@@ -32,6 +32,31 @@ Phil Elson rewrote of the documentation and userguide for both Legend and PathEf
32
32
New plotting features
33
33
---------------------
34
34
35
+ Fully customizable boxplots
36
+ ````````````````````````````
37
+ Paul Hobson overahuled the :func: `~matplotlib.pyplot.boxplot ` method such
38
+ that it is now completely customizable in terms of the styles and positions
39
+ of the individual artists. Under the hood, :func: `~matplotlib.pyplot.boxplot `
40
+ relies on a new function (:func: `~matplotlib.cbook.boxplot_stats `), which
41
+ accepts any data structure currently compatible with
42
+ :func: `~matplotlib.pyplot.boxplot `, amd returns a list of dictaries
43
+ containing the positions of each of artists for the boxplots. Then
44
+ a second method, :func: `~matplotlib.Axes.bxp ` is called to actually
45
+ return the stats onto the figure.
46
+
47
+ Therefore, it is now possible for the user can use
48
+ :func: `~matplotlib.pyplot.boxplot ` in the same fashion or generate her own
49
+ list of dictionaries of statistics and feed those directio to
50
+ :func: `~matplotlib.Axes.bxp `. Similarly, the output from
51
+ :func: `~matplotlib.cbook.boxplot_stats ` can easily be modified by the user.
52
+
53
+ Lastly, each artist (e.g., the box, outliers, cap, notches) can now be
54
+ toggled on or off and their styles can be passed in through individual
55
+ kwargs. See the examples:
56
+ :ref: `~examples/statistics/boxplot_demo.py ` and
57
+ :ref: `~examples/statistics/bxp_demo.py `
58
+
59
+
35
60
Support for datetime axes in 2d plots
36
61
`````````````````````````````````````
37
62
Andrew Dawson added support for datetime axes to
0 commit comments