User Guide

Installation

graphviz provides a simple pure-Python interface for the Graphviz graph-drawing software. It runs under Python 3.6+. To install it with pip, run the following:

$ pip install graphviz

For a system-wide install, this typically requires administrator access. For an isolated install, you can run the same inside a venv or a virtualenv.

The only dependency is a working installation of Graphviz (download page, archived versions, installation procedure for Windows).

After installing Graphviz, make sure that its bin/ subdirectory containing the dot layout command for rendering graph descriptions is on your systems’ PATH (sometimes done by the installer; setting PATH on Linux, Mac, and Windows): On the command-line, dot -V should print the version of your Graphiz installation.

Platform: Windows

Windows users might want to check the status of known issues (gvedit.exe, sfdp, commands) and consider trying an older archived version as a workaround (e.g. graphviz-2.38.msi).

Platform: Anaconda

See the downstream conda-forge distribution conda-forge/python-graphviz (feedstock), which should automatically conda install conda-forge/graphviz (feedstock) as dependency.

Basic usage

The graphviz package provides two main classes: graphviz.Graph and graphviz.Digraph. They create graph descriptions in the DOT language for undirected and directed graphs respectively. They have the same API.

Hint

Graph and Digraph produce different DOT syntax and have different values for directed.

Create a graph by instantiating a new Graph or Digraph object:

>>> import graphviz

>>> dot = graphviz.Digraph('round-table', comment='The Round Table')  

>>> dot  
<graphviz.graphs.Digraph object at 0x...>

Their constructors allow to set the graph’s name identifier, the filename for the DOT source and the rendered graph, an optional comment for the first source code line, etc.

Add nodes and edges to the graph object using its node() and edge() or edges() methods:

>>> dot.node('A', 'King Arthur')  
>>> dot.node('B', 'Sir Bedevere the Wise')
>>> dot.node('L', 'Sir Lancelot the Brave')

>>> dot.edges(['AB', 'AL'])
>>> dot.edge('B', 'L', constraint='false')

The node() method takes a name identifier as first argument and an optional label. The edge() method takes the names of start node and end node, while edges() takes an iterable of name pairs. Keyword arguments are turned into (node and edge) attributes (see extensive Graphviz docs on available attributes).

Check the generated DOT source code:

>>> print(dot.source)  
// The Round Table
digraph "round-table" {
    A [label="King Arthur"]
    B [label="Sir Bedevere the Wise"]
    L [label="Sir Lancelot the Brave"]
    A -> B
    A -> L
    B -> L [constraint=false]
}

Use the render() method to save the DOT source code and render it with the default dot layout engine (see below for using other layout engines).

>>> doctest_mark_exe()

>>> dot.render(directory='doctest-output').replace('\\', '/')
'doctest-output/round-table.gv.pdf'

Passing view=True will automatically open the resulting (PDF, SVG, PNG, etc.) file with your system’s default viewer application for the rendered file type.

>>> doctest_mark_exe()

>>> dot.render(directory='doctest-output', view=True)  
'doctest-output/round-table.gv.pdf'
_images/round-table.svg

Formats

To use a different output file format than the default PDF, you canuse the format argument when creating your Graph or Digraph object:

>>> import graphviz

>>> g = graphviz.Graph(format='png')  

You can also change the format attribute on an existing graph object:

>>> doctest_mark_exe()

>>> dot = graphviz.Digraph('hello')
>>> dot.edge('hello', 'world')
>>> dot.format = 'svg'

>>> dot.render(directory='doctest-output').replace('\\', '/')
'doctest-output/hello.gv.svg'

Piped output

To directly access the raw results from the Graphviz dot layout command as binary bytes or as decoded str (for plain-text formats like SVG) instead of writing to a file, use the pipe() method of your Graph or Digraph object:

>>> import graphviz

>>> h = graphviz.Graph('hello', format='svg')  

>>> h.edge('Hello', 'World')
>>> doctest_mark_exe()

>>> h.pipe(format='pdf')[:4]
b'%PDF'

>>> print(h.pipe(encoding='utf-8'))  
<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<!DOCTYPE svg
...
</svg>

Tip

Because pipe() returns the raw stdout from the layout subprocess by default (bytes), you usually want to decode the return value when piping into formats like 'svg' or 'plain',

Caution

The output for pipe() is buffered in memory, so avoid this method if the data size is large.

Jupyter notebooks

Graph and Digraph objects have a _repr_mimebundle_() method so they can be rendered and displayed directly inside a Jupyter notebook. For an example, check the examples/graphviz-notebook.ipynb file in the source repository/distribution (or the same notebook in nbviewer).

This also allows direct displaying within the Jupyter Qt Console (also the one inside Spyder IDE):

_images/qtconsole.png

By default _repr_mimebundle_() uses 'svg' format. You can use the graphviz.set_jupyter_format() to override the default format that is used for displaying in IPython/Jupyter. (example, nbviewer).

Hint

You can also use display_svg(), display_png(), or .display_jpeg() from IPython.display to display the rendered Graph or Digraph as SVG, PNG or JPEG in IPython/Jupyter.

Styling

Use the graph_attr, node_attr, and edge_attr arguments of the Graph and Digraph constuctors to change the default attributes for your graph, nodes, and edges.

>>> import graphviz  

>>> ps = graphviz.Digraph('pet-shop', node_attr={'shape': 'plaintext'})

>>> ps.node('parrot')
>>> ps.node('dead')
>>> ps.edge('parrot', 'dead')

After creation, the graph_attr, node_attr, and edge_attr attributes be edited on instances:

>>> ps.graph_attr['rankdir'] = 'LR'  
>>> ps.edge_attr.update(arrowhead='vee', arrowsize='2')

>>> print(ps.source)  
digraph "pet-shop" {
    graph [rankdir=LR]
    node [shape=plaintext]
    edge [arrowhead=vee arrowsize=2]
    parrot
    dead
    parrot -> dead
}
_images/pet-shop.svg

Attributes

To directly add DOT att_stmt attribute statements, call the attr() method of the Graph or Digraph instance with the wanted target as first argument and the attributes as keyword args.

Hint

Attribute statements affect all later graphs, nodes, or edges within the same (sub-)graph.

>>> import graphviz

>>> ni = graphviz.Graph('ni')  

>>> ni.attr('node', shape='rarrow')
>>> ni.node('1', 'Ni!')
>>> ni.node('2', 'Ni!')

>>> ni.node('3', 'Ni!', shape='egg')

>>> ni.attr('node', shape='star')
>>> ni.node('4', 'Ni!')
>>> ni.node('5', 'Ni!')

If you omit the first attr() argument, the method can be used to set arbitrary attributes as key-value pairs targeting the current (sub-)graph (e.g. for rankdir, label, or setting rank='same' within a subgraph context, example):

>>> ni.attr(rankdir='LR')  

>>> ni.edges(['12', '23', '34', '45'])

>>> print(ni.source)  
graph ni {
    node [shape=rarrow]
    1 [label="Ni!"]
    2 [label="Ni!"]
    3 [label="Ni!" shape=egg]
    node [shape=star]
    4 [label="Ni!"]
    5 [label="Ni!"]
    rankdir=LR
    1 -- 2
    2 -- 3
    3 -- 4
    4 -- 5
}
_images/ni.svg

Node ports & compass

The edge() and edges() methods use the colon-separated node[:port[:compass]] format for tail and head nodes. This allows to specify an optional node port plus an optional compass point the edge should aim at for the given tail or head node (example).

Caution

As colons are used to indicate port and compass for edges, node names containing one or more literal colons : are currently not supported. GH #54

Tip

There is no such restriction for the label argument, so you can work around by choosing a colon-free name together with the wanted label as demonstrated below

>>> import graphviz  

>>> cpp = graphviz.Digraph('C++')

>>> cpp.node('A', 'std::string')
>>> cpp.node('B', '"spam"')

>>> cpp.edge('A', 'B')

>>> print(cpp.source)  
digraph "C++" {
    A [label="std::string"]
    B [label="\"spam\""]
    A -> B
}

Backslash escapes

The Graphviz layout engines support a number of escape sequences such as \n, \l, \r (for placement of multi-line labels: centered, left-justified, right-justified) and \N, \G, \L (expanded to the current node name, graph name, object label). To be able to use them from this library (e.g. for labels), backslashes in strings are (mostly) passed on as is.

Attention

This means that literal backslashes need to be escaped (doubled) by the user. As the backslash is also special in Python string literals a second level of doubling is needed. E.g. label='\\\\' for a label that is rendered as single literal backlash: \.

Tip

Doubling of backslashes can be avoided by using raw string literals (r'...') instead. This is similar to the solution proposed for the stdlib re module. See also https://en.wikipedia.org/wiki/Leaning_toothpick_syndrome.

>>> import graphviz  

>>> e = graphviz.Digraph('escapes')

>>> e.node('backslash', label=r'\\')
>>> e.node('multi_line', label=r'centered\nleft\lright\r')

>>> print(e.source)  
digraph escapes {
    backslash [label="\\"]
    multi_line [label="centered\nleft\lright\r"]
}
_images/escapes.svg

To disable any special character meaning in a string (e.g. from user input to be rendered literally), use the graphviz.escape() function (similar to the re.escape() function):

>>> bs = graphviz.Digraph('literal_backslash')  

>>> bs.node(graphviz.escape('\\'))

>>> print(bs.source)  
digraph literal_backslash {
    "\\"
}
>>> doctest_mark_exe()

>>> bs.render(format='svg', directory='doctest-output').replace('\\', '/')
'doctest-output/literal_backslash.gv.svg'
_images/literal_backslash.svg

Historical note

To prevent breaking the internal quoting mechanism, the special meaning of \" as a backslash-escaped quote has been disabled since version 0.14 of this library. E.g. both label='"' and label='\\"' now produce the same DOT source [label="\""] (a label that renders as a literal quote). See also examples/graphviz-escapes.ipynb (nbviewer).

Quoting and HTML-like labels

The graph-building methods of Graph and Digraph objects automatically take care of quoting (and escaping quotes) where needed (whitespace, keywords, double quotes, etc.):

>>> import graphviz  

>>> q = graphviz.Digraph()

>>> q.edge('spam', 'eggs eggs')
>>> q.edge('node', '"here\'s a quote"')

>>> print(q.source)  
digraph {
    spam -> "eggs eggs"
    "node" -> "\"here's a quote\""
}

If a string starts with '<' and ends with '>', it is passed on as is, i.e. without quoting/escaping: The content between the angle brackets is treated by the Graphviz layout engine as special HTML string that can be used for HTML-like labels:

>>> h = graphviz.Graph('html_table')  

>>> h.node('tab', label='''<<TABLE>
...  <TR>
...    <TD>left</TD>
...    <TD>right</TD>
...  </TR>
... </TABLE>>''')
_images/html_table.svg

For strings that should literally begin with '<' and end with '>', use the graphviz.nohtml() function to disable the special meaning of angled parenthesis and apply normal quoting/escaping:

>>> d = graphviz.Digraph('diamond', format='svg')  

>>> d.node('diamond', label=graphviz.nohtml('<>'))

>>> print(d.source)  
digraph diamond {
    diamond [label="<>"]
}
>>> doctest_mark_exe()

>>> d.render(directory='doctest-output').replace('\\', '/')
'doctest-output/diamond.gv.svg'
_images/diamond.svg

Historical note

Before version 0.8.2, the only workaround was to add leading or trailing space (label=' <>'):

Subgraphs & clusters

Graph and Digraph objects have a subgraph() method for adding a subgraph to the instance.

There are two ways to use it: Either with a ready-made instance of the same kind as the only argument (whose content is added as a subgraph) or omitting the graph argument (returning a context manager for defining the subgraph content more elegantly within a with-block).

First option, with graph as the only argument:

>>> import graphviz  

>>> p = graphviz.Graph(name='parent')
>>> p.edge('spam', 'eggs')

>>> c = graphviz.Graph(name='child', node_attr={'shape': 'box'})
>>> c.edge('foo', 'bar')

>>> p.subgraph(c)

Second usage, with a with-block (omitting the graph argument):

>>> p = graphviz.Graph('parent')  
>>> p.edge('spam', 'eggs')

>>> with p.subgraph(name='child', node_attr={'shape': 'box'}) as c:
...    c.edge('foo', 'bar')

Both produce the same result:

>>> print(p.source)  
graph parent {
    spam -- eggs
    subgraph child {
        node [shape=box]
        foo -- bar
    }
}

Tip

If the name of a subgraph begins with 'cluster' (all lowercase), the layout engine treats it as a special cluster subgraph (example ). See the Subgraphs and Clusters section in DOT language.

When subgraph() is used as a context manager, the new graph instance is created with strict=None copying the parent graph values for directory, engine, format, renderer, formatter, and encoding:

>>> doctest_mark_exe()

>>> p = graphviz.Graph('parent', directory='doctest-output')

>>> with p.subgraph(name='child') as c:
...    c.edge('bacon', 'eggs')
...    c.render().replace('\\', '/')
'doctest-output/child.gv.pdf'

Note

These copied attributes are only relevant for rendering the subgraph independently (i.e. as a stand-alone graph) from within the with-block.

Engines

To use a different layout engine than the default dot when rendering your graph, you can use the engine argument on the constructor of Graph or Digraph.

>>> import graphviz  

>>> g = graphviz.Graph(engine='neato')

You can also change the engine attribute on an existing instance:

>>> g.engine = 'circo'  

Unflatten

To prepocess the DOT source of a Graph or Digraph with the unflatten preprocessor (manpage, PDF), use the unflatten() method.

>>> import graphviz  

>>> w = graphviz.Digraph('wide')

>>> w.edges(('0', str(i)) for i in range(1, 10))
>>> doctest_mark_exe()

>>> w.view()  
_images/wide.svg

Hint

unflatten() improves the aspect ratio of graphs with many leaves or disconnected nodes.

>>> u = w.unflatten(stagger=3)  
>>> doctest_mark_exe()

>>> u.view()  
_images/wide-unflatten-stagger-3.svg

The method returns a Source object that you can render(), view(), etc. with the same basic API as Graph or Digraph objects (minus modification, see details below).

>>> u = w.unflatten(stagger=2)  
>>> u  
<graphviz.sources.Source object at 0x...>
>>> doctest_mark_exe()

>>> u.view()  
_images/wide-unflatten-stagger-2.svg

Custom DOT statements

To add arbitrary statements to the created DOT source, you can use the body attribute of Graph and Digraph objects. It holds the verbatim list of (str) lines to be written to the source file (including their final newline). Use its append() or extend() method:

>>> import graphviz

>>> rt = graphviz.Digraph(comment='The Round Table')  

>>> rt.body.append('\t"King Arthur" -> {\n\t\t"Sir Bedevere", "Sir Lancelot"\n\t}\n')
>>> rt.edge('Sir Bedevere', 'Sir Lancelot', constraint='false')

>>> print(rt.source)  
// The Round Table
digraph {
    "King Arthur" -> {
        "Sir Bedevere", "Sir Lancelot"
    }
    "Sir Bedevere" -> "Sir Lancelot" [constraint=false]
}

Attention

Note that you might need to correctly quote/escape identifiers and strings containing whitespace or other special characters when using this method.

Using raw DOT

To render a ready-made DOT source code string (instead of assembling one with the higher-level interface of Graph or Digraph), create a graphviz.Source object holding your DOT string:

>>> import graphviz  

>>> src = graphviz.Source('digraph "the holy hand grenade" { rankdir=LR; 1 -> 2 -> 3 -> lob }')

>>> src  
<graphviz.sources.Source object at 0x...>

Use the render() method to save and render it:

>>> doctest_mark_exe()

>>> src.render('doctest-output/holy-grenade.gv').replace('\\', '/')
'doctest-output/holy-grenade.gv.pdf'
>>> doctest_mark_exe()

>>> src.render('doctest-output/holy-grenade.gv', view=True).replace('\\', '/')  
'doctest-output/holy-grenade.gv.pdf'
_images/holy-grenade.svg

Hint

Apart from lacking editing methods, Source objects have the same basic API as the higher-level Graph and Digraph objects (e.g. save(), render(), view(), pipe() methods, engine and format attributes, Jupyter notebook _repr_mimebundle_(), etc. See API docs).

Existing files

To directly render an existing DOT source file (e.g. created with other tools), you can use the graphviz.render() function.

>>> doctest_mark_exe()

>>> import pathlib
>>> import graphviz

>>> src = 'digraph "the holy hand grenade" { rankdir=LR; 1 -> 2 -> 3 -> lob }'
>>> filepath = pathlib.Path('doctest-output/the-holy-hand-grenade.gv')
>>> filepath.write_text(src, encoding='ascii')
66

>>> graphviz.render('dot', 'png', filepath).replace('\\', '/')
'doctest-output/the-holy-hand-grenade.gv.png'

To directly display the rendered visualization of an existing DOT source file inside a Jupyter notebook or Qt Console, you can use graphviz.Source.from_file() (alternative constructor):

_images/qtconsole-source.png

Note that render() and view() on Source instances returned by graphviz.Source.from_file() skip writing the loaded file back. The same holds for save(). The instances resolve default .save(skip_existing=None) to .save(skip_existing_run=True) to skip writing the read source back into the same file (specifically the same path that it was loaded from). Call .save(skip_existing=False) if you want to re-write the loaded source.

Historical note

Before version 0.18 of this library, Source.save(), Source.render(), and Source.view(), wrote the content read into source back into the file. It was advised to use graphviz.render() and graphviz.view() to directly work on files if the superflous saving needed to be avoided.

Integration with viewers

On platforms such as Windows, viewer programs opened by render() with view=True (or eqivalently with the view() shortcut-method) might lock the (PDF, PNG, etc.) file for as long as the viewer is open (blocking re-rendering it with a Permission denied error).

Tip

You can use the tempfile.mktemp() function from the stdlib tempfile module to render to a different file for each invocation. This avoids needing to close the viewer window each time within such an incremental workflow (and also serves to preserves the intermediate steps).

>>> import tempfile  
>>> import graphviz

>>> g = graphviz.Graph()

>>> g.node('spam')
>>> doctest_mark_exe()

>>> g.view(tempfile.mktemp('.gv'))  
'C:\\Users\\User\\AppData\\Local\\Temp\\tmp3aoie8d0.gv.pdf'

>>> g.view(tempfile.mktemp('.gv'))  
'C:\\Users\\User\\AppData\\Local\\Temp\\tmphh4ig7a_.gv.pdf'

Other options: