Easiest way to create variables by substracting elements from group? e.g., value of each row of each group – first row of group value

In the following example, what would be the best to group so that there could be a new column that is formed by taking first year in each group and subtracting by current year. For example in in row with index 0 it would be NaN, row with index 1 , it would = 1, row with index 2 it would = 3, row with index 4 = 1 and so forth.

>>> import pandas as pd
>>> df = pd.DataFrame({'id': ['1', '1', '1', '2', '2', '3', '4', '4'],
...                    'Year': [2000, 2001, 2003, 2004, 2005, 2002, 2001, 2003]})
>>> print(df)
  id  Year
0  1  2000
1  1  2001
2  1  2003
3  2  2004
4  2  2005
5  3  2002
6  4  2001
7  4  2003

>Solution :

Transform Year with first to get the first year per id, then subtract this from Year column to get difference, finally mask the values where difference is 0:

s = df['Year'] - df.groupby('id')['Year'].transform('first')
df['col'] = s.mask(s == 0)

  id  Year  col
0  1  2000  NaN
1  1  2001  1.0
2  1  2003  3.0
3  2  2004  NaN
4  2  2005  1.0
5  3  2002  NaN
6  4  2001  NaN
7  4  2003  2.0