18. Pandas 面板数据#
除 Anaconda 中已有的库外，本讲座还需要以下库：
!pip install --upgrade seaborn
我们使用以下导入语句。
import matplotlib.pyplot as plt
import matplotlib as mpl # i18n
import matplotlib.font_manager # i18n
FONTPATH = "_fonts/SourceHanSerifSC-SemiBold.otf" # i18n
mpl.font_manager.fontManager.addfont(FONTPATH) # i18n
mpl.rcParams['font.family'] = ['Source Han Serif SC'] # i18n
import seaborn as sns
sns.set_theme()
18.1. 概述#
在早先关于 pandas 的讲座中，我们研究了如何处理简单的数据集。
计量经济学家通常需要处理更复杂的数据集，例如面板数据。
常见任务包括：
导入数据、清洗数据并跨多个轴对其进行重塑。
从面板中选择时间序列或横截面。
对数据进行分组和汇总。
pandas（源自”panel”和”data”）包含强大且易于使用的工具，专门用于解决此类问题。
在下文中，我们将使用经合组织（OECD）的实际最低工资面板数据集来创建：
跨多个数据维度的汇总统计
数据集中各国平均最低工资的时间序列
按大洲划分的工资核密度估计
我们将首先从 CSV 文件中读取长格式面板数据，然后使用 pivot_table 对生成的 DataFrame 进行重塑，以构建 MultiIndex。
我们将使用 pandas 的 merge 函数向 DataFrame 添加更多细节，并使用 groupby 函数对数据进行汇总。
18.2. 数据切片与重塑#
我们将读入一个来自 OECD 的数据集，该数据集包含 32 个国家的实际最低工资，并将其赋值给 realwage。
可通过以下链接访问该数据集：
url1 = 'https://raw.githubusercontent.com/QuantEcon/lecture-python/master/source/_static/lecture_specific/pandas_panel/realwage.csv'
import pandas as pd
# 出于显示目的，设置最多显示 6 列
pd.set_option('display.max_columns', 6)
# 将小数位数减少到 2 位
pd.options.display.float_format = '{:,.2f}'.format
realwage = pd.read_csv(url1)
让我们看看数据的内容
realwage.head() # 显示前 5 行
| Unnamed: 0 | Time | Country | Series | Pay period | value | |
|---|---|---|---|---|---|---|
| 0 | 0 | 2006-01-01 | Ireland | In 2015 constant prices at 2015 USD PPPs | Annual | 17,132.44 |
| 1 | 1 | 2007-01-01 | Ireland | In 2015 constant prices at 2015 USD PPPs | Annual | 18,100.92 |
| 2 | 2 | 2008-01-01 | Ireland | In 2015 constant prices at 2015 USD PPPs | Annual | 17,747.41 |
| 3 | 3 | 2009-01-01 | Ireland | In 2015 constant prices at 2015 USD PPPs | Annual | 18,580.14 |
| 4 | 4 | 2010-01-01 | Ireland | In 2015 constant prices at 2015 USD PPPs | Annual | 18,755.83 |
数据目前为长格式，当数据有多个维度时，这种格式难以分析。
我们将使用 pivot_table 创建宽格式面板，并使用 MultiIndex 来处理高维数据。
pivot_table 的参数应指定数据（values）、索引（index）以及结果数据框中所需的列（columns）。
通过在 columns 中传入列表，我们可以在列轴上创建 MultiIndex
realwage = realwage.pivot_table(values='value',
index='Time',
columns=['Country', 'Series', 'Pay period'])
realwage.head()
| Country | Australia | ... | United States | ||||
|---|---|---|---|---|---|---|---|
| Series | In 2015 constant prices at 2015 USD PPPs | In 2015 constant prices at 2015 USD exchange rates | ... | In 2015 constant prices at 2015 USD PPPs | In 2015 constant prices at 2015 USD exchange rates | ||
| Pay period | Annual | Hourly | Annual | ... | Hourly | Annual | Hourly |
| Time | |||||||
| 2006-01-01 | 20,410.65 | 10.33 | 23,826.64 | ... | 6.05 | 12,594.40 | 6.05 |
| 2007-01-01 | 21,087.57 | 10.67 | 24,616.84 | ... | 6.24 | 12,974.40 | 6.24 |
| 2008-01-01 | 20,718.24 | 10.48 | 24,185.70 | ... | 6.78 | 14,097.56 | 6.78 |
| 2009-01-01 | 20,984.77 | 10.62 | 24,496.84 | ... | 7.58 | 15,756.42 | 7.58 |
| 2010-01-01 | 20,879.33 | 10.57 | 24,373.76 | ... | 7.88 | 16,391.31 | 7.88 |
5 rows × 128 columns
为了便于后续过滤时间序列数据，我们将索引转换为 DateTimeIndex
realwage.index = pd.to_datetime(realwage.index)
type(realwage.index)
pandas.core.indexes.datetimes.DatetimeIndex
列包含多层索引，称为 MultiIndex，层级按层次排列（国家 > 系列 > 支付周期）。
MultiIndex 是在 pandas 中管理面板数据最简单、最灵活的方式
type(realwage.columns)
pandas.core.indexes.multi.MultiIndex
realwage.columns.names
FrozenList(['Country', 'Series', 'Pay period'])
与之前一样，我们可以选择国家（MultiIndex 的顶层）
realwage['United States'].head()
| Series | In 2015 constant prices at 2015 USD PPPs | In 2015 constant prices at 2015 USD exchange rates | ||
|---|---|---|---|---|
| Pay period | Annual | Hourly | Annual | Hourly |
| Time | ||||
| 2006-01-01 | 12,594.40 | 6.05 | 12,594.40 | 6.05 |
| 2007-01-01 | 12,974.40 | 6.24 | 12,974.40 | 6.24 |
| 2008-01-01 | 14,097.56 | 6.78 | 14,097.56 | 6.78 |
| 2009-01-01 | 15,756.42 | 7.58 | 15,756.42 | 7.58 |
| 2010-01-01 | 16,391.31 | 7.88 | 16,391.31 | 7.88 |
在本讲座中，对 MultiIndex 层级进行堆叠和展开将被频繁使用，以将数据框重塑为所需格式。
.stack() 将列 MultiIndex 的最低层级旋转到行索引（.unstack() 方向相反——可以尝试一下）
realwage.stack(future_stack=True).head()
| Country | Australia | Belgium | ... | United Kingdom | United States | |||
|---|---|---|---|---|---|---|---|---|
| Series | In 2015 constant prices at 2015 USD PPPs | In 2015 constant prices at 2015 USD exchange rates | In 2015 constant prices at 2015 USD PPPs | ... | In 2015 constant prices at 2015 USD exchange rates | In 2015 constant prices at 2015 USD PPPs | In 2015 constant prices at 2015 USD exchange rates | |
| Time | Pay period | |||||||
| 2006-01-01 | Annual | 20,410.65 | 23,826.64 | 21,042.28 | ... | 20,376.32 | 12,594.40 | 12,594.40 |
| Hourly | 10.33 | 12.06 | 10.09 | ... | 9.81 | 6.05 | 6.05 | |
| 2007-01-01 | Annual | 21,087.57 | 24,616.84 | 21,310.05 | ... | 20,954.13 | 12,974.40 | 12,974.40 |
| Hourly | 10.67 | 12.46 | 10.22 | ... | 10.07 | 6.24 | 6.24 | |
| 2008-01-01 | Annual | 20,718.24 | 24,185.70 | 21,416.96 | ... | 20,902.87 | 14,097.56 | 14,097.56 |
5 rows × 64 columns
我们也可以传入参数来选择要堆叠的层级
realwage.stack(level='Country', future_stack=True).head() # pandas>3.0 之前需要 future_stack=True
| Series | In 2015 constant prices at 2015 USD PPPs | In 2015 constant prices at 2015 USD exchange rates | |||
|---|---|---|---|---|---|
| Pay period | Annual | Hourly | Annual | Hourly | |
| Time | Country | ||||
| 2006-01-01 | Australia | 20,410.65 | 10.33 | 23,826.64 | 12.06 |
| Belgium | 21,042.28 | 10.09 | 20,228.74 | 9.70 | |
| Brazil | 3,310.51 | 1.41 | 2,032.87 | 0.87 | |
| Canada | 13,649.69 | 6.56 | 14,335.12 | 6.89 | |
| Chile | 5,201.65 | 2.22 | 3,333.76 | 1.42 | |
使用 DatetimeIndex 可以方便地选择特定时间段。
选择某一年并堆叠 MultiIndex 的两个较低层级，可以创建面板数据的横截面
realwage.loc['2015'].stack(level=(1, 2), future_stack=True).transpose().head() # pandas>3.0 之前需要 future_stack=True
| Time | 2015-01-01 | |||
|---|---|---|---|---|
| Series | In 2015 constant prices at 2015 USD PPPs | In 2015 constant prices at 2015 USD exchange rates | ||
| Pay period | Annual | Hourly | Annual | Hourly |
| Country | ||||
| Australia | 21,715.53 | 10.99 | 25,349.90 | 12.83 |
| Belgium | 21,588.12 | 10.35 | 20,753.48 | 9.95 |
| Brazil | 4,628.63 | 2.00 | 2,842.28 | 1.21 |
| Canada | 16,536.83 | 7.95 | 17,367.24 | 8.35 |
| Chile | 6,633.56 | 2.80 | 4,251.49 | 1.81 |
在本讲座的其余部分，我们将使用一个包含各国和各时间段每小时实际最低工资的数据框，以 2015 年美元计价。
为了创建过滤后的数据框（realwage_f），我们可以使用 xs 方法在多级索引的较低层级中选择值，同时保留较高层级（此处为国家）
realwage_f = realwage.xs(('Hourly', 'In 2015 constant prices at 2015 USD exchange rates'),
level=('Pay period', 'Series'), axis=1)
realwage_f.head()
| Country | Australia | Belgium | Brazil | ... | Turkey | United Kingdom | United States |
|---|---|---|---|---|---|---|---|
| Time | |||||||
| 2006-01-01 | 12.06 | 9.70 | 0.87 | ... | 2.27 | 9.81 | 6.05 |
| 2007-01-01 | 12.46 | 9.82 | 0.92 | ... | 2.26 | 10.07 | 6.24 |
| 2008-01-01 | 12.24 | 9.87 | 0.96 | ... | 2.22 | 10.04 | 6.78 |
| 2009-01-01 | 12.40 | 10.21 | 1.03 | ... | 2.28 | 10.15 | 7.58 |
| 2010-01-01 | 12.34 | 10.05 | 1.08 | ... | 2.30 | 9.96 | 7.88 |
5 rows × 32 columns
18.3. 合并数据框与填充 NaN 值#
与 SQL 等关系型数据库类似，pandas 内置了合并数据集的方法。
使用来自 WorldData.info 的国家信息，我们将通过 merge 函数向 realwage_f 添加每个国家所属的大洲。
可通过以下链接访问该数据集：
url2 = 'https://raw.githubusercontent.com/QuantEcon/lecture-python/master/source/_static/lecture_specific/pandas_panel/countries.csv'
worlddata = pd.read_csv(url2, sep=';')
worlddata.head()
| Country (en) | Country (de) | Country (local) | ... | Deathrate | Life expectancy | Url | |
|---|---|---|---|---|---|---|---|
| 0 | Afghanistan | Afghanistan | Afganistan/Afqanestan | ... | 13.70 | 51.30 | https://www.laenderdaten.info/Asien/Afghanista... |
| 1 | Egypt | Ägypten | Misr | ... | 4.70 | 72.70 | https://www.laenderdaten.info/Afrika/Aegypten/... |
| 2 | Åland Islands | Ålandinseln | Åland | ... | 0.00 | 0.00 | https://www.laenderdaten.info/Europa/Aland/ind... |
| 3 | Albania | Albanien | Shqipëria | ... | 6.70 | 78.30 | https://www.laenderdaten.info/Europa/Albanien/... |
| 4 | Algeria | Algerien | Al-Jaza’ir/Algérie | ... | 4.30 | 76.80 | https://www.laenderdaten.info/Afrika/Algerien/... |
5 rows × 17 columns
首先，我们从 worlddata 中只选取国家和大洲变量，并将列重命名为”Country”
worlddata = worlddata[['Country (en)', 'Continent']]
worlddata = worlddata.rename(columns={'Country (en)': 'Country'})
worlddata.head()
| Country | Continent | |
|---|---|---|
| 0 | Afghanistan | Asia |
| 1 | Egypt | Africa |
| 2 | Åland Islands | Europe |
| 3 | Albania | Europe |
| 4 | Algeria | Africa |
我们希望将新的数据框 worlddata 与 realwage_f 合并。
pandas 的 merge 函数允许按行连接数据框。
我们的数据框将通过国家名称进行合并，因此需要对 realwage_f 进行转置，使两个数据框中的行均对应国家名称
realwage_f.transpose().head()
| Time | 2006-01-01 | 2007-01-01 | 2008-01-01 | ... | 2014-01-01 | 2015-01-01 | 2016-01-01 |
|---|---|---|---|---|---|---|---|
| Country | |||||||
| Australia | 12.06 | 12.46 | 12.24 | ... | 12.67 | 12.83 | 12.98 |
| Belgium | 9.70 | 9.82 | 9.87 | ... | 10.01 | 9.95 | 9.76 |
| Brazil | 0.87 | 0.92 | 0.96 | ... | 1.21 | 1.21 | 1.24 |
| Canada | 6.89 | 6.96 | 7.24 | ... | 8.22 | 8.35 | 8.48 |
| Chile | 1.42 | 1.45 | 1.44 | ... | 1.76 | 1.81 | 1.91 |
5 rows × 11 columns
我们可以使用左连接、右连接、内连接或外连接来合并数据集：
左连接只包含左侧数据集中的国家
右连接只包含右侧数据集中的国家
外连接包含左侧或右侧数据集中的所有国家
内连接只包含两个数据集共有的国家
默认情况下，merge 使用内连接。
这里我们将传入 how='left'，以保留 realwage_f 中的所有国家，但丢弃 worlddata 中在 realwage_f 中没有对应数据条目的国家。
下图中的红色阴影部分说明了这一点
我们还需要指定每个数据框中国家名称的位置，这将是用于合并数据框的 key。
我们的”左”数据框（realwage_f.transpose()）在索引中包含国家名称，因此设置 left_index=True。
我们的”右”数据框（worlddata）在”Country”列中包含国家名称，因此设置 right_on='Country'
merged = pd.merge(realwage_f.transpose(), worlddata,
how='left', left_index=True, right_on='Country')
merged.head()
| 2006-01-01 00:00:00 | 2007-01-01 00:00:00 | 2008-01-01 00:00:00 | ... | 2016-01-01 00:00:00 | Country | Continent | |
|---|---|---|---|---|---|---|---|
| 17.00 | 12.06 | 12.46 | 12.24 | ... | 12.98 | Australia | Australia |
| 23.00 | 9.70 | 9.82 | 9.87 | ... | 9.76 | Belgium | Europe |
| 32.00 | 0.87 | 0.92 | 0.96 | ... | 1.24 | Brazil | South America |
| 100.00 | 6.89 | 6.96 | 7.24 | ... | 8.48 | Canada | North America |
| 38.00 | 1.42 | 1.45 | 1.44 | ... | 1.91 | Chile | South America |
5 rows × 13 columns
出现在 realwage_f 中但未出现在 worlddata 中的国家，其 Continent 列将显示 NaN。
为检查是否存在这种情况，我们可以对 Continent 列使用 .isnull() 并过滤合并后的数据框
merged[merged['Continent'].isnull()]
| 2006-01-01 00:00:00 | 2007-01-01 00:00:00 | 2008-01-01 00:00:00 | ... | 2016-01-01 00:00:00 | Country | Continent | |
|---|---|---|---|---|---|---|---|
| NaN | 3.42 | 3.74 | 3.87 | ... | 5.28 | Korea | NaN |
| NaN | 0.23 | 0.45 | 0.39 | ... | 0.55 | Russian Federation | NaN |
| NaN | 1.50 | 1.64 | 1.71 | ... | 2.08 | Slovak Republic | NaN |
3 rows × 13 columns
有三个缺失值！
处理 NaN 值的一种方法是创建一个包含这些国家及其对应大洲的字典。
.map() 将把 merged['Country'] 中的国家与字典中的大洲进行匹配。
注意，不在字典中的国家将被映射为 NaN
missing_continents = {'Korea': 'Asia',
'Russian Federation': 'Europe',
'Slovak Republic': 'Europe'}
merged['Country'].map(missing_continents)
17.00 NaN
23.00 NaN
32.00 NaN
100.00 NaN
38.00 NaN
108.00 NaN
41.00 NaN
225.00 NaN
53.00 NaN
58.00 NaN
45.00 NaN
68.00 NaN
233.00 NaN
86.00 NaN
88.00 NaN
91.00 NaN
NaN Asia
117.00 NaN
122.00 NaN
123.00 NaN
138.00 NaN
153.00 NaN
151.00 NaN
174.00 NaN
175.00 NaN
NaN Europe
NaN Europe
198.00 NaN
200.00 NaN
227.00 NaN
241.00 NaN
240.00 NaN
Name: Country, dtype: object
我们不想用此映射覆盖整个系列。
.fillna() 只用映射填充 merged['Continent'] 中的 NaN 值，而保持列中其他值不变
merged['Continent'] = merged['Continent'].fillna(merged['Country'].map(missing_continents))
# 检查大洲是否被正确映射
merged[merged['Country'] == 'Korea']
| 2006-01-01 00:00:00 | 2007-01-01 00:00:00 | 2008-01-01 00:00:00 | ... | 2016-01-01 00:00:00 | Country | Continent | |
|---|---|---|---|---|---|---|---|
| NaN | 3.42 | 3.74 | 3.87 | ... | 5.28 | Korea | Asia |
1 rows × 13 columns
我们还将把美洲合并为单一大洲——这将使后续可视化效果更好。
为此，我们将使用 .replace() 并遍历要替换的大洲值列表
replace = ['Central America', 'North America', 'South America']
merged['Continent'] = merged['Continent'].replace(to_replace=replace, value='America')
现在我们已将所需的全部数据整合到一个 DataFrame 中，我们将使用 MultiIndex 将其重塑回面板形式。
我们还应使用 .sort_index() 对索引进行排序，以便后续能高效过滤数据框。
默认情况下，层级将从上到下排序
merged = merged.set_index(['Continent', 'Country']).sort_index()
merged.head()
| 2006-01-01 | 2007-01-01 | 2008-01-01 | ... | 2014-01-01 | 2015-01-01 | 2016-01-01 | ||
|---|---|---|---|---|---|---|---|---|
| Continent | Country | |||||||
| America | Brazil | 0.87 | 0.92 | 0.96 | ... | 1.21 | 1.21 | 1.24 |
| Canada | 6.89 | 6.96 | 7.24 | ... | 8.22 | 8.35 | 8.48 | |
| Chile | 1.42 | 1.45 | 1.44 | ... | 1.76 | 1.81 | 1.91 | |
| Colombia | 1.01 | 1.02 | 1.01 | ... | 1.13 | 1.13 | 1.12 | |
| Costa Rica | NaN | NaN | NaN | ... | 2.41 | 2.56 | 2.63 |
5 rows × 11 columns
合并过程中，由于合并的列不是日期时间格式，我们丢失了 DatetimeIndex
merged.columns
Index([2006-01-01 00:00:00, 2007-01-01 00:00:00, 2008-01-01 00:00:00,
2009-01-01 00:00:00, 2010-01-01 00:00:00, 2011-01-01 00:00:00,
2012-01-01 00:00:00, 2013-01-01 00:00:00, 2014-01-01 00:00:00,
2015-01-01 00:00:00, 2016-01-01 00:00:00],
dtype='object')
现在我们已将合并后的列设置为索引，可以使用 .to_datetime() 重新创建 DatetimeIndex
merged.columns = pd.to_datetime(merged.columns)
merged.columns = merged.columns.rename('Time')
merged.columns
DatetimeIndex(['2006-01-01', '2007-01-01', '2008-01-01', '2009-01-01',
'2010-01-01', '2011-01-01', '2012-01-01', '2013-01-01',
'2014-01-01', '2015-01-01', '2016-01-01'],
dtype='datetime64[ns]', name='Time', freq=None)
DatetimeIndex 在行轴中工作更为顺畅，因此我们将对 merged 进行转置
merged = merged.transpose()
merged.head()
| Continent | America | ... | Europe | ||||
|---|---|---|---|---|---|---|---|
| Country | Brazil | Canada | Chile | ... | Slovenia | Spain | United Kingdom |
| Time | |||||||
| 2006-01-01 | 0.87 | 6.89 | 1.42 | ... | 3.92 | 3.99 | 9.81 |
| 2007-01-01 | 0.92 | 6.96 | 1.45 | ... | 3.88 | 4.10 | 10.07 |
| 2008-01-01 | 0.96 | 7.24 | 1.44 | ... | 3.96 | 4.14 | 10.04 |
| 2009-01-01 | 1.03 | 7.67 | 1.52 | ... | 4.08 | 4.32 | 10.15 |
| 2010-01-01 | 1.08 | 7.94 | 1.56 | ... | 4.81 | 4.30 | 9.96 |
5 rows × 32 columns
18.4. 数据分组与汇总#
对于理解大型面板数据集而言，数据分组和汇总特别有用。
汇总数据的一种简单方法是对数据框调用聚合方法，例如 .mean() 或 .max()。
例如，我们可以计算 2006 年至 2016 年间每个国家的平均实际最低工资（默认按行聚合）
merged.mean().head(10)
Continent Country
America Brazil 1.09
Canada 7.82
Chile 1.62
Colombia 1.07
Costa Rica 2.53
Mexico 0.53
United States 7.15
Asia Israel 5.95
Japan 6.18
Korea 4.22
dtype: float64
利用此系列，我们可以绘制数据集中每个国家过去十年的平均实际最低工资
merged.mean().sort_values(ascending=False).plot(kind='bar',
title="2006 - 2016 年平均实际最低工资")
# 设置国家标签
country_labels = merged.mean().sort_values(ascending=False).index.get_level_values('Country').tolist()
plt.xticks(range(0, len(country_labels)), country_labels)
plt.xlabel('国家')
plt.show()
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 22269 (\N{CJK UNIFIED IDEOGRAPH-56FD}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 23478 (\N{CJK UNIFIED IDEOGRAPH-5BB6}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24180 (\N{CJK UNIFIED IDEOGRAPH-5E74}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24179 (\N{CJK UNIFIED IDEOGRAPH-5E73}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 22343 (\N{CJK UNIFIED IDEOGRAPH-5747}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 23454 (\N{CJK UNIFIED IDEOGRAPH-5B9E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 38469 (\N{CJK UNIFIED IDEOGRAPH-9645}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 26368 (\N{CJK UNIFIED IDEOGRAPH-6700}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20302 (\N{CJK UNIFIED IDEOGRAPH-4F4E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24037 (\N{CJK UNIFIED IDEOGRAPH-5DE5}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 36164 (\N{CJK UNIFIED IDEOGRAPH-8D44}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
向 .mean() 传入 axis=1 将按列聚合（给出所有国家随时间变化的平均最低工资）
merged.mean(axis=1).head()
Time
2006-01-01 4.69
2007-01-01 4.84
2008-01-01 4.90
2009-01-01 5.08
2010-01-01 5.11
dtype: float64
我们可以将此时间序列绘制为折线图
merged.mean(axis=1).plot()
plt.title('2006 - 2016 年平均实际最低工资')
plt.ylabel('2015 年美元')
plt.xlabel('年份')
plt.show()
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24180 (\N{CJK UNIFIED IDEOGRAPH-5E74}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20221 (\N{CJK UNIFIED IDEOGRAPH-4EFD}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 32654 (\N{CJK UNIFIED IDEOGRAPH-7F8E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20803 (\N{CJK UNIFIED IDEOGRAPH-5143}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24179 (\N{CJK UNIFIED IDEOGRAPH-5E73}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 22343 (\N{CJK UNIFIED IDEOGRAPH-5747}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 23454 (\N{CJK UNIFIED IDEOGRAPH-5B9E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 38469 (\N{CJK UNIFIED IDEOGRAPH-9645}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 26368 (\N{CJK UNIFIED IDEOGRAPH-6700}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20302 (\N{CJK UNIFIED IDEOGRAPH-4F4E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24037 (\N{CJK UNIFIED IDEOGRAPH-5DE5}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 36164 (\N{CJK UNIFIED IDEOGRAPH-8D44}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24180 (\N{CJK UNIFIED IDEOGRAPH-5E74}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20221 (\N{CJK UNIFIED IDEOGRAPH-4EFD}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 32654 (\N{CJK UNIFIED IDEOGRAPH-7F8E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20803 (\N{CJK UNIFIED IDEOGRAPH-5143}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24179 (\N{CJK UNIFIED IDEOGRAPH-5E73}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 22343 (\N{CJK UNIFIED IDEOGRAPH-5747}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 23454 (\N{CJK UNIFIED IDEOGRAPH-5B9E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 38469 (\N{CJK UNIFIED IDEOGRAPH-9645}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 26368 (\N{CJK UNIFIED IDEOGRAPH-6700}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20302 (\N{CJK UNIFIED IDEOGRAPH-4F4E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24037 (\N{CJK UNIFIED IDEOGRAPH-5DE5}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 36164 (\N{CJK UNIFIED IDEOGRAPH-8D44}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24180 (\N{CJK UNIFIED IDEOGRAPH-5E74}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20221 (\N{CJK UNIFIED IDEOGRAPH-4EFD}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 32654 (\N{CJK UNIFIED IDEOGRAPH-7F8E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20803 (\N{CJK UNIFIED IDEOGRAPH-5143}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24179 (\N{CJK UNIFIED IDEOGRAPH-5E73}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 22343 (\N{CJK UNIFIED IDEOGRAPH-5747}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 23454 (\N{CJK UNIFIED IDEOGRAPH-5B9E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 38469 (\N{CJK UNIFIED IDEOGRAPH-9645}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 26368 (\N{CJK UNIFIED IDEOGRAPH-6700}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20302 (\N{CJK UNIFIED IDEOGRAPH-4F4E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24037 (\N{CJK UNIFIED IDEOGRAPH-5DE5}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 36164 (\N{CJK UNIFIED IDEOGRAPH-8D44}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
我们还可以指定 MultiIndex 的某个层级（在列轴上）进行聚合。
对于 groupby，由于 pandas 已弃用在 groupby 方法中使用 axis=1，我们需要使用 .T 将列转置为行。
merged.T.groupby(level='Continent').mean().head()
| Time | 2006-01-01 | 2007-01-01 | 2008-01-01 | ... | 2014-01-01 | 2015-01-01 | 2016-01-01 |
|---|---|---|---|---|---|---|---|
| Continent | |||||||
| America | 2.80 | 2.85 | 2.99 | ... | 3.22 | 3.26 | 3.30 |
| Asia | 4.29 | 4.44 | 4.45 | ... | 4.86 | 5.10 | 5.44 |
| Australia | 10.25 | 10.73 | 10.76 | ... | 11.25 | 11.52 | 11.73 |
| Europe | 4.80 | 4.94 | 4.99 | ... | 5.17 | 5.48 | 5.57 |
4 rows × 11 columns
我们可以将各大洲的平均最低工资绘制为时间序列
merged.T.groupby(level='Continent').mean().T.plot()
plt.title('平均实际最低工资')
plt.ylabel('2015 年美元')
plt.xlabel('年份')
plt.show()
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24180 (\N{CJK UNIFIED IDEOGRAPH-5E74}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20221 (\N{CJK UNIFIED IDEOGRAPH-4EFD}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 32654 (\N{CJK UNIFIED IDEOGRAPH-7F8E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20803 (\N{CJK UNIFIED IDEOGRAPH-5143}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24179 (\N{CJK UNIFIED IDEOGRAPH-5E73}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 22343 (\N{CJK UNIFIED IDEOGRAPH-5747}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 23454 (\N{CJK UNIFIED IDEOGRAPH-5B9E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 38469 (\N{CJK UNIFIED IDEOGRAPH-9645}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 26368 (\N{CJK UNIFIED IDEOGRAPH-6700}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20302 (\N{CJK UNIFIED IDEOGRAPH-4F4E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24037 (\N{CJK UNIFIED IDEOGRAPH-5DE5}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 36164 (\N{CJK UNIFIED IDEOGRAPH-8D44}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24180 (\N{CJK UNIFIED IDEOGRAPH-5E74}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20221 (\N{CJK UNIFIED IDEOGRAPH-4EFD}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 32654 (\N{CJK UNIFIED IDEOGRAPH-7F8E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20803 (\N{CJK UNIFIED IDEOGRAPH-5143}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24179 (\N{CJK UNIFIED IDEOGRAPH-5E73}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 22343 (\N{CJK UNIFIED IDEOGRAPH-5747}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 23454 (\N{CJK UNIFIED IDEOGRAPH-5B9E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 38469 (\N{CJK UNIFIED IDEOGRAPH-9645}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 26368 (\N{CJK UNIFIED IDEOGRAPH-6700}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20302 (\N{CJK UNIFIED IDEOGRAPH-4F4E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24037 (\N{CJK UNIFIED IDEOGRAPH-5DE5}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 36164 (\N{CJK UNIFIED IDEOGRAPH-8D44}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24180 (\N{CJK UNIFIED IDEOGRAPH-5E74}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20221 (\N{CJK UNIFIED IDEOGRAPH-4EFD}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 32654 (\N{CJK UNIFIED IDEOGRAPH-7F8E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20803 (\N{CJK UNIFIED IDEOGRAPH-5143}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24179 (\N{CJK UNIFIED IDEOGRAPH-5E73}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 22343 (\N{CJK UNIFIED IDEOGRAPH-5747}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 23454 (\N{CJK UNIFIED IDEOGRAPH-5B9E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 38469 (\N{CJK UNIFIED IDEOGRAPH-9645}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 26368 (\N{CJK UNIFIED IDEOGRAPH-6700}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20302 (\N{CJK UNIFIED IDEOGRAPH-4F4E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24037 (\N{CJK UNIFIED IDEOGRAPH-5DE5}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 36164 (\N{CJK UNIFIED IDEOGRAPH-8D44}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
出于绘图目的，我们将删除澳大利亚这一大洲
merged = merged.drop('Australia', level='Continent', axis=1)
merged.T.groupby(level='Continent').mean().T.plot()
plt.title('平均实际最低工资')
plt.ylabel('2015 年美元')
plt.xlabel('年份')
plt.show()
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24180 (\N{CJK UNIFIED IDEOGRAPH-5E74}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20221 (\N{CJK UNIFIED IDEOGRAPH-4EFD}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 32654 (\N{CJK UNIFIED IDEOGRAPH-7F8E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20803 (\N{CJK UNIFIED IDEOGRAPH-5143}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24179 (\N{CJK UNIFIED IDEOGRAPH-5E73}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 22343 (\N{CJK UNIFIED IDEOGRAPH-5747}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 23454 (\N{CJK UNIFIED IDEOGRAPH-5B9E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 38469 (\N{CJK UNIFIED IDEOGRAPH-9645}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 26368 (\N{CJK UNIFIED IDEOGRAPH-6700}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20302 (\N{CJK UNIFIED IDEOGRAPH-4F4E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24037 (\N{CJK UNIFIED IDEOGRAPH-5DE5}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 36164 (\N{CJK UNIFIED IDEOGRAPH-8D44}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24180 (\N{CJK UNIFIED IDEOGRAPH-5E74}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20221 (\N{CJK UNIFIED IDEOGRAPH-4EFD}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 32654 (\N{CJK UNIFIED IDEOGRAPH-7F8E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20803 (\N{CJK UNIFIED IDEOGRAPH-5143}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24179 (\N{CJK UNIFIED IDEOGRAPH-5E73}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 22343 (\N{CJK UNIFIED IDEOGRAPH-5747}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 23454 (\N{CJK UNIFIED IDEOGRAPH-5B9E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 38469 (\N{CJK UNIFIED IDEOGRAPH-9645}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 26368 (\N{CJK UNIFIED IDEOGRAPH-6700}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20302 (\N{CJK UNIFIED IDEOGRAPH-4F4E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24037 (\N{CJK UNIFIED IDEOGRAPH-5DE5}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 36164 (\N{CJK UNIFIED IDEOGRAPH-8D44}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24180 (\N{CJK UNIFIED IDEOGRAPH-5E74}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20221 (\N{CJK UNIFIED IDEOGRAPH-4EFD}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 32654 (\N{CJK UNIFIED IDEOGRAPH-7F8E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20803 (\N{CJK UNIFIED IDEOGRAPH-5143}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24179 (\N{CJK UNIFIED IDEOGRAPH-5E73}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 22343 (\N{CJK UNIFIED IDEOGRAPH-5747}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 23454 (\N{CJK UNIFIED IDEOGRAPH-5B9E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 38469 (\N{CJK UNIFIED IDEOGRAPH-9645}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 26368 (\N{CJK UNIFIED IDEOGRAPH-6700}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20302 (\N{CJK UNIFIED IDEOGRAPH-4F4E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24037 (\N{CJK UNIFIED IDEOGRAPH-5DE5}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 36164 (\N{CJK UNIFIED IDEOGRAPH-8D44}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
.describe() 可用于快速检索多个常用汇总统计量
merged.stack(future_stack=True).describe()
| Continent | America | Asia | Europe |
|---|---|---|---|
| count | 69.00 | 44.00 | 200.00 |
| mean | 3.19 | 4.70 | 5.15 |
| std | 3.02 | 1.56 | 3.82 |
| min | 0.52 | 2.22 | 0.23 |
| 25% | 1.03 | 3.37 | 2.02 |
| 50% | 1.44 | 5.48 | 3.54 |
| 75% | 6.96 | 5.95 | 9.70 |
| max | 8.48 | 6.65 | 12.39 |
这是使用 groupby 的一种简化方式。
使用 groupby 通常遵循”拆分-应用-合并”流程：
拆分：根据一个或多个键对数据进行分组
应用：对每个组独立调用函数
合并：将函数调用的结果合并到新的数据结构中
groupby 方法完成此流程的第一步，创建一个新的 DataFrameGroupBy 对象，其中数据被分成若干组。
让我们再次按大洲对 merged 进行拆分，这次使用 groupby 函数，并将结果对象命名为 grouped
grouped = merged.T.groupby(level='Continent')
grouped
<pandas.core.groupby.generic.DataFrameGroupBy object at 0x7faea32d6780>
对对象调用聚合方法会对每个组应用该函数，结果将合并到新的数据结构中。
例如，我们可以使用 .size() 返回数据集中每个大洲的国家数量。
在此情况下，新的数据结构是一个 Series
grouped.size()
Continent
America 7
Asia 4
Europe 19
dtype: int64
通过调用 .get_group() 只返回单个组中的国家，我们可以为每个大洲创建 2016 年实际最低工资分布的核密度估计。
grouped.groups.keys() 将返回 groupby 对象中的键
continents = grouped.groups.keys()
for continent in continents:
sns.kdeplot(grouped.get_group(continent).T.loc['2015'].unstack(), label=continent, fill=True)
plt.title('2015 年实际最低工资')
plt.xlabel('美元')
plt.legend()
plt.show()
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 32654 (\N{CJK UNIFIED IDEOGRAPH-7F8E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20803 (\N{CJK UNIFIED IDEOGRAPH-5143}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24180 (\N{CJK UNIFIED IDEOGRAPH-5E74}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 23454 (\N{CJK UNIFIED IDEOGRAPH-5B9E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 38469 (\N{CJK UNIFIED IDEOGRAPH-9645}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 26368 (\N{CJK UNIFIED IDEOGRAPH-6700}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20302 (\N{CJK UNIFIED IDEOGRAPH-4F4E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24037 (\N{CJK UNIFIED IDEOGRAPH-5DE5}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 36164 (\N{CJK UNIFIED IDEOGRAPH-8D44}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
18.5. 结语#
本讲座介绍了 pandas 的一些高级功能，包括多级索引、合并、分组和绘图。
在面板数据分析中，其他可能有用的工具包括 xarray，这是一个将 pandas 扩展到 N 维数据结构的 Python 包。
18.6. 练习#
Exercise 18.1
在这些练习中，你将使用来自 Eurostat 的欧洲按年龄和性别划分的就业率数据集。
可通过以下链接访问该数据集：
url3 = 'https://raw.githubusercontent.com/QuantEcon/lecture-python/master/source/_static/lecture_specific/pandas_panel/employ.csv'
读入 CSV 文件会返回长格式的面板数据集。使用 .pivot_table() 构建列中带有 MultiIndex 的宽格式数据框。
首先探索数据框以及 MultiIndex 层级中可用的变量。
编写一个程序，快速返回 MultiIndex 中的所有值。
Solution
employ = pd.read_csv(url3)
employ = employ.pivot_table(values='Value',
index=['DATE'],
columns=['UNIT','AGE', 'SEX', 'INDIC_EM', 'GEO'])
employ.index = pd.to_datetime(employ.index) # 确保日期为 datetime 格式
employ.head()
| UNIT | Percentage of total population | ... | Thousand persons | ||||
|---|---|---|---|---|---|---|---|
| AGE | From 15 to 24 years | ... | From 55 to 64 years | ||||
| SEX | Females | ... | Total | ||||
| INDIC_EM | Active population | ... | Total employment (resident population concept - LFS) | ||||
| GEO | Austria | Belgium | Bulgaria | ... | Switzerland | Turkey | United Kingdom |
| DATE | |||||||
| 2007-01-01 | 56.00 | 31.60 | 26.00 | ... | NaN | 1,282.00 | 4,131.00 |
| 2008-01-01 | 56.20 | 30.80 | 26.10 | ... | NaN | 1,354.00 | 4,204.00 |
| 2009-01-01 | 56.20 | 29.90 | 24.80 | ... | NaN | 1,449.00 | 4,193.00 |
| 2010-01-01 | 54.00 | 29.80 | 26.60 | ... | 640.00 | 1,583.00 | 4,186.00 |
| 2011-01-01 | 54.80 | 29.80 | 24.80 | ... | 661.00 | 1,760.00 | 4,164.00 |
5 rows × 1440 columns
这是一个大型数据集，因此探索可用的层级和变量非常有用
employ.columns.names
FrozenList(['UNIT', 'AGE', 'SEX', 'INDIC_EM', 'GEO'])
层级中的变量可以通过循环快速获取
for name in employ.columns.names:
print(name, employ.columns.get_level_values(name).unique())
UNIT Index(['Percentage of total population', 'Thousand persons'], dtype='object', name='UNIT')
AGE Index(['From 15 to 24 years', 'From 25 to 54 years', 'From 55 to 64 years'], dtype='object', name='AGE')
SEX Index(['Females', 'Males', 'Total'], dtype='object', name='SEX')
INDIC_EM Index(['Active population', 'Total employment (resident population concept - LFS)'], dtype='object', name='INDIC_EM')
GEO Index(['Austria', 'Belgium', 'Bulgaria', 'Croatia', 'Cyprus', 'Czech Republic',
'Denmark', 'Estonia', 'Euro area (17 countries)',
'Euro area (18 countries)', 'Euro area (19 countries)',
'European Union (15 countries)', 'European Union (27 countries)',
'European Union (28 countries)', 'Finland',
'Former Yugoslav Republic of Macedonia, the', 'France',
'France (metropolitan)',
'Germany (until 1990 former territory of the FRG)', 'Greece', 'Hungary',
'Iceland', 'Ireland', 'Italy', 'Latvia', 'Lithuania', 'Luxembourg',
'Malta', 'Netherlands', 'Norway', 'Poland', 'Portugal', 'Romania',
'Slovakia', 'Slovenia', 'Spain', 'Sweden', 'Switzerland', 'Turkey',
'United Kingdom'],
dtype='object', name='GEO')
Exercise 18.2
对上述数据框进行过滤，只保留就业占”活跃人口”百分比的数据。
使用 seaborn 创建 2015 年按年龄组和性别划分的就业率分组箱线图。
Hint
GEO 包含区域和国家两类地理单元。
Solution
为便于按国家过滤，将 GEO 交换到顶层并对 MultiIndex 进行排序
employ.columns = employ.columns.swaplevel(0,-1)
employ = employ.sort_index(axis=1)
我们需要去掉 GEO 中一些非国家的条目。
快速去掉欧盟区域的方法是使用列表推导式，找到 GEO 中以”Euro”开头的层级值
geo_list = employ.columns.get_level_values('GEO').unique().tolist()
countries = [x for x in geo_list if not x.startswith('Euro')]
employ = employ[countries]
employ.columns.get_level_values('GEO').unique()
Index(['Austria', 'Belgium', 'Bulgaria', 'Croatia', 'Cyprus', 'Czech Republic',
'Denmark', 'Estonia', 'Finland',
'Former Yugoslav Republic of Macedonia, the', 'France',
'France (metropolitan)',
'Germany (until 1990 former territory of the FRG)', 'Greece', 'Hungary',
'Iceland', 'Ireland', 'Italy', 'Latvia', 'Lithuania', 'Luxembourg',
'Malta', 'Netherlands', 'Norway', 'Poland', 'Portugal', 'Romania',
'Slovakia', 'Slovenia', 'Spain', 'Sweden', 'Switzerland', 'Turkey',
'United Kingdom'],
dtype='object', name='GEO')
从数据框中只选取活跃人口就业百分比
employ_f = employ.xs(('Percentage of total population', 'Active population'),
level=('UNIT', 'INDIC_EM'),
axis=1)
employ_f.head()
| GEO | Austria | ... | United Kingdom | ||||
|---|---|---|---|---|---|---|---|
| AGE | From 15 to 24 years | ... | From 55 to 64 years | ||||
| SEX | Females | Males | Total | ... | Females | Males | Total |
| DATE | |||||||
| 2007-01-01 | 56.00 | 62.90 | 59.40 | ... | 49.90 | 68.90 | 59.30 |
| 2008-01-01 | 56.20 | 62.90 | 59.50 | ... | 50.20 | 69.80 | 59.80 |
| 2009-01-01 | 56.20 | 62.90 | 59.50 | ... | 50.60 | 70.30 | 60.30 |
| 2010-01-01 | 54.00 | 62.60 | 58.30 | ... | 51.10 | 69.20 | 60.00 |
| 2011-01-01 | 54.80 | 63.60 | 59.20 | ... | 51.30 | 68.40 | 59.70 |
5 rows × 306 columns
在创建分组箱线图之前删除”Total”值
employ_f = employ_f.drop('Total', level='SEX', axis=1)
box = employ_f.loc['2015'].unstack().reset_index()
sns.boxplot(x="AGE", y=0, hue="SEX", data=box, palette=("husl"), showfliers=False)
plt.xlabel('')
plt.xticks(rotation=35)
plt.ylabel('占人口百分比（%）')
plt.title('欧洲就业状况（2015 年）')
plt.legend(bbox_to_anchor=(1,0.5))
plt.show()
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 21344 (\N{CJK UNIFIED IDEOGRAPH-5360}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20154 (\N{CJK UNIFIED IDEOGRAPH-4EBA}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 21475 (\N{CJK UNIFIED IDEOGRAPH-53E3}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 30334 (\N{CJK UNIFIED IDEOGRAPH-767E}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20998 (\N{CJK UNIFIED IDEOGRAPH-5206}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 27604 (\N{CJK UNIFIED IDEOGRAPH-6BD4}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 65288 (\N{FULLWIDTH LEFT PARENTHESIS}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 65289 (\N{FULLWIDTH RIGHT PARENTHESIS}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 27431 (\N{CJK UNIFIED IDEOGRAPH-6B27}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 27954 (\N{CJK UNIFIED IDEOGRAPH-6D32}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 23601 (\N{CJK UNIFIED IDEOGRAPH-5C31}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 19994 (\N{CJK UNIFIED IDEOGRAPH-4E1A}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 29366 (\N{CJK UNIFIED IDEOGRAPH-72B6}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 20917 (\N{CJK UNIFIED IDEOGRAPH-51B5}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
/home/runner/miniconda3/envs/quantecon/lib/python3.13/site-packages/IPython/core/pylabtools.py:170: UserWarning: Glyph 24180 (\N{CJK UNIFIED IDEOGRAPH-5E74}) missing from font(s) DejaVu Sans.
fig.canvas.print_figure(bytes_io, **kw)
