GDL: Graph Attributes<>index

This section describes the whole list of graph attributes. Each attribute is listed together with its type, default value and where it can be used, i.e. in the top-level graph, a subgraph specification or in both.

Index - B - C - D - E - F - G - H - I - L - M - N - O - P - R - S - T - U - V - W - X/Y

amax: Int, default: calculated automatically
Attribute of top-level graph
Specifies the number of iterations that are animated after relayout. Specifying 0 means animation is turned off. This value can be changed interactively in the View dialog box (see general view parameters).

Top

Arrow Modes and Port Sharingarrow_mode: free or fixed, default is free as of aiSee 2.1.89
Attribute of top-level graph, subgraphs
Specifies the mode for drawing edge arrow heads. fixed is useful if port_sharing is used because then only a fixed set of rotations for the arrow heads is used, the arrow head being rotated only in increments of 45 degrees. In the free mode, each arrow head is rotated individually for each edge. If a node has many incoming edges and port_sharing is used, the free mode can lead to a somewhat confusing image. See example.

Top

attraction: Int, default is 60
Attribute of top-level graph
Applies only to the forcedir layout algorithm, where it is part of the force-directed spring embedder during attractive impulse calculation. It specifies the constant proportional to the attractive forces acting on a node. This constant and its repulsive counterpart repulsion enable the length of edges to be controlled. For example, if only attractive and repulsive forces are working on a node, an edge length of n pixels is achieved for edges with priority of 1 by specifying n2 for the attributes attraction and repulsion. Usually the values for these two attributes are of the same order of magnitude. For details, see force-directed layout.

Top

bmax: Int, default is 100
Attribute of top-level graph, subgraphs
Sets the maximum number of iterations of the phase reducing edge bends. Edge bends are used to prevent edges from being drawn across nodes. Reducing the number of iterations reduces layout calculation time, however the layout quality may suffer.

Top

border x: Int, default is 600 (pixels)
border y: Int, default is 600 (pixels)
Attributes of top-level graph
Apply only to the forcedir layout algorithm. Depending on the specification of the layout parameters for the forcedir layout algorithm it is possible for nodes to move far away from one another (especially single nodes or unconnected components). These two attributes enable a rectangle to be specified within which the graph is drawn, preventing nodes from being placed "infinitely" far from one another. For details, see force-directed layout.

Top

bordercolor: Color Entry, default: the value of textcolor
Attribute of subgraphs
Specifies the color for borders of summary nodes, boxes and frames of clusters. For available Color Entries, see section Colors.

Top

borderstyle: Style, default is solid
Attribute of subgraphs
Specifies the line style used for drawing the borders of a summary node. The following Styles are available:

  • solid
  • continuous
  • dashed
  • dotted
  • double
  • triple

Top

borderwidth: Int, default is 2
Attribute of subgraphs
Specifies the thickness of the border of a summary node in pixels.

Top

classname Int:"String", default is 1:"1", 2:"2", 3:"3", etc.
Attributes of top-level graph
Enable the names of edges classes to be introduced. These names appear in the Select Edge Classes dialog box.

Top

cmax: Int, default is infinite
cmin: Int, default is 0
Attributes of top-level graph, subgraphs
cmin sets the minimum number of iterations that are performed for reducing crossings using crossing weights. The normal method stops when two consecutive checks no longer cause the number of crossings to be reduced. However, this number of crossings might be a local minimum, meaning the number of crossing might decrease even more after some more iterations.
cmax sets the maximum number of iterations of the crossing reduction phase. A reduction of this value causes the layout process to be speeded up. The default value is infinite meaning that the method is iterated as long as any improvement is possible.

Top

color: Color Entry,
default for top-level graph is white,
default for subgraphs is the default value of node.color
Attribute of top-level graph, subgraphs
Specifies the background color of the graph window or the background color of subgraphs respectively. This color is valid as the background color in summary nodes, boxes, clusters and as a wrapping color. For available Color Entries, see section Colors. See also textcolor and bordercolor.

Top

colorentry Int: Int Int Int, no default value defined
Attribute of top-level graph, subgraphs
Enables the default color map to be filled and changed. A color is a triple of integer values for the red, green and blue part (RGB). Each integer ranges from 0 (color part turned off) to 255 (color part turned on), e.g. 0 0 0 specifies the color black and 255 255 255 specifies the color white. For instance, colorentry 75: 70 130 180 sets map entry 75 to steel blue. This color can then be used by merely specifying 75 wherever a Color Entry is expected. More details can be found in the section Colors.

Top

crossing_optimization: yes or no, default is yes
Attribute of top-level graph, subgraphs
yes activates the crossing optimization phase, which works locally. It is a postprocessing phase after normal crossing reduction. It tries to optimize locally by exchanging pairs of nodes to reduce the number of crossings.

Top

crossing_phase2: yes or no, default is yes
Attribute of top-level graph, subgraphs
yes activates crossing reduction phase two. In this phase, nodes having equal crossing weights are permuted. Note that this is the most time-consuming phase of crossing reduction.

Top

crossing_weight: Weight, default is bary
Attribute of top-level graph, subgraphs
Specifies the weight to be used for crossing reduction. The following Weights are available:

  • bary
    The barycenter is used for calculating the weights during crossing reduction. This is the fastest method for graphs with nodes whose average degree is very large.
  • median
    The median center is used for calculating the weights during crossing reduction.
  • barymedian
    These weights are the combination of barycenter and mediancenter weights, with barycenter having priority and mediancenter only being used for nodes whose barycenter weights are equal.
  • medianbary
    These weights are the combination of barycenter and mediancenter weights, with the mediancenter having priority.

There is no general recommendation as to which is the best method. A guideline might be to use bary if the degree of the nodes is large and median or one of the hybrid methods barymedian or medianbary if the degree is small, the degree of a node being the total of incoming and outgoing edges at a node. See crossing reduction for details.

Top

dirty_edge_labels: yes or no, default is no
Attribute of top-level graph, subgraphs
yes forces a fast layout of edge labels, which may result in overlapping of labels. Dirty edge labels cannot be used if splines are used for edge drawing.

Top

display_edge_labels: yes or no, default is no
Attribute of top-level graph, subgraphs
Enables/disables displaying edge labels.

Top

edges: yes or no, default is yes
Attribute of top-level graph
no suppresses the drawing of edges in the top-level graph and in all nested subgraphs.

Top

energetic: yes or no, default is no
Attribute of top-level graph
Applies only to the forcedir layout algorithm. Apart from the forces of the spring embedder and the magnetic fields, the local energy level of a node can be taken into account in deciding whether the node should move or not. Setting this attribute to yes causes the local energy present at a node to be considered during layout. For details, see force-directed layout.

Top

energetic attraction: Float, default is 70.0
energetic repulsion: Float, default is 70.0
energetic gravity: Float, default is 0.3
energetic crossing: Float, default is 80.0
energetic overlapping: Float, default is 80.0
energetic border: Float, default is 70.0
Attributes of top-level graph
Apply only to the forcedir layout algorithm. Apart from the forces of the spring embedder and the magnetic fields, the local energy level of a node can be taken into account in deciding whether the node should move or not. The behavior of the local energy present at a node can be influenced via these attributes as follows:

  • energetic attraction is the weight of the attractive energy of edges
  • energetic repulsion is the weight of the repulsive energy between nodes
  • energetic gravity is the weight of the gravitational energy of a node
  • energetic crossing is the weight of the global energy of an edge crossing
  • energetic overlapping is the weight of the global energy of a node overlapping
  • energetic border is the weight of the border energy of a node

Gauging layout quality can be done as follows: The better the layout, the lower the total of all the energy values mentioned above. For details, see force-directed layout.

Top

equal_y_dist: yes or no, default is no
Attribute of top-level graph, subgraphs
If this attribute is enabled (yes), then the vertical distance in a hierarchical layout is equal among all levels.

Top

fast_icons: yes or no, default is no
Attribute of top-level graph
yes causes icon file loading to be faster, which may negatively impact the quality of the drawing if not all the icon colors are present. For details on pictures in nodes, see icons.

Top

fdmax: Int, default is 300
Attribute of top-level graph
Applies only to the forcedir layout algorithm, where it is used in the simulated annealing part of the algorithm for specifying the upper hard limit for the number of iterations performed. The algorithm stops when the global temperature drops below a threshold value or when the limit specified here is reached. For details, see force-directed layout.

Top

finetuning: yes or no, default is yes
Attribute of top-level graph, subgraphs
no switches off the fine-tuning phase of the graph layout algorithm. The fine-tuning phase tries to give all edges the same length. It tries to improve the ranks of nodes in order to avoid very long edges (see assignment of ranks).

Top

focus, no value, no default value
Attribute of subgraphs
Sets the focus for the summary node of a subgraph, i.e. if the status of the subgraph is folded at startup, then the summary node of the subgraph for which focus was specified is centered in the graph window. The focus can also be specified for nodes. It goes without saying that the focus attribute should appear only once in a graph specification.

Top

fontname: "Font File",
default is the default vector font drawn by turtle graphics routines
Attribute of summary nodes,
Attribute of subgraphs as of aiSee 2.1.96
Specifies a pixel font different from the default vector font and used for drawing the text labels of summary nodes. This font is given by the name of the aiSee Font File containing the font description, e.g. a 12-point Helvetica Bold font can be specified via fontname: "helvB12". Note: If the font file is not in the current directory the environment variable AISEEFONTS has to be set to the directory containing the font description files. For the available Font Files, see additional fonts.

Top

fstraight_phase: yes or no, default is no
Attribute of top-level graph
yes forces straight edges that are not anchored at the same position on the border of the nodes. This is useful only if no port sharing is selected, because bends are avoided by correcting the port position.

Top

gravity: Float, default is 0.0625
Attribute of top-level graph
Applies only to the forcedir layout algorithm, where it is used for impulse calculation. Only using a simulation of a spring embedder would force unconnected components of a graph to move further and further apart from one another, as there would be no attractive forces acting between them. That is why gravity is introduced as a counterforce. gravity: Float specifies the constant which is proportional to the gravitational force acting on a node. This constant controls the strength of the gravitational force, e.g. a value of zero cancels out the influence of any gravitational force. For details, see force-directed layout.

Top

height: Int
default for top-level graph is (height of root screen - 100) pixels,
default for subgraphs is (height of the label for summary nodes) pixels
Attribute of top-level graph, subgraphs
In the top-level graph, this attribute specifies the height of the display window in pixels. For a subgraph, it specifies the height of the summary node. See also width.

Top

hidden: Int, no default value
Attribute of top-level graph
Specifies the edge class to be hidden. To hide more than one edge class, repeat this attribute for each additional edge class. Edges in such a class are ignored during layout calculation and are not drawn. Nodes that are only accessible (forward or backward) via edges of a hidden class are not drawn, either. However, nodes that are not accessible at all may be drawn (see ignore_singles).

Note the difference between hiding edge classes and the edge line style invisible. Hidden edges do not exist in the layout. Edges with the invisible line style do exist, i.e. they influence the layout, meaning they need space and may produce crossings, for example.

Top

horizontal_order: Int, default is -1 (i.e. no default)
Attribute of subgraphs
In a hierarchical layout, this attribute specifies the horizontal position of the summary node within a level (see vertical_order). Nodes specified by horizontal positions are ordered according to these positions within levels. Nodes without this attribute are inserted into this ordering by the crossing reduction mechanism (see crossing reduction).

Note: Connected components are handled separately during crossing reduction, thus it is not possible to intermix nodes of different connected components in one ordering sequence. For example, one connected component consists of nodes A, B, C and another of nodes D, E, all nodes being positioned at the same level. Then, for instance, it is not possible to specify the following horizontal order at level 0: A, D, C, E.
Note further: If the algorithm for downward laid-out trees is used the specified horizontal order is retained only within nodes that are children of the same node, i.e. in case of downward laid-out trees it is not possible to specify a horizontal order for the entire level.

Top

iconcolors: Int, default is 32
Attribute of top-level graph
Specifies the size of the color map used for colors in bitmap files. For details on pictures in nodes, see icons.

Top

iconfile: "File",
no default value
Attribute of subgraphs
Specifies the bitmap file (in PNG, PBM, PPM, or raw PPM format) to be displayed in the summary node of the folded subgraph. If the bitmap file to be displayed is not in the current directory the environment variable AISEEICONS can be set to the directory containing the bitmap file. For details, see icons.

Top

icons: yes or no, default is yes
Attribute of top-level graph
no disables displaying of icons in nodes. Displaying of icons can be enabled again interactively from the View dialog box.

Top

ignore_singles: yes or no, default is no
Attribute of top-level graph, subgraphs
yes hides all nodes of the graph which would appear singly and unconnected. These nodes have no edges at all and drawing them sometimes results in an ugly layout of the remaining graph. The default setting is to show all nodes.

Top

importance: Int, default is 0 (which actually means infinity)
Attribute of subgraphs
This is the central attribute when it comes to filtering in fish-eye views as it enables the importance of a summary node of a folded subgraph to be specified via an integer. Low integers signify less important nodes which are filtered out first by a filtering fish-eye view. High integer numbers signify nodes that are important, their being rarely filtered out. A value of 0 represents an infinite importance, the result being that these nodes are never filtered out. This attribute exists for nodes too, see importance. See also view.

Top

info1: "String", default is "" (empty string)
info2: "String", default is "" (empty string)
info3: "String", default is "" (empty string)
Attributes of subgraphs
Enable three additional text fields to be specified for a subgraph (and its summary node). The same set of attributes exists for nodes (see node attribute info1). These additional information fields can be selected interactively from the Information submenu.

Top

infoname1: "String", default is "0"
infoname2: "String", default is "1"
infoname3: "String", default is "2"
Attributes of top-level graph
Enable names for the additional information fields available for each node to be introduced. These names appear in the submenu of the menu item Information in the main menu. See also info1, info2, info3.

Top

inport_sharing: yes or no, default is no
Attribute of top-level graph, subgraphs as of aiSee 2.1.89
See port_sharing.

Top

invisible: Int, no default value
Attribute of top-level graph
This is a synonym for hidden.

Top

label: "String", default is "" (empty string)
Attribute of subgraphs
Specifies the text to be displayed inside the summary node of a folded subgraph. This text may contain control characters, e.g. \n (newline character), that influence the size of the node. See character set for more details. If no label is specified for a subgraph, the value of the title of the subgraph is used.

Top

late_edge_labels: yes or no, default is no
Attribute of top-level graph, subgraphs
Controls the moment when edge labels are drawn. yes: The graph is first partitioned and then edge labels are introduced. no: The algorithm first creates labels and then partitions the graph. The latter option yields a more compact layout, but may result in more crossings.

Top

layout_algorithm: Algorithm, default is normal
Attribute of top-level graph, subgraphs
Specifies the basic layout Algorithm, there being two main categories. The first fourteen algorithms describe variations of a hierarchical layout, whereas the last algorithm implements a force-directed layout. The variations differ in the way nodes are selected for the various levels in the hierarchical layout.

  • normal
    Tries to give all edges the same orientation and is based on the calculation of strongly connected components. The algorithms based on depth first search are faster.
  • dfs
    Uses a depth first search for layout calculation, but does not enforce additional constraints pertaining to the degree of nodes. It is faster than the normal layout algorithm. The resulting layout is heavily dependent on the initial order of the nodes in the graph specification.
  • maxdepth / mindepth
    These two algorithms are based on depth first search and are both fast heuristics. maxdepth tries to increase the depth of the layout, mindepth tries to increase the width of the layout.
  • maxdepthslow / mindepthslow
    These slower algorithms might be better if the fast heuristics offered by the algorithms maxdepth and maxdepth do not provide satisfying results. maxdepthslow tries to increase the depth of the layout and mindepthslow the width of the layout.
  • maxindegree / minindegree
    Schedule nodes with a maximum/minimum of incoming edges first, i.e. these nodes are positioned early.
  • maxoutdegree / minoutdegree
    Schedule nodes with a maximum/minimum of outgoing edges first, i.e. these nodes are positioned early.
  • maxdegree / mindegree
    Schedule nodes with a maximum/minimum of the sum of incoming and outgoing edges first, i.e. these nodes are positioned early.
  • minbackward
    Instead of calculating strongly connected components, this algorithm performs a topological sorting to assign ranks to the nodes. It is fast if the graph is acyclic.
  • tree
    A specialized method for downward laid out trees, see rank assignment. It is very fast for trees and tree-like graphs and results in a balanced layout.
  • forcedir
    Computes a force-directed placement. It is a nonhierarchical layout method that works well for undirected graphs. Some features of hierarchical layout like near edges are disabled for this layout algorithm. The layout calculation for this algorithm can be controlled via the following graph attributes:

Top

layout_downfactor: Int, default is 1
layout_nearfactor: Int, default is 1
layout_upfactor: Int, default is 1
Attributes of top-level graph, subgraphs
Have no effect if the layout_algorithm tree or forcedir is used.

A layout algorithm partitions the set of edges into edges pointing upward, edges pointing downward, and edges pointing sidewards. The last type of edges is also called near edges. If the layout_downfactor is large as compared to the layout_upfactor and layout_nearfactor, then the positions of the nodes are mainly determined by the edges pointing downwards.

If the layout_upfactor is large as compared to the layout_downfactor and layout_nearfactor, then the positions of the nodes are mainly determined by the edges pointing upwards.

If the layout_nearfactor is large, then the positions of the nodes are mainly determined by the edges pointing sidewards.

Top

level: maxlevel or Int, default is -1 (i.e. no default)
Attribute of subgraphs
This is a synonym for vertical_order.

Top

linear_segments: yes or no, default is no
Attribute of top-level graph, subgraphs
yes switches linear segment layout on. This layout favors straight long vertical edges. See also the command line options -linseg and -linsegmax.

Top

loc: { x: Int y: Int },
default is {x:0 y:0} for top-level graph, unspecified for subgraphs
Attribute of top-level graph, subgraphs
See x / y.

Top

magnetic_field1: Orientation, default is no
magnetic_field2: Orientation, default is no
Attributes of top-level graph
Apply only to the forcedir layout algorithm, where magnetic fields are part of impulse calculation. Forces that originate from a simulation of a spring embedder neglect the directions of edges. In directed graphs edges should point in a uniform direction, consequently magnetic forces are introduced, with edges being interpreted as magnetic needles that align according to the Orientation of a magnetic field. The following Orientations are available:

  • top_to_bottom
  • bottom_to_top
  • left_to_right
  • right_to_left
  • polar
  • circular
  • polcircular
  • orthogonal
  • no

Two independent magnetic fields are possible. If two fields are specified, the edges are influenced by both. The attributes magnetic_force1 / magnetic_force2 influence the strength of each field. For details, see force-directed layout.

Top

magnetic_force1: Int, default is 1
magnetic_force2: Int, default is 1
Attributes of top-level graph
Apply only to the forcedir layout algorithm, where these attributes specify the constant factors that are multiplied by the corresponding magnetic forces of the two magnetic fields magnetic_field1 / magnetic_field2. For details, see force-directed layout.

Top

manhattan_edges: yes or no, default is no
Attribute of top-level graph, subgraphs
yes switches Manhattan Layout on, meaning that all edges consist of horizontal or vertical line segments. Vertical edge segments might be shared by several edges, while horizontal edge segments are never shared. This results in aesthetic layouts for flowcharts. If orthogonal layout is used, the priority_phase and straight_phase are also used by default.

Top

margin: Int, as of aiSee 2.1.89,
default is 3 if the value of borderwidth is > 0,
0 if the value of borderwidth is 0
Attribute of subgraphs
Specifies the horizontal and vertical offset between the border of a summary node and its label in pixels. Useful for rectangular nodes only.

Top

near_edges: yes or no, default is yes
Attribute of top-level graph, subgraphs
no: All near edges are treated as normal edges in the graph layout.

Top

nodes: yes or no, default is yes
Attribute of top-level graph
no suppresses the drawing of nodes in the top-level graph and all nested subgraphs.

Top

node_alignment: top, or center, or bottom, default is center
Attribute of top-level graph, subgraphs
For hierarchical layout, this attribute specifies the vertical alignment of nodes at the horizontal reference line of levels. top means the tops of all nodes of a level have the same y coordinate. center means all nodes of a level are centered. bottom means the bottoms of all nodes of a level have the same y coordinate.

Top

orientation: Orientation, default is top_to_bottom
Attribute of top-level graph, subgraphs
Specifies the Orientation of the graph, the available Orientations being top_to_bottom, bottom_to_top, left_to_right, and right_to_left. All explanations in this section are given in relation to the default orientation.

Top

outport_sharing: yes or no, default is no
Attribute of top-level graph, subgraphs as of aiSee 2.1.89
See port_sharing.

Top

pmax: Int, default is 100
pmin: Int, default is 0
Attributes of top-level graph, subgraphs
Set the maximum/minimum number of iterations of the pendulum method. Like crossing reduction, this method stops when the "imbalance weight" stops decreasing. However, an increase in imbalance weight might be a local phenomenon, meaning that the imbalance might decrease much more after a few more iterations. Reducing the number of iterations increases layout calculation speed.

Top

port_sharing: yes or no, default is no as of aiSee 2.1.89
inport_sharing: yes or no, default is no
outport_sharing: yes or no, default is no
Attributes of top-level graph, subgraphs as of aiSee 2.1.89
no suppresses the sharing of ports by edges at nodes. inport_sharing enables the port sharing of incoming edges only, with outport_sharing enabling the port sharing of outgoing edges only.

Arrow Modes and Port SharingGenerally speaking, if multiple edges are adjacent to the same node, and the arrow heads of all these edges have the same appearance (color, size, etc.), these edges may share a port at a node. This means that only one arrow head is drawn, and all edges meet at this arrow head. This enables many edges to be located adjacent to one node without getting confused by too many arrow heads. If no port sharing is used, each edge gets its own port. See example.

Top

priority_phase: yes or no, default is no
Attribute of top-level graph, subgraphs
yes switches on the priority phase. This phase replaces the normal pendulum method with a specialized method: It forces long vertical edges to be straight, just like the straight_phase. In fact, the straight phase is a fine-tuning of the priority phase, the priority phase being recommended for an orthogonal layout (see manhattan_edges).

Top

randomfactor: Int, default is 70
Attribute of top-level graph
Applies only to the forcedir layout algorithm. If the number of randomrounds has been specified, then a node is placed with a probability of randomfactor percent during a round. This factor should be close to 100 in order to prevent the process from stopping too early. For details, see force-directed layout.

Top

randomimpulse: Int, default is 32
Attribute of top-level graph
Applies only to the forcedir layout algorithm, where it specifies the strength of the random impulse vector. If the forcedir algorithm should behave like a simulated annealing algorithm, this constant should be large and a slow tempscheme should be chosen. Otherwise a small value is preferable for the randomimpulse attribute. For details, see force-directed layout.

Top

randomrounds: Int, default is -1
Attribute of top-level graph
Applies only to the forcedir layout algorithm, where it specifies the number of randomized rounds during impulse calculation. It should only be used for the first few rounds so as to add a randomimpulse. Afterwards, the random impulse would delay completion of calculation. For details, see force-directed layout.

Top

repulsion: Int, default is 60
Attribute of top-level graph
Applies only to the forcedir layout algorithm, where it is part of the force-directed spring embedder during repulsive impulse calculation. It specifies the constant that is inversely proportional to the attractive forces acting on a node. This constant and its attractive counterpart attraction enable the length of edges to be controlled. Usually the values for these two attributes are of the same order of magnitude. For details, see force-directed layout.

Top

rmax: Int, default is 100
rmin: Int, default is 0
Attributes of top-level graph, subgraphs
Set the maximum/minimum number of iterations for rubberbanding. This works in a manner similar to the pendulum method. Reducing the number of iterations increases layout calculation speed.

Top

scaling: maxspect or Float, default is 1.0
Attribute of top-level graph, subgraphs
Specifies the scaling factor for graph representation. A scaling factor of 1.0 means normal size. maxspect scales a graph so that the entire graph fits into the graph window.

When specified for the top-level graph, this attribute determines the size of the entire graph including all the subgraphs. When specified for a subgraph it determines the scaling factor of the summary node of the folded subgraph. The size of a boxed subgraph is not affected, however the size of the subgraph nodes may still be affected (see node attribute shrink).

See also graph attributes shrink / stretch.

Top

shape: Shape, default is box
Attribute of subgraphs
Specifies the Shape of the summary node of a folded subgraph. The following Shapes are available

  • box
  • triangle
  • circle
  • ellipse
  • rhomb
  • hexagon
  • trapeze
  • uptrapeze
  • lparallelogram
  • rparallelogram

Top

shrink: Int, default is 1
stretch: Int, default is 1
Attributes of top-level graph, subgraphs
Specify the shrinking and stretching factors for the scaling. The scaling of the graph as a percentage is given by the formula ((stretch / shrink) * 100).

For instance, (stretch, shrink) = (1,1), or (2,2), etc., is normal size, (1,2) is half size, and (2,1) is double size.

When these attributes are specified for the top-level graph, they determine the size of the entire graph including all the subgraphs. When specified for a subgraph they determine the scaling factor of the summary node of the folded subgraph. The size of a boxed subgraph is not affected, however the size of the subgraph nodes may still be affected (see node attribute shrink).

Top

smanhattan_edges: yes or no, default is no
Attribute of top-level graph, subgraphs
yes selects a specialized orthogonal layout: All horizontal edge segments between two levels share the same horizontal line, i.e. not only vertical edge segments are shared (as in the Manhattan Layout). However, horizontal edge segments are shared by several edges, too. This looks nice for trees but might be confusing in general.

Top

smax: Int, default is 100
Attribute of top-level graph, subgraphs
Sets the maximum number of iterations of the straight-line recognition phase. This value is not of any use unless the straight-line recognition phase is switched on, see straight_phase. It can be used to improve the Manhattan Layout or the layout with the priority phase turned on.

Top

splinefactor: Int, default is 70
Attribute of top-level graph
Determines the bending of splines. A factor of 100 indicates very sharp bending, a factor of 1 indicating very flat bending. Useful values range from 30 to 80.

Top

splines: yes or no, default is no
Attribute of top-level graph
Specifies whether splines are used to draw edges. Polygon segments are used to draw edges by default, because this is much faster. Note that the spline drawing routine is very slow. Splines are mainly used to prepare high-quality PostScript or SVG output for small graphs. See also splinefactor.

Top

spreadlevel: Int, default is 1
Attribute of top-level graph, subgraphs
Influences only the tree algorithm. Spreading of the uppermost nodes of large balanced trees would increase the width of the tree to such an extent that the tree would no longer fit in a window. Consequently, the spread level specifies the minimum level (rank) where nodes are spread. Nodes of levels above the spread level are not spread.

Top

state: State, default is unfolded
Attribute of subgraphs
Specifies the initial state of a subgraph, i.e. the way a subgraph is displayed the first time a graph is visualized. The appearance of the subgraph (its state) can be changed interactively. The following States are available:

  • folded
    The nodes of a subgraph are hidden. They are represented by a single node, called a summary node. For details, see folding.
  • boxed
    The subgraph is surrounded by a frame, i.e. drawn in a box. The nodes inside the box are independent of the rest of the graph, i.e. there are no edges connecting nodes outside the box with nodes inside the box and vice versa. See also box operations.
  • clustered
    The subgraph is surrounded by a frame. In contrast to a box, edges from nodes outside the frame are drawn to nodes inside the frame and vice versa. This is an experimental feature. See also cluster operations.
  • unfolded
    This is the default setting.
  • wrapped
    All nodes and edges belonging to the subgraph are wrapped using the same color. See also wrapping operations.
  • exclusive
    The subgraph is shown exclusively. All other nodes of the graph are not visible. Only edges between nodes of a group are visible. Of course, this value should appear only once in a graph specification. See also displaying node groups exclusively.

Top

straight_phase: yes or no, default is no
Attribute of top-level graph, subgraphs
yes switches on the straight phase. This is an additional phase that tries to avoid bends in long edges. Long edges are drawn as long straight vertical lines. Thus, this phase is not very appropriate for normal layout, however it is recommended when an orthogonal layout is selected (see manhattan_edges).

Top

stretch: Int, default is 1
Attribute of top-level graph, subgraphs
See shrink / stretch.

Top

subgraph_labels: yes or no, default is yes
Attribute of top-level graph
no switches off the displaying of subgraph labels. This can also be done interactively from the View dialog box.

Top

tempfactor: Float, default is 1.3
Attribute of top-level graph
Applies only to the forcedir layout algorithm, where it is used in the simulated annealing part of the algorithm. It specifies the temperature scheme factor used for exponential and reverse exponential temperature schemes. For details, see force-directed layout.

Top

tempmax: Int, default is 128
tempmin: Int, default is 1
Attributes of top-level graph
Apply only to the forcedir layout algorithm, where they are used in the simulated annealing part of the algorithm. They specify the upper/lower limit of the temperature range. For details, see force-directed layout.

Top

tempscheme: Scheme, default is 1
Attribute of top-level graph
Applies only to the forcedir layout algorithm. There are local and global temperature schemes. In global temperature schemes all nodes have the same temperature. The following Schemes are available:

  • 1, local temperature temp_speed, local adaptive temperature scheme with speedup during cooling
  • 2, local temperature temp_normal, local adaptive temperature scheme with no speedup
  • 3, global temperature temp_linear, linear curve
  • 4, global temperature temp_hyperbolical, hyperbolic curve: very fast descent, then a low temperature for an extended period of time
  • 5, global temperature temp_exponential, exponential descending temperature, i.e. a small temperature for an extended period of time
  • 6, global temperature temp_logarithmic, logarithmic descending, i.e. a small temperature for an extended period of time.
  • 7, global temperature temp_reverse_exponential, reverse exponential descending temperature, i.e. a high temperature for an extended period of time.
  • 8, global temperature temp_reverse_logarithmic, reverse logarithmic descent, i.e. a high temperature for an extended period of time.

For details, see force-directed layout.

Top

temptreshold: Int, default is 3
Attribute of top-level graph
Applies only to the forcedir layout algorithm, where it is used in the simulated annealing part of the algorithm. It specifies the threshold value for the global temperature. The algorithm stops if the global temperature drops below the value specified here. For details, see force-directed layout.

Top

textcolor: Color Entry, default is black
Attribute of subgraphs
Specifies the color for text labels of summary nodes. If no bordercolor is specified, then this attribute also specifies the color for borders of summary nodes, boxes and frames of clusters. For the available Color Entries, see section Colors. See also graph attribute color.

Top

textmode: center, or left_justify, or right_justify, default is center
Attribute of subgraphs
Specifies the alignment of text within a summary node frame.

Top

title: "String", default is "name of the graph specification file"
Attribute of subgraphs
Specifies the name associated with the subgraph. If no title is specified the name of the file containing the graph specification is used. Since titles have to be unique throughout a graph specification, there can be only one subgraph at most without a title specification.

The name of a subgraph is used to identify it, so that the subgraph can be the source and target of an edge specification. These edges start or end at the summary nodes of folded subgraphs. If the subgraph is visualized unfolded, these edges start or end at the root of the subgraph or at the root of the first subgraph in the subgraph.

Top

treefactor: Float, default is 0.5
Attribute of top-level graph, subgraphs
The tree algorithm for downward laid-out trees tries to produce a medium dense balanced tree-like layout. If the tree factor is greater than 0.5 the tree edges are spread, i.e. they have a larger gradient, thus possibly improving the readability of the tree. It is not obvious whether spreading results in a denser or wider layout. A tree factor exists for each tree, enabling maximum density of the entire tree.

Top

useractioncmd1: "String", default is "" (empty string)
useractioncmd2: "String", default is "" (empty string)
useractioncmd3: "String", default is "" (empty string)
useractioncmd4: "String", default is "" (empty string)
Attributes of top-level graph
Enable four commands to be specified in a graph specification, that are executed when one of the User Action menu entries is invoked by the user. The commands differ in the way that command line arguments are supplied to the command. For details and downloadable demos, see User Actions and DDE. See also useractionname.

Top

useractionname1: "String", default is "User Action 1"
useractionname2: "String", default is "User Action 2"
useractionname3: "String", default is "User Action 3"
useractionname4: "String", default is "User Action 4"
Attributes of top-level graph
Enable names for the User Actions menu entries to be introduced. These names appear in the submenu of the menu item Auxiliaries in the main menu. For details and downloadable demos, see User Actions and DDE. See also useractioncmd.

Top

vertical_order: maxlevel or Int, default is -1 (i.e. no default)
Attribute of subgraphs
In a hierarchical layout, this attribute specifies the vertical position of a summary node of a folded subgraph. maxlevel tries to position the node at the maximum calculated level. Generally for all nodes, their vertical position is called their level or rank (see rank assignment). A synonym for vertical_order is level.

All nodes of level 0 form the uppermost layer (if the orientation is top_to_bottom). Nodes of level 1 form the second layer, etc. The level specification is not in effect unless automatic layout is being calculated. Layout is calculated automatically if there is at least one node without a specified location (see loc attribute for nodes and loc attribute for summary nodes).

The level specification may conflict with a near edge specification, because the source and target node of a near edge have to have the same level. In this case, the level specification of the source or the target node of the near edge is ignored.

Top

view: Mode, default is normal, i.e. no fish-eye view
Attribute of top-level graph
Enables one of the six fish-eye view Modes to be selected. If a graph is large only a small amount of it is visible in the graph window because of the fixed size of the window. Scaling the graph down so that it fits into the window causes details to no longer be recognizable. The solution to this is fish-eye views. A fish-eye view is a coordinate transformation, causing the view of the graph to be distorted. There is a focus which is magnified so that all the details can be seen. Parts of the graph that are far away from the focus are scaled down. The following fish-eye view Modes are available:

  • pfish
    Self-adaptable polar fish-eye view. The plane containing the graph layout is projected onto a sphere. Polar fish-eye view is a 3D look onto this sphere.
  • cfish
    Self-adaptable Cartesian fish-eye view. The effect of the Cartesian fish eye is similar to polar fish-eye view, only that the Cartesian coordinate system is transformed instead of the polar system. In this view horizontal and vertical lines remain horizontal and vertical after transformation, i.e. they are not bent as in the case of polar fish-eye view.
  • fpfish
    Polar fish-eye view with a fixed radius. Here the fish-eye view only shows a fixed radius around the focus, meaning the entire graph may no longer be visible.
  • fcfish
    Cartesian fish-eye view with a fixed radius. Here the fish eye only shows a fixed radius around the focus, meaning the entire graph may no longer be visible.
  • dpfish
    Polar fish-eye view with a double focus.
  • dcfish
    Cartesian fish-eye view with a double focus.

See also importance and fish-eye view modes and parameters

Top

width: Int
default for top-level graph is (width of root screen - 100) pixels,
default for subgraphs is (width of the label for summary nodes) pixels
Attribute of top-level graph, subgraphs
In the top-level graph, this attribute specifies the width of the display window in pixels. In a subgraph specification, it specifies the width of the summary node in pixels. See also height.

Top

x: Int
y: Int
default for top-level graph is 0 for both,
default for subgraphs is unspecified for both
Attributes of top-level graph, subgraphs
In the top-level graph, these attributes specify the position of the graph window in relation to the root screen, i.e. the x and y coordinates of the upper left corner of the graph window are specified in pixels. The origin of the root screen is in the upper left corner. In a subgraph specification, these attributes specify the x and y coordinates (in pixels) of the summary node in relation to the upper left corner of the graph window.

The positions can also be specified via loc.

Top

xbase: Int, default is 5
ybase: Int, default is 5
Attributes of top-level graph, subgraphs
In the top-level graph, these attributes specify the horizontal and vertical offset between the graph window and the upper left-hand corner of the graph, i.e. the position of the origin of the coordinate system in relation to the upper left-hand corner of the virtual window. In subgraph specifications, these are the offsets from the frame of the box containing the subgraph.

When exporting the graph to SVG or PS, these attributes can be used to control horizontal and vertical offsets between the image border and the actual graph. Also, these attributes come in handy for working around the lack of image background color attribute in SVG. See example.

Top

Displayed Window and Virtual Windowxmax: Int, default is (width of the root screen - 90) pixels
ymax: Int, default is (height of the root screen - 90) pixels
Attributes of top-level graph
Specify the maximum size of the virtual window used to display the graph (see picture). This is usually larger than the displayed part. The width and height of the displayed part cannot be larger than xmax and ymax. Only the parts of the graph inside the virtual window are drawn. The virtual window can be moved over the potentially infinite coordinate system by special positioning commands (see navigation).

Note: It is advisable to set xmax and ymax so they do not exceed the size of the root screen so as to get good performance.

Top

xraster: Int, default is 1
yraster: Int, default is 1
Attributes of top-level graph, subgraphs
In a hierarchical layout, these attributes specify raster distances for positioning nodes. The center of each node is aligned to this raster (grid layout).

Top

xlraster: Int, default is 1
Attribute of top-level graph, subgraphs
Specifies a horizontal raster distance for xlspace in pixels. The value of xlraster should be a divisor of the value of xraster.

Top

xlspace: Int,
default is 1/2 * xspace pixels, if polygons are used for edge drawing
default is 4/5 * yspace pixels, if splines are used for edge drawing
Attribute of top-level graph, subgraphs
Describes the horizontal distance between lines at the points where they cross levels. It is advisable to set xlspace to a larger value, if splines are used in order to prevent sharp bends.

Top

xspace: Int, default is 20
yspace: Int, default is 70
Attributes of top-level graph, subgraphs
Specify the minimum horizontal and vertical distance between nodes in pixels


HomeSitemapai
Last modified on 22 April 2004 by webmaster. © 2000-2004 AbsInt.
URL: http://www.aisee.com/manual/unix/44list.htm


Home
About
Examples
Free trial

 Support
» FAQs
» Quick Guide
» User Manual
   » Windows
   » Mac
» GDL
» Options
» Changelog
» GDLedit

Store
Legal
Contact
Extras
Sitemap