFunction -> (If[!MemberQ[{Left, Right}, #2[[1]]] && !MemberQ[{Left, Right}, #2[[2]]], Line[#1], {}]&)]When the bipartite graph is connected, it works even better without augmenting it with "Left" and "Right" anchors:
GraphPlot[bipartite, VertexCoordinates -> Flatten[Map[({#[[1]] -> {-1, Automatic}, #[[2]] -> {1, Automatic}})&, bipartite]], VertexLabels -> Automatic]VertexLabels
The option VertexLabels specifies whether to show vertex names as labels. Possible values for this option are All, None and Automatic (the default). VertexLabels->All shows the labels. For graphs specified by an adjacency matrix, vertex labels are taken to be successive integers 
, 
, …, 
, where 
is the size of the matrix. For graphs specified by a list of rules, labels are the expressions used in the rules. VertexLabels->None displays each vertex as a point. You can also use Tooltip[vk,vlbl] anywhere in the list of rules to specify an alternative tooltip for a vertex vk.
, , …, , where is the size of the matrix. For graphs specified by a list of rules, labels are the expressions used in the rules. VertexLabels->None displays each vertex as a point. You can also use Tooltip[vk,vlbl] anywhere in the list of rules to specify an alternative tooltip for a vertex vk. GraphPlot[(| | | | | |
| :- | :- | :- | :- | :- |
| 0 | 1 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 | 0 |
| 1 | 1 | 0 | 0 | 0 |
| 0 | 1 | 0 | 0 | 0 |
| 0 | 1 | 0 | 0 | 0 |), VertexLabels -> Automatic]GraphPlot[{"A" -> "B", "A" -> "a", "B" -> "C", "C" -> "A"}, VertexLabels -> Automatic]GraphPlot[{5 -> 4, 6 -> 2, 6 -> 3, 6 -> 5, 7 -> 1, 7 -> 3, 7 -> 4, 7 -> 6}, VertexLabels -> {All, 3 -> a, 5 -> b}]VertexShapeFunction
The option VertexShapeFunction specifies graphical representation of the graph edges. Possible values for this option are Automatic, None, or a function that gives a proper combination of graphics primitives and directives. With the default setting of Automatic, vertices are displayed as points.
GraphPlot[{5 -> 3, 5 -> 4, 6 -> 2, 6 -> 4, 7 -> 1, 7 -> 4, 7 -> 5, 7 -> 6}]GraphPlot[{5 -> 3, 5 -> 4, 6 -> 2, 6 -> 4, 7 -> 1, 7 -> 4, 7 -> 5, 7 -> 6}, VertexShapeFunction -> None]With VertexShapeFunction->g, each vertex is rendered with the graphics primitives given by g[ri,vi,…], where ri is the coordinate of the vertex and vi is the label of the vertex. Explicit settings for VertexShapeFunction->g override settings for VertexLabels.
GraphPlot[{1 -> 2, 2 -> 3, 3 -> 4, 4 -> 5, 5 -> 6, 6 -> 7, 7 -> 8, 8 -> 1, 1 -> 9, 2 -> 9, 3 -> 10, 4 -> 10, 6 -> 11, 5 -> 11, 7 -> 12, 8 -> 12}, VertexShapeFunction -> ({EdgeForm[Black], Yellow, Disk[#1, 0.2], Black, Text[#2, #1]}&)]gr = [image];GraphPlot[{4 -> 2, 4 -> 3, 5 -> 1, 5 -> 3, 6 -> 1, 6 -> 2, 6 -> 4, 6 -> 5}, VertexShapeFunction -> (Inset[gr, #1]&)]All graph drawing methods accept the method suboption "Rotation", which specifies the desired amount of clockwise rotation in positive radians from the default orientation. The option takes any numeric values, or False. The default is 0.
For GraphPlot and GraphPlot3D, the default orientation is derived by an alignment step where the principal axis is found and the graph drawing is aligned with the 
coordinate. However if "Rotation"->False is specified, this step is skipped.
coordinate. However if "Rotation"->False is specified, this step is skipped.
GraphicsRow[Table[GraphPlot[{4 -> 3, 5 -> 3, 5 -> 4, 6 -> 3, 6 -> 4, 6 -> 5, 7 -> 1, 7 -> 2, 7 -> 3, 7 -> 4, 7 -> 5, 7 -> 6}, GraphLayout -> {"SpringElectricalEmbedding", "Rotation" -> rot}], {rot, {0, Pi / 2, Pi}}]]Grid[Partition[Table[GraphPlot[{4 -> 3, 5 -> 3, 5 -> 4, 6 -> 3, 6 -> 4, 6 -> 5, 7 -> 1, 7 -> 2, 7 -> 3, 7 -> 4, 7 -> 5, 7 -> 6}, GraphLayout -> {"SpringElectricalEmbedding", "MaxIteration" -> i}, ImageSize -> 50], {i, 15}], {5}]]Both the SpringEmbedding and SpringElectricalEmbedding methods belong to the family of so-called force-directed methods. These methods work by calculating the force on each vertex, and iteratively moving the vertex along the force in an effort to minimize the overall system's energy. See [8] for algorithmic details. These two methods have the following common options.
option name | default value | |
| "EnergyControl" | Automatic | how the energy function is controlled during minimization |
| "InferentialDistance" | Automatic | cutoff distance beyond which the force calculation ignores inference from faraway vertices |
| MaxIterations | Automatic | maximum number of iterations to be used in attempting to minimize the energy |
| "RandomSeed" | Automatic | seed to use in the random generator for initial vertex placement |
| "RecursionMethod" | Automatic | whether a multilevel algorithm is used to lay out the graph |
| "StepControl" | Automatic | how step lengths are modified during energy minimization |
| "StepLength" | Automatic | initial step length used in moving the vertices |
| "Tolerance" | Automatic | tolerance used in terminating the energy minimization process |
"EnergyControl"
The suboption "EnergyControl" specifies limitations on the total energy of the system during minimization. Possible values are Automatic (the default), "Monotonic", or "NonMonotonic". When the value is "Monotonic", a step along the force will only be accepted if the energy is lowered. When the value is "NonMonotonic", a step along the force will be accepted even if the energy is not lowered.
"InferentialDistance"
The suboption "InferentialDistance" specifies a cutoff distance beyond which the interaction between vertices is assumed to be nonexistent. Possible values are Automatic (the default) or a positive numeric value. For the "SpringEmbedding" method, if the graph distance between a vertex i and a vertex j is greater than the option value of "InferentialDistance", the repulsive and attractive spring force between i and j is ignored. For the "SpringElectricalEmbedding" method, if the Euclidean distance between a vertex i and a vertex j is greater than the option value of "InferentialDistance", the repulsive force between i and j is ignored.
This draws a random tree using the "SpringElectricalEmbedding" method:
g = RandomInteger[#] -> # + 1& /@ Range[0, 1000];GraphPlot[g, GraphLayout -> "SpringElectricalEmbedding"]Using a smaller (more negative) "RepulsiveForcePower" option value (see the next section), the graph now fills more space:
GraphPlot[g, GraphLayout -> {"SpringElectricalEmbedding", "RepulsiveForcePower" -> -2}]A similar effect can be achieved using a small "InferentialDistance" option value:
GraphPlot[g, GraphLayout -> {"SpringElectricalEmbedding", "InferentialDistance" -> .5}]MaxIterations
The option MaxIterations specifies the maximum number of iterations to be used in attempting to minimize the energy. Possible values are Automatic (the default) or a positive integer.
"RandomSeed"
The option "RandomSeed" specifies a seed for the random number generator that computes the initial vertex placement. Changing this option usually affects the orientation of the drawing of the graph, but it can also change the layout. Possible values are Automatic or an integer.
g = {1 -> 3, 1 -> 4, 2 -> 4, 2 -> 5, 3 -> 5, 6 -> 7, 7 -> 8, 8 -> 9, 9 -> 10, 6 -> 10, 1 -> 6, 2 -> 7, 3 -> 8, 4 -> 9, 5 -> 10};GraphPlot[g, GraphLayout -> {"SpringElectricalEmbedding", "RandomSeed" -> Automatic}, VertexLabels -> Automatic]GraphPlot[g, GraphLayout -> {"SpringElectricalEmbedding", "RandomSeed" -> 4321}, VertexLabels -> Automatic]"RecursionMethod"
The option "RecursionMethod" specifies whether the graph layout should be produced by a recursive procedure. Possible values are Automatic (the default), "Multilevel", or None. In a "Multilevel" algorithm, the graph is successively coarsened into graphs with a smaller and smaller number of vertices. The coarser graphs are laid out first, and those layouts are interpolated into the finer graphs and then further refined.
suboption name | default value | |
| "Randomization" | Automatic | whether to inspect vertices in random order |
| "MinSize" | Automatic | minimal number of vertices in a coarsened graph |
| "CoarseningScheme" | Automatic | how graphs are coarsened |
For the option "Randomize", possible values are Automatic, True, and False. For "MinSize", possible values are Automatic or a positive number. For "CoarseningScheme", the implemented algorithms are based on either a maximal independent vertex set, which forms the coarse vertices, or a maximal independent edge set, also called a matching. In a matching, two vertices that form an edge are merged to form a coarse graph vertex. The following are possible values for "CoarseningScheme".
| "MaximalIndependentVertexSet" | link vertices in the maximal independent set if their graph distance is 3 or less |
| "MaximalIndependentVertexSetInjection" | |
link vertices in the maximal independent set if their graph distance is 1 or 2 | |
| "MaximalIndependentVertexSetRugeStuben" | |
generate the maximal independent vertex set, giving priority to vertices with more neighbors not in the set, then link vertices in the set if their graph distance is 3 or less
| |
| "MaximalIndependentVertexSetRugeStubenInjection" | |
link vertices if their graph distance is 1 or 2, giving priority to vertices with more neighbors
| |
| "MaximalIndependentEdgeSet" | consider edges in their natural order when matching |
| "MaximalIndependentEdgeSetHeavyEdge" | |
give priority to edges with higher edge weight (i.e., edges that represent a larger number of edges in the original graph) when matching
| |
| "MaximalIndependentEdgeSetSmallestVertexWeight" | |
give priority to matchings of vertices with neighbors that have the smallest vertex weight | |
"StepControl"
The option "StepControl" defines how step length is modified during energy minimization. It can be Automatic (the default), "Monotonic" (where step length can only be decreased), "NonMonotonic" (where step length can be made larger or smaller), or "StrictlyMonotonic" (where step length is strictly reduced between iterations).
"StepLength"
The option "StepLength" gives the initial step length used in moving the vertices around. Possible values are Automatic (the default) or a positive real number.
Tolerance
The option Tolerance specifies the tolerance used in terminating the energy minimization process. If the average change of coordinates of each vertex is less than the tolerance, the energy minimization process is terminated and the current coordinates are given as output. Possible values are Automatic or a positive real number.
option name | value | |
| "Octree" | Automatic |
whether to use an octree data structure (in three dimensions) or a quadtree data structure (in two dimensions) in the calculation of repulsive force
|
| "RepulsiveForcePower " | -1 | how fast the repulsive force decays over distance |
"Octree"
The "Octree" option specifies whether to use an octree data structure (in three dimensions) or a quadtree data structure (in two dimensions) in the calculation of repulsive force. Possible values are Automatic (the default), True, or False. Use of an octree/quadtree data structure minimizes the complexity of computation by approximating the long-range repulsive force. However, it introduces an approximation to the force calculation. Therefore, in a few cases the result may not be as good.
"RepulsiveForcePower"
Possible values are negative real numbers, with -1 as the default. In the spring-electrical embedding, the repulsive force between two vertices 
and 
is 
by default. If the value of RepulsiveForcePower is 
(with 
), then the repulsive force is defined as 
, where 
is the distance between the vertices and 
is a constant coefficient.
and is by default. If the value of RepulsiveForcePower is (with ), then the repulsive force is defined as , where is the distance between the vertices and is a constant coefficient.A strong long-range repulsive force over long distance often has the boundary effect that vertices in the periphery are closer to each other than those in the center are. Specifying a weaker long-range repulsive force can sometimes alleviate this effect. This option can also be useful in drawing a graph so that it fills up more space. (See the "InferentialDistance" method option for details.)
With a repulsive force power of 
, the boundary vertices are not as close to each other as they are with the default value of 
:
, the boundary vertices are not as close to each other as they are with the default value of :GraphPlot[GridGraph[{20, 20}], GraphLayout -> "SpringElectricalEmbedding"]GraphPlot[GridGraph[{20, 20}], GraphLayout -> {"SpringElectricalEmbedding", "RepulsiveForcePower" -> -2}]Drawing a Graph to Fill Up More Space
While the default setting for GraphPlot works well in general, for graphs that have a wide range of values for the vertex degree, it is often necessary to use a setting that helps the vertices to occupy less space.
SeedRandom[321];g = Table[i -> RandomInteger[{1, 52}], {i, 52}]GraphPlot[g]GraphPlot[g, GraphLayout -> "SpringEmbedding"]A similar effect can be achieved with a repulsive force power smaller than the default (-1), so that the repulsive force decays more quickly over distance:
GraphPlot[g, GraphLayout -> {"SpringElectricalEmbedding", "RepulsiveForcePower" -> -1.8}]GraphPlot[g, GraphLayout -> {"SpringElectricalEmbedding", "InferentialDistance" -> 4}]For such tree-like graphs, a tree-drawing algorithm may be preferable. See "Tree Drawing" for greater control over tree layout:
GraphPlot[g, GraphLayout -> "RadialEmbedding"]GraphPlot[ExampleData[{"Matrix", "HB/1138_bus"}, "Matrix"], GraphLayout -> {"SpringElectricalEmbedding", "InferentialDistance" -> 4, "RepulsiveForcePower" -> -1.8}, PlotRangePadding -> 0]Improving Performance for Drawing Very Large Graphs
Although the default option set usually ensures a very good performance, it is often possible to further increase drawing speed and reduce memory usage by selecting specific option combinations for a particular task. For example, speed and/or memory usage can be improved using a smaller number of iterations, a smaller inferential distance, or a lower tolerance. These settings tend to reduce quality, but still frequently offer an acceptable compromise.
g = Import["LinearAlgebraExamples/Data/nos6.mtx"];GraphPlot[g]//TimingA coarsening scheme based on the maximal independent vertex set is often faster and uses less memory, and yet offers a comparable layout quality:
GraphPlot[g, GraphLayout -> {"SpringElectricalEmbedding", "RecursionMethod" -> {"Multilevel", "CoarseningScheme" -> "MaximalIndependentVertexSetRugeStuben"}}]//TimingGraphPlot[g, GraphLayout -> {"SpringElectricalEmbedding", "MaxIteration" -> 30}]//TimingSetting the inferential distance to 2 and the number of iterations to 40 is also faster than the default:
GraphPlot[g, GraphLayout -> {"SpringElectricalEmbedding", "InferentialDistance" -> 2, "MaxIteration" -> 40}]//TimingGraphPlot[g, GraphLayout -> {"SpringElectricalEmbedding", "InferentialDistance" -> 2, "MaxIteration" -> 20}]//TimingGraphPlot[g, GraphLayout -> {"SpringElectricalEmbedding", "InferentialDistance" -> 2, "MaxIteration" -> 20, "RecursionMethod" -> {"Multilevel", "CoarseningScheme" -> "MaximalIndependentVertexSetRugeStuben"}}]//Timing"HighDimensionalEmbedding" tends to be the fastest method, but the quality of the drawing often suffers:
GraphPlot[g, GraphLayout -> "HighDimensionalEmbedding"]//TimingFor comparison, "SpringEmbedding" is the slowest method, but it is the only one that draws the mesh using orthogonal lines:
GraphPlot[g, GraphLayout -> "SpringEmbedding"]//TimingE. coli Transcription Networks
In a graph representation of a transcriptional regulation network that controls gene expression in cells, nodes (vertices) are operons, which are one or more genes transcribed on the same messenger ribonucleic acid (mRNA). Edges of the graph are directed from an operon that encodes a transcription factor to an operon that it directly regulates [1].
Data
g = {{"acrR" -> "acrAB", 2}, {"ada_alkB" -> "aidB", 1}, {"ada_alkB" -> "alkA", 1}, {"adiA_adiY" -> "adiA", 1}, {"alpA" -> "slp", 1}, {"appY" -> "appCBA", 1}, {"araC" -> "araBAD", 3}, {"araC" -> "araE", 3}, {"araC" -> "araFG_araH_1H_2", 3}, {"araC" -> "araJ", 1}, {"arcA" -> "aceBAK", 2}, {"arcA" -> "appCBA", 1}, {"arcA" -> "appY", 1}, {"arcA" -> "betIBA", 2}, {"arcA" -> "cydAB", 1}, {"arcA" -> "cyoABCDE", 2}, {"arcA" -> "dctA", 2}, {"arcA" -> "fadBA", 2}, {"arcA" -> "focA_pflB", 1}, {"arcA" -> "fumA", 2}, {"arcA" -> "fumC", 2}, {"arcA" -> "glcDEFGB", 2}, {"arcA" -> "glpACB", 2}, {"arcA" -> "gltA", 2}, {"arcA" -> "icdA", 2}, {"arcA" -> "lctPRD", 2}, {"arcA" -> "mdh", 2}, {"arcA" -> "nuoABCEFGHIJKLMN", 2}, {"arcA" -> "sdhCDAB_b0725_sucABCD", 2}, {"arcA" -> "sodA", 2}, {"argR" -> "argCBH", 2}, {"argR" -> "argD", 2}, {"argR" -> "argE", 2}, {"argR" -> "argF", 2}, {"argR" -> "argI", 2}, {"argR" -> "carAB", 2}, {"arsR" -> "arsEFG", 2}, {"asnC" -> "asnA", 1}, {"atoC" -> "atoDAB", 1}, {"atoC" -> "atoDAE", 1}, {"betIBA" -> "betT", 2}, {"birA_murA" -> "bioA", 2}, {"birA_murA" -> "bioBFCD", 2}, {"cadC" -> "cadBA", 1}, {"caiF" -> "caiTABCDE", 1}, {"caiF" -> "fixABCX", 1}, {"cbl" -> "ssuEADCB", 1}, {"cbl" -> "tauABCD", 1}, {"cpxAR" -> "cpxP", 1}, {"cpxAR" -> "dsbA", 1}, {"cpxAR" -> "ecfI", 1}, {"cpxAR" -> "htrA", 1}, {"cpxAR" -> "motABcheAW", 2}, {"cpxAR" -> "rotA", 1}, {"cpxAR" -> "skp_lpxDA_fabZ", 1}, {"cpxAR" -> "tsr", 2}, {"cpxAR" -> "xprB_dsbC_recJ", 1}, {"crp" -> "acs", 1}, {"crp" -> "aldB", 1}, {"crp" -> "ansB", 1}, {"crp" -> "araBAD", 1}, {"crp" -> "araC", 1}, {"crp" -> "araE", 1}, {"crp" -> "araFG_araH_1H_2", 1}, {"crp" -> "araJ", 1}, {"crp" -> "caiF", 1}, {"crp" -> "caiTABCDE", 1}, {"crp" -> "cirA", 1}, {"crp" -> "cpdB", 1}, {"crp" -> "cyaA", 2}, {"crp" -> "dadAX", 1}, {"crp" -> "dctA", 2}, {"crp" -> "dcuB_fumB", 2}, {"crp" -> "deoCABD", 3}, {"crp" -> "dsdXA", 1}, {"crp" -> "ebgAC", 1}, {"crp" -> "epd_pgk", 1}, {"crp" -> "fadL", 2}, {"crp" -> "fixABCX", 1}, {"crp" -> "flhDC", 1}, {"crp" -> "focA_pflB", 1}, {"crp" -> "fucAO", 1}, {"crp" -> "fucPIKUR", 1}, {"crp" -> "fur", 1}, {"crp" -> "galETKM", 3}, {"crp" -> "galS", 1}, {"crp" -> "glgCAP", 1}, {"crp" -> "glgS", 1}, {"crp" -> "glnALG", 1}, {"crp" -> "glpACB", 1}, {"crp" -> "glpD", 1}, {"crp" -> "glpFK", 1}, {"crp" -> "glpTQ", 1}, {"crp" -> "gltA", 1}, {"crp" -> "gntKU", 1}, {"crp" -> "gntT", 1}, {"crp" -> "ivbL_ilvBN", 1}, {"crp" -> "lacZYA", 1}, {"crp" -> "malEFG", 1}, {"crp" -> "malI", 2}, {"crp" -> "malK_lamB_malM", 1}, {"crp" -> "malS", 1}, {"crp" -> "malT", 1}, {"crp" -> "malXY", 3}, {"crp" -> "manXYZ", 1}, {"crp" -> "melAB", 1}, {"crp" -> "melR", 1}, {"crp" -> "mglBAC", 1}, {"crp" -> "nagBACD", 1}, {"crp" -> "nagE", 1}, {"crp" -> "nupG", 3}, {"crp" -> "ompA", 2}, {"crp" -> "ppiA", 3}, {"crp" -> "proP", 3}, {"crp" -> "ptsHI_crr", 3}, {"crp" -> "rhaBAD", 3}, {"crp" -> "rhaT", 1}, {"crp" -> "rpoH", 3}, {"crp" -> "sdhCDAB_b0725_sucABCD", 1}, {"crp" -> "speC", 2}, {"crp" -> "srlAEBD_gutM_srlR_gutQ", 1}, {"crp" -> "tdcABCDEFG", 1}, {"crp" -> "tnaLAB", 1}, {"crp" -> "tsx", 3}, {"crp" -> "ubiG", 1}, {"crp" -> "udp", 1}, {"crp" -> "uhpT", 1}, {"crp" -> "yhfA", 3}, {"crp" -> "yiaKLMNOPQRS", 1}, {"csgDEFG" -> "csgBA", 1}, {"cspA" -> "gyrA", 1}, {"cspA" -> "hns", 1}, {"cynR" -> "cynTSX", 1}, {"cysB" -> "cysDNC", 1}, {"cysB" -> "cysJIH", 1}, {"cysB" -> "cysK", 1}, {"cysB" -> "cysPUWAM", 1}, {"cysB" -> "tauABCD", 1}, {"cytR" -> "deoCABD", 2}, {"cytR" -> "nupC", 2}, {"cytR" -> "nupG", 2}, {"cytR" -> "ppiA", 2}, {"cytR" -> "rpoH", 2}, {"cytR" -> "tsx", 3}, {"cytR" -> "udp", 2}, {"deoR" -> "deoCABD", 2}, {"deoR" -> "nupG", 2}, {"deoR" -> "tsx", 2}, {"dnaA" -> "nrdAB", 1}, {"dnaA" -> "rpoH", 2}, {"dsdC" -> "dsdXA", 1}, {"ebgR" -> "ebgAC", 2}, {"envY_ompT" -> "ompC", 1}, {"envY_ompT" -> "ompF", 1}, {"evgA" -> "ompC", 1}, {"exuR" -> "exuT", 2}, {"exuR" -> "uxaCA", 2}, {"exuR" -> "uxuABR", 2}, {"fadR" -> "fabA", 1}, {"fadR" -> "fadBA", 2}, {"fadR" -> "fadL", 2}, {"fadR" -> "iclMR", 2}, {"fadR" -> "uspA", 2}, {"fecIR" -> "fecABCDE", 1}, {"fhlA" -> "fdhF", 1}, {"fhlA" -> "hycABCDEFGH", 1}, {"fhlA" -> "hypABCDE", 1}, {"flhDC" -> "flgAMN", 1}, {"flhDC" -> "flgBCDEFGHIJK", 1}, {"flhDC" -> "flhBAE", 1}, {"flhDC" -> "fliAZY", 1}, {"flhDC" -> "fliE", 1}, {"flhDC" -> "fliFGHIJK", 1}, {"flhDC" -> "fliLMNOPQR", 1}, {"fliAZY" -> "flgBCDEFGHIJK", 1}, {"fliAZY" -> "flgKL", 1}, {"fliAZY" -> "flgMN", 1}, {"fliAZY" -> "flhBAE", 1}, {"fliAZY" -> "fliC", 1}, {"fliAZY" -> "fliDST", 1}, {"fliAZY" -> "fliE", 1}, {"fliAZY" -> "fliFGHIJK", 1}, {"fliAZY" -> "fliLMNOPQR", 1}, {"fliAZY" -> "motABcheAW", 1}, {"fliAZY" -> "tarTapcheRBYZ", 1}, {"fliAZY" -> "tsr", 1}, {"fnr" -> "acs", 1}, {"fnr" -> "ansB", 1}, {"fnr" -> "arcA", 1}, {"fnr" -> "aspA", 1}, {"fnr" -> "caiF", 1}, {"fnr" -> "cydAB", 2}, {"fnr" -> "cyoABCDE", 2}, {"fnr" -> "dcuB_fumB", 1}, {"fnr" -> "dmsABC", 1}, {"fnr" -> "fdnGHI", 1}, {"fnr" -> "focA_pflB", 1}, {"fnr" -> "frdABCD", 1}, {"fnr" -> "glpACB", 1}, {"fnr" -> "hypABCDE", 1}, {"fnr" - ... -> "adhE", 2}, {"narL" -> "caiF", 2}, {"narL" -> "dcuB_fumB", 2}, {"narL" -> "dmsABC", 2}, {"narL" -> "fdnGHI", 1}, {"narL" -> "focA_pflB", 2}, {"narL" -> "frdABCD", 2}, {"narL" -> "narGHJI", 1}, {"narL" -> "narK", 1}, {"narL" -> "nirBDC_cysG", 1}, {"narL" -> "nrfABCDEFG", 1}, {"narL" -> "nuoABCEFGHIJKLMN", 1}, {"narL" -> "torCAD", 2}, {"nhaR" -> "nhaA", 1}, {"nlpD_rpoS" -> "acs", 1}, {"nlpD_rpoS" -> "adhE", 1}, {"nlpD_rpoS" -> "aldB", 1}, {"nlpD_rpoS" -> "alkA", 1}, {"nlpD_rpoS" -> "appY", 1}, {"nlpD_rpoS" -> "cpxAR", 1}, {"nlpD_rpoS" -> "dps", 1}, {"nlpD_rpoS" -> "ftsQAZ", 1}, {"nlpD_rpoS" -> "katG", 1}, {"nlpD_rpoS" -> "narZYWV", 1}, {"nlpD_rpoS" -> "nhaA", 1}, {"nlpD_rpoS" -> "osmC", 1}, {"nlpD_rpoS" -> "osmY", 1}, {"nlpD_rpoS" -> "proP", 1}, {"ompR_envZ" -> "csgBA", 1}, {"ompR_envZ" -> "csgDEFG", 1}, {"ompR_envZ" -> "fadL", 2}, {"ompR_envZ" -> "flhDC", 2}, {"ompR_envZ" -> "ompC", 1}, {"ompR_envZ" -> "ompF", 3}, {"oxyR" -> "ahpCF", 1}, {"oxyR" -> "dps", 1}, {"oxyR" -> "gorA", 1}, {"oxyR" -> "katG", 1}, {"phoBR" -> "phnCDE_f73_phnFGHIJKLMNOP", 1}, {"phoBR" -> "phoA", 1}, {"phoBR" -> "phoE", 1}, {"phoBR" -> "pstSCAB_phoU", 1}, {"pspF" -> "pspABCDE", 1}, {"purR" -> "codBA", 2}, {"purR" -> "cvpA_purF_ubiX", 2}, {"purR" -> "gcvTHP", 2}, {"purR" -> "glnB", 2}, {"purR" -> "glyA", 2}, {"purR" -> "guaBA", 2}, {"purR" -> "prsA", 2}, {"purR" -> "purC", 2}, {"purR" -> "purEK", 2}, {"purR" -> "purHD", 2}, {"purR" -> "purL", 2}, {"purR" -> "purMN", 2}, {"purR" -> "pyrC", 2}, {"purR" -> "pyrD", 2}, {"purR" -> "speA", 2}, {"purR" -> "ycfC_purB", 2}, {"rbsR" -> "rbsDACBK", 2}, {"rcsA" -> "ftsQAZ", 1}, {"rcsA" -> "wza_wzb_b2060_wcaA_wcaB", 1}, {"rhaSR" -> "rhaBAD", 1}, {"rhaSR" -> "rhaT", 1}, {"rob" -> "aslB", 1}, {"rob" -> "fumC", 1}, {"rob" -> "galETKM", 2}, {"rob" -> "inaA", 1}, {"rob" -> "marRAB", 1}, {"rob" -> "mdlA", 1}, {"rob" -> "nfo", 1}, {"rob" -> "sodA", 1}, {"rob" -> "ybaO", 1}, {"rob" -> "ybiS", 1}, {"rob" -> "yfhD", 1}, {"rob" -> "zwf", 1}, {"rpiR_alsBACEK" -> "rpiB", 2}, {"rpoE_rseABC" -> "cutC", 1}, {"rpoE_rseABC" -> "dapA_nlpB_purA", 1}, {"rpoE_rseABC" -> "ecfABC", 1}, {"rpoE_rseABC" -> "ecfD", 1}, {"rpoE_rseABC" -> "ecfF", 1}, {"rpoE_rseABC" -> "ecfG", 1}, {"rpoE_rseABC" -> "ecfH", 1}, {"rpoE_rseABC" -> "ecfI", 1}, {"rpoE_rseABC" -> "ecfJ", 1}, {"rpoE_rseABC" -> "ecfK", 1}, {"rpoE_rseABC" -> "ecfLM", 1}, {"rpoE_rseABC" -> "fkpA", 1}, {"rpoE_rseABC" -> "htrA", 1}, {"rpoE_rseABC" -> "ksgA_epaG_epaH", 1}, {"rpoE_rseABC" -> "lpxDA_fabZ", 1}, {"rpoE_rseABC" -> "mdoGH", 1}, {"rpoE_rseABC" -> "nlpB_purA", 1}, {"rpoE_rseABC" -> "ostA_surA_pdxA", 1}, {"rpoE_rseABC" -> "rfaDFCL", 1}, {"rpoE_rseABC" -> "rpoD", 1}, {"rpoE_rseABC" -> "rpoH", 1}, {"rpoE_rseABC" -> "skp_lpxDA_fabZ", 1}, {"rpoE_rseABC" -> "uppS_cdsA_ecfE", 1}, {"rpoE_rseABC" -> "xprB_dsbC_recJ", 1}, {"rpoH" -> "clpP", 1}, {"rpoH" -> "dnaKJ", 1}, {"rpoH" -> "grpE", 1}, {"rpoH" -> "hflB", 1}, {"rpoH" -> "htpG", 1}, {"rpoH" -> "htpY", 1}, {"rpoH" -> "ibpAB", 1}, {"rpoH" -> "lon", 1}, {"rpoH" -> "mopA", 1}, {"rpoH" -> "mopB", 1}, {"rpoN" -> "atoC", 1}, {"rpoN" -> "dctA", 1}, {"rpoN" -> "fdhF", 1}, {"rpoN" -> "fhlA", 1}, {"rpoN" -> "glnALG", 1}, {"rpoN" -> "glnHPQ", 1}, {"rpoN" -> "hycABCDEFGH", 1}, {"rpoN" -> "hypA", 1}, {"rpoN" -> "nac", 1}, {"rpoN" -> "nycA", 1}, {"rpoN" -> "pspABCDE", 1}, {"rpoN" -> "rtcR", 1}, {"rpoN" -> "zraP", 1}, {"rtcR" -> "rtcAB", 2}, {"soxR" -> "soxS", 1}, {"soxS" -> "acnA", 1}, {"soxS" -> "fpr", 1}, {"soxS" -> "fumC", 1}, {"soxS" -> "nfo", 1}, {"soxS" -> "sodA", 1}, {"soxS" -> "zwf", 1}, {"tdcAR" -> "tdcABCDEFG", 1}, {"torR" -> "torCAD", 1}, {"treR" -> "treBC", 2}, {"trpR" -> "aroH", 2}, {"trpR" -> "aroL_yaiA_aroM", 2}, {"trpR" -> "mtr", 2}, {"trpR" -> "trpLEDCBA", 2}, {"tyrR" -> "aroF_tyrA", 2}, {"tyrR" -> "aroG", 2}, {"tyrR" -> "aroL_yaiA_aroM", 2}, {"tyrR" -> "aroP", 2}, {"tyrR" -> "mtr", 1}, {"tyrR" -> "tyrB", 2}, {"tyrR" -> "tyrP", 3}, {"uhpA" -> "uhpT", 1}, {"uidR" -> "uidRABC", 2}, {"uxuABR" -> "uidRABC", 2}, {"xapR" -> "xapAB", 1}, {"xylFGHR" -> "xylAB", 1}, {"yhdG_fis" -> "adhE", 1}, {"yhdG_fis" -> "alaWX", 1}, {"yhdG_fis" -> "aldB", 3}, {"yhdG_fis" -> "argU", 1}, {"yhdG_fis" -> "argW", 1}, {"yhdG_fis" -> "argX_hisR_leuT_proM", 1}, {"yhdG_fis" -> "aspV", 1}, {"yhdG_fis" -> "leuQPV", 1}, {"yhdG_fis" -> "leuX", 1}, {"yhdG_fis" -> "lysT_valT_lysW", 1}, {"yhdG_fis" -> "metT_leuW_glnUW_metU_glnVX", 1}, {"yhdG_fis" -> "metY_yhbC_nusA_infB", 1}, {"yhdG_fis" -> "nrdAB", 1}, {"yhdG_fis" -> "pdhR_aceEF_lpdA", 1}, {"yhdG_fis" -> "pheU", 1}, {"yhdG_fis" -> "pheV", 1}, {"yhdG_fis" -> "proK", 1}, {"yhdG_fis" -> "proL", 1}, {"yhdG_fis" -> "proP", 1}, {"yhdG_fis" -> "sdhCDAB_b0725_sucABCD", 1}, {"yhdG_fis" -> "serT", 1}, {"yhdG_fis" -> "serX", 1}, {"yhdG_fis" -> "thrU_tyrU_glyT_thrT", 1}, {"yhdG_fis" -> "thrW", 1}, {"yhdG_fis" -> "tyrTV", 1}, {"yhdG_fis" -> "valUXY_lysV", 1}, {"yiaJ" -> "yiaKLMNOPQRS", 2}, {"yjbK" -> "znuABC", 2}, {"yjdHG" -> "dctA", 1}, {"yjdHG" -> "dcuB_fumB", 1}, {"yjdHG" -> "frdABCD", 1}, {"zntR" -> "zntA", 1}};Drawing the Network
GraphPlot[g, EdgeLabels -> None]GraphPlot[g, EdgeLabels -> None, GraphLayout -> {"PackingLayout" -> "ClosestPackingCenter"}]GraphPlot[g, EdgeLabels -> None, GraphLayout -> {"SpringElectricalEmbedding", "RepulsiveForcePower" -> -2}]GraphPlot[g, EdgeLabels -> None, GraphLayout -> {"SpringElectricalEmbedding", "InferentialDistance" -> .5}]Protein: An Oxidoreductase
GraphPlot[{{1 -> 2, 1}, {1 -> 9, 2}, {1 -> 17, 2}, {2 -> 3, 1}, {2 -> 9, 2}, {3 -> 49, 2}, {4 -> 5, 2}, {4 -> 42, 2}, {5 -> 6, 2}, {5 -> 9, 2}, {5 -> 17, 2}, {5 -> 42, 1}, {6 -> 7, 1}, {6 -> 8, 2}, {6 -> 17, 2}, {6 -> 28, 2}, {6 -> 42, 2}, {7 -> 8, 2}, {7 -> 16, 2}, {7 -> 28, 2}, {7 -> 44, 2}, {8 -> 10, 2}, {8 -> 11, 2}, {8 -> 14, 2}, {8 -> 16, 2}, {8 -> 18, 2}, {8 -> 19, 2}, {8 -> 28, 2}, {9 -> 17, 2}, {10 -> 11, 1}, {10 -> 12, 2}, {10 -> 13, 2}, {10 -> 14, 2}, {10 -> 17, 2}, {10 -> 18, 2}, {11 -> 12, 2}, {11 -> 18, 2}, {12 -> 13, 2}, {12 -> 14, 2}, {12 -> 18, 2}, {13 -> 14, 2}, {13 -> 15, 2}, {13 -> 16, 2}, {13 -> 18, 2}, {13 -> 32, 2}, {14 -> 15, 1}, {14 -> 16, 2}, {14 -> 18, 2}, {14 -> 32, 1}, {14 -> 36, 2}, {15 -> 16, 2}, {15 -> 32, 2}, {15 -> 36, 2}, {15 -> 37, 2}, {16 -> 17, 2}, {16 -> 18, 2}, {16 -> 32, 2}, {16 -> 33, 2}, {16 -> 37, 2}, {18 -> 19, 2}, {19 -> 20, 2}, {19 -> 43, 2}, {20 -> 21, 2}, {20 -> 43, 2}, {21 -> 22, 2}, {22 -> 27, 2}, {22 -> 28, 2}, {22 -> 29, 2}, {23 -> 24, 2}, {23 -> 25, 2}, {23 -> 26, 2}, {23 -> 27, 2}, {23 -> 30, 2}, {23 -> 31, 2}, {23 -> 39, 2}, {24 -> 25, 2}, {24 -> 31, 2}, {24 -> 39, 2}, {25 -> 26, 2}, {25 -> 31, 2}, {26 -> 27, 2}, {26 -> 30, 2}, {26 -> 31, 2}, {26 -> 32, 2}, {27 -> 30, 2}, {28 -> 29, 2}, {29 -> 85, 2}, {30 -> 31, 2}, {30 -> 32, 2}, {30 -> 33, 2}, {32 -> 36, 2}, {32 -> 85, 2}, {33 -> 36, 2}, {35 -> 36, 2}, {35 -> 79, 2}, {35 -> 82, 2}, {36 -> 37, 2}, {37 -> 38, 2}, {39 -> 40, 2}, {39 -> 41, 2}, {39 -> 47, 2}, {40 -> 41, 2}, {40 -> 46, 2}, {40 -> 48, 2}, {43 -> 44, 2}, {44 -> 45, 1}, {45 -> 46, 1}, {47 -> 48, 2}, {50 -> 51, 1}, {50 -> 52, 2}, {50 -> 59, 1}, {50 -> 67, 2}, {51 -> 52, 2}, {51 -> 59, 2}, {51 -> 67, 2}, {52 -> 53, 1}, {53 -> 93, 2}, {53 -> 99, 2}, {54 -> 55, 2}, {54 -> 92, 2}, {55 -> 56, 2}, {55 -> 59, 2}, {55 -> 67, 2}, {55 -> 92, 1}, {56 -> 57, 1}, {56 -> 58, 2}, {56 -> 67, 2}, {56 -> 78, 2}, {56 -> 92, 2}, {57 -> 58, 2}, {57 -> 66, 2}, {57 -> 78, 2}, {57 -> 94, 2}, {58 -> 60, 2}, {58 -> 61, 2}, {58 -> 64, 2}, {58 -> 66, 2}, {58 -> 68, 2}, {58 -> 69, 2}, {58 -> 78, 2}, {59 -> 67, 2}, {60 -> 61, 2}, {60 -> 62, 2}, {60 -> 64, 2}, {60 -> 67, 2}, {60 -> 68, 2}, {61 -> 62, 2}, {61 -> 68, 2}, {62 -> 63, 2}, {62 -> 68, 2}, {63 -> 64, 2}, {63 -> 65, 2}, {63 -> 66, 2}, {63 -> 68, 2}, {63 -> 82, 2}, {64 -> 65, 1}, {64 -> 66, 2}, {64 -> 68, 2}, {64 -> 82, 1}, {64 -> 86, 2}, {65 -> 66, 2}, {65 -> 82, 2}, {65 -> 86, 2}, {65 -> 87, 2}, {66 -> 67, 2}, {66 -> 68, 2}, {66 -> 82, 2}, {66 -> 83, 2}, {66 -> 87, 2}, {68 -> 69, 2}, {69 -> 70, 2}, {69 -> 93, 2}, {70 -> 71, 2}, {71 -> 72, 2}, {72 -> 77, 2}, {72 -> 78, 2}, {72 -> 79, 2}, {73 -> 74, 2}, {73 -> 75, 2}, {73 -> 76, 2}, {73 -> 77, 2}, {73 -> 80, 2}, {73 -> 81, 2}, {73 -> 89, 2}, {74 -> 75, 2}, {74 -> 81, 2}, {74 -> 89, 2}, {75 -> 76, 2}, {75 -> 81, 2}, {76 -> 77, 2}, {76 -> 80, 2}, {76 -> 81, 2}, {76 -> 82, 2}, {77 -> 80, 2}, {78 -> 79, 2}, {80 -> 81, 2}, {80 -> 82, 2}, {80 -> 83, 2}, {82 -> 86, 2}, {85 -> 86, 2}, {86 -> 87, 2}, {87 -> 88, 2}, {89 -> 90, 2}, {89 -> 91, 2}, {89 -> 97, 2}, {90 -> 91, 2}, {90 -> 96, 2}, {90 -> 98, 2}, {93 -> 94, 2}, {94 -> 95, 1}, {95 -> 96, 1}, {97 -> 98, 2}}, EdgeLabels -> None, MultiedgeStyle -> False]Social Network
GraphPlot[{"Li" -> "Hu", "Wang" -> "Hu", "Li" -> "Wang", "Carol" -> "Andre", "Carol" -> "Fernando", "Carol" -> "Diane", "Andre" -> "Diane", "Andre" -> "Fernando", "Andre" -> "Beverly", "Fernando" -> "Diane", "Fernando" -> "Garth", "Fernando" -> "Heather", "Diane" -> "Beverly", "Diane" -> "Garth", "Diane" -> "Ed", "Beverly" -> "Garth", "Beverly" -> "Ed", "Garth" -> "Ed", "Garth" -> "Heather", "Jane" -> "Li", "Jane" -> "Hu", "Heather" -> "Ike", "Ike" -> "Jane", "Li" -> "Yang", "Yang" -> "Liu", "Liu" -> "Wang"}, VertexLabels -> Automatic, PlotRangePadding -> Automatic]Graphs from Words and Texts
This plots a network of words all starting with "din". Here, two nearest words are found for each word and linked to it with an edge:
words = DictionaryLookup["din*"];nearbys = Flatten[Map[(Thread[# -> DeleteCases[Nearest[words, #, 3], #]])&, words]];GraphPlot[nearbys, GraphLayout -> {"SpringElectricalEmbedding", "RepulsiveForcePower" -> -3}, VertexShapeFunction -> (Text[#2, #1, Background -> White]&)]This generates a graph by linking each letter in a word to all the letters that follow it in the word:
WordPlot[w_String] := GraphPlot[(x = Characters[w];
Thread[Drop[x, -1] -> Drop[x, 1]]), VertexLabels -> Automatic, DirectedEdges -> True];WordPlot["Shakespeare"]TextPlot[w_String] := GraphPlot[(x = Map[ToLowerCase, StringCases[w, WordCharacter..]];
g = Thread[Drop[x, -1] -> Drop[x, 1]];
g), DirectedEdges -> True, VertexLabels -> Automatic];TextPlot["to be or not to be, that is the question."]Torus
torus[m_, n_] :=
SparseArray[
Flatten[Table[{{i * n + j + 1, Mod[i + 1, m] * n + j + 1} ->
1, {i * n + j + 1, Mod[i - 1, m] * n + j + 1} ->
1, {i * n + j + 1, i * n + Mod[j + 1, n] + 1} ->
1, {i * n + j + 1, i * n + Mod[j - 1, n] + 1} -> 1}, {i, 0,
m - 1}, {j, 0, n - 1}]], {m * n, m * n}];GraphPlot3D[torus[40, 40], VertexShapeFunction -> None][1] Milo, R., S. Shen-Orr, S. Itzkovitz, N. Kashtan, D. Chklovskii, and U. Alon. "Network Motifs: Simple Building Blocks of Complex Networks." Science 298, no. 5594 (2002): 824–827.
[2] Alon, U. "Collection of Complex Networks." Uri Alon Homepage 2007.
[3] Milo, R., S. Itzkovitz, N. Kashtan, et al. "Superfamilies of Designed and Evolved Networks." Science 303, no. 5663 (2004): 1538–1542.
[4] Alon, U. "1AORInter." network Motifs 2007.
[5] National Institute of Standards and Technology. "DWA512: Square Dielectric Waveguide." Matrix Market 2007.
[6] Freivalds, K., U. Dogrusoz, and P. Kikusts, "Disconnected Graph Layout and the Polyomino Packing Approach." Lecture Notes in Computer Science: Revised Papers from the 9th International Symposium on Graph Drawing 2265 (2001): 378–391.
[7] Hu, Y. F. "Graph Drawing of Square Matrices from University of Florida Sparse Matrix Collection." (2007).
[8] Hu, Y. F. "Efficient, High-Quality Force-Directed Graph Drawing." The Mathematica Journal 10, no. 1 (2006): 37–71.