The internal numeric representation of a date corresponding to your Death_dt of 20130420 (20APR2013) is 19468. To have a value of death_dt - index = 20111309 as shown above then Index would have to have a value of -20091841 which has a value prior to the earliest valid SAS date values (by about 19700000 days). HOWEVER that value of Index would correspond to a datetime of 13MAY59:10:55:59. If index is supposed to be 20120428 (28APR2012) then your value of Death_dt is 20130420 and is not a SAS date variable at all.   So what is the actual value of your Index variable? and if the SAS format assigned is actually YYMMDDN8. then show the unformatted values for Death_dt and Index for a few records. Proc print data=have;    var death_dt index;    format death_dt index  best12.; run; for example. I believe from what you have shown that at least one of your variables does not actually have the SAS format you ... View more

You still a need a time for the people who didn't recur otherwise this analysis doesn't make sense. You're not comparing anything valid. If that's what you're after, just split the time using a format on it and use PROC FREQ to calculate the percentages in each interval.    ... View more

The following modification of your current code would do what you want:   DATA for_count (keep=_icd rename=(_icd=icd)); SET ccaei121; ARRAY ICD {15} DX1-DX15; DO i = 1 to 15; IF ICD {i} in ("45341", "45340", "45342", "45350", "45351", "45352", "45181", "45111", "45119", "41511", "41512", "41513", "41519") THEN DO; _icd=icd(i); OUTPUT; END; END; RUN; proc freq data=for_count; tables icd; run; Art, CEO, AnalystFinder.com   ... View more

Here's an SQL version that might do it:   data have; input a b c d $; datalines; 1 2 3 x 1 2 2 x 1 2 2 y 2 1 1 x 1 2 4 y 1 2 4 x ; proc sql; CREATE TABLE want AS SELECT DISTINCT a, b, c, CASE count(DISTINCT d) WHEN 1 THEN d ELSE 'z' END AS d FROM have GROUP BY a, b, c; quit; I'm not sure how you'd easily get the final sort order you have in your image, but I think this gets the data right. ... View more

Sorry, I can't give the variable names. The second picture each column can be represented as X, Y, Z etc.  The first picture is the final output I want to have. Thanks ... View more

Post a mock up sample please. If your files have the same variables what are you expecting SAS to do, automatically rename them? How do you identify which is from which dataset?  ... View more

@Sujithpeta That is correct behaviour. 2 dashes include all variables between the two. A single dash includes only variables with a consecutive suffix.  ... View more

@Sujithpeta That is correct behaviour. 2 dashes include all variables between the two. A single dash includes only variables with a consecutive suffix.  ... View more