Cumulative ID Tracking with Date summatization

[shasank]

Hi Community, 

I have a requirement to summarize a table with dates. Any help is appreciated.  

Raw data:

SubjectName	QueryID	ParentQueryID	OpenDate
0001-S002	818930	818929	21-Nov-22
0001-S002	818929	782251	21-Nov-22
0001-S002	782251		24-Oct-22
01659-S031	962154		7-May-24
01659-S031	962200	962154	8-May-24
01659-S031	962462	962200	17-May-24

Each Query has a unique ID and any Re-query has a new Query ID with the parent Query ID in a second column. I am trying to summarize the information based on the master Query ID and Query Start and Query Stop dates. 

Expected output:

SubjectName	Master QueryID	Query Start	QueryEnd
0001-S002	782251	24-Oct-22	21-Nov-22
0001-S031	962154	7-May-24	17-May-24

Please advise with any approach/logic.

Any help is appreaciated. Thank you for your time. 

[PeterClemmensen]

If your actual data is as simple as the posted sample, then why not just set querystart = min(OpenDate) and queryend = max(OpenDate) ? No need to bother with some fancy logic traversing query ID's and Parent query ID's.

[ballardw]

If you have any SubjectName that have more than one "master" query you should include an example.

Otherwise perhaps:

data work.have;
   input SubjectName :$10.	QueryID	ParentQueryID	OpenDate :date11.;
   format opendate date11.;
datalines;
0001-S002	818930	818929	21-Nov-22
0001-S002	818929	782251	21-Nov-22
0001-S002	782251	. 	24-Oct-22
01659-S031	962154	. 	7-May-24
01659-S031	962200	962154	8-May-24
01659-S031	962462	962200	17-May-24
;


proc summary data=work.have idmin nway;
   class subjectname;
   id queryid;
   var opendate  ;
   output out=work.want (drop= _:) min=querystart max=queryend;
run;

[Ksharp · Accepted solution]

What you gonna do ? If there were not link in ID . Like:

0001-S002	818930	818929	21-Nov-22
0001-S002	818929	782251	21-Nov-22
0001-S002	782251	. 	24-Oct-22
01659-S031	962154	. 	7-May-24
01659-S031	962200	962154	8-May-24
01659-S031	962462	962200	17-May-24
01659-S031	222222	333333	27-May-24

Assuming there are always links in your data.

data work.have;
   input SubjectName :$10.	QueryID	ParentQueryID	OpenDate :date11.;
   format opendate date11.;
datalines;
0001-S002	818930	818929	21-Nov-22
0001-S002	818929	782251	21-Nov-22
0001-S002	782251	. 	24-Oct-22
01659-S031	962154	. 	7-May-24
01659-S031	962200	962154	8-May-24
01659-S031	962462	962200	17-May-24
;

proc sql;
create table want as
select SubjectName,
(select QueryID from have where SubjectName=a.SubjectName and ParentQueryID is  missing) as Master_QueryID,
min(OpenDate) as Query_Start format=date11.,max(OpenDate) as Query_End format=date11.
 from have as a
  group by SubjectName;
quit;