How can I group this data and make this calculations and drops?

I have this data:

0       1.514333e+12
1       1.514419e+12
2       1.514506e+12
3       1.514592e+12
4       1.514678e+12
            ...
1881    1.676851e+12
1882    1.676938e+12
1883    1.677024e+12
1884    1.677110e+12
1885    1.677146e+12
Name: volume1d, Length: 1886, dtype: float64

The data is in timeperiods, so, I need to group this data every 4 rows (every 4 hours (each row is an hour)).
for example:

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

after this, I would have to add the groups.

a b c
1 2 2
1 3 5
1 4 9
1 4 13
2 3 3
2 4 7
2 5 12
2 4 16
3 4 4

to finish leaving just the las row of the groups:

a b c
1 4 13
2 4 16
3 4 4

How can I all do this?
(what I need is the last table, so if you find another way to do this, perfect)

Edited: if the last group is only contains 1,2 or 3 rows, I would like it to have those rows added.

You can transform records using cumulative sum (cumsum function) and eventually extract the last record from each group:

df.assign(c=df.groupby('a').transform('cumsum')).groupby('a').last().reset_index()

   a  b   c
0  1  4  13
1  2  4  16
2  3  4   4