尝试一下，

>>> df[~df.assign(MONEY=df.MONEY.abs()).duplicated(keep=False)]

输出：

   NUMBER  MONEY
2  123456   10.0
4  123457    3.0

来自评论（@piRSquared） ：

From comments(by @piRSquared):

通过使用assign，在 df 中添加了新列，并随后添加了 drop_duplicates 同时考虑了 NUMBER 和 MONEY

By using assign, a new column was added to df and the subsequent drop_duplicates takes into account both NUMBER and the absolute value of MONEY

df [〜df.assign（MONEY = df.MONEY.abs（））。duplicated（keep = False）]