确实有可能在任何尺寸的图中创建双轴。一个人必须确保了解代码的写作。即请勿使用figure创建新数字,然后抱怨显示第二个数字。
坚持到matplotlib状态机接口,一个解决办法是这样的:
import matplotlib.pyplot as plt
import numpy as np
#get data
x=np.arange(40)
y=np.random.rand(len(x))*20000+30000
y2=np.random.rand(len(x))*0.5
#create a figure
plt.figure(figsize=(10,10))
#plot to first axes
plt.plot(x,y,color='blue',label="label1")
plt.ylim(0,50000)
plt.ylabel('ylabel1')
plt.xticks(rotation=90)
#create twin axes
ax2=plt.gca().twinx()
#plot to twin axes
plt.plot(x,y2,color='purple',label='label2')
plt.ylabel('ylabel2')
plt.legend(loc='upper right')
plt.show()
或者,如果你喜欢的matplotlib API:
import matplotlib.pyplot as plt
import numpy as np
#get data
x=np.arange(40)
y=np.random.rand(len(x))*20000+30000
y2=np.random.rand(len(x))*0.5
#create a figure
fig = plt.figure(figsize=(10,10))
ax = fig.add_subplot(111)
#plot to first axes
ax.plot(x,y,color='blue',label="label1")
ax.set_ylim(0,50000)
ax.set_ylabel('ylabel1')
ax.set_xticklabels(ax.get_xticklabels(),rotation=90)
#create twin axes
ax2=ax.twinx()
#plot to twin axes
ax2.plot(x,y2,color='purple',label='label2')
ax2.set_ylabel('ylabel2')
h1, l1 = ax.get_legend_handles_labels()
h2, l2 = ax2.get_legend_handles_labels()
ax.legend(handles=h1+h2, labels=l1+l2, loc='upper right')
plt.show()