topic Re: how to know that you have replicated obs in SAS Procedures

how to know that you have replicated obs

Radwan — Wed, 03 Apr 2019 07:09:42 GMT

Hello

usually i use nodupkey option to eliminate the replicated observations, but i think this option removes all observations that occur next year and so on.

other words this option keeps only first year-observations while i have ten years observations, so I need to know how much replicated obs i have in my datafile.

thanks

Re: how to know that you have replicated obs

PeterClemmensen — Wed, 03 Apr 2019 07:18:05 GMT

You can use the DUPOUT= Option in the PROC SORT Statement.

Re: how to know that you have replicated obs

andreas_lds — Wed, 03 Apr 2019 07:19:37 GMT

Please post the code you are using and an excerpt of the data and explain which obs are "accidentally" removed. The "nodupkey" option keeps only the first observation within a group defined by BY-statement.

Re: how to know that you have replicated obs

Radwan — Wed, 03 Apr 2019 07:25:46 GMT

proc sort data=real_var1 nodupkey; by code year; run;

the observations were 34000 then became 1931 observations

when i checked the number of firms it was 1931 firms.

while i have data for 9 years.

Re: how to know that you have replicated obs

andreas_lds — Wed, 03 Apr 2019 07:27:51 GMT

You should add the year-variable to the by-statement, if you want obs for each year.

Re: how to know that you have replicated obs

Astounding — Wed, 03 Apr 2019 12:09:19 GMT

Check your data. It's probably messed up. The result you describe would only be possible if the YEAR values were the same for all observations.

Re: how to know that you have replicated obs

Radwan — Thu, 04 Apr 2019 07:27:10 GMT

ok i will
thanks