Scanning multiple variables to create new variable

[KathyD1]

Hi,

I need to assess whether data across multiple variables matches.  At the same time I need to bypass missings (not take them into account when determining whether there's a match or non-matching data.  For example, In the data below I would want to note in the first data row that data do not match.  The second data row would be a 'match'.  Thank you!

REF1    REF2     REF3     REF4     REF5

123          .               123      124          .

    .        124             124         ,            ,

   123     123              .            .            .

[Astounding]

With numeric variables, the comparison is easy:

if min(of ref1-ref5) ne max(of ref1-ref5) then result = 'no match';

else result='match';

[ballardw]

---

@Astounding wrote:

With numeric variables, the comparison is easy:

if min(of ref1-ref5) ne max(of ref1-ref5) then result = 'no match';

else result='match';

---

Simpler

if range (of ref1-ref5) = 0 then result='match';

else result = 'no match'.

Though I find the though of using a character result cringeworthy:

result = range(of ref1-ref5) = 0;

[PaigeMiller · Accepted solution]

If these are numeric then

data want;
     set have;
     nonmatches = var (of ref1 - ref5);
run;

If the variance is zero, then the non-missings are all equal, they all match. If the variance is greater than zero, then there is at least one mismatch among the non-missing values.

[KathyD1]

Thank you everyone!