BookmarkSubscribeRSS Feed
LarissaW
Obsidian | Level 7

I have a table like below, there are only 4 distinct IDs but have 6 observations. I'd like to count how many people are in type 1 only, in type 2 only, in type 3 only, and in Combination of 3 types...

Is there a code capable of fulfilling this process?

 

ID

Type Length
1 1 108
1 2 54
1 3 87
2 1 46
3 1 93
4 2

29

4 REPLIES 4
mkeintz
PROC Star

I guess what you want is to make a profile of each ID  (7 profiles possibles: 3 single types, 3 pairs, and 1 triple), and then do a frequency of those profiles.   But then what do you want to do about the LENGTH variable?  Is it to be aggregated? averaged? ignored?  Or do you simply want a frequency of profiles?

 

The code below does nothing but report the frequency of each type-profile:

 

data have;
  input ID Type	Length;
datalines;
1	1	108
1	2	54
1	3	87
2	1	46
3	1	93
4	2	29
run;

data profiles (keep=id type1-type3); 
  do until (last.id);
    set have;
    by id;
    array dummy {3} type1-type3;
    dummy{type}=1;
    if first.id then put id= ;
  end;
  do _n_=1 to dim(dummy);
    if dummy{_n_}=. then dummy{_n_}=0;
  end;
run; 
proc freq data=profiles;
  table type1*type2*type3 / list;
run;

But if you are interested in corresponding statistics about other vars, like LENGTH, you have to describe what you want.

 

BTW: this code assumes that your data are sorted by ID.

--------------------------
The hash OUTPUT method will overwrite a SAS data set, but not append. That can be costly. Consider voting for Add a HASH object method which would append a hash object to an existing SAS data set

Would enabling PROC SORT to simultaneously output multiple datasets be useful? Then vote for
Allow PROC SORT to output multiple datasets

--------------------------
FreelanceReinh
Jade | Level 19

Alternatively, you can use COUNT(DISTINCT ...) in PROC SQL:

proc sql;
create table want as
select comb, count(distinct id) as n from
  (select id, compress(cat(max(type=1),2*max(type=2),3*max(type=3)),'0') as comb 
   from have
   group by id)
group by comb
order by input(comb,3.);
quit;
PaigeMiller
Diamond | Level 26
data have;
input id type length;
cards;
1	1	108
1	2	54
1	3	87
2	1	46
3	1	93
4	2	29
;

proc transpose data=have out=have_t;
    by id;
    var type;
run;

data want;
    set have_t;
    length category $ 10;
    call sortn(of col:);
    category=compress(cats(of col:),'.');
run;
--
Paige Miller
Sajid01
Meteorite | Level 14

Hello @LarissaW 
I see you what are looking for is "I'd like to count how many people are in type 1 only, in type 2 only, in type 3 only, and in Combination of 3 types..."
The following code create two tables Types for the first case and combinations for the second case. 
The table Types answers the question how many users there in a are given type?

The table Combination answers the question   in how many type's is a user present.? 

data have;
input id type length;
cards;
1 1 108
1 2 54
1 3 87
2 1 46
3 1 93
4 2 29
;
proc sql;
create table Type as
Select TYPE,Count(ID) as users from have
group by Type
order by Type;
Create table combination as
Select ID, Count(Type) as TYPES from have
group by ID
order by ID;
quit;

 

 

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
  • 4 replies
  • 353 views
  • 0 likes
  • 5 in conversation