python - 如何通过 pandas.factorize 识别多条件？

Question

我读到这篇文章是关于 pd.factorize 来识别和创建用户识别的唯一值。

但是，就我而言，我想应用多重条件，即OR识别用户的条件，并且条件具有重要性排序。

例如：df：

cond_1(email)  cond_2(phone)  cond_3(other)
[e-mail]  12345678       qwe
[e-mail]  789456123      rty
[e-mail]  905132312      zxc
[e-mail] 789456123       po
[e-mail]  789456123      special

预期：

cond_1(email)  cond_2(phone)  cond_3(other) unique_id
[e-mail]  12345678       qwe            1
[e-mail]  789456123      rty            2  
[e-mail]  905132312      zxc            1
[e-mail] 789456123       po            2
[e-mail]  789456123      special        1

标签： pythonpandasdataframeuniqueidentifier

Answer 1

IIUC，你可以这样做：

df['unique_id']=df.apply(lambda x: pd.factorize(x)[0]+1).min(axis=1)
print(df)

---

    cond_1(email)  cond_2(phone) cond_3(other)  unique_id
0   [e-mail]       12345678           qwe          1
1   [e-mail]      789456123           rty          2
2   [e-mail]      905132312           zxc          1
3  [e-mail]      789456123            po          2
4   [e-mail]      789456123       special          1