create three X Y scatter for each group

[Alexxxxxxx]

Hello all,

I expect to create a figure 

for the following data,

data have;
INFILE CARDS DSD DLM=',';
input
CTRY :$10.
INDUSTRY :$29.
R&D :8
PATENTS :8
;
cards;
AT,Manufacure,4316550,2033058
AU,Manufacure,2701929,961452
BE,Manufacure,1935804,564041
AT,Service,2069617,483619
AU,Service,1058213,235353
BE,Service,1024675,60304
AT,Oil,1725920,242679
AU,Oil,2178870,543554
BE,Oil,1024675,60304
;
run;

I expect to create three X Y scatter for each INDUSTRY

I would like to put the value of "R&D" at Y-axis and value of "Patent" at X-axis and the value of "INDUSTRY" as the title of each scatter. SO the value of each country is expected to be displayed as a point on each scatter chart.

Could you please give me some suggestion about this?

thanks in advance.

[Reeza]

Have you tried the first example in SGPLOT documentation?
https://documentation.sas.com/?docsetId=grstatproc&docsetTarget=n06fb9ichgg6akn18060gbtvw9pw.htm&doc...

To split by industry instead of group it, use a BY statement and don’t use the GROUP option. 

---

@Alexxxxxxx wrote:

Hello all,

I expect to create a figure 

for the following data,

data have;
INFILE CARDS DSD DLM=',';
input
CTRY :$10.
INDUSTRY :$29.
R&D :8
PATENTS :8
;
cards;
AT,Manufacure,4316550,2033058
AU,Manufacure,2701929,961452
BE,Manufacure,1935804,564041
AT,Service,2069617,483619
AU,Service,1058213,235353
BE,Service,1024675,60304
AT,Oil,1725920,242679
AU,Oil,2178870,543554
BE,Oil,1024675,60304
;
run;

I expect to create three X Y scatter for each INDUSTRY

I would like to put the value of "R&D" at Y-axis and value of "Patent" at X-axis and the value of "INDUSTRY" as the title of each scatter. SO the value of each country is expected to be displayed as a point on each scatter chart.

Could you please give me some suggestion about this?

thanks in advance.

---

[RichardDeVen · Accepted solution]

For three plots in one image, you will want to use Proc SGPANEL. 

• PANELBY statement lets you specify the variables (1 or 2) that control the data subset going each panels.
• SCATTER statement lets you specify the X and Y variables, and the data point labeling variable.

Proper data is important, the sample in your question does not input properly.  The sample code below fixes the input:

data have;
attrib 
  country_code length=$2
  industry     length=$20 label = "INDUSTRY"
  rnd          format=dollar12. label = 'R & D'
  patents      format=dollar12. label = 'Patent'
;

INFILE CARDS DSD DLM=','; input
country_code industry rnd patents;
datalines;
AT,Manufacture,4316550,2033058
AU,Manufacture,2701929,961452
BE,Manufacture,1935804,564041
AT,Service,2069617,483619
AU,Service,1058213,235353
BE,Service,1024675,60304
AT,Oil,1725920,242679
AU,Oil,2178870,543554
BE,Oil,1024675,60304
;

proc sgpanel data=have;
  panelby industry / columns=1;
  scatter x=patents y=rnd / datalabel=country_code;  
run;