# SAS: When using “by” statement, what does “var1 = .” mean in output?

Let's say I have some simple code like what is below.

 proc sort; by  var1 var2;
run;

proc gplot;
plot y*x ;
by var1 var2;
run;


In the output, there are a bunch of graphs. In addition to all the combinations of values of var1 and var2, I also get charts where var1=. in combination with each value of var2 (etc). Does the var1=. mean the plot is including all values of var1? Or data where var1 is somehow blank? Or something else? Is there a way I can get these to not print out (i.e. if I want exactly as many plots as there are combinations of var1 and var2)? It's not a big deal for only two variables, but creates a lot of extra plots to scroll through when using more variables.

Thanks!

-
I looks like some values of 'var1' are missing values... –  ocram Feb 29 '12 at 14:58
Are var1 and var2 character or numeric variables? –  Glen Feb 29 '12 at 15:04
@Glen They are numeric –  Jand Feb 29 '12 at 15:10
Hmmm.. I don't think there are values missing because this happens no matter what variables I use (including var2). I think it is just collapsing across all values of the variable. But, I would like to figure out how to get it to not do that. –  Jand Feb 29 '12 at 15:30
Could you post a subset of your data to reproduce the problem? –  Glen Feb 29 '12 at 16:35

It the case of gplot it means what it says, that the graph in question is limited to observations where var1=., i.e. where var1 is missing. I tried but was unable to break this behavior, and as far as I can see the manual doesn't specify any other behavior.

If these graphs are unwanted then a simple where statement at any of several places would remove them. E.g. place this somewhere before proc gplot:

data;
set;
where not missing (var1);
run;


If these graphs are unexpected then I advise you to look through your data and programs to see if something went wrong somewhere. For me personally unexpected missing data tends to indicate I misspelled a variable name somewhere.

For a more specific answer you'll have to give more specific information.

-