Days Per Year Between Begin and End Date (Python)

Question

I have a process I run in a Juptyter notebook where in one step, a calculation is done to basically says what is the amount of a loan and spread it over the years daily. So say a loan begins on '2024-04-01' and ends on '2028-04-01' it would apply the partial accrual in 2024 (from April on) and 2028 (from the beginning of the year to April) and the full amounts for the other years.

Where I am hoping to improve it is to dynamically generate the yearly data versus my current method of specifically defining every year in the process below. What I would like to keep is defining the year begin/end as an accounting year can vary from a calendar year. Any input is appreciated and if you see a much simpler way to achieve what I've written, I would very much appreciate it.

data['2024 Year Start'] = datetime.strptime('2024-01-01', '%Y-%m-%d')
data['2024 Year End'] = datetime.strptime('2024-12-31', '%Y-%m-%d')
data['2024 Days'] = pd.to datetime(data[['2024 Year End'], ['Last Loan Date']].min(axis=1)) - pd.to_datetime(data[['2024 Year Start', 'Loan From Date']].max(axis=1))
data['2024 Days'] = data['2024 Days'].astype('timedelta64[D]')-0
data['2024 Accrual'] = np.where((data['Last Loan Date'] >= data['2024 Year Start'])
            & (data['Loan From Date'] < data['2024 Year End']),
            data[ 'Daily Accrual'] * data['2024 Days'],
            0)

data['2025 Year Start'] = datetime.strptime('2025-01-01', '%Y-%m-%d')
data['2025 Year End'] = datetime.strptime('2025-12-31', '%Y-%m-%d')
data['2025 Days'] = pd.to datetime(data[['2025 Year End'], ['Last Loan Date']].min(axis=1)) - pd.to_datetime(data[['2025 Year Start', 'Loan From Date']].max(axis=1))
data['2025 Days'] = data['2025 Days'].astype('timedelta64[D]')-0
data['2025 Accrual'] = np.where((data['Last Loan Date'] >= data['2025 Year Start'])
            & (data['Loan From Date'] < data['2025 Year End']),
            data[ 'Daily Accrual'] * data['2025 Days'],
            0)

Answer 1

How about this:

from dateutil.relativedelta import relativedelta

# fiscal year settings
m_start = 3
d_start = 15

for y in range(2021,2028):
    year_start = datetime(year=y, month=m_start, day=d_start)
    year_end = year_start + relativedelta(years=1,days=-1)
    # if you like, put it in data cols
    data[str(y)+' Year Start'] = year_start
    data[str(y)+' Year End'] = year_end

The rest can be calculated as you are already doing, if necessary using a dynamically generated string (like 'str(y)' above) as key. If this is not what you need, maybe you can give us some source data and desired output.