0

我正在尝试创建一种搜索列的方法，如果它是重复值，请替换一个字符串。这是我到目前为止所拥有的

date
November 1st 2020
November 2nd 2020
November 3rd 2020
November 1st 2020
November 2nd 2020
November 3rd 2020
November 1st 2020
November 2nd 2020
November 3rd 2020

我想做的是这个

date
November 1st 2020 first instance
November 2nd 2020 first instance
November 3rd 2020 first instance
November 1st 2020 second instance
November 2nd 2020 second instance
November 3rd 2020 second instance
November 1st 2020 third instance
November 2nd 2020 third instance
November 3rd 2020 third instance

有没有办法做到这一点？ 在 Pandas 中跨列替换重复值 这只会将“1 个字符”添加到数据框中，因此您会创建一个循环以继续添加 1 或您想要的任何内容吗？就像是

pd.Series(['date']).duplicated()
is_duplicate = df.apply(pd.Series.duplicated, axis=1)
for is_duplicate
   df.where(~is_duplicate, +1)

我真的不明白如何遍历列以获得所需的结果。

4

1 回答 1

2

IIUC，您可以使用 和 来做到groupby这cumcount一点map。

df['date'] += ' ' + df.groupby('date').cumcount().map({0:'first ', 1:'second ', 2: 'third '}) + 'instance'

输出：

                                date
0   November 1st 2020 first instance
1   November 2nd 2020 first instance
2   November 3rd 2020 first instance
3  November 1st 2020 second instance
4  November 2nd 2020 second instance
5  November 3rd 2020 second instance
6   November 1st 2020 third instance
7   November 2nd 2020 third instance
8   November 3rd 2020 third instance
于 2020-11-18T21:30:18.813 回答