合并不同大小的数据框并绘制列的差异

Question

我有两个数据框 Region_education_0 和 Region_education_1 Region_education_0

index	Region	ConvertedComp
1	Australia/New Zealand	122573.834171
2	Caribbean	53562.111111
3	Central Asia	134422.000000
4	East Asia	112492.507042
5	Melanesia	605

Region_education_1

index	Region	ConvertedComp
1	Australia/New Zealand	122573.834171
2	Caribbean	53562.111111
3	Central Asia	134422.000000
4	East Asia	112492.507042

索引 5，Region_education_1 中不存在美拉尼西亚，因为某种情况，我想比较它们并绘制，所以我尝试了这个

from matplotlib.pyplot import *

Region_education_combined=Region_education_0.merge(Region_education_1,left_on="Region",right_on="Region")
Region_education_combined.columns=["Region","Max of Bachelors Higher Ed","Higher Formal Education"]
Region_education_combined['Diff_HigherEd_Vals'] = Region_education_combined['Higher Formal Education'] - Region_education_combined['Max of Bachelors Higher Ed']
print(Region_education_combined)
comp_df.style.bar(subset=['Diff_HigherEd_Vals'], align='mid', color=['#d65f5f', '#5fba7d'])

index	Max of Bachelors Higher Ed	Higher Formal Education	Diff_HigherEd_Vals
1	151698.500659	122573.834171	-29124.666488
2	28413.753425	53562.111111	53562.111111
3	3944.750000	5883.000000	1938.250000
4	45091.041667	27052.384615	-18038.657051

输出中缺少区域列，包括我尝试过的区域

comp_df.style.bar(subset=['Diff_HigherEd_Vals','Region'], align='mid', color=['#d65f5f', '#5fba7d'])

和

comp_df.style.bar(Region_education_combined, align='mid', color=['#d65f5f', '#5fba7d'])

有没有办法在最终输出中包含区域？ 我从“Region_education_0”数据框中省略了“索引 5，美拉尼西亚”，有没有办法将它也包含在输出中？

Answer 1

当您像这样调用merge 时，您可以使用how="outer" 来维护丢失的区域

Region_education_combined=Region_education_0.merge(Region_education_1,left_on="Region",right_on="Region")

请注意，在这种情况下，您将有一个包含 Nan 的表，其中无法合并，在您的情况下，Melanesia 将在 Higher Formal Education 列中有一个 Nan。为了避免问题，您可以使用此设置默认值

Region_education_combined["Higher Formal Education"].fillna(0, inplace=True)