series statement

[himself]

      series x=col2 y=col3 /   markers markerattrs=(color=blue size=8 symbol=CircleFilled)
              lineattrs=(pattern=1 color=blue) name="lsc" legendlabel="Language score" break;

	  series x=col2 y=col4 /  markers markerattrs=(color=orange symbol=SquareFilled size=8) 
            lineattrs=(pattern=1 color=orange) name="psc" legendlabel="Performance score" break;
      
      series x=col2 y=col5 /  markers markerattrs=(symbol=TriangleFilled size=8  color=red ) 
		      lineattrs=(pattern=1 color=red) name="psic" legendlabel="Processing speed Index" break ; 

	  series x=col2 y=col6 / markers markerattrs=(color=black symbol=DiamondFilled size=8) 
            lineattrs=(pattern=1 color=black) name="tiqc" legendlabel="Total Intelligence quotient (IQ)" break ;
    
      series x=col2 y=col7 /  markers markerattrs=(symbol=StarFilled size=8  color=green ) 
		      lineattrs=(pattern=1 color=green) name="wmic" legendlabel="Working memory Index" break;  

Hi, am doing  a proc sg panel , I have missing values for col5-col7, is there a way i can tell sas not to perform task at the point where the columns are missing? so am getting a warning as shown below 

[PaigeMiller]

I don't think there is such an option.

You could certainly write a macro that first tests to see if a variable is all missing, and then excludes it from the PROC SGPANEL code. But that seems like a lot of work just to avoid a WARNING in the log, in my opinion.

[himself]

Hi thanks, for the reply, I have tried to use proc sql, to count only the
columns with non missing values for each group,but its to vain

[ballardw]

Example data and full sgpanel code would likely be the best place to start.

One would ask why you have included a variable with all missing values in the first place. If the reason comes down to "it worked for a previous data set" then the process likely needs to be addressed to account for changing data. Which would mean explaining what data comes up missing.