Comparing two arrays element wise in pandas-CodePudding

I have two data frames. My final goal is to compare a column in both data frames and return those values which donot match with each other example:

df_1["column_1"]= ["A45", "kl24", "mhg", "tz22" ]
df_2["column_2"]= ["KL24", "tz22", "mhg", "A 45"]

I need a code that comparing two array values in the respective dataframe["column"] and returns those values from df_1 which did not match in df_2(Ex: from our example "A45" and "kl24" will return because there is a space and upper and lower case error)

Can anyone kindly please help me with this !

CodePudding user response：

Use pd.merge(..., how='left', indicator=True) and select the rows with left_only:

>>> df = df_1.merge(df_2, how='outer', left_on='column_1', right_on='column_2', indicator=True)
>>> df
  column_1 column_2      _merge
0      A45      NaN   left_only
1     kl24      NaN   left_only
2      mhg      mhg        both
3     tz22     tz22        both
4      NaN     KL24  right_only
5      NaN     A 45  right_only
>>> df[df._merge == 'left_only']['column_1']
0     A45
1    kl24
Name: column_1, dtype: object

CodePudding user response：

You can use df.isin and df.loc to do the trick

df_1.loc[~df_1["column_1"].isin(df_2["column_2"])]

  column_1
0      A45
1     kl24