从seaborn保存情节

Question

当我尝试使用seaborn保存我的情节时，像这样：

import seaborn as sn
import pandas as pd
import matplotlib.pyplot as plt
from pylab import savefig

array = [[100,0], 
        [33,67]]

df_cm = pd.DataFrame(array)

svm = sn.heatmap(df_cm, annot=True,cmap='coolwarm', linecolor='white', linewidths=1)

svm.savefig('svm_conf.png', dpi=400)

我收到此错误

AttributeError                            Traceback (most recent call last)
<ipython-input-71-5c0ae9cda020> in <module>()
----> 1 svm.savefig('svm_conf.png', dpi=400)

AttributeError: 'AxesSubplot' object has no attribute 'savefig'

我之前用相同的代码保存了一些箱图，但这一次，它不起作用。

Answer 1

实际上你需要做的是：

• 从sn.heatmap
返回的对象中检索图形
• 然后只保存图

见下面最后两行：

import seaborn as sn
import pandas as pd
import matplotlib.pyplot as plt
from pylab import savefig

array = [[100,0], 
        [33,67]]

df_cm = pd.DataFrame(array)

svm = sn.heatmap(df_cm, annot=True,cmap='coolwarm', linecolor='white', linewidths=1)

figure = svm.get_figure()    
figure.savefig('svm_conf.png', dpi=400)

Answer 2

最简单的方法是使用

plt.savefig('svm_conf.png', dpi=400)

而不是

svm.savefig('svm_conf.png', dpi=400)

Answer 3

保存当前数字的命令是

plt.savefig()

因为这显然会引起一些混乱，所以这是完整的工作示例：

import seaborn as sn
import pandas as pd
import matplotlib.pyplot as plt

array = [[100,0], 
        [33,67]]

df_cm = pd.DataFrame(array)

svm = sn.heatmap(df_cm, annot=True,cmap='coolwarm', linecolor='white', linewidths=1)

plt.savefig('svm_conf.png', dpi=400)