proc sort nodupkey question?

[Mirisage]

Hi SAS Forum,

To show my proc sort question, I have taken the below data set from SUGI paper 037-30 of Britta Kelsey, and is acknowledged.

data best;

input patient 1-2 arm $ 4-5 bestres $ 6-7 delay 9-10;

datalines;

01 A CR 0

02 A PD 1

03 B PR 1

04 B CR 2

05 C SD 1

06 C SD 3

07 C PD 2

01 A CR 0

03 B PD 1

;

run;

I have seen in many papers nodupkey in proc sort is used like in mtheod 1 below but my boss is always using it

like in method 2.

I have checked using proc compare and found that identical resutls are produced by both methods.

Method 1:

proc sort data=best  nodupkey out=ex1;

by arm patient;

run;

Method 2:

proc sort data=best out=ex2 nodupkey;

by arm patient;

run;

proc compare base=ex1 compare=ex2; run;

Question:

1. Do both methods produce identical results for larger data sets too?

2. Which method is efficient?

Could some one help me.

Thanks

Mirisa

[LinusH]

Perhaps I'm blind, but I can't see any different methods.

Ordering of options does not impact performance, and nothing other as well.

[SASKiwi · Accepted solution]

NODUPKEY is an option on the PROC SORT statement. The placement of it is simply a matter of coding style and this doesn't change what it does or affect processing efficiency at all.

As long as NODUPKEY is placed after PROC SORT it can go in any order with the DATA = and OUT = options.

[Mirisage]

Hi SASKiwi and LinusH,

Many thanks to both of you.

Regards

Mirisa