I'm working on a statistical report whose objective is the detection of differences among treatment A, treatment B and a control group after 6 and 8 weeks.
I have this two kinds of bar plots:
As you can see the bar color depends on the group therefore I need a conditional bar color.
Nevertheless I would like to have the same color for each group regardless of the plot, that is , to have "control group" always in red, group "A" always in green and group "B" always in blue (or vice versa, It doesn't matter).
Therefore, I need a conditional color depending on the group and also some kind of manual instruction since ggplot automatically sets the conditionals colors.
Here is the code I've used:
First plot:
Treatment Week Mean
1 A 8 weeks 77.42
2 B 8 weeks 75.09
3 A 6 weeks 88.39
4 B 6 weeks 89.04
ggplot(data=d, aes(x=Week, y=as.numeric(as.character(Mean)), fill=Treatment))
+ geom_bar(stat="identity", position=position_dodge()) +ylab("Estimated Mean")
+ geom_errorbar(aes(
ymin=as.numeric(as.character(Mean))-as.numeric(as.character(sd)),
ymax=as.numeric(as.character(Mean))+as.numeric(as.character(sd))), width=.2,
position=position_dodge(.9))
Second plot:
Treatment Week Mean
1 Control 8 weeks 77.4
2 A 8 weeks 75.09
3 B 8 weeks 88.39
4 Control 6 weeks 89.04
5 A 6 weeks 77.42
6 B 6 weeks 75.09
ggplot(data=d, aes(x=Week, y=as.numeric(as.character(Mean)), fill=Treatment)) +
geom_bar(stat="identity", position=position_dodge()) +ylab("Estimated Mean")+
geom_errorbar(aes(
ymin=as.numeric(as.character(Mean))-as.numeric(as.character(sd)),
ymax=as.numeric(as.character(Mean))+as.numeric(as.character(sd))), width=.2,
position=position_dodge(.9))
Thank you very much.
