在分组后使用mean和std绘制错误栏

Question

我有以下数据框：

                    mean       std
insert quality                    
0.0    good     0.009905  0.003662
0.1    good     0.450190  0.281895
       poor     0.376818  0.306806
0.2    good     0.801856  0.243288
       poor     0.643859  0.322378
0.3    good     0.833235  0.172025
       poor     0.698972  0.263266
0.4    good     0.842288  0.141925
       poor     0.706708  0.241269
0.5    good     0.853634  0.118604
       poor     0.685716  0.208073
0.6    good     0.845496  0.118609
       poor     0.675907  0.207755
0.7    good     0.826335  0.133820
       poor     0.656934  0.222823
0.8    good     0.829707  0.130154
       poor     0.627111  0.213046
0.9    good     0.816636  0.137371
       poor     0.589331  0.232756
1.0    good     0.801211  0.147864
       poor     0.554589  0.245867

如果想要绘制2条曲线（点+误差），使用索引列"Insert"作为X轴并用"Quality" [好，差]区分两条曲线，我该怎么办？它们也应该有不同的颜色。

我有点被困，我制作了各种各样的情节。

Answer 1

您可以遍历df.groupby('quality')中的群组，并在每个群组中调用group.plot。

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

df = pd.DataFrame({
    'insert': [0.0, 0.1, 0.1, 0.2, 0.2, 0.3, 0.3, 0.4, 0.4, 0.5, 0.5, 0.6, 0.6,
    0.7, 0.7, 0.8, 0.8, 0.9, 0.9, 1.0, 1.0],
    'mean': [0.009905, 0.45019, 0.376818, 0.801856, 0.643859, 0.833235,
    0.698972, 0.842288, 0.706708, 0.853634, 0.685716, 0.845496, 0.675907,
    0.826335, 0.656934, 0.829707, 0.627111, 0.816636, 0.589331, 0.801211,
    0.554589],
    'quality': ['good', 'good', 'poor', 'good', 'poor', 'good', 'poor', 'good',
    'poor', 'good', 'poor', 'good', 'poor', 'good', 'poor', 'good', 'poor',
    'good', 'poor', 'good', 'poor'], 
    'std': [0.003662, 0.281895, 0.306806, 0.243288, 0.322378, 0.172025,
    0.263266, 0.141925, 0.241269, 0.118604, 0.208073, 0.118609, 0.207755,
    0.13382, 0.222823, 0.130154, 0.213046, 0.137371, 0.232756, 0.147864,
    0.245867]})

fig, ax = plt.subplots()    # 1

for key, group in df.groupby('quality'):
    group.plot('insert', 'mean', yerr='std', label=key, ax=ax)   # 2

plt.show()

要使两个图出现在同一轴上：

1. 创建您自己的axes对象，ax。
2. 在每次调用ax
时将axes参数设置为group.plot个对象

作为条形图可能看起来更好：

# fill in missing data with 0, so the bar plots are aligned
df = df.pivot(index='insert', columns='quality').fillna(0).stack().reset_index()

colors = ['green', 'red']
positions = [0, 1]

for group, color, pos in zip(df.groupby('quality'), colors, positions):
    key, group = group
    print(group)
    group.plot('insert', 'mean', yerr='std', kind='bar', width=0.4, label=key, 
               position=pos, color=color, alpha=0.5, ax=ax)

ax.set_xlim(-1, 11)  
plt.show()