How to drop duplicates between two columns, but keep unique values in respective columns?

Question

I have a df that looks like this;

col1     col2  
aa       aa
cc       bb
dd       dd 

How do I compare both columns in the same df but drop the duplicates and keep the unique in their respective columns?

new df:

col1     col2  
cc       bb

Answer 1

We can construct a filter that checks if the value of df.col1 is different from df.col2 , and then filter, like:

df[df.col1 != df.col2]

For example:

>>> df = pd.DataFrame([['aa', 'aa'], ['cc', 'bb'], ['dd', 'dd']], columns=['col1', 'col2'])
>>> df
  col1 col2
0   aa   aa
1   cc   bb
2   dd   dd
>>> df[df.col1 != df.col2]
  col1 col2
1   cc   bb

We here construct a new dataframe, but we can thus set df to the new dataframe, like:

df = df[df.col1 != df.col2]

Answer 2

这应该可以解决问题：

df[df[col1] != df[col2]]

Answer 3

If you just want to extract the rows that have the same value across columns, this should do.

import pandas as pd

data = {'a':[40, 30, 10],
       'b':[40, 20, 10]}
df = pd.DataFrame(data)
df = df[~(df['a']==df['b'])]

Output

>>> df
    a   b
0  10  40
2  30  10