1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
|
.. _matplotlib_sec:
Integration with matplotlib
===========================
The drawing capabilities of ``graph-tool`` (see :mod:`~graph_tool.draw`
module) can be integrated with `matplotlib <https://matplotlib.org>`_,
as we demonstrate in the following.
.. note::
:class: margin
Integration with :mod:`matplotlib` works with every backend, but `vector
drawing <https://en.wikipedia.org/wiki/Vector_graphics>`_ only works with a
`cairo <https://cairographics.org>`_-based backend (e.g. ``cairo`` or
``GTK3Cairo``). The backend can be changed by calling
:func:`matplotlib.pyplot.switch_backend`:
.. code::
import matplotlib.pyplot as plt
plt.switch_backend("cairo")
When using a backend not based on cairo, `rasterization
<https://en.wikipedia.org/wiki/Raster_graphics>`_ will be used instead. In
this case, the resolution can be controlled via the ``dpi`` parameter of
:class:`matplotlib.figure.Figure`.
Drawing with matplotlib is done by calling :func:`~graph_tool.draw.graph_draw`
and passing a container (e.g. :class:`matplotlib.axes.Axes`) as the ``mplfig``
parameter. When this option is passed, the function will return a
:func:`~graph_tool.draw.GraphArtist` object that has been added to the figure.
.. tip::
:class: margin
`Axis autoscaling
<https://matplotlib.org/stable/users/explain/axes/autoscale.html>`_ will work
as expected with :func:`~graph_tool.draw.GraphArtist`, but the aspect ratio
will be set the by the figure axis shape, rather than the node positions themselves.
In order to restore the aspect ratio independently of the axis shape,
:func:`~graph_tool.draw.GraphArtist` offers a
:meth:`~graph_tool.draw.GraphArtist.fit_view` method that does this automatically.
Furthermore, when calling :func:`~graph_tool.draw.graph_draw` without
integrating with matplotlib, the node positions correspond to cairo
coordinates, which have an origin in the upper left corner, and with the y
axis increasing from top to bottom.
In order for the visualization to be the same when matplotlib is being used,
the y axis needs to be flipped by inverting the limits with
:meth:`matplotlib.axes.Axes.set_ylim`.
Alternatively, the option ``yflip`` can be passed to
:meth:`~graph_tool.draw.GraphArtist.fit_view` for this to be done
automatically.
The example below shows how to plot several graphs in different subplots of the
same figure.
.. testsetup::
import matplotlib
backend = matplotlib.get_backend()
gt.seed_rng(44)
.. testcode::
import graph_tool.all as gt
import matplotlib.pyplot as plt
plt.switch_backend("cairo") # to enable vector drawing
fig, ax = plt.subplots(2, 2, figsize=(12, 11.5))
g = gt.collection.data["polbooks"]
a = gt.graph_draw(g, g.vp.pos, vertex_size=1.5, mplfig=ax[0,0])
a.fit_view(yflip=True)
ax[0,0].set_xlabel("$x$ coordinate")
ax[0,0].set_ylabel("$y$ coordinate")
state = gt.minimize_nested_blockmodel_dl(g)
a = state.draw(mplfig=ax[0,1])[0]
a.fit_view(yflip=True)
ax[0,1].set_xlabel("$x$ coordinate")
ax[0,1].set_ylabel("$y$ coordinate")
g = gt.collection.data["lesmis"]
a = gt.graph_draw(g, g.vp.pos, vertex_size=1.5, mplfig=ax[1,0])
a.fit_view(yflip=True)
ax[1,0].set_xlabel("$x$ coordinate")
ax[1,0].set_ylabel("$y$ coordinate")
state = gt.minimize_nested_blockmodel_dl(g)
a = state.draw(mplfig=ax[1,1])[0]
a.fit_view(yflip=True)
ax[1,1].set_xlabel("$x$ coordinate")
ax[1,1].set_ylabel("$y$ coordinate")
plt.subplots_adjust(left=0.08, right=0.99, top=0.99, bottom=0.06)
fig.savefig("gt-mpl.svg")
.. testcleanup::
plt.switch_backend(backend)
.. figure:: gt-mpl.svg
:width: 90%
Four subplots showing networks drawn using graph-tool.
Integration with :mod:`~mpl_toolkits.basemap`
+++++++++++++++++++++++++++++++++++++++++++++
As a slightly more elaborate example, below we show how we can draw the
:ns:`European airline <eu_airlines>` graph on a map using
:mod:`mpl_toolkits.basemap`.
.. testcode::
from itertools import chain
from mpl_toolkits.basemap import Basemap
fig, ax = plt.subplots(1, 1, figsize=(8, 8))
g = gt.collection.ns["eu_airlines"]
pos = gt.group_vector_property([g.vp.nodeLong, g.vp.nodeLat])
m = Basemap(projection='ortho', resolution=None,
lat_0=g.vp.nodeLat.fa.mean(), lon_0=g.vp.nodeLong.fa.mean())
m.shadedrelief(scale=.2)
lats = m.drawparallels(np.linspace(-90, 90, 13))
lons = m.drawmeridians(np.linspace(-180, 180, 13))
lat_lines = chain(*(tup[1][0] for tup in lats.items()))
lon_lines = chain(*(tup[1][0] for tup in lons.items()))
all_lines = chain(lat_lines, lon_lines)
for line in all_lines:
line.set(linestyle='-', alpha=0.3, color='w')
a = gt.graph_draw(g, pos=pos.t(lambda x: m(*x)), # project positions
edge_color=(.1,.1,.1,.1), mplfig=ax)
a.fit_view()
a.set_zorder(10)
tight_layout()
fig.savefig("gt-map.svg")
.. figure:: gt-map.svg
:width: 90%
:ns:`Network of European fligths <eu_airlines>` drawn on the globe with an
`orthographic projection
<https://en.wikipedia.org/wiki/Orthographic_projection>`_.
|
