在 Matplotlib 中使用左右轴绘制条形图和线形图

Question

给一个数据框如下：

          date    gdp  tertiary_industry  gdp_growth  tertiary_industry_growth
0    2015/3/31   3768               2508        10.3                      11.3
1    2015/6/30   8285               5483        10.9                      12.0
2    2015/9/30  12983               8586        11.5                      12.7
3   2015/12/31  18100              12086        10.5                      13.2
4    2016/3/31   4118               2813        13.5                      14.6
5    2016/6/30   8844               6020        13.3                      14.3
6    2016/9/30  14038               9513        14.4                      13.9
7   2016/12/31  19547              13557        16.3                      13.3
8    2017/3/31   4692               3285        13.3                      12.4
9    2017/6/30   9891               6881        12.9                      12.5
10   2017/9/30  15509              10689        12.7                      12.3
11  2017/12/31  21503              15254        14.8                      12.7
12   2018/3/31   4954               3499        12.4                      11.3
13   2018/6/30  10653               7520        12.9                      12.4
14   2018/9/30  16708              11697        13.5                      13.0
15  2018/12/31  22859              16402        14.0                      13.2
16   2019/3/31   5508               3983        13.5                      13.9
17   2019/6/30  11756               8556        10.2                      13.4
18   2019/9/30  17869              12765        10.2                      14.8
19  2019/12/31  23629              16923        11.6                      15.2
20   2020/3/31   5229               3968        11.9                      14.9

我已应用以下代码为gdp 和tertiary_industry 绘制条形图。

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import matplotlib.ticker as ticker
import matplotlib.style as style
style.available
style.use('fivethirtyeight')
from pylab import rcParams
plt.rcParams["figure.figsize"] = (20, 10)
plt.rcParams['font.sans-serif']=['SimHei'] 
plt.rcParams['axes.unicode_minus']=False 
import matplotlib
matplotlib.matplotlib_fname() 
plt.rcParams.update({'font.size': 25})

colors = ['#c23531','#2f4554', '#61a0a8', '#d48265', '#91c7ae','#749f83', '#ca8622', '#bda29a', '#6e7074', '#546570', '#c4ccd3']

df = df.sort_values(by = 'date')
df['date'] = pd.to_datetime(df['date']).dt.to_period('M')

df = df.set_index('date')
df.columns
cols = ['gdp', 'tertiary_industry']
df[cols] = df[cols].apply(pd.to_numeric, errors='coerce')
color_dict = dict(zip(cols, colors))
plt.figure(figsize=(20, 10))
df[cols].plot(color=[color_dict.get(x, '#333333') for x in df.columns], kind='bar', width=0.8)
plt.xticks(rotation=45)
plt.xlabel("")
plt.ylabel("million dollar")
fig = plt.gcf()
plt.show()
plt.draw()
fig.savefig("./gdp.png", dpi=100, bbox_inches = 'tight')
plt.clf()

上面代码的输出：

现在我想使用线型和右轴在同一个图上绘制gdp_growth 和tertiary_industry_growth，它们是百分比值。

请注意，我想在代码中使用自定义颜色列表中的颜色，而不是默认颜色。

如何根据上面的代码做到这一点？非常感谢您的热心帮助。

Answer 1

这就是我要做的：

#convert to datetime
df['date'] = pd.to_datetime(df['date']).dt.to_period('M')

cols = ['gdp', 'tertiary_industry']
colors = ['#c23531','#2f4554', '#61a0a8', '#d48265', '#91c7ae','#749f83', '#ca8622', '#bda29a', '#6e7074', '#546570', '#c4ccd3']

df[cols] = df[cols].apply(pd.to_numeric, errors='coerce')

# modify color_dict here:
color_dict = dict(zip(cols, colors))


# initialize an axis instance
fig, ax = plt.subplots(figsize=(10,6))

# plot on new instance
df.plot.bar(y=cols,ax=ax, 
            color=[color_dict.get(x, '#333333') for x in cols])

# create a twinx axis
ax1 = ax.twinx()

# plot the other two columns on this axis
df.plot.line(y=['gdp_growth','tertiary_industry_growth'], ax=ax1,
            color=[color_dict.get(x, '#333333') for x in line_cols])
ax.set_xticklabels(df['date'])

# set y-axes labels:
ax.set_ylabel('Million Dollar')
ax1.set_ylabel('%')

# set x-axis label
ax.set_xlabel('Quarter')

plt.show()

输出：

如果您将上述代码中的两个colors=[...] 替换为原始color=[color_dict.get(x, '#333333') for x in df.columns]，您将得到