【发布时间】：2016-12-10 04:25:23
【问题描述】：
我有两个长度不同的数据框，一个是关于人口的，另一个是关于死亡的。我需要合并它们。这是人口表的结构...
Year Age Female Male Total
1933 0 984472.26 1015361.55 1999833.81
1933 1 1040496.02 1064088.29 2104584.31
1933 2 1093043.81 1117527.14 2210570.95
1933 3 1107994.31 1135046.59 2243040.90
1933 4 1130624.43 1179513.62 2310138.05
1933 5 1168930.56 1228225.14 2397155.70
1933 6 1190706.56 1238800.33 2429506.89
1933 7 1203816.58 1245575.51 2449392.09
1933 8 1224285.20 1255721.28 2480006.48
1933 9 1230968.73 1254639.67 2485608.40
1933 10 1243608.10 1262739.94 2506348.04
死亡表的结构与人口相同，但值不同。如果您注意到每一行的年龄递增。人口表的行数多于死亡表。合并两张表后，我希望在死亡行中有NaN's。但是，在运行代码合并表后，我得到以下输出...
year,p_age,p_female,p_male,p_total,d_age,d_female,d_male,d_total
0,1933,0,984472.26,1015361.55,1999833.81,0,52615.77,68438.11,121053.88
1,1933,0,984472.26,1015361.55,1999833.81,1,8917.13,10329.16,19246.29
2,1933,0,984472.26,1015361.55,1999833.81,2,4336.92,5140.05,9476.97
3,1933,0,984472.26,1015361.55,1999833.81,3,3161.59,3759.88,6921.47
4,1933,0,984472.26,1015361.55,1999833.81,4,2493.84,2932.59,5426.43
5,1933,0,984472.26,1015361.55,1999833.81,5,2139.87,2537.53,4677.4
6,1933,0,984472.26,1015361.55,1999833.81,6,1939.7,2337.76,4277.46
7,1933,0,984472.26,1015361.55,1999833.81,7,1760.47,2163.9,3924.37
8,1933,0,984472.26,1015361.55,1999833.81,8,1602.2,2015.97,3618.17
9,1933,0,984472.26,1015361.55,1999833.81,9,1464.88,1893.96,3358.84
10,1933,0,984472.26,1015361.55,1999833.81,10,1357.91,1805.52,3163.43
如果您注意到年龄正在重复，并且数据框从 900 多到 100 万。这是我正在使用的合并代码...
df_usa = usa_population.merge(usa_death, how='left', on='year')
我也用过……
df_usa = pd.merge(usa_population, usa_death, how='left', on='year')
或者...
df_usa = pd.merge(usa_population, usa_death, how='inner', on='year')
如何修复此代码？
【问题讨论】：
标签： python-3.x pandas merge