topic Re: How do I specify which values SYMBOL should be applied to in the GPLOT procedure? in SAS Procedures

How do I specify which values SYMBOL should be applied to in the GPLOT procedure?

Hadrien — Fri, 03 Jun 2016 16:46:05 GMT

I have a table with the following fields:

Happiness	Income	Happy
4	55000	Yes
7	90000	Yes
-3	120000	No
8	42000	Yes
-6	17000	No

I am trying to create a scatter plot of the results where all Happy="Yes" points show up as blue and all Happy="Red" points show up as red. Unfortunately, my code produces the inverse (happy folks show up as red). How do I specify that SYMBOL1 should be applied to Happy="No" values and SYMBOL2 to Happy="Yes"?

/*Graph results*/
TITLE 'Happiness vs. Income';
SYMBOL1 VALUE=dot HEIGHT=0.3 CO=RED;
SYMBOL2 VALUE=dot HEIGHT=0.3 CO=STEEL;
PROC GPLOT DATA=PL_H3_DRC_POLS;
	PLOT Happiness*Income=Happy ;
RUN;

Thanks in advance for the help

Re: How do I specify which values SYMBOL should be applied to in the GPLOT procedure?

ballardw — Fri, 03 Jun 2016 17:32:21 GMT

In GPLOT you do not have this functionality. The Assignments of symbol to value are generally the first VALUE for the variable found is assigned the first Symbol, second to second, etc.

So sort your data by HAPPY so that the value you want to be red is the first value.

Or move over to SGPLOT. With SAS 9.3 they added a feature to assign attributes to values for consistency.

Re: How do I specify which values SYMBOL should be applied to in the GPLOT procedure?

FreelanceReinh — Fri, 03 Jun 2016 20:25:36 GMT

Hi Hadrien,

I think it will work if you just write COLOR= instead of CO= in your SYMBOL statements. (At least it worked in my SAS session.) CO= is not an alias for COLOR=, but "specifies color for confidence limit lines and area outlines" (documentation).

Re: How do I specify which values SYMBOL should be applied to in the GPLOT procedure?

Hadrien — Mon, 06 Jun 2016 13:28:25 GMT

Thanks for the help, I was confused about what CO did. I got it to work using COLOR and changing the order of the SYMBOL variables.