Python绘制Venn图

pyvenn

这里介绍一个使用python绘制Venn图的方法，常见的绘制Venn图有两种方法，本文介绍第二种，它提供了2-6个子集时的Venn图绘制方法。

下面是demo：

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
import matplotlib
matplotlib.use('Agg')

import matplotlib.pyplot as plt
import venn

labels = venn.get_labels([range(10), range(5, 15)], fill=['number', 'logic'])
fig, ax = venn.venn2(labels, names=['list 1', 'list 2'])
fig.savefig('betterpyvenn/images/venn2.png', bbox_inches='tight')
plt.close()

labels = venn.get_labels([range(10), range(5, 15), range(3, 8)], fill=['number', 'logic'])
fig, ax = venn.venn3(labels, names=['list 1', 'list 2', 'list 3'])
fig.savefig('betterpyvenn/images/venn3.png', bbox_inches='tight')
plt.close()

labels = venn.get_labels([range(10), range(5, 15), range(3, 8), range(8, 17)], fill=['number', 'logic'])
fig, ax = venn.venn4(labels, names=['list 1', 'list 2', 'list 3', 'list 4'])
fig.savefig('betterpyvenn/images/venn4.png', bbox_inches='tight')
plt.close()

labels = venn.get_labels([range(10), range(5, 15), range(3, 8), range(8, 17), range(10, 20)], fill=['number', 'logic'])
fig, ax = venn.venn5(labels, names=['list 1', 'list 2', 'list 3', 'list 4', 'list 5'])
fig.savefig('betterpyvenn/images/venn5.png', bbox_inches='tight')
plt.close()

labels = venn.get_labels([range(10), range(5, 15), range(3, 8), range(8, 17), range(10, 20), range(13, 25)], fill=['number', 'logic'])
fig, ax = venn.venn6(labels, names=['list 1', 'list 2', 'list 3', 'list 4', 'list 5', 'list 6'])
fig.savefig('betterpyvenn/images/venn6.png', bbox_inches='tight')
plt.close()

两个子集

其他方式展示Venn图

使用python的upsetplot库进行可视化

安装

1
pip install upsetplot

简单例子

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
from upsetplot import generate_counts
example = generate_counts()
example 

out:
    cat0   cat1   cat2 
False  False  False      56
              True      283
       True   False    1279
              True     5882
True   False  False      24
              True       90
       True   False     429
              True     1957
Name: value, dtype: int64

1
2
3
4
5
6
7
from upsetplot import plot
plot(example) 
out:
    {'matrix': <Axes: >,
 'shading': <Axes: >,
 'totals': <Axes: >,
 'intersections': <Axes: ylabel='Intersection size'>}

X轴：各个集合有无两种逻辑状态，

y轴：展示数量

复杂例子

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
import pandas as pd
from sklearn.datasets import load_diabetes
from matplotlib import pyplot as plt
from upsetplot import UpSet

# 将数据集加载到DataFrame中
diabetes = load_diabetes()
diabetes_df = pd.DataFrame(diabetes.data, columns=diabetes.feature_names)

# 获取与中位数房屋价值最相关的五个特征
correls = diabetes_df.corrwith(pd.Series(diabetes.target), method='spearman').sort_values()
top_features = correls.index[-5:]

# 获取每个顶级特征是否高于平均值的二进制指示符
diabetes_above_avg = diabetes_df > diabetes_df.median(axis=0)
diabetes_above_avg = diabetes_above_avg[top_features]
diabetes_above_avg = diabetes_above_avg.rename(columns=lambda x: x + '>')

# 将该指示符掩码应用到diabetes_df的索引中
diabetes_df = pd.concat([diabetes_df, diabetes_above_avg], axis=1)
diabetes_df = diabetes_df.set_index(list(diabetes_above_avg.columns))

# 同时提供目标变量（中位数房屋价值）
diabetes_df = diabetes_df.assign(progression=diabetes.target)

# 使用库进行可视化
upset = UpSet(diabetes_df, subset_size='count', intersection_plot_elements=3)
upset.add_catplot(value='progression', kind='strip', color='blue')
print(diabetes_df)
upset.add_catplot(value='bmi', kind='strip', color='black')
upset.plot()
plt.title("UpSet with catplots, for orientation='horizontal'")
plt.savefig('./upset2.jpg')
plt.show()

out:
    s6>   bp>   s4>   bmi>  s5>                                             
False True  False True  True   0.038076  0.050680  0.061696  0.021872   
      False False False False -0.001882 -0.044642 -0.051474 -0.026328   
                  True  True   0.085299  0.050680  0.044451 -0.005670   
            True  False True  -0.089063 -0.044642 -0.011595 -0.036656   
      True  False False False  0.005383 -0.044642 -0.036385  0.021872   
...                                 ...       ...       ...       ...   
True  True  False True  True   0.041708  0.050680  0.019662  0.059744   
      False True  False False -0.005515  0.050680 -0.015906 -0.067642   
      True  False False False  0.041708  0.050680 -0.015906  0.017293   
False True  True  True  True  -0.045472 -0.044642  0.039062  0.001215   
True  False False False False -0.045472 -0.044642 -0.073030 -0.081413   

                                     s1        s2        s3        s4  \
s6>   bp>   s4>   bmi>  s5>                                             
False True  False True  True  -0.044223 -0.034821 -0.043401 -0.002592   
      False False False False -0.008449 -0.019163  0.074412 -0.039493   
                  True  True  -0.045599 -0.034194 -0.032356 -0.002592   
            True  False True   0.012191  0.024991 -0.036038  0.034309   
      True  False False False  0.003935  0.015596  0.008142 -0.002592   
...                                 ...       ...       ...       ...   
True  True  False True  True  -0.005697 -0.002566 -0.028674 -0.002592   
      False True  False False  0.049341  0.079165 -0.028674  0.034309   
      True  False False False -0.037344 -0.013840 -0.024993 -0.011080   
...
False True  True  True  True   0.044529 -0.025930        220.0  
True  False False False False -0.004222  0.003064         57.0  

[442 rows x 11 columns]

竖直例子

1
2
3
4
5
6
7
8
9
10
# 竖直方向

upset = UpSet(diabetes_df, subset_size='count', intersection_plot_elements=3,
              orientation='vertical')
upset.add_catplot(value='progression', kind='strip', color='blue')
upset.add_catplot(value='bmi', kind='strip', color='black')
upset.plot()
plt.title("UpSet with catplots, for orientation='vertical'")
plt.savefig('./upset3.jpg')
plt.show()

总结

在实际应用过程中，可以结合两个，使用Venn图可以让人眼前一亮，获得定性层面的感知，而绘制upset图可以使读者知道集合中交集量的关系，更加科学严谨。