如何使用多索引数据框绘制图表

Question

我正在尝试使用 pandas 和 matplotlib 在折线图中绘制各个国家的人均 GDP。我正在处理的数据框有一个关于国家和日期的多索引，例如：

这是我试图运行以生成情节的代码：

countries = df_no_na.index.levels[0].to_list()
dates = df_no_na.index.levels[1].to_list()
# Create Figure (empty canvas)
fig = plt.figure()
# Add set of axes to figure
axes = fig.add_axes([0.1, 0.1, 0.8, 0.8]) # left, bottom, width, height (range 0 to 1)
axes.plot(dates,df_no_na["GDP per capita"]["United States"])

我收到一个错误： ValueError: x 和 y 必须具有相同的第一维，但形状为 (61,) 和 (5,)

我认为这是因为df_no_na["GDP per capita"]["United States"] 返回：

date
2019-01-01    65297.517508
2018-01-01    62996.471285
2017-01-01    60062.222313
2016-01-01    57951.584082
2015-01-01    56839.381774
Name: GDP per capita, dtype: float64

如何从具有多索引的数据框中绘制数据，或者最好的方法是什么？

Answer 1

• 绘图的主要问题是将 DataFrame 设置为绘图 API 的正确形状。
  • 在这种情况下，重置索引可能是“最简单”，然后使用seaborn.lineplot 进行绘图。然而，这是离散的，而不是连续的数据，所以它“更好”显示为seaborn.barplot。

import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt

# test data
data = {('Aruba', '1986-01-01'): {'gdp': 6472.50202920407}, ('Aruba', '1987-01-01'): {'gdp': 7885.79654466735}, ('Aruba', '1988-01-01'): {'gdp': 9764.789978793291}, ('Aruba', '1989-01-01'): {'gdp': 11392.455810576399}, ('Aruba', '1990-01-01'): {'gdp': 12307.311737831398}, ('Aruba', '1991-01-01'): {'gdp': 13496.003142641799}, ('Aruba', '1992-01-01'): {'gdp': 14046.5037643078}, ('Aruba', '1993-01-01'): {'gdp': 14936.8272187795}, ('Aruba', '1994-01-01'): {'gdp': 16241.0465209443}, ('Aruba', '1995-01-01'): {'gdp': 16439.3563609282}, ('Aruba', '1996-01-01'): {'gdp': 16586.068435754198}, ('Aruba', '1997-01-01'): {'gdp': 17927.749635208602}, ('Aruba', '1998-01-01'): {'gdp': 19078.3431907515}, ('Aruba', '1999-01-01'): {'gdp': 19356.2033894901}, ('Aruba', '2000-01-01'): {'gdp': 20620.7006259175}, ('Aruba', '2001-01-01'): {'gdp': 20669.0319688645}, ('Aruba', '2002-01-01'): {'gdp': 20436.8871286309}, ('Aruba', '2003-01-01'): {'gdp': 20833.7616116694}, ('Aruba', '2004-01-01'): {'gdp': 22569.9749851801}, ('Aruba', '2005-01-01'): {'gdp': 23300.0395575696}, ('Aruba', '2006-01-01'): {'gdp': 24045.272483354704}, ('Aruba', '2007-01-01'): {'gdp': 25835.132667628397}, ('Aruba', '2008-01-01'): {'gdp': 27084.7036903653}, ('Aruba', '2009-01-01'): {'gdp': 24630.4537141023}, ('Aruba', '2010-01-01'): {'gdp': 23512.602595639702}, ('Aruba', '2011-01-01'): {'gdp': 24985.9932813737}, ('Aruba', '2012-01-01'): {'gdp': 24713.6980451285}, ('Aruba', '2013-01-01'): {'gdp': 26189.4355088129}, ('Aruba', '2014-01-01'): {'gdp': 26647.938100985}, ('Aruba', '2015-01-01'): {'gdp': 27980.880695275097}, ('Aruba', '2016-01-01'): {'gdp': 28281.35048163}, ('Aruba', '2017-01-01'): {'gdp': 29007.6930034887}, ('Afghanistan', '1960-01-01'): {'gdp': 59.7731938409853}, ('Afghanistan', '1961-01-01'): {'gdp': 59.8608738790779}, ('Afghanistan', '1962-01-01'): {'gdp': 58.458014949543895}, ('Afghanistan', '1963-01-01'): {'gdp': 78.7063875407802}, ('Afghanistan', '1964-01-01'): {'gdp': 82.0952307131832}, ('Afghanistan', '1965-01-01'): {'gdp': 101.10830485337699}, ('Afghanistan', '1966-01-01'): {'gdp': 137.594352053111}, ('Afghanistan', '1967-01-01'): {'gdp': 160.89858887243798}, ('Afghanistan', '1968-01-01'): {'gdp': 129.108323102596}, ('Afghanistan', '1969-01-01'): {'gdp': 129.329712876621}, ('Afghanistan', '1970-01-01'): {'gdp': 156.518939442982}, ('Afghanistan', '1971-01-01'): {'gdp': 159.567578521888}, ('Afghanistan', '1972-01-01'): {'gdp': 135.31730831433}, ('Afghanistan', '1973-01-01'): {'gdp': 143.14464950008102}, ('Afghanistan', '1974-01-01'): {'gdp': 173.653764639169}, ('Afghanistan', '1975-01-01'): {'gdp': 186.510897140201}, ('Afghanistan', '1976-01-01'): {'gdp': 197.44550755114497}, ('Afghanistan', '1977-01-01'): {'gdp': 224.224797281134}, ('Afghanistan', '1978-01-01'): {'gdp': 247.354106347038}, ('Afghanistan', '1979-01-01'): {'gdp': 275.738197619262}, ('Afghanistan', '1980-01-01'): {'gdp': 272.65528565023203}, ('Afghanistan', '1981-01-01'): {'gdp': 264.11131745306096}, ('Afghanistan', '2002-01-01'): {'gdp': 179.426610967229}, ('Afghanistan', '2003-01-01'): {'gdp': 190.683814295088}, ('Afghanistan', '2004-01-01'): {'gdp': 211.382116942655}, ('Afghanistan', '2005-01-01'): {'gdp': 242.031284871985}, ('Afghanistan', '2006-01-01'): {'gdp': 263.733691663044}, ('Afghanistan', '2007-01-01'): {'gdp': 359.69323750139506}, ('Afghanistan', '2008-01-01'): {'gdp': 364.6607447985}, ('Afghanistan', '2009-01-01'): {'gdp': 438.076034406941}, ('Afghanistan', '2010-01-01'): {'gdp': 543.303041863931}, ('Afghanistan', '2011-01-01'): {'gdp': 591.162759035926}, ('Afghanistan', '2012-01-01'): {'gdp': 641.8714791575389}, ('Afghanistan', '2013-01-01'): {'gdp': 637.165523187024}, ('Afghanistan', '2014-01-01'): {'gdp': 613.856689167623}, ('Afghanistan', '2015-01-01'): {'gdp': 578.466352941708}, ('Afghanistan', '2016-01-01'): {'gdp': 547.228110150363}, ('Afghanistan', '2017-01-01'): {'gdp': 556.30200240406}, ('Afghanistan', '2018-01-01'): {'gdp': 524.162880925404}, ('Afghanistan', '2019-01-01'): {'gdp': 502.115486913067}, ('Angola', '1980-01-01'): {'gdp': 710.981648140027}, ('Angola', '1981-01-01'): {'gdp': 642.383857952257}, ('Angola', '1982-01-01'): {'gdp': 619.9613575311099}, ('Angola', '1983-01-01'): {'gdp': 623.440584831564}, ('Angola', '1984-01-01'): {'gdp': 637.715230700475}, ('Angola', '1985-01-01'): {'gdp': 758.237576171151}, ('Angola', '1986-01-01'): {'gdp': 685.270085316704}, ('Angola', '1987-01-01'): {'gdp': 756.2618530274119}, ('Angola', '1988-01-01'): {'gdp': 792.3031202186439}, ('Angola', '1989-01-01'): {'gdp': 890.5541364590081}, ('Angola', '1990-01-01'): {'gdp': 947.7041820853709}, ('Angola', '1991-01-01'): {'gdp': 865.69272959239}, ('Angola', '1992-01-01'): {'gdp': 656.361755960006}, ('Angola', '1993-01-01'): {'gdp': 441.200673252825}, ('Angola', '1994-01-01'): {'gdp': 328.673294707808}, ('Angola', '1995-01-01'): {'gdp': 397.17945076947194}, ('Angola', '1996-01-01'): {'gdp': 522.643807265256}, ('Angola', '1997-01-01'): {'gdp': 514.295223223424}, ('Angola', '1998-01-01'): {'gdp': 423.59366023208}, ('Angola', '1999-01-01'): {'gdp': 387.784316047502}, ('Angola', '2000-01-01'): {'gdp': 556.836318086553}, ('Angola', '2001-01-01'): {'gdp': 527.333528536691}, ('Angola', '2002-01-01'): {'gdp': 872.4944915928411}, ('Angola', '2003-01-01'): {'gdp': 982.960899291112}, ('Angola', '2004-01-01'): {'gdp': 1255.5640447149099}, ('Angola', '2005-01-01'): {'gdp': 1902.42234554625}, ('Angola', '2006-01-01'): {'gdp': 2599.56646397608}, ('Angola', '2007-01-01'): {'gdp': 3121.99563726236}, ('Angola', '2008-01-01'): {'gdp': 4080.94140992346}, ('Angola', '2009-01-01'): {'gdp': 3122.78076649385}, ('Angola', '2010-01-01'): {'gdp': 3587.88379824396}, ('Angola', '2011-01-01'): {'gdp': 4615.46802807906}, ('Angola', '2012-01-01'): {'gdp': 5100.095808097671}, ('Angola', '2013-01-01'): {'gdp': 5254.8823379961605}, ('Angola', '2014-01-01'): {'gdp': 5408.41049555432}, ('Angola', '2015-01-01'): {'gdp': 4166.97968386501}, ('Angola', '2016-01-01'): {'gdp': 3506.07288506966}, ('Angola', '2017-01-01'): {'gdp': 4095.8129415585704}, ('Angola', '2018-01-01'): {'gdp': 3289.64666408633}, ('Angola', '2019-01-01'): {'gdp': 2973.5911597986797}}

# setup dataframe
df = pd.DataFrame.from_dict(data, orient='index')
df.index.set_names(['country', 'date'], inplace=True)

# display(df.head())
                              gdp
country     date                 
Afghanistan 1960-01-01  59.773194
            1961-01-01  59.860874
            1962-01-01  58.458015
            1963-01-01  78.706388
            1964-01-01  82.095231

# reset the index
df.reset_index(inplace=True)

# set the date column to a datetime format
df.date = pd.to_datetime(df.date).dt.date

# sort values
df.sort_values(['date', 'country'], inplace=True)

# display(df.head())
       country        date        gdp
0  Afghanistan  1960-01-01  59.773194
1  Afghanistan  1961-01-01  59.860874
2  Afghanistan  1962-01-01  58.458015
3  Afghanistan  1963-01-01  78.706388
4  Afghanistan  1964-01-01  82.095231

# plot the date with seaborn.lineplot
plt.figure(figsize=(10, 8))
sns.lineplot(x='date', y='gdp', hue='country', data=df)
plt.yscale('log')
plt.legend(title='country', bbox_to_anchor=(1.05, 1), loc='upper left')

seaborn.barplot

plt.figure(figsize=(8, 15))
sns.barplot(x='gdp', y='date', data=df, orient='h', hue='country')
plt.xscale('log')
plt.legend(title='country', bbox_to_anchor=(1.05, 1), loc='upper left')

seaborn.catplot

sns.catplot(data=df, x='date', y='gdp', col='country', col_wrap=2, kind='bar', height=2.5, aspect=4).set_xticklabels(rotation=90)