1[](https://www.travis-ci.com/pydot/pydot) 2[](https://pypi.org/project/pydot/) 3 4 5About 6===== 7 8`pydot`: 9 10 - is an interface to [Graphviz][1] 11 - can parse and dump into the [DOT language][2] used by GraphViz, 12 - is written in pure Python, 13 14and [`networkx`][3] can convert its graphs to `pydot`. 15 16Development occurs at [GitHub][11], where you can report issues and 17contribute code. 18 19 20Examples 21======== 22 23The examples here will show you the most common input, editing and 24output methods. 25 26Input 27----- 28 29No matter what you want to do with `pydot`, it will need some input to 30start with. Here are 3 common options: 31 321. Import a graph from an existing DOT-file. 33 34 Use this method if you already have a DOT-file describing a graph, 35 for example as output of another program. Let's say you already 36 have this `example.dot` (based on an [example from Wikipedia][12]): 37 38 ```dot 39 graph my_graph { 40 bgcolor="yellow"; 41 a [label="Foo"]; 42 b [shape=circle]; 43 a -- b -- c [color=blue]; 44 } 45 ``` 46 47 Just read the graph from the DOT-file: 48 49 ```python 50 import pydot 51 52 graphs = pydot.graph_from_dot_file('example.dot') 53 graph = graphs[0] 54 ``` 55 562. or: Parse a graph from an existing DOT-string. 57 58 Use this method if you already have a DOT-string describing a 59 graph in a Python variable: 60 61 ```python 62 import pydot 63 64 dot_string = """graph my_graph { 65 bgcolor="yellow"; 66 a [label="Foo"]; 67 b [shape=circle]; 68 a -- b -- c [color=blue]; 69 }""" 70 71 graphs = pydot.graph_from_dot_data(dot_string) 72 graph = graphs[0] 73 ``` 74 753. or: Create a graph from scratch using pydot objects. 76 77 Now this is where the cool stuff starts. Use this method if you 78 want to build new graphs from Python. 79 80 ```python 81 import pydot 82 83 graph = pydot.Dot('my_graph', graph_type='graph', bgcolor='yellow') 84 85 # Add nodes 86 my_node = pydot.Node('a', label='Foo') 87 graph.add_node(my_node) 88 # Or, without using an intermediate variable: 89 graph.add_node(pydot.Node('b', shape='circle')) 90 91 # Add edges 92 my_edge = pydot.Edge('a', 'b', color='blue') 93 graph.add_edge(my_edge) 94 # Or, without using an intermediate variable: 95 graph.add_edge(pydot.Edge('b', 'c', color='blue')) 96 ``` 97 98 Imagine using these basic building blocks from your Python program 99 to dynamically generate a graph. For example, start out with a 100 basic `pydot.Dot` graph object, then loop through your data while 101 adding nodes and edges. Use values from your data as labels, to 102 determine shapes, edges and so forth. This way, you can easily 103 build visualizations of thousands of interconnected items. 104 1054. or: Convert a NetworkX graph to a pydot graph. 106 107 NetworkX has conversion methods for pydot graphs: 108 109 ```python 110 import networkx 111 import pydot 112 113 # See NetworkX documentation on how to build a NetworkX graph. 114 115 graph = networkx.drawing.nx_pydot.to_pydot(my_networkx_graph) 116 ``` 117 118Edit 119---- 120 121You can now further manipulate your graph using pydot methods: 122 123- Add further nodes and edges: 124 125 ```python 126 graph.add_edge(pydot.Edge('b', 'd', style='dotted')) 127 ``` 128 129- Edit attributes of graph, nodes and edges: 130 131 ```python 132 graph.set_bgcolor('lightyellow') 133 graph.get_node('b')[0].set_shape('box') 134 ``` 135 136Output 137------ 138 139Here are 3 different output options: 140 1411. Generate an image. 142 143 To generate an image of the graph, use one of the `create_*()` or 144 `write_*()` methods. 145 146 - If you need to further process the output in Python, the 147 `create_*` methods will get you a Python bytes object: 148 149 ```python 150 output_graphviz_svg = graph.create_svg() 151 ``` 152 153 - If instead you just want to save the image to a file, use one of 154 the `write_*` methods: 155 156 ```python 157 graph.write_png('output.png') 158 ``` 159 1602. Retrieve the DOT string. 161 162 There are two different DOT strings you can retrieve: 163 164 - The "raw" pydot DOT: This is generated the fastest and will 165 usually still look quite similar to the DOT you put in. It is 166 generated by pydot itself, without calling Graphviz. 167 168 ```python 169 # As a string: 170 output_raw_dot = graph.to_string() 171 # Or, save it as a DOT-file: 172 graph.write_raw('output_raw.dot') 173 ``` 174 175 - The Graphviz DOT: You can use it to check how Graphviz lays out 176 the graph before it produces an image. It is generated by 177 Graphviz. 178 179 ```python 180 # As a bytes literal: 181 output_graphviz_dot = graph.create_dot() 182 # Or, save it as a DOT-file: 183 graph.write_dot('output_graphviz.dot') 184 ``` 185 1863. Convert to a NetworkX graph. 187 188 Here as well, NetworkX has a conversion method for pydot graphs: 189 190 ```python 191 my_networkx_graph = networkx.drawing.nx_pydot.from_pydot(graph) 192 ``` 193 194More help 195--------- 196 197For more help, see the docstrings of the various pydot objects and 198methods. For example, `help(pydot)`, `help(pydot.Graph)` and 199`help(pydot.Dot.write)`. 200 201More [documentation contributions welcome][13]. 202 203 204Installation 205============ 206 207From [PyPI][4] using [`pip`][5]: 208 209`pip install pydot` 210 211From source: 212 213`python setup.py install` 214 215 216Dependencies 217============ 218 219- [`pyparsing`][6]: used only for *loading* DOT files, 220 installed automatically during `pydot` installation. 221 222- GraphViz: used to render graphs as PDF, PNG, SVG, etc. 223 Should be installed separately, using your system's 224 [package manager][7], something similar (e.g., [MacPorts][8]), 225 or from [its source][9]. 226 227 228License 229======= 230 231Distributed under an [MIT license][10]. 232 233 234Contacts 235======== 236 237Maintainers: 238- Sebastian Kalinowski <sebastian@kalinowski.eu> (GitHub: @prmtl) 239- Peter Nowee <peter@peternowee.com> (GitHub: @peternowee) 240 241Original author: Ero Carrera <ero.carrera@gmail.com> 242 243 244[1]: https://www.graphviz.org 245[2]: https://en.wikipedia.org/wiki/DOT_%28graph_description_language%29 246[3]: https://github.com/networkx/networkx 247[4]: https://pypi.python.org/pypi 248[5]: https://github.com/pypa/pip 249[6]: https://github.com/pyparsing/pyparsing 250[7]: https://en.wikipedia.org/wiki/Package_manager 251[8]: https://www.macports.org 252[9]: https://gitlab.com/graphviz/graphviz 253[10]: https://github.com/pydot/pydot/blob/master/LICENSE 254[11]: https://github.com/pydot/pydot 255[12]: https://en.wikipedia.org/w/index.php?title=DOT_(graph_description_language)&oldid=1003001464#Attributes 256[13]: https://github.com/pydot/pydot/issues/130 257
){kind=link}
){kind=link}
