Hi,
I have following type of data and now I want to clean data such that I my data retains only those observations of dr when frequency of 0 <= 4 in any given month. If frequency of dr=0 s greater than 4 I would like to delete that entire month from my data.
Fir example, in the following data I would want to delete the entire month of Apr-00 as count(0) is greater than 4 but would want to retain May-00 as count (0) is only 1
permno | month | dr |
1 | Apr-00 | 0 |
1 | Apr-00 | 0 |
1 | … | 0 |
1 | Apr-00 | 0 |
1 | Apr-00 | 1.2 |
1 | May-00 | 3.2 |
1 | May-00 | 1.3 |
1 | … | 4.5 |
1 | May-00 | 0 |
Any suggestions?
Please post a better and comprehensive sample so responders can test their codes. In your example, count of dr=0 is 4 and it is not greater than 4 for apr00
proc sql;
create table want as
select *
from have
group by permno, month;
having sum(dr=0)<=4;
quit;
Please post a better and comprehensive sample so responders can test their codes. In your example, count of dr=0 is 4 and it is not greater than 4 for apr00
proc sql;
create table want as
select *
from have
group by permno, month;
having sum(dr=0)<=4;
quit;
Thanks a bunch, it works.
It's finally time to hack! Remember to visit the SAS Hacker's Hub regularly for news and updates.
Learn how use the CAT functions in SAS to join values from multiple variables into a single value.
Find more tutorials on the SAS Users YouTube channel.
Ready to level-up your skills? Choose your own adventure.