matplotlib画波动很小的图

今天测试一个画图时，有一个很神奇的发现

import matplotlib.pyplot as plt
import numpy as np 
import pandas as pd 
x = [10,20,30,40,50,60,70,80]

y=[-23.99534833975495,
 -23.9999998600783,
 -24.000000070633167,
 -24.000000068469788,
 -24.00000006672905,
 -24.000000065189436,
 -24.00000006373562,
 -24.00000006232896]

plt.plot(x,y,color="r")
plt.show()

结果如下

看左上角那个标识，-2.399e1到底是什么鬼，我的y值波动很小的呀，为什么还是会出现这种突然下降的情况，这个就是matplotlib默认的问题了，我查了之后大致是可以认为的，-2.399e1就是顶部的标度(即x=10)时，往下的-0.006代表的真实值就是-2.399e1-0.006，依次类推，因为最大值和最小值刚好差0.005左右

正确的画图思路是下面的，假如y轴的显示范围

import matplotlib.pyplot as plt
import numpy as np 
import pandas as pd 
x = [10,20,30,40,50,60,70,80]

y=[-23.99534833975495,
 -23.9999998600783,
 -24.000000070633167,
 -24.000000068469788,
 -24.00000006672905,
 -24.000000065189436,
 -24.00000006373562,
 -24.00000006232896]

plt.ylim([-25,-23])
plt.plot(x,y,color="r")

plt.show()

结果如下

这个结果才是我的预期

注意这种标识不是所有情况都会出现

import matplotlib.pyplot as plt
import numpy as np 
import pandas as pd 
x = [5,10,20,30,40,50,60,70,80]

y=[-3.99534833975495,
 -3.99534833975495,
 -3.9999998600783,
 -4.000000070633167,
 -4.000000068469788,
 -4.00000006672905,
 -4.000000065189436,
 -4.00000006373562,
 -4.00000006232896]

#plt.ylim([-25,-23])
plt.plot(x,y,color="r")

plt.show()

结果如下

import matplotlib.pyplot as plt
import numpy as np 
import pandas as pd 
x = [5,10,20,30,40,50,60,70,80]

y=[-13.99534833975495,
 -13.99534833975495,
 -13.9999998600783,
 -14.000000070633167,
 -14.000000068469788,
 -14.00000006672905,
 -14.000000065189436,
 -14.00000006373562,
 -14.00000006232896]

#plt.ylim([-25,-23])
plt.plot(x,y,color="r")

plt.show()

结果如下

散点图也会出现这种情况

import matplotlib.pyplot as plt
import numpy as np 
import pandas as pd 
x = [10,20,30,40,50,60,70,80]

y=[-3.99534833975495,
 -3.9999998600783,
 -24.000000070633167,
 -24.000000068469788,
 -24.00000006672905,
 -24.000000065189436,
 -24.00000006373562,
 -24.00000006232896]


plt.scatter(x,y,color="r")
plt.show()

norm_vector=[2.663689859547342e-08,
7.999154916053879e-09,
7.99915525716199e-09,
7.999155333718048e-09]
plt.figure(figsize=(10,12))
plt.grid()
plt.yscale('log')
plt.scatter([1,2,3,4],norm_vector)