Question

我有两个数据帧

         colA    colB   amount1    amount2     key  
maindf = xx      yy     12         1           55_ab
         qs      rr     13         2           14_rs

        colA    colB   amount1     amount2     key           
df =    dd      tt     10          5           55_ab

我想根据字段键将数据帧df的字段amount1和数量2的值添加到maindf以获取以下内容

         colA    colB   amount1    amount2     key  
maindf = xx      yy     22         6           55_ab
         qs      rr     13         2           14_rs

Answer 1

如何使用pandas - 合并df而不是创建新的列/列，如下所述（只需两列） join - left = maindf，right = df，how =＆＃39; left＆＃39;，left_on =＆＃39; key＆＃39;，right_on =＆＃39; key＆＃39;

import pandas as pd
df = pd.DataFrame({'$a':[1,2], '$b': [10,20]}) # creating fake df
df2 = pd.DataFrame({'$a':[1,2], '$b': [10,20]}

dfMerge = pd.merge(left=df, right=df2, how='left', left_on='$a', right_on='$a')
dfMerge["SUM"] = dfMergeTemp["$b_x"] + dfMergeTemp["$b_y"]

# print(dfMergeTemp)

在您的情况下（也许重命名您的列不要混淆正在发生的事情）

mergedDf = pd.merge(left = maindf, right = df, how = 'left', left_on='key', right_on = 'key')

Answer 2

这个怎么样？

# make main df
main = pd.DataFrame({'ColA' : ['xx', 'qs'],
                     'ColB' : ['yy', 'rr'],
                     'amount1' : [12, 13],
                     'amount2' : [1, 2],
                     'key' : ['55_ab', '14_rs']})

# make other df
df = pd.DataFrame({'ColA' : ['dd'],
                   'ColB' : ['tt'],
                   'amount1' : [10],
                   'amount2' : [5],
                   'key' : ['55_ab']})

# merge dataframes
df1 = main.merge(df,
                 how = 'left',
                 left_on = 'key',
                 right_on = 'key').fillna(0)

# sum amount1
df1.insert(df1.shape[1],
           'amount1',
           df1.amount1_x + df1.amount1_y)

# sum amount2
df1.insert(df1.shape[1],
           'amount2',
           df1.amount2_x + df1.amount2_y)

# drop unneeded columns
df1.drop(['amount1_x',
          'amount2_x',
          'ColA_y',
          'ColB_y',
          'amount1_y',
          'amount2_y'],
         axis = 1,
         inplace = True)

Answer 3

Pôle        Libellé entité          PNB N-1         PNB N initial       Key
IBFS        PJSC ROSBANK            -200            0                   zz_t55
IBFS        PJSC ROSBANK            0               1500                bb_y44  
IBFS        PJSC ROSBANK            1000            2000                az_89   
IBFS        PJSC ROSBANK            5000            -250                rr_14

所以这就是maindf 让我们说有更多的列，我不能发布一切。所以我想要从另一个数据帧中加总值，这些数据帧与字段“PNB N-1”和“PNB N-1”相同。和＆＃39; PNB N初始＆＃39;

总结两个数据帧pandas python

3 个答案: