topic filtering data set using proc freq count in SAS Programming

filtering data set using proc freq count

kparker — Tue, 26 May 2015 19:17:12 GMT

I'm fairly new to SAS but was wondering if there is a way to delete observations by frequency count? For example, I used proc freq to generate a count of all diagnosis codes but all frequency counts<=5, I want to remove from the dataset. Is this possible?

Re: filtering data set using proc freq count

Steelers_In_DC — Tue, 26 May 2015 19:37:40 GMT

There's going to be a lot of different ways to get this result. One way is to use the out= statement in your proc freq. This will get you a dataset with the result that you can manipulate anyway you want:

proc freq data= have;table variable /out=want;

Re: filtering data set using proc freq count

kparker — Tue, 26 May 2015 19:45:44 GMT

For example,

I have over 500,000 observations in this dataset.

I already have my data filtered for the columns I need but for the diagnosis column, I ran a proc freq.

My Proc Freq data looks like:

DIAGNOSIS_PRIN_CD

Frequency

DIAGNOSIS_PRIN_CD

Frequency

V3000

32677

V3001

21109

486

15683

389

10455

V3000

32677

V3001

21109

486

15683

389

10455

7778

7915

7672

861

I want to suppress or result any diagnosis code that has a frequency count <=5 with Blank or '.'

I am thinking about creating a data step and using IF-THEN statements to accomplish this task but it will be very time consuming.

Re: filtering data set using proc freq count

PGStats — Tue, 26 May 2015 19:46:16 GMT

proc freq might not be the best tool for data management. I would use proc SQL :

proc sql;

create table want as

select * from have

group by diagnosisCode

having count(*) > 5;

quit;