Python Pandas - 使用反向值删除重复项

Question

假设df就像：

pd.DataFrame({"col1": ["banana", "apple", "grapes", "banana"],
              "col2": ["apple", "banana", "apple", "grapes"]})
col1    col2
banana  apple
apple   banana
grapes  apple
banana  grapes

我们如何删除反复制品，即：banana - apple和apple - banana组合？

我试过

df["col_to_check"] = df.apply(lambda x: set(x), axis=1).astype(str)
idx_to_remove = df[df.duplicated(["col_to_check"])].index

其他建议？