CODEWITHSUNDEEP
pythonpandas
DevEx
DevEx

Reputation: 4571

Adding columns of different index

I know this adds consecutive columns of same index:

df['C'] = df['A'] + df['B']

But how to add columns of different index:

Lets say I have a dataframe like this:

df
   A   B    
0  9  40  
1  1  70  
2  5  80  
3  8  30  
4  7  50

I need to create another column C which is the addition of current index row of column A(i.e. idx) and previous index row of column B(i.e. idx-1), such as:

df
   A   B   C  
0  9  40  (9)
1  1  70  (40 + 1)
2  5  80  (70 + 5)
3  8  30  (80 + 8)
4  7  50  (30 + 7)

final result should look like this:

df
   A   B  C  
0  9  40  9
1  1  70  41
2  5  80  75
3  8  30  88
4  7  50  37

Upvotes: 2

Views: 786

Answers (3)

jezrael
jezrael

Reputation: 863166

Amother solution with Series.add:

df['C'] = df['A'].add(df['B'].shift(), fill_value=0)
print (df)
   A   B     C
0  9  40   9.0
1  1  70  41.0
2  5  80  75.0
3  8  30  88.0
4  7  50  37.0

Timings:

In [2]: %timeit df['C'] = df['A'].add(df['B'].shift(), fill_value=0)
1000 loops, best of 3: 457 µs per loop

In [3]: %timeit df['C1'] = df['A'] + df['B'].shift(1).fillna(0)
1000 loops, best of 3: 544 µs per loop

If need function, you can add column names (similar use for another solution):

def f(a,b,c):
    df[c] = df[a].add(df[b].shift(), fill_value=0)
    return df

print (f('A','B','C'))
  A   B     C
0  9  40   9.0
1  1  70  41.0
2  5  80  75.0
3  8  30  88.0
4  7  50  37.0

Upvotes: 1

BrenBarn
BrenBarn

Reputation: 251428

Use shift:

df['C'] = df.A + df.B.shift().fillna(0)

Upvotes: 0

Alex
Alex

Reputation: 19114

You can use Series.shift.

df['C'] = df['A'] + df['B'].shift(1).fillna(0)

Upvotes: 4

Related Questions