An untested guess: change your ORDER BY clause. Remove "t1." so it simply reads: ORDER BY CLUSTER_ID; ... View more

This is a good start, spelling out the input.  It looks like your data is sorted by sales_number which is helpful to know.  However, it is a mistake to try to guide what the steps should be.  Programmers on this forum are much more experienced and can guide you as to a good method to get the final result you want.  In the same way that you are showing the input data, it would help if you could show the analysis you would like to perform once you have selected the right data points. ... View more

Is your data in a sensible order, such as by customer date? Or by date? We can sort it if need be, but it helps to know what we are starting with. ... View more

Since VAR is already character, you can replace it without needing to create VARX. var = put ( input(var, 6.), z6.); ... View more

Looking at the data, your objective is clear as mud. So here is an approach that might work, but will at least gets the dialog moving in a productive direction. First, create a data set with your "combined" variable: data want; set have; comb = cat(cat1, cat2, cat3); run; So both the original variables as well as the combined variable are now in your data. Or perhaps move in the other direction: data want: set have; v = var1; output; v = var2; output; v = var3; output; run; Next, conduct your analysis but use any variables you like. At the end, drop the variables or observations that you no longer need to return the data to its original form. ... View more

When  you create a format, the name of the format cannot end with a number.  Change the name slightly and the program should work: proc format; value eot_test_ce1_ '82160-3'='Adenovirus' '82161-1'='Cov HKU1' '82162-9'='CoV NL63' '82163-7'='CoV 229E'; run; Proc freq data=Sample.; table spec_id*eot_test_ce1 / nopercent norow nocol; format eot_test_ce1 eot_test_ce1_.; run; ... View more

There's an easy way to get a result similar to what you ask. Perhaps the extra periods won't bother you. Try it and see. When printing the data, add: format volume1 commax21.9; Not knowing more about your data, I had to guess that the width of 21 would be a good choice. ... View more

The biggest problem is that these programs replace the original version of TEMP.  If you need it (such as if you kater discover a logic error), you no longer have the original input available. Depending on the size of the data sets, you can be running up the bill.  Each step needs to read the entire data set and output all the observations.  It's easy to get into bad habits, such as using more steps that you actually need.  Typically, input and output are much more costly than data manipulation statements. ... View more

A DATA step can do all the calculations, if you keep track of what information is where.  I would use arrays for this: data want; array lebron {6} _temporary_; array stats {6} pts oreb dreb ast stl blk; array similar {6} similar_pts similar_oreb simil_dreb similar_ast similar_stl similar_blk; if _n_=1 then do; set player_traditional_usage_merged (where=(player="LeBron James")); lebron{1} = pts; lebron{2} = oreb; lebron{3} = dreb; lebron{4} = ast; lebron{5} = stl; lebron{6} = blk; end; set player_traditional_usage_merged; do _n_=1 to 6; similar{_n_} = stats{_n_} - lebron{_n_}; end; run; The temporary array cleans up after itself.  None of the variables fields get stored permanently once the DATA step ends.  You can pick any names you would like in the SIMILAR array.   It's untested code, but you have the data so easy enough for you to try it. ... View more

A minor tweak: since the LENGTH statement defines VISITCD, get rid of SUBSTR. visitcd = _name_; There's room to store only 3 characters, which is exactly the right outcome. ... View more

What combination of fields (if any) form a unique identifier for your records? ... View more

First, get rid of the "3" in the INPUT statement: input id $3 Just a dollar sign would be sufficient.  The "3" tells SAS to take the contents of column 3, which it always blank. Then sort your data: proc sort data=have; by id pnr pid time cn var1 var2 var3 var4 var5; run; Then pick out the observations that need to be investigated: data want; set have; by id pnr pid time cn var1 var2 var3 var4 var5; if first.var5 and last.var5 then delete; run; It's untested code, but easy enough to test since you have the data.  This just picks out the observations, but doesn't tell you where the differences lie. Follow-up:  right idea but wrong logic  will post an update later.   ... View more