"Constructs an empty, anonymous " { $link subgraph } " by automatically generating a (somewhat) unique " { $slot "id" } "."
}
{ $notes
"Each " { $slot "id" } " has the form " { $snippet "\"_anonymous_n\"" } ", where " { $snippet "n" } " is a counter incremented by 1 each time an anonymous " { $slot "id" } " is generated (e.g., each time you call " { $link <anon> } " or " { $link <graph> } "). This is also how the Graphviz DOT parser internally handles anonymous graphs and subgraphs."
$nl
"Thus, while it's possible to manually create a " { $link subgraph } " whose " { $slot "id" } " conflicts with an " { $link <anon> } "'s , in practice it's unlikely to happen by accident."
}
{ $examples
"Each " { $link <anon> } " will generate a " { $link subgraph } " with a new " { $slot "id" } ", such as:"
"Constructs a cluster, which is a " { $link subgraph } " whose " { $slot "id" } " begins with the word " { $emphasis "\"cluster\"" } "."
$nl
{ $snippet "id" } " must be an object supported by the " { $link present } " word. The string " { $snippet "\"cluster_\"" } " is automatically prefixed to the " { $slot "id" } " of the resulting " { $link subgraph } "."
}
{ $notes
"Clusters are just a syntactic convention. Not all Graphviz layout engines treat clusters any differently from regular subgraphs. See the Graphviz documentation (" { $url "http://graphviz.org/Documentation.php" } ") for more information."
"If you want, you can still give the resulting " { $link graph } " a specific " { $slot "id" } " using standard words like " { $link >>id } ". For example,"
"Constructs an " { $link edge } " with the given " { $slot "tail" } " and " { $slot "head" } ", each of which must be either:"
{ $list
{ "an " { $link array } " of objects supported by the " { $link present } " word, which is treated as an anonymous " { $link subgraph } " of " { $link node } "s with corresponding " { $slot "id" } "s;" }
{ "a " { $link subgraph } "; or" }
{ "any object supported by the " { $link present } " word, which is taken to be the " { $slot "id" } " of a " { $link node } "." }
}
}
{ $notes
"There is more detailed information about how different " { $slot "tail" } " and " { $slot "head" } " types interact in the documentation for " { $link edge } "."
"If you want, you can still give the resulting " { $link graph } " a specific " { $slot "id" } " using standard words like " { $link >>id } ". For example,"
"If you want, you can still give the resulting " { $link graph } " a specific " { $slot "id" } " using standard words like " { $link >>id } ". For example,"
"In " { $emphasis "strict" } " " { $link graph } "s, there is at most one " { $link edge } " between any two " { $link node } "s, so duplicates are ignored by Graphviz."
"If you want, you can still give the resulting " { $link graph } " a specific " { $slot "id" } " using standard words like " { $link >>id } ". For example,"
"In " { $emphasis "strict" } " " { $link graph } "s, there is at most one " { $link edge } " between any two " { $link node } "s, so duplicates are ignored by Graphviz."
"Adds an arbitrary object to the " { $slot "statements" } " slot of a " { $link graph } " or " { $link subgraph } ", leaving the updated tuple on the stack. This is the most basic way to construct a " { $link graph } "."
}
{ $notes ! $warning
{ $link add } " does not check the type of " { $snippet "statement" } ". You should ensure that " { $link graph } "s and " { $link subgraph } "s only contain instances of:"
"Adds a " { $link node } " to " { $snippet "graph" } " for each element in " { $snippet "nodes" } ", which must be a " { $link sequence } " of objects that are supported by the " { $link present } " word. Thus, the following two lines are equivalent:"
"USING: accessors graphviz kernel math prettyprint sequences ;"
"<graph>"
" 100 [ \"spam\" ] replicate add-nodes"
"statements>> [ id>> \"spam\" = ] all? ."
"t"
}
}
;
HELP:add-path
{ $values
{ "graph" { $or graph subgraph } }
{ "nodes"sequence }
{ "graph'" { $or graph subgraph } }
}
{ $description
"Adds " { $link edge } "s to " { $snippet "graph" } " corresponding to a path through " { $snippet "nodes" } "."
$nl
"That is, an " { $link edge } " is added between each object and the one immediately following it in " { $snippet "nodes" } ". Thus, the following two lines are equivalent:"
{ $code
"{ A B C D E } add-path"
"A B add-edge B C add-edge C D add-edge D E add-edge"
"Represents a Graphviz edge. Each " { $link edge } " is defined by its " { $slot "tail" } " slot and its " { $slot "head" } " slot. Each slot must be either"
{ "edge from " { $slot "tail" } " node\nto each node in " { $slot "head" } }
}
{
{ { $slot "tail" } " is a " { $link subgraph } "..." }
{ "edge from each node in " { $slot "tail" } "\nto " { $slot "head" } " node" }
{ "edge from each node in " { $slot "tail" } "\nto each node in " { $slot "head" } }
}
}
$nl
"In addition, an " { $link edge } " may store local attributes in its " { $slot "attributes" } " slot (" { $instance edge-attributes } " tuple)."
}
{ $notes
"By convention, an " { $link edge } " orders its endpoints \"from\" " { $slot "tail" } " \"to\" " { $slot "head" } ", even if it belongs to an undirected " { $link graph } ", where such a distinction is generally meaningless. See the Graphviz documentation (" { $url "http://graphviz.org/Documentation.php" } "), and specifically the notes about ambiguous attributes (in " { $url "http://graphviz.org/content/attrs" } ") for more information."
} ;
HELP:graph
{ $class-description
"Represents the top-level (or " { $emphasis "root" } ") graph used in Graphviz. Its structure is modeled after the DOT language (see " { $url "http://graphviz.org/Documentation.php" } "):"
$nl
{ $table
{
{ $strong "Slot name" }
{ $strong "Value" }
{ $strong "Meaning in DOT" }
}
{
{ $slot "id" }
{ $instance string }
{ "the reference name of a graph, as in " { $strong "graph" } " " { $slot "id" } " " { $strong "{" } " ... " { $strong "}" } }
}
{
{ $slot "strict?" }
{ $instance boolean }
{ "indicates strictness, as in " { $strong "strict graph {" } " ... " { $strong "}" } }
"Represents a single Graphviz node. Each " { $link node } " is uniquely determined by an " { $slot "id" } " (" { $instance string } ") and may have per-node attributes stored in its " { $slot "attributes" } " slot (" { $instance node-attributes } " tuple)."
"Represents a logical grouping of nodes and edges within a Graphviz graph. See " { $url "http://graphviz.org/Documentation.php" } " for more information."
"To use the " { $vocab-link "graphviz" } " vocabulary, we construct Factor objects that can be converted to data understood by Graphviz (see " { $vocab-link "graphviz.dot" } ")."
"In graph theory, a " { $emphasis "bipartite graph" } " is one in which the nodes can be divided into exactly two independent sets (i.e., there are no edges between nodes in the same set)."
"In graph theory, a " { $emphasis "wheel graph" } " on " { $emphasis "n" } " nodes is composed of a single node connected to each node of a cycle of " { $emphasis "n-1" } " nodes."
"This example is adapted (and slightly altered) from " { $url "http://graphviz.org/content/datastruct" } "."
$nl
"As it shows, special label syntax is still parsed, like escape sequences (see " { $url "http://graphviz.org/content/attrs#kescString" } ") or, in this case, record syntax (see " { $url "http://graphviz.org/content/node-shapes#record" } "). However, there is no equivalent to Graphviz's headport/tailport syntax, so we set the " { $link edge } " attributes explicitly."
"Below are some examples of the typical usage of the " { $vocab-link "graphviz" } " vocabulary."
$nl
"The images in the gallery were pre-compiled using Graphviz version 2.26.3. Depending on your particular Graphviz installation, these examples may not actually work for you, especially if you have a non-standard installation."
$nl
"Also, while most of the images have a reasonable size, some of these examples may be slow to load in the UI listener."
$nl
{ $subsections
{ "graphviz""gallery""complete" }
{ "graphviz""gallery""bipartite" }
{ "graphviz""gallery""cycle" }
{ "graphviz""gallery""wheel" }
{ "graphviz""gallery""cluster" }
{ "graphviz""gallery""circles" }
{ "graphviz""gallery""fsm" }
{ "graphviz""gallery""record" }
}
;
ARTICLE: "graphviz""Graphviz"
"The " { $vocab-link "graphviz" } " vocabulary provides an interface to your existing Graphviz installation, thus allowing you to create, edit, and render Graphviz graphs using Factor. For more information about Graphviz, see " { $url "http://graphviz.org" } "."
$nl
"This vocabulary provides the basic tools to construct Factor representations of graphs. For more details, see:"
{ $subsections { "graphviz""data" } }
"Other vocabularies let you change a graph's look & feel, write cleaner code to represent it, and (of course) generate its Graphviz output:"
