topic Re: Bar Charts- SGPLOT in SAS Programming

Bar Charts- SGPLOT

david27 — Mon, 03 Jun 2019 17:06:12 GMT

Hello,

We have this code from -- https://support.sas.com/rnd/datavisualization/yourGraphs/businessQuick/vbar/#s1=2

ods listing style=listing;
ods graphics / width=5in height=2.81in;

title 'Mileage by Origin and Type';
proc sgplot data=sashelp.cars(where=(type ne 'Hybrid'));
  vbar origin / response=mpg_city group=type groupdisplay=cluster 
    stat=mean dataskin=gloss;
  xaxis display=(nolabel noticks);
  yaxis grid;
run;

Can we have the exact same chart but with vbar=origin one followed by other.-- Not Stacked.

So a bar chart for Asia followed by Europe followed by USA.

Please advise.

Thanks

Re: Bar Charts- SGPLOT

Reeza — Mon, 03 Jun 2019 17:12:28 GMT

Try adding a BY statement for the ORIGIN. You may need to sort the data by origin before you can do that.

ods listing style=listing;
ods graphics / width=5in height=2.81in;

proc sort data=sashelp.cars out=cars; by origin;
run;

title 'Mileage by Origin and Type';
proc sgplot data=cars(where=(type ne 'Hybrid'));
by origin;
  vbar type/ response=mpg_city
    stat=mean dataskin=gloss;
  xaxis display=(nolabel noticks);
  yaxis grid;
run;

Making a random guess, that's what I think you want.

@david27 wrote:

Hello,

We have this code from -- https://support.sas.com/rnd/datavisualization/yourGraphs/businessQuick/vbar/#s1=2
ods listing style=listing;
ods graphics / width=5in height=2.81in;

title 'Mileage by Origin and Type';
proc sgplot data=sashelp.cars(where=(type ne 'Hybrid'));
  vbar origin / response=mpg_city group=type groupdisplay=cluster 
    stat=mean dataskin=gloss;
  xaxis display=(nolabel noticks);
  yaxis grid;
run;
Can we have the exact same chart but with vbar=origin one followed by other.-- Not Stacked.

So a bar chart for Asia followed by Europe followed by USA.

Please advise.

Thanks

Re: Bar Charts- SGPLOT

ballardw — Mon, 03 Jun 2019 17:28:02 GMT

Are you asking for 3 separate charts with each one showing Asia, one Europe and one USA or something else?

Maybe

proc sgpanel data=sashelp.cars(where=(type ne 'Hybrid'));
  panelby origin/ rows=3;
  vbar type / response=mpg_city groupdisplay=cluster 
    stat=mean dataskin=gloss;
  colaxis display=(nolabel noticks);
  rowaxis grid;
run;