Sunburst Charts

🔖 Check out the full documentation!

🚧 This is an older project that is currently being brushed up 🚧

Introduction

sunburst creates "sunburst charts" (also called "Ring Charts", "Hierarchical Pie Chargs" or "Multilevel Pie Charts") together with the matplotlib package. Quoting Wikipedia:

A ring chart, also known as a sunburst chart or a multilevel pie chart, is used to visualize hierarchical data, depicted by concentric circles. The circle in the centre represents the root node, with the hierarchy moving outward from the center. A segment of the inner circle bears a hierarchical relationship to those segments of the outer circle which lie within the angular sweep of the parent segment.

A prominent example are disk usage charts (see examples/disk_usage_plot.py for this example):

✨ Features

sunburst tries to be

• Intuitive & Easy to use: After setting up your data and matplotlib, not more than two lines are necessary to generate first plots (minimal example).
• Flexible & Robust: Wherever it makes sense, the methods of the SunburstPlot class are intended to be overwritten. Methods that are responsible for spacing, coloring, styling etc. of the wedges take the corresponding data point (path) as an argument, allowing to set most properties independently for each wedge.

More specifically:

• Clever positioning of labels, which - depending on space constraints are positioned in a tangentially or radially.
• An easy way to "explode" the plot by redefining sunburst.wedge_spacing (example, example).

📦 Installation

To install, run

pip3 install sunburst

🚧 Migration notice

For more information about recent changes, see the changelog.

• The 2.0 version will have a different API without the Path class
• Before the first release to pypi, this package was called hpie.
  • There have since bean many non-backwards compatible changes to the API: Most importantly, the HPie class is now SunburstPlot and the package is called sunburst
  • To get the old version back, check out the old-hpie branch. It will however not be maintained any longer.

🔥 Minimal example

You can find several examples at examples/. The most basic example is minimal_example_sunburst.py:

import matplotlib.pyplot as plt
from sunburst import SunburstPlot, stringvalues_to_pv

# set up some data
data = stringvalues_to_pv({
    'ipsum':                      40.45,
    'ipsum/eirmod':               29.34,
    'ipsum/eirmod/dolor':         94.4,
    'lorem':                      36.12,
    'lorem/sadipscing/dolor':     44.32,
    'lorem/sadipscing/lorem':     37.15,
    'lorem/sadipscing/nonumy':    23.98,
    'lorem/eirmod':               11.12,
    'lorem/eirmod/lorem':         45.65,
    'lorem/sadipscing':           79.67,
})


fig, ax = plt.subplots()
sbp = SunburstPlot(data, ax)
sbp.plot(setup_axes=True)
ax.set_title('Example SunburstPlot')

plt.show()

Running this script with python3 minimal_example_sunburst.py will produce the following plot:

The Data

Note that the value corresponding to path is always the value excluding the values of the children of the path. Therefore plotting the SunburstPlot object computes a "completed" version of the "pathvalue dictionary". You can check this with the SunburstPlot._completed_pv instance variable which gets initialized after calling SunburstPlot.plot(*args). Running our minimal example prints the following:

sbp._completed_pv.items() = {
    Path((, )): 442.2,  # = the total sum of all items =
                        # = 36.12 + 44.32 + 37.15 + 23.98 + ...
    Path(('ipsum', )): 164.19000000000003,  # = sum of "ipsum" and all of its children =
                                            # = 40.45 + 29.34 + 94.4
    Path(('ipsum', 'eirmod', )): 123.74000000000001, # = sum of ipsum/eirmod and all of its children =
                                                     # = 29.34 + 94.4
    Path(('ipsum', 'eirmod', 'dolor', )): 94.4,
    Path(('lorem', )): 278.01,
    Path(('lorem', 'eirmod', )): 56.769999999999996,
    Path(('lorem', 'eirmod', 'lorem', )): 45.65,
    Path(('lorem', 'sadipscing', )): 185.12,
    Path(('lorem', 'sadipscing', 'dolor', )): 44.32,
    Path(('lorem', 'sadipscing', 'lorem', )): 37.15,
    Path(('lorem', 'sadipscing', 'nonumy', )): 23.98,
}

sunburst/tests/test_calc.py contains an explicit test of this calculation based on a similar example.

Ring Charts

Thus you get ring charts, if and only if all of the non-zero values correspond to paths with the same length. E.g. if we change the above data as follows (by lengthening the paths with question marks and removing the entry for the empty path):

data = stringvalues_to_pv({
    'ipsum/?/?':                 40.45,
    'ipsum/eirmod/?':            29.34,
    'ipsum/eirmod/dolor':        94.4,
    'lorem/?/?':                 36.12,
    'lorem/sadipscing/dolor':    44.32,
    'lorem/sadipscing/lorem':    37.15,
    'lorem/sadipscing/nonumy':   23.98,
    'lorem/eirmod/?':            11.12,
    'lorem/eirmod/lorem':        45.65,
    'lorem/sadipscing/?':        79.67,
})

we should get a classical ring chart. This is minimal_example_rings.py. Running it via python3 minimal_example_rings.py yields the following plot, which indeed just fills up the white space of the above plot with wedges labeled ?.

License

This project is licensed under the BSD 3-Clause License, see LICENSE.txt.