Hello everyone, this is my first post on here so my apologies if the format is not up to the usual standards.

I have a data set covering a time span between 1998 and 2016 with about 65.000 observations that may be divided into 15 groups (Dutch political parties) The aspects that are relevant for this question are listed below:
date partyid sponsorp~y success cumsuc~s
1. 26aug1998 08:11:41 5 CDA 0 0
2. 26aug1998 08:12:12 10 GroenLinks 0 0
3. 26aug1998 08:18:51 5 CDA 0 0
4. 26aug1998 08:46:35 10 GroenLinks 0 0
5. 26aug1998 09:00:21 10 GroenLinks 0 0
6. 08sep1998 07:41:39 5 CDA 1 1
7. 08sep1998 08:50:57 5 CDA 0 1
8. 08sep1998 09:05:53 24 SP 0 0
9. 17sep1998 07:39:19 10 GroenLinks 0 0
10. 17sep1998 07:44:20 25 VVD 0 0
11. 17sep1998 07:59:09 7 D66 0 0
12. 17sep1998 08:03:12 20 PvdA 0 0
13. 17sep1998 08:12:32 10 GroenLinks 0 0
14. 17sep1998 08:12:55 23 SGP 1 1
15. 17sep1998 08:14:15 24 SP 0 0
16. 17sep1998 08:16:44 5 CDA 0 1
17. 17sep1998 08:21:38 20 PvdA 0 0
18. 17sep1998 08:22:40 5 CDA 1 2
19. 17sep1998 08:24:20 5 CDA 0 2
20. 17sep1998 08:27:57 20 PvdA 1 1
21. 17sep1998 08:42:09 10 GroenLinks 0 0
22. 17sep1998 08:44:11 7 D66 0 0
23. 17sep1998 08:54:14 20 PvdA 0 1
24. 17sep1998 08:55:03 10 GroenLinks 0 0
25. 17sep1998 09:02:38 24 SP 0 0
26. 17sep1998 09:08:47 7 D66 0 0
27. 22sep1998 07:25:52 24 SP 1 0
28. 22sep1998 09:37:04 10 GroenLinks 1 1
29. 29sep1998 07:15:41 20 PvdA 0 1
30. 29sep1998 07:22:35 5 CDA 1 3
What I want to do is create a variable that is the sum of the variable 'success' per 'partyid'(numeric reflection of sponsorparty).

I manually added the variable 'cumsuccess' in the table above manually to better illustrate what I mean. As you can see, the variable 'success' is just 0's and 1's to illustrate rejection (0) or success (1). The numbers in cumsuccess should reflect the cumulative sum of success per party. Which is why 'CDA' is at 3 and other parties are not.

I hope this makes sense. Please tell me if I could make some things more clear. Thank you!