SGPLOT VS SGPANEL

[cd2011]

Hi Guys,

So I have this question for you. I have a stacked bargraph that I produced using SGPLOT, where I created textured fill to distinguish between subgroups using patternid. But, I could not replicate that in SGPANEL, where bars are stacked by subgroup, and then there are different plots for differenet groups, thats why using sgpanel. 

I think I can do it using template, but is there an easier way ? Maybe, a optin like patternid in SGPLOT? My search did not yield any results. 

Thanks a bunch in advance !

[Cynthia_sas]

Hi:
PATTERNID is for PROC GCHART. You use different methods with SGPLOT. Can you post the SGPLOT code or code showing SGPLOT and PATTERNID that worked?

cynthia

[cd2011]

You might be right. I dont have the code file with me so I cant check. But it could be gchart. I will check if it worked with SGPLOT and post an update. But definitely did not work with SGPANEL. 

[Cynthia_sas]

Hi:

  You cannot "mix" SAS/GRAPH controls or SAS/GRAPH procedures with ODS GRAPHICS controls and ODS GRAPHICS procedures. So PROC GCHART will not work with SGPANEL.

  However, a modification of this Tech Support example, http://support.sas.com/kb/45/663.html worked for me with SGPLOT and SGPANEL.

cynthia

*** the code;

ods path work.tmp(update) sasuser.templat(update) sashelp.tmplmst(read);
  
proc template;
   define style mypatterns;
   parent=styles.rtf;

   /* The GraphBar element must be included with the
      FILLPATTERN option in order to use fill patterns. */
   style GraphBar from GraphComponent /                                 
         displayopts = "outline fillpattern";  
   
   /* Fill patterns are defined using the FILLPATTERN 
	  style element attribute. 
	  COLOR is used for the bar fill colors
	  CONTRASTCOLOR is used for the line colors */

      style GraphData1 from GraphData1 /                                      
            fillpattern = "L3"
            color=cxDADAEB
            contrastcolor=cx252525;                                                  
      style GraphData2 from GraphData2 /
            fillpattern = "R3"
            color=cxBCBDDC
            contrastcolor=cx636363;                                                  
      style GraphData3 from GraphData3 /                                      
            fillpattern = "X5"
            color=cx9E9AC8
            contrastcolor=cxCCCCCC;                                                  
      style GraphData4 from GraphData4 /                                      
            fillpattern = "L5"
            color=cx756BB1
            contrastcolor=cxBDBDBD;                                                  
      style GraphData5 from GraphData5 /                                      
            fillpattern = "X2"
            color=cx6A51A3
            contrastcolor=grayaa;    
     end;
run;

options nodate number;
ods pdf file='c:\temp\pat_sg_pan_ex1a.pdf' style=mypatterns;
ods html gpath='c:\temp' (url=none)
         path='c:\temp' (url=none) 
         file='pat_sg_pan_ex1b.html' style=mypatterns;
ods rtf file='c:\temp\pat_sg_pan_ex1c.rtf' style=mypatterns;

title 'Using fill patterns with a bar chart';

data sample;
   length grpvar $8;
   input mid $ value;
   do supergroup = 'Z1', 'Z2';
     do grpvar='One', 'Two', 'Three';
       if grpvar = 'One' then output;
	   else if grpvar = 'Two' then do;
	      value = value + 3;
		  output;
	   end;
	   else do;
	     value = value + 5;
	     output;
	   end;
     end;
   end;
return;
datalines;
A 10
B 20
C 25
;
run;

** plot one "supergroup";
proc sgplot data=sample noautolegend;
   where supergroup = 'Z1';
   vbar mid / response=value stat=mean group=grpvar groupdisplay=cluster 
              grouporder=data name="Bar";
xaxis label="Grouped";
yaxis label="Mean" values=(0 to 35 by 5);
keylegend "Bar" / title="Wombat" ;
run;

** plot both "supergroups" with SGPANEL;
proc sgpanel data=sample;
   panelby supergroup;
   vbar mid / response=value stat=mean group=grpvar groupdisplay=cluster 
              grouporder=data name="Bar2";
colaxis label="Grouped";
rowaxis label="Mean" values=(0 to 50 by 5);
run;
ods _all_ close;

*** the output;