Matplotlib Toolkits：python高级绘图库seaborn

迷南。 2022-08-09 15:52 257阅读 0赞

http://[blog.csdn.net/pipisorry/article/details/49515745][blog.csdn.net_pipisorry_article_details_49515745]

# Seaborn介绍 #

### seaborn ###

(Not distributed with matplotlib)

[seaborn][] is a highlevel interface for drawing statistical graphics with matplotlib. Itaims to make visualization a central part of exploring andunderstanding complex datasets.

\[seaborn [¶][Link 1]\]

Matplotlib是Python主要的绘图库。但是不建议你直接使用它，原因与不推荐你使用NumPy是一样的。虽然Matplotlib很强大，它本身就很复杂，你的图经过大量的调整才能变精致。因此，作为替代推荐一开始使用Seaborn。

Seaborn本质上使用Matplotlib作为核心库（就像Pandas对NumPy一样）。

seaborn的优点：

1.  默认情况下就能创建赏心悦目的图表。（只有一点，默认不是[jet colormap][]）
2.  创建具有统计意义的图
3.  能理解pandas的DataFrame类型，所以它们一起可以很好地工作。

安装pip install seaborn[  
][Link 2]

Note: lz发现，就算你不用seaborn绘图，只要在matplotlib绘图中加上seaborn的import语句，就会以seaborn的图形方式展示图片，具有seaborn的效果。如：[  
][Link 2]

import seaborn
    import matplotlib.pyplot as plt

注意要显示出图形，需要引入matplotlib并plt.show()出来。[  
][Link 2]

[皮皮blog][Link 2]

# Seaborn使用 #

Style functions: [API][] | [Tutorial][]

Color palettes: [API][API 1] | [Tutorial][Tutorial 1]

Distribution plots: [API][API 2] | [Tutorial][Tutorial 2]

Regression plots: [API][API 3] | [Tutorial][Tutorial 3]

Categorical plots: [API][API 4] | [Tutorial][Tutorial 4]

Axis grid objects: [API][API 5] | [Tutorial][Tutorial 5]

## 分布图绘制Distribution plots    ##

<table> 
 <tbody> 
  <tr> 
   <td><a href="http://seaborn.pydata.org/generated/seaborn.jointplot.html#seaborn.jointplot" title="seaborn.jointplot" rel="nofollow"><code><span>jointplot</span></code></a>(x,&nbsp;y[,&nbsp;data,&nbsp;kind,&nbsp;stat_func,&nbsp;...])</td> 
   <td>Draw a plot of two variables with bivariate and univariate graphs.</td> 
  </tr> 
  <tr> 
   <td><a href="http://seaborn.pydata.org/generated/seaborn.pairplot.html#seaborn.pairplot" title="seaborn.pairplot" rel="nofollow"><code><span>pairplot</span></code></a>(data[,&nbsp;hue,&nbsp;hue_order,&nbsp;palette,&nbsp;...])</td> 
   <td>Plot pairwise relationships in a dataset.</td> 
  </tr> 
  <tr> 
   <td><a href="http://seaborn.pydata.org/generated/seaborn.distplot.html#seaborn.distplot" title="seaborn.distplot" rel="nofollow"><code><span>distplot</span></code></a>(a[,&nbsp;bins,&nbsp;hist,&nbsp;kde,&nbsp;rug,&nbsp;fit,&nbsp;...])</td> 
   <td>Flexibly plot a univariate distribution of observations.</td> 
  </tr> 
  <tr> 
   <td><a href="http://seaborn.pydata.org/generated/seaborn.kdeplot.html#seaborn.kdeplot" title="seaborn.kdeplot" rel="nofollow"><code><span>kdeplot</span></code></a>(data[,&nbsp;data2,&nbsp;shade,&nbsp;vertical,&nbsp;...])</td> 
   <td>Fit and plot a univariate or bivariate kernel density estimate.</td> 
  </tr> 
  <tr> 
   <td><a href="http://seaborn.pydata.org/generated/seaborn.rugplot.html#seaborn.rugplot" title="seaborn.rugplot" rel="nofollow"><code><span>rugplot</span></code></a>(a[,&nbsp;height,&nbsp;axis,&nbsp;ax])</td> 
   <td>Plot datapoints in an array as sticks on an axis.</td> 
  </tr> 
 </tbody> 
</table>

### 单变量绘制distplot ###

seaborn.[distplot][](a, bins=None, hist=True, kde=True, rug=False, fit=None, hist\_kws=None, kde\_kws=None, rug\_kws=None, fit\_kws=None, color=None, vertical=False, norm\_hist=False, axlabel=None, label=None, ax=None)

Note:

1 如果想显示统计个数而不是概率，需要同时设置norm\_hist=False, kde=False。

2 自己指定fit的函数from scipy import stats ... fit=stats.norm

>>> import seaborn as sns, numpy as np
    >>> sns.set(rc={
       "figure.figsize": (8, 4)}); np.random.seed(0)
    >>> x = np.random.randn(100)
    >>> ax = sns.distplot(x)

### 双变量+单变量统一绘制jointplot ###

import   seaborn  as   sns

\# Load one of the data sets that come with seaborn

tips  =  sns . load\_dataset ( "tips" )

sns . jointplot ( "total\_bill" ,  "tip" ,  tips ,  kind = 'reg' ) ;

![7cc829d3gw1enqrs5xbbqj20bp0brq3r.jpg][]

如你所见，仅通过一行代码，我们就创建了一个漂亮复杂的统计图，其中包含拥有置信区间的最拟合回归直线、边界图，以及相关系数。  
使用matplotlib重新绘制这幅图的话需要相当多的（丑陋）代码，包括调用scipy执行线性回归并手动利用线性回归方程绘制直线（我甚至想不出怎么在边界绘图，怎么计算置信区间）。  
\[ [the tutorial on quantitative linear models][]\]

### 与Pandas的DataFrame很好地工作 ###

数据有自己的结构。通常我们感兴趣的包含不同的组或类（这种情况下使用pandas中groupby的功能会让人感到很神奇）。比如tips（小费）的数据集是这样的：

tips . head ( )

Out\[9\]:

<table> 
 <thead> 
  <tr> 
   <th>&nbsp;</th> 
   <th>total_bill</th> 
   <th>tip</th> 
   <th>sex</th> 
   <th>smoker</th> 
   <th>day</th> 
   <th>time</th> 
   <th>size</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <th>0</th> 
   <td>16.99</td> 
   <td>1.01</td> 
   <td>Female</td> 
   <td>No</td> 
   <td>Sun</td> 
   <td>Dinner</td> 
   <td>2</td> 
  </tr> 
  <tr> 
   <th>1</th> 
   <td>10.34</td> 
   <td>1.66</td> 
   <td>Male</td> 
   <td>No</td> 
   <td>Sun</td> 
   <td>Dinner</td> 
   <td>3</td> 
  </tr> 
  <tr> 
   <th>2</th> 
   <td>21.01</td> 
   <td>3.50</td> 
   <td>Male</td> 
   <td>No</td> 
   <td>Sun</td> 
   <td>Dinner</td> 
   <td>3</td> 
  </tr> 
  <tr> 
   <th>3</th> 
   <td>23.68</td> 
   <td>3.31</td> 
   <td>Male</td> 
   <td>No</td> 
   <td>Sun</td> 
   <td>Dinner</td> 
   <td>2</td> 
  </tr> 
  <tr> 
   <th>4</th> 
   <td>24.59</td> 
   <td>3.61</td> 
   <td>Female</td> 
   <td>No</td> 
   <td>Sun</td> 
   <td>Dinner</td> 
   <td>4</td> 
  </tr> 
 </tbody> 
</table>

我们可能想知道吸烟者给的小费是否与不吸烟的人不同。没有seaborn的话，这需要使用pandas的groupby功能，并通过复杂的代码绘制线性回归直线。使用seaborn的话，我们可以给col参数提供列名，按我们的需要划分数据：

## 回归图绘制Regression plots ##

<table>     
 <tbody> 
  <tr> 
   <td><a href="http://seaborn.pydata.org/generated/seaborn.lmplot.html#seaborn.lmplot" title="seaborn.lmplot" rel="nofollow"><code><span>lmplot</span></code></a>(x,&nbsp;y,&nbsp;data[,&nbsp;hue,&nbsp;col,&nbsp;row,&nbsp;palette,&nbsp;...])</td> 
   <td>Plot data and regression model fits across a FacetGrid.</td> 
  </tr> 
  <tr> 
   <td><a href="http://seaborn.pydata.org/generated/seaborn.regplot.html#seaborn.regplot" title="seaborn.regplot" rel="nofollow"><code><span>regplot</span></code></a>(x,&nbsp;y[,&nbsp;data,&nbsp;x_estimator,&nbsp;x_bins,&nbsp;...])</td> 
   <td>Plot data and a linear regression model fit.</td> 
  </tr> 
  <tr> 
   <td><a href="http://seaborn.pydata.org/generated/seaborn.residplot.html#seaborn.residplot" title="seaborn.residplot" rel="nofollow"><code><span>residplot</span></code></a>(x,&nbsp;y[,&nbsp;data,&nbsp;lowess,&nbsp;x_partial,&nbsp;...])</td> 
   <td>Plot the residuals of a linear regression.</td> 
  </tr> 
  <tr> 
   <td><a href="http://seaborn.pydata.org/generated/seaborn.interactplot.html#seaborn.interactplot" title="seaborn.interactplot" rel="nofollow"><code><span>interactplot</span></code></a>(x1,&nbsp;x2,&nbsp;y[,&nbsp;data,&nbsp;filled,&nbsp;...])</td> 
   <td>Visualize a continuous two-way interaction with a contour plot.</td> 
  </tr> 
  <tr> 
   <td><a href="http://seaborn.pydata.org/generated/seaborn.coefplot.html#seaborn.coefplot" title="seaborn.coefplot" rel="nofollow"><code><span>coefplot</span></code></a>(formula,&nbsp;data[,&nbsp;groupby,&nbsp;...])</td> 
   <td>Plot the coefficients from a linear model.</td> 
  </tr> 
 </tbody> 
</table>

sns . lmplot ( "total\_bill" ,  "tip" ,  tips ,  col = "smoker" ) ;

![7cc829d3gw1enqrs6o858j20jr09r75b.jpg][]

很整洁吧？随着你研究得越深，你可能想更细粒度地控制这些图表的细节。因为seaborn只是调用了matplotlib，那时你可能会想学习这个库。

from: [http://blog.csdn.net/pipisorry/article/details/49515745][blog.csdn.net_pipisorry_article_details_49515745]

ref: \[[seaborn API reference][]\]\*

\[[Seaborn tutorial][]\]\*

[Python和数据科学的起步指南][Python]

[Example gallery][][  
][Link 3]

[Python数据可视化模块：Seaborn][Link 3]

[blog.csdn.net_pipisorry_article_details_49515745]: http://blog.csdn.net/pipisorry/article/details/49515745
[seaborn]: http://web.stanford.edu/%7Emwaskom/software/seaborn
[Link 1]: http://matplotlib.org/mpl_toolkits/index.html#seaborn
[jet colormap]: https://jakevdp.github.io/blog/2014/10/16/how-bad-is-your-colormap/
[Link 2]: http://blog.csdn.net/pipisorry
[API]: http://seaborn.pydata.org/api.html#style-api
[Tutorial]: http://seaborn.pydata.org/tutorial/aesthetics.html#aesthetics-tutorial
[API 1]: http://seaborn.pydata.org/api.html#palette-api
[Tutorial 1]: http://seaborn.pydata.org/tutorial/color_palettes.html#palette-tutorial
[API 2]: http://seaborn.pydata.org/api.html#distribution-api
[Tutorial 2]: http://seaborn.pydata.org/tutorial/distributions.html#distribution-tutorial
[API 3]: http://seaborn.pydata.org/api.html#regression-api
[Tutorial 3]: http://seaborn.pydata.org/tutorial/regression.html#regression-tutorial
[API 4]: http://seaborn.pydata.org/api.html#categorical-api
[Tutorial 4]: http://seaborn.pydata.org/tutorial/categorical.html#categorical-tutorial
[API 5]: http://seaborn.pydata.org/api.html#grid-api
[Tutorial 5]: http://seaborn.pydata.org/tutorial/axis_grids.html#grid-tutorial
[distplot]: http://seaborn.pydata.org/generated/seaborn.distplot.html#seaborn.distplot
[7cc829d3gw1enqrs5xbbqj20bp0brq3r.jpg]: /images/20220731/071b76fe2c8e48ef803eec2e0474a8f7.png
[the tutorial on quantitative linear models]: http://web.stanford.edu/%7Emwaskom/software/seaborn/tutorial/quantitative_linear_models.html
[7cc829d3gw1enqrs6o858j20jr09r75b.jpg]: /images/20220731/d2f61e5e7e2b4bb8b9c646a2ed2a7124.png
[seaborn API reference]: http://web.stanford.edu/%7Emwaskom/software/seaborn/api.html
[Seaborn tutorial]: http://web.stanford.edu/%7Emwaskom/software/seaborn/tutorial.html
[Python]: http://python.jobbole.com/80853/
[Example gallery]: http://web.stanford.edu/~mwaskom/software/seaborn/examples/index.html
[Link 3]: http://segmentfault.com/a/1190000002789457