Replace each occurrence with a separate value

Question

I was seeking a pandas solution to the following problem

Dataframe

col1,col2
1,a
2,a
3,a
4,a

Desired output: Replace every instance of a in col2 with a different value

col1,col2
1,cat
2,elephant
3,monkey
4,tiger

I've tried the replace() function, but that replace all occurrences of a

df = df['col2'].replace('a','cat')

But doing this results in

col1,col2
1,cat
2,cat
3,cat
4,cat

https://stackoverflow.com/questions/9647202/ordinal-numbers-replacement — BENY, Sep 02 '20 at 17:36
I'm guessing ordinality may skew the ask, so I am revising the desired output — rs79, Sep 02 '20 at 17:40
You need a dictionary to map 1, 2,3 to cat,elephant,monkey etc.., then you can use `series.map`, — anky, Sep 02 '20 at 17:41

score 1 · Answer 1 · answered Sep 02 '20 at 17:42

1

You need to create the addtional key with cumcount

df['new'] = df.groupby('col2').cumcount().add(1).astype(str).add(df.col2)
df['new'].replace({'1a':'cat'})
Out[375]: 
0    cat
1     2a
2     3a
3     4a
dtype: object

answered Sep 02 '20 at 17:42

BENY

317,841
20
164
234

Mustafa Aydın · Accepted Answer · 2020-09-02T17:58:09.913

1

This should do it:

replacing_values = ["cat", "elephant", "monkey", "tiger"]
a_indices = (df["col2"] == "a").index
df.loc[a_indices, "col2"] = replacing_values

edited Sep 02 '20 at 17:58

answered Sep 02 '20 at 17:51

Mustafa Aydın

17,645
4
15
38

Replace each occurrence with a separate value

2 Answers2