pandas merge same name columns group by

Question

i have the following frame:

col1 col1 

null  hi
hello null

I am trying to merge the two columns as they have the same name:

col1
hi
hello

I am using:

result= result.groupby(result.columns, axis=1).first()

But the first() applies it to the whole column so i get:

col1
hi
null

Anyone to merge the columns and avoid the null values.

Do you have only 2 columns? Do you have only one value in each row, while the others are null? — A. Wolf, Oct 23 '18 at 12:19
Your solution working nice with your sample data, `null` are always `NaN`s or `None`s ? — jezrael, Oct 23 '18 at 12:19

score 1 · Accepted Answer · answered Oct 23 '18 at 12:26

1

Your solution working nice if null are converted to NaNs.

So maybe problem is null are sometimes strings, so first replace.

df = df.replace('null', np.nan).groupby(df.columns, axis=1).first()
print (df)
    col1
0     hi
1  hello

answered Oct 23 '18 at 12:26

jezrael

1 Answers1