Pandas : Add column to another column-CodePudding

I'm confused on how do I add a column to another in pandas

Here is what I'm trying to do :

from pandas import DataFrame
df1 = DataFrame({'a':[1,2], 'b':[3,4]})
concat((df1['a'], df1['b'].rename({'b':'a'}))).reset_index(drop=True)

Which do return what I want : A serie with my 4 values. What I don't understand is : Why I can't assign it to column 'a' ?

>>> from pandas import DataFrame
>>> df1 = DataFrame({'a':[1,2], 'b':[3,4]})
>>> concat((df1['a'], df1['b'].rename({'b':'a'}))).reset_index(drop=True)
0    1
1    2
2    3
3    4
dtype: int64
>>> df1['a'] = concat((df1['a'], df1['b'].rename({'b':'a'}))).reset_index(drop=True)
>>> df1
   a  b
0  1  3
1  2  4

Is there any way to make it more readable by the way? I'm confused on how it should worked... Note that I don't need column 'b' afterward.

Thanks for your help :)

Sam

CodePudding user response：

pandas series don have columns.

if you want use column by Dataframe, use df[['a']] instead df['a']

& you want change column's name need axis or columns

pd.concat([df1[['a']], df1[['b']].rename(columns={'b':'a'})]).reset_index(drop=True)

output

If i create your output using your code, code like above. But I wouldn't use the above code.

i will use following code:

pd.concat([df1['a'], df1['b']]).to_frame('a').reset_index(drop='True')

CodePudding user response：

When you assign, you are not creating new rows/indices (except with a single value, which is not the case here).

pd.concat((df1['a'], df1['b'].rename({'b':'a'}))).reset_index(drop=True)

Gives you:

0    1
1    2
2    3
3    4
dtype: int64

Pandas aligns the indices before assignment. So, only the indices matching the existing df index are used, here 0 and 1, the rest is discarded