如何将不同的列值映射到一列

Question

我下面有一个数据框：

import pandas as pd
df = pd.DataFrame({"SK":["EYF","EYF","RMK","MB","RMK","GYF","RMK","MYF"],
               "SA":["a","b","tm","tmb","tm","cd","tms","alb"],
               "C":["","11","12","13","","15","16","17"]})
df

我想将“SK”、“SA”和“C”的一些值映射到新列：

df["D"]= df["SK"].map({"EYF":1,"MB":2,"GYF":3})
df

df["D"]= df["SA"].map({"tm":4})
df

df["D"]= df["C"].map({"16":5,"17":6})
df

但是当我运行下一个映射函数时，前一个映射函数映射的“D”列值变为NaN。我想在下面获得 df:

我们将不胜感激。

Answer 1

您可以创建 3 个系列，然后将之前 Series 中的缺失值替换为 Series.fillna or Series.combine_first:

a = df["SK"].map({"EYF":1,"MB":2,"GYF":3})
b = df["SA"].map({"tm":4})
c = df["C"].map({"16":5,"17":6})

df["D"] = a.fillna(b).fillna(c)
#alternative
df["D"] = a.combine_first(b).combine_first(c)
print (df)
   SK   SA   C    D
0  EYF    a      1.0
1  EYF    b  11  1.0
2  RMK   tm  12  4.0
3   MB  tmb  13  2.0
4  RMK   tm      4.0
5  GYF   cd  15  3.0
6  RMK  tms  16  5.0
7  MYF  alb  17  6.0

如果某些值相同，则顺序对于优先级很重要：

df = pd.DataFrame({"SK":["EYF","EYF"],
               "SA":["a","tm"],
               "C":["16","17"]})

a = df["SK"].map({"EYF":1,"MB":2,"GYF":3})
b = df["SA"].map({"tm":4})
c = df["C"].map({"16":5,"17":6})

df["D1"] = a.fillna(b).fillna(c)
df["D2"] = b.fillna(a).fillna(c)
df["D3"] = c.fillna(b).fillna(a)
print (df)
    SK  SA   C  D1   D2  D3
0  EYF   a  16   1  1.0   5
1  EYF  tm  17   1  4.0   6

如何将不同的列值映射到一列

How to map different column values to one column

mapping

pandas