How to loop through and create a rollover sum in python?

Looking at the table, I am looking to replicate the Rollover column in a Python dataframe. The Rollover looks at:

Rollover is created when production > max capacity. The following date will look at the previous day and will either add the delta of production - max capacity or will subtract the delta of max capacity - production.

Rollover will never be less than 0.

You can identify the index of the first time production > max_capacity, set all previous entries equal to 0, then perform a cumulative sum (rounding any negative values up to 0, see this answer). Here is a reproducible example and result:

import numpy as np
import pandas as pd

df = pd.DataFrame({'DATE': pd.date_range('2024-01-01','2024-01-13'),'PRODUCTION': [10,8,8,15,15,12,10,9,14,16,2,2,2],'MAX_CAPACITY':[11]*13})
df['EXTRA'] = df['PRODUCTION'] - df['MAX_CAPACITY']
first_pos_idx = df['EXTRA'].gt(0).idxmax()
df.loc[:first_pos_idx-1, "EXTRA"] = 0

## cumulative sum where negative results are rounded up to 0
sumlm = np.frompyfunc(lambda a,b: 0 if a+b < 0 else a+b,2,1)
df['ROLLOVER'] = sumlm.accumulate(df['EXTRA'].values, dtype=np.object_)

Result:

         DATE  PRODUCTION  MAX_CAPACITY  EXTRA ROLLOVER
0  2024-01-01          10            11      0        0
1  2024-01-02           8            11      0        0
2  2024-01-03           8            11      0        0
3  2024-01-04          15            11      4        4
4  2024-01-05          15            11      4        8
5  2024-01-06          12            11      1        9
6  2024-01-07          10            11     -1        8
7  2024-01-08           9            11     -2        6
8  2024-01-09          14            11      3        9
9  2024-01-10          16            11      5       14
10 2024-01-11           2            11     -9        5
11 2024-01-12           2            11     -9        0
12 2024-01-13           2            11     -9        0