Seaborn:管理图表的艺术

原文链接
管理图表的艺术

画一个吸引人注意的图表相当重要。当你探索一个数据集，需要画图表，图表看起来令人愉悦是件很高兴的事。在与你的观众交流观点时，可视化同样重要，同时，也很有必要去让图表吸引注意力和印入脑海里。Matplotlib自动化程度非常高，但是，掌握如何设置系统以便获得一个吸引人的图是相当困难的事。为了控制matplotlib图表的外观，Seaborn模块自带许多定制的主题和高级的接口。

%matplotlib inline

import numpy as np
import matplotlib as mpl
import matplotlib.pyplot as plt
np.random.seed(sum(map(ord, "aesthetics")))

让我们先定义一个函数用来画正弦函数，这将帮助我们了解我们可以控制的不同风格的参数

def sinplot(flip=1):
    x = np.linspace(0, 14, 100)
    for i in range(1, 7):
        plt.plot(x, np.sin(x + i * .5) * (7 - i) * flip)

默认情况下matplotlib的画的图是这样的：
sinplot()

转换成Seaborn模块，只需要引入seaborn模块。

import seaborn as sns
sinplot()

seaborn默认浅灰色背景与白色网络线的灵感来源于matplotlib，却比matplotlib的颜色更多柔和。我们发现，网络线对于传播信息很有用，几乎在所有情况下，人们喜欢图甚于表。默认情况下白灰网格的形式可以避免过于刺眼。在多面作图的情况下，网络形式显得相当的有利，提供了一种作图结构，这对模块中的一些复杂工具非常重要。

seaborn将matplotlib的参数划分为两个组。第一组控制图表的样式和图的度量尺度元素，这样就可以轻易在纳入到不同的上下文中。

操控这些参数由两个函数提供接口。控制样式，用axes_style()和set_style()这两个函数。度量图则用plotting_context()和set_context()这两个函数。在这两种情况下，第一组函数返回一系列的参数，第二组则设置matplotlib的默认属性。

图样式函数axes_style()和set_style()

There are five preset seaborn themes: darkgrid, whitegrid, dark, white, and ticks. They are each suited to different applications and personal preferences. The default theme is darkgrid. As mentioned above, the grid helps the plot serve as a lookup table for quantitative information, and the white-on grey helps to keep the grid from competing with lines that represent data. The whitegrid theme is similar, but it is better suited to plots with heavy data elements:

sns.set_style("whitegrid")
data = np.random.normal(size=(20, 6)) + np.arange(6) / 2
sns.boxplot(data=data);


For many plots, (especially for settings like talks, where you primarily want to use figures to provide impressions of patterns in the data), the grid is less necessary.

sns.set_style("dark")
sinplot()

sns.set_style("white")
sinplot()

Sometimes you might want to give a little extra structure to the plots, which is where ticks come in handy:

sns.set_style("ticks")
sinplot()

Removing spines with despine()

Both the white and ticks styles can benefit from removing the top and right axes spines, which are not needed. It’s impossible to do this through the matplotlib parameters, but you can call the seaborn function despine() to remove them:
Both the white and ticks styles can benefit from removing the top and right axes spines, which are not needed. It’s impossible to do this through the matplotlib parameters, but you can call the seaborn function despine() to remove them:

sinplot()
sns.despine()

Some plots benefit from offsetting the spines away from the data, which can also be done when calling despine(). When the ticks don’t cover the whole range of the axis, the trim parameter will limit the range of the surviving spines.

f, ax = plt.subplots()
sns.violinplot(data=data)
sns.despine(offset=10, trim=True);

You can also control which spines are removed with additional arguments to despine():

sns.set_style("whitegrid")
sns.boxplot(data=data, palette="deep")
sns.despine(left=True)

Temporarily setting figure style

Although it’s easy to switch back and forth, you can also use the axes_style() function in a with statement to temporarily set plot parameters. This also allows you to make figures with differently-styled axes:

with sns.axes_style("darkgrid"):
    plt.subplot(211)
    sinplot()
plt.subplot(212)
sinplot(-1)

http://seaborn.pydata.org/_images/aesthetics_27_0.png
Overriding elements of the seaborn styles

If you want to customize the seaborn styles, you can pass a dictionary of parameters to the rc argument of axes_style() and set_style(). Note that you can only override the parameters that are part of the style definition through this method. (However, the higher-level set() function takes a dictionary of any matplotlib parameters).

If you want to see what parameters are included, you can just call the function with no arguments, which will return the current settings:

sns.axes_style()

{'axes.axisbelow': True,
 'axes.edgecolor': '.8',
 'axes.facecolor': 'white',
 'axes.grid': True,
 'axes.labelcolor': '.15',
 'axes.linewidth': 1.0,
 'figure.facecolor': 'white',
 'font.family': [u'sans-serif'],
 'font.sans-serif': [u'Arial',
  u'Liberation Sans',
  u'Bitstream Vera Sans',
  u'sans-serif'],
 'grid.color': '.8',
 'grid.linestyle': u'-',
 'image.cmap': u'Greys',
 'legend.frameon': False,
 'legend.numpoints': 1,
 'legend.scatterpoints': 1,
 'lines.solid_capstyle': u'round',
 'text.color': '.15',
 'xtick.color': '.15',
 'xtick.direction': u'out',
 'xtick.major.size': 0.0,
 'xtick.minor.size': 0.0,
 'ytick.color': '.15',
 'ytick.direction': u'out',
 'ytick.major.size': 0.0,
 'ytick.minor.size': 0.0}

You can then set different versions of these parameters:

sns.set_style("darkgrid", {"axes.facecolor": ".9"})
sinplot()

Scaling plot elements with plotting_context() and set_context()

A separate set of parameters control the scale of plot elements, which should let you use the same code to make plots that are suited for use in settings where larger or smaller plots are appropriate.

First let’s reset the default parameters by calling set():

sns.set()

The four preset contexts, in order of relative size, are paper, notebook, talk, and poster. The notebook style is the default, and was used in the plots above.

sns.set_context("paper")
plt.figure(figsize=(8, 6))
sinplot()

sns.set_context("talk")
plt.figure(figsize=(8, 6))
sinplot()

../_images/aesthetics_36_0.png

sns.set_context("poster")
plt.figure(figsize=(8, 6))
sinplot()

Most of what you now know about the style functions should transfer to the context functions.

You can call set_context() with one of these names to set the parameters, and you can override the parameters by providing a dictionary of parameter values.

You can also independently scale the size of the font elements when changing the context. (This option is also available through the top-level set() function).

sns.set_context("notebook", font_scale=1.5, rc={"lines.linewidth": 2.5})
sinplot()

Similarly (although it might be less useful), you can temporarily control the scale of figures nested under a with statement.

Both the style and the context can be quickly configured with the set() function. This function also sets the default color palette, but that will be covered in more detail in the next section of the tutorial.