topic date formats in SAS Programming

date formats

vraj1 — Wed, 20 Jan 2016 15:02:44 GMT

I have 2 datasets one has subject with start_date and other with subject with few missing first drug date.

I want to get the start_date of first dataset if my second dataset first drug date is missing(only if missing), else use the same one.

can any one help me in this. i am using the below code but not wiorking

proc sql;
create table subj_dose2 as
select distinct trl_id
, subj_id
,src_start_dt
from ddm.subj_comp_dos
where topic_cd in ("BIASP")
and dose > 0
order by trl_id
, subj_id
,src_start_dt;
quit;

proc sort data=ddm.subj out=subj_tmp;
by trl_id subj_id;
run;

***Update exposed flag in ddm.subj***;
data subj(drop=src_start_dt);
merge subj_tmp(in=subj)
subj_dose2(in=dose2);
by trl_id subj_id;
if subj;
if fdrg_dt ne . then do;
rename src_start_dt=fdrg;
end;
run;

Re: date formats

RW9 — Wed, 20 Jan 2016 15:12:43 GMT

Posting test data in the form of a datastep would help. At a guess something like:

proc sql;
  create table WANT as
  select  A.TRT_ID,
            A.SUBJID,
            coalesce(A.DRUG_DATE,B.START_DATE) as DRUG_DATE
  from    HAVE A
  left join START_DATES B
  on       A.SUBJID=B.SUBJID;
quit;

The coalesce() function takes the first non-missing from the list of variables. The same logic works in datastep.

Re: date formats

FreelanceReinh — Wed, 20 Jan 2016 15:32:47 GMT

The RENAME statement is a declarative statement. Hence, you can't execute it conditionally (with IF/THEN). Its purpose is to change variable names, not to transfer values from one variable to another.

To impute a missing value of variable X with the value of variable Y in the same observation (without overwriting non-missing values of X), you can use a conditional assignment statement like if x=. then x=y; (but there are alternative approaches as well, see @RW9's post).

Re: date formats

vraj1 — Thu, 21 Jan 2016 08:50:16 GMT

Thanks,

But in this case the drug date is in date9.(numeric) and start date is in char.

i get the below error

ERROR: The COALESCE function requires its arguments to be of the same data type.

Re: date formats

RW9 — Thu, 21 Jan 2016 09:27:37 GMT

You will of coursehave to change any code to match your data as you didn't supply any. Depending on what you want the output variable type to be:

Numeric

coalesce(A.DRUG_DATE,input(B.START_DATE,date9.)) as DRUG_DATE

Char

coalesce(put(A.DRUG_DATE,date9.),B.START_DATE) as DRUG_DATE