BookmarkSubscribeRSS Feed
☑ This topic is solved. Need further help from the community? Please sign in and ask a new question.
luvscandy27
Quartz | Level 8

Hello everyone, 

I have a dataset with millions of records. What I would like to do is keep the most recent date by id unless 

the other 4 rating variables are more complete. Any assistance would be appreciated .This is the sample dataset:

 

id

date

rate_1

rate_2

rate_3

rate_4

100

3/3/2023

1

1

1

1

100

2/27/2023

1

1

1

1

100

1/21/2023

1

1

1

1

100

12/28/2022

1

1

1

1

200

3/4/2023

.

.

.

.

200

1/21/2023

1

1

1

1

 

and this is what I would like to end up with :

id

date

rate_1

rate_2

rate_3

rate_4

100

3/3/2023

1

1

1

1

200

1/21/2023

1

1

1

1

 

The code for the sample dataset is: 

data test1;
infile datalines delimiter = ',';
input id $ date:mmddyy10. rate_1 rate_2 rate_3 rate_4 ;
format date mmddyy10.;
datalines;
100,03/03/2023,1,1,1,1,
100,02/27/2023,1,1,1,1,
100,01/21/2023,1,1,1,1,
100,12/28/2022,1,1,1,1,
200,03/04/2023,.,.,.,.,
200,01/21/2023,1,1,1,1
;
run;
1 ACCEPTED SOLUTION

Accepted Solutions
Quentin
Super User

If it's not too big to sort, you could compute N and then sort the data by ID N Date, and grab the last record per ID.  Something like:

 

data test2 ;
  set test1 ;
  n=n(of rate: ) ;
run ;

proc sort data=test2 ;
  by id n date ;
run ;

data want ;
  set test2 ;
  by id n date ;
  if last.id ;
  drop n ;
run ;
BASUG is hosting free webinars Next up: Mark Keintz presenting History Carried Forward, Future Carried Back: Mixing Time Series of Differing Frequencies on May 8. Register now at the Boston Area SAS Users Group event page: https://www.basug.org/events.

View solution in original post

3 REPLIES 3
Quentin
Super User

If it's not too big to sort, you could compute N and then sort the data by ID N Date, and grab the last record per ID.  Something like:

 

data test2 ;
  set test1 ;
  n=n(of rate: ) ;
run ;

proc sort data=test2 ;
  by id n date ;
run ;

data want ;
  set test2 ;
  by id n date ;
  if last.id ;
  drop n ;
run ;
BASUG is hosting free webinars Next up: Mark Keintz presenting History Carried Forward, Future Carried Back: Mixing Time Series of Differing Frequencies on May 8. Register now at the Boston Area SAS Users Group event page: https://www.basug.org/events.
FreelanceReinh
Jade | Level 19

Hello @luvscandy27,

 

Alternatively, you can use a single PROC SQL step:

proc sql;
create table want(drop=c) as
select *, 1e6*n(rate_1,rate_2,rate_3,rate_4)+date as c from test1
group by id
having c=max(c);
quit;
vijaypratap0195
Obsidian | Level 7

I believe you can achieve this by using proc sort and last.id-

 

Try this-

proc sort data=test1; by id date; run;
 
data tmp;
set test1;
by id;
if sum(of rate_:) = 4;
run;
 
proc sort data=tmp; by id ; run;
  
data tmp1;
set tmp;
by id;
if last.id;
run;

-Vijay

sas-innovate-2024.png

Available on demand!

Missed SAS Innovate Las Vegas? Watch all the action for free! View the keynotes, general sessions and 22 breakouts on demand.

 

Register now!

How to Concatenate Values

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.

Click image to register for webinarClick image to register for webinar

Classroom Training Available!

Select SAS Training centers are offering in-person courses. View upcoming courses for:

View all other training opportunities.

Discussion stats
  • 3 replies
  • 260 views
  • 1 like
  • 4 in conversation