将 function 应用到引用前一行数据的 dataframe 行

Question

I'm trying to set col 'b' of my dataframe based on it's previous value from the row above. Is there any way to do this without iterating through the rows or using decorators to the pd.apply function?

Psuedo code:

if row != 0:
    curr_row['b'] = prev_row['b'] + curr_row['a']
else: 
    curr_row['b'] = curr_row['a']

Here's what i've tried:

df = pd.DataFrame({'a': [1,2,3,4,5],
                   'b': [0,0,0,0,0]})

df.b = df.apply(lambda row: row.a if row.name < 1 else (df.iloc[row.name-1].b + row.a), axis=1)

output:

    a   b
0   1   1
1   2   2
2   3   3
3   4   4
4   5   5

Desired output:

    a   b
0   1   1
1   2   3
2   3   6
3   4   10
4   5   15

if I run the apply function a second time on the new df one more row value of c is correct.:

    a   b
0   1   1
1   2   3
2   3   5
3   4   7
4   5   9

This pattern continues if I continue to re-run the apply function until the output is correct.

I'm guessing the issue has something to do with the mechanics of how the apply function works which makes it break when you use a value from the same column you are 'applying' on. That or I'm just being an idiot somehow (very plausible). Can someone explain this?

Do I have to use decorators to store the previous row or is there a cleaner way to do this?

Answer 1

Your requirement is cumsum()

df = pd.DataFrame({'a': [1,2,3,4,5],
                   'b': [0,0,0,0,0]})
df.assign(b=df.a.cumsum())