Create a multi-colored block graph in ODS Graphics Designer

[mariko5797]

I am trying to use ODS Graphics Designer to create the following graph:

I tried to use the "Block" plot, but it seems to only allow two colors: missing/not missing or multi-colors with no control over color assignment. Is there a way to create such a plot using ODS Graphics? I would prefer using designer since I need to stack a lot of graphs, but it's not a requirement.

[DanH_sas]

Yes, this can be done; but can you tell more about your plot axis? Is it continuous or discrete? Do you want each of your block plots to be labeled?

[mariko5797]

Data is discrete and stored numerically: 1 (mild), 2 (moderate), 3 (severe), 0 (none), 5 (not collected). I do not need the boxes labelled with the values nor axis labelled as this plot will be under a scatter plot with the same axis.

[DanH_sas]

Normally, I would use PROC SGPLOT to do this; but for your case, I needed the REPEATEDVALUES option on the BLOCK plot, which is not currently surfaced in the SG procedures. To control the value/color association, modify the attrmap data set. To add more BLOCK plots, just add them inside the INNERMARGIN to stack them.

Hope this helps!

Dan

proc template;
define statgraph example;
begingraph;
layout overlay / xaxisopts=(discreteOpts=(tickvaluefitpolicy=SplitRotate));
   ScatterPlot X=Name Y=Weight / primary=true;
   InnerMargin / align=bottom;
      BlockPlot X=Name Block=age / Display=( Fill Outline Label ) NAME="block" repeatedvalues=true;
   EndInnerMargin;
   DiscreteLegend "block" / Location=Outside;
endlayout;
endgraph;
end;
run;

data attrmap;
retain ID "myid";
length fillcolor $ 6;
input value $ fillcolor $;
cards;
12 blue
13 green
14 yellow
15 orange
16 red
;
run;

proc sgrender data=sashelp.class template=example dattrmap=attrmap;
dattrvar age="myid";
run;

[Rick_SAS]

If you are willing to eliminate the gap between each row, this can be done by using a heat map. For example, see the article, "Create a discrete heat map with PROC SGPLOT"

You can use a discrete attribute map to assign colors to values.

data Clinical;
input SiteID @;
do Week = 0 to 13;
   input Count @;
   output;
end;
/* ID Wk1  Wk2  Wk3  Wk4 ... Wk14*/
datalines;
001  1 0 0 0 0 0 0 3 1 3 3 0 3 0
002  0 0 0 1 1 2 1 2 2 1 1 0 2 2
003  1 . . 1 0 1 0 3 . 1 0 3 2 1 
004  1 1 . 1 0 1 2 2 3 2 1 0 . 0
005  1 1 1 . 0 0 0 1 0 1 2 3 3 1
;

/* https://blogs.sas.com/content/iml/2019/07/15/create-discrete-heat-map-sgplot.html */
data DiscreteAttrOrder;                /* create discrete attribute map */
length Value $1 FillColor $15;
input Value FillColor;
retain ID 'Malaria'                    /* name of map */
     Show 'AttrMap';                   /* always show all groups in legend */
datalines;
.  Gray
0  White
1  CXFFFFB2
2  CXFD8D3C
3  CXBD0026
;

title "Heat Map of Malaria Data";
proc sgplot data=Clinical DATTRMAP=DiscreteAttrOrder;
   heatmapparm x=Week y=SiteID colorgroup=Count / outline outlineattrs=(color=gray)
               ATTRID=Malaria; 
   discretelegend;
   refline (1.5 to 5.5) / axis=Y lineattrs=(color=black thickness=2);
   xaxis integer values=(0 to 13) valueshint;
run;

 

[Rick_SAS]

I wrote more details and comments about using formats to bin numeric data, discrete attribute maps, and using heat maps for longitudinal data: https://blogs.sas.com/content/iml/2022/04/20/viz-ordinal-response-longitudinal.html