top of page
Search

Half-whiskers plot of difference in treatment effect

Below an alternative graphical representation of difference in treatment effects.



data ests;
missing z;
infile datalines delimiter=',' MISSOVER;
input behandling time VASe VASs lowere uppere lowers uppers;
datalines;
1,0,100,.z,100,100,.z,.z
1,1,78.88888889,.z,78.88888889,120,.z,.z
1,2,53.23552736,.z,53.23552736,100,.z,.z
1,3,27.7394636,.z,27.7394636,72.34042553,.z,.z
1,4,13.67521368,.z,13.67521368,78.72340426,.z,.z
1,5,7.948717949,.z,7.948717949,76.59574468,.z,.z
1,6,7.948717949,.z,7.948717949,77.65957447,.z,.z
1,7,2.948717949,.z,2.948717949,82.9787234,.z,.z
1,8,2.948717949,.z,2.948717949,81.91489362,.z,.z
2,0,.z,100,.z,.z,100,100
2,1,.z,64.40092166,.z,.z,11.11111111,64.40092166
2,2,.z,39.42105263,.z,.z,0,39.42105263
2,3,.z,22,.z,.z,0,22
2,4,.z,17.44444444,.z,.z,0,17.44444444
2,5,.z,9.545454545,.z,.z,0,9.545454545
2,6,.z,6.666666667,.z,.z,0,6.666666667
2,7,.z,5.409356725,.z,.z,0,5.409356725
2,8,.z,2.631578947,.z,.z,0,2.631578947
;;
run;

ods graphics on /noborder;
title 'Treatment effects';
proc sgplot data=ests noautolegend;
highlow x=time low=lowere high=uppere/HIGHCAP=Serif LINEATTRS=(color=cx003399); 
highlow x=time low=lowers high=uppers/LOWCAP=Serif LINEATTRS=(color=cxA23A2E); 
vbox VASe/category=time NOOUTLIERS EXTREME NOMEAN NOCAPS BOXWIDTH=0 CONNECT=MEAN meanattrs=(size=9) CONNECTATTRS=(color=cx003399);
vbox VASs/category=time NOOUTLIERS EXTREME NOMEAN NOCAPS BOXWIDTH=0 CONNECT=MEAN meanattrs=(size=9) CONNECTATTRS=(color=cxA23A2E);
scatter x=time y=VASe/MARKERATTRS=(symbol=SquareFilled color=cx003399);
scatter x=time y=VASs/MARKERATTRS=(symbol=CircleFilled color=cxA23A2E);
yaxis label="Treatment";
xaxis label="Time";
run;
0 views0 comments

Recent Posts

See All

dplyr or base R

dplyr and tidyverse are convenient frameworks for data management and technical analytic programming. With more than 25 years of R experience, I have a tendency to analyze programmatic problems before

bottom of page