How do I reorder a single column within ggplot2?

Question

I have created and plotted a small dataset using ggplot2. Given my overall experimental design it was suggested to me that I reorder the columns in my graph such that the ones in the fifth/final position are moved into the first position (keeping the order of the other bars the same). Unfortunately, I have not been able to figure out how to do this.

My data looks like this:

emean_Sleep_State<-c('WAKE','SWS','REM','WAKE','SWS','REM','WAKE','SWS','REM','WAKE','SWS','REM','WAKE','SWS','REM')
emean_Treatment<-c('Disturbed_Feed','Disturbed_Feed','Disturbed_Feed','Disturbed_Foot','Disturbed_Foot','Disturbed_Foot','Disturbed_Temp','Disturbed_Temp','Disturbed_Temp','Recovery','Recovery','Recovery','UnDisturbed','UnDisturbed','UnDisturbed')
emean_Means<-c(27.68,52.64,21.36,35.38,47.96,15.59,42.28,45.84,6.03,27.56,52.2,22.36,28.62,52.84,21.01)
emean_SE<-c(3.94,2.99,3.08,3.97,3.01,3.1,3.84,2.91,3.03,2.94,2.34,2.51,5.75,2.92,4.18)
Estimated_Max_Means_emean<-emeanframe$emean_Means+emeanframe$emean_SE
Estimated_Min_Means_emean<-emeanframe$emean_Means-emeanframe$emean_SE
emeanframe<-data.frame(emean_Sleep_State,emean_Treatment,emean_Means,emean_SE,Estimated_Max_Means_emean,Estimated_Min_Means_emean)
realvalueemean<-apply(emeanframe[,c(3,5,6)],2,FUN=backtransANG)
emeanframe$realmean<-realvalueemean[,1]
emeanframe$realmaxmean<-realvalueemean[,2]
emeanframe$realminmean<-realvalueemean[,3]

Of additional note is that the function contained within is the following:

backtransANG<-function(ANG_value){
  real_value<-(sin(ANG_value/(180/pi)))^2
  return(real_value)
}

The above data was then plotted using the below code:

emeanplot2<-ggplot(emeanframe,
                   aes(x=emean_Sleep_State,
                       y=realmean,
                       fill=emean_Treatment))+
  geom_col(position='dodge',
           color='black')+
  geom_errorbar(aes(ymin=realminmean,
                    ymax=realmaxmean),
                color='black',width=0.2,position=position_dodge(width=0.9),linewidth=1)+
  scale_x_discrete(limits=c('WAKE','SWS','REM'))+
  labs(x='Sleep State',
       y= 'Proportion of Behaviour',
       color="Type of Night",size=14)+
  ylim(0,1)+
  guides(fill=guide_legend(title='Type of Night'))+
  theme(axis.text.x=element_text(size=12),
        axis.text.y=element_text(size=12),
        axis.title.x=element_text(size=14),
        axis.title.y=element_text(size=14))+
  scale_fill_manual(values=c('pink','#F8766D','red','#00BA38','#619CFF'))

I thought it might be able to quickly solve this problem by altering the order of the contents in the dataset itself, moving the 'Undisturbed' terms and related values (e.g. emean_Means and emean_SE) to the front of their respective lines within the code (see below), but this did not change the location of the corresponding bars on the graph.

emean_Sleep_State<-c('WAKE','SWS','REM','WAKE','SWS','REM','WAKE','SWS','REM','WAKE','SWS','REM','WAKE','SWS','REM')
emean_Treatment<-c('UnDisturbed','UnDisturbed','UnDisturbed','Disturbed_Feed','Disturbed_Feed','Disturbed_Feed','Disturbed_Foot','Disturbed_Foot','Disturbed_Foot','Disturbed_Temp','Disturbed_Temp','Disturbed_Temp','Recovery','Recovery','Recovery')
emean_Means<-c(28.62,52.84,21.01,27.68,52.64,21.36,35.38,47.96,15.59,42.28,45.84,6.03,27.56,52.2,22.36)
emean_SE<-c(5.75,2.92,4.18,3.94,2.99,3.08,3.97,3.01,3.1,3.84,2.91,3.03,2.94,2.34,2.51)
Estimated_Max_Means_emean<-emeanframe$emean_Means+emeanframe$emean_SE
Estimated_Min_Means_emean<-emeanframe$emean_Means-emeanframe$emean_SE
emeanframe<-data.frame(emean_Sleep_State,emean_Treatment,emean_Means,emean_SE,Estimated_Max_Means_emean,Estimated_Min_Means_emean)
realvalueemean<-apply(emeanframe[,c(3,5,6)],2,FUN=backtransANG)
emeanframe$realmean<-realvalueemean[,1]
emeanframe$realmaxmean<-realvalueemean[,2]
emeanframe$realminmean<-realvalueemean[,3]

I have tried various commands such as levels, and scale_x_manual but these have either had no effect, have resulted in an error, or have overwritten my x axis.

You need to make the variable a factor, and specify the order of levels you want: `emeanframe$emean_Sleep_State <- factor(emeanframe$emean_Sleep_State, c("WAKE", "SWS", "REM"))`. — zephryl, Jun 13 '23 at 12:27
If you just want to move one level to the front, you could also use `relevel(factor(x, "FIRST"))` or `forcats::fct_relevel(x, "FIRST")`. — zephryl, Jun 13 '23 at 12:35

score 1 · Accepted Answer · answered Jun 13 '23 at 12:33

You just need to change the variable emean_Treatment from character to factor, and specify the factor levels in the order you want them. Then ggplot2 will respect that order in the factor variable.

So the relevant part in the code below that does this is:

type_of_night_levels <- c(
  "UnDisturbed",
  "Disturbed_Feed",
  "Disturbed_Foot",
  "Disturbed_Temp",
  "Recovery"
)

emeanframe2 <- 
  emeanframe |>
  mutate(emean_Treatment = factor(emean_Treatment, levels = type_of_night_levels))

library(tidyverse)
  
emean_Sleep_State<-c('WAKE','SWS','REM','WAKE','SWS','REM','WAKE','SWS','REM','WAKE','SWS','REM','WAKE','SWS','REM')
emean_Treatment<-c('Disturbed_Feed','Disturbed_Feed','Disturbed_Feed','Disturbed_Foot','Disturbed_Foot','Disturbed_Foot','Disturbed_Temp','Disturbed_Temp','Disturbed_Temp','Recovery','Recovery','Recovery','UnDisturbed','UnDisturbed','UnDisturbed')
emean_Means<-c(27.68,52.64,21.36,35.38,47.96,15.59,42.28,45.84,6.03,27.56,52.2,22.36,28.62,52.84,21.01)
emean_SE<-c(3.94,2.99,3.08,3.97,3.01,3.1,3.84,2.91,3.03,2.94,2.34,2.51,5.75,2.92,4.18)

Estimated_Max_Means_emean<-emean_Means+emean_SE
Estimated_Min_Means_emean<-emean_Means-emean_SE

emeanframe<-data.frame(emean_Sleep_State,emean_Treatment,emean_Means,emean_SE,Estimated_Max_Means_emean,Estimated_Min_Means_emean)

backtransANG<-function(ANG_value){
  real_value<-(sin(ANG_value/(180/pi)))^2
  return(real_value)
}

realvalueemean<-apply(emeanframe[,c(3,5,6)],2,FUN=backtransANG)
emeanframe$realmean<-realvalueemean[,1]
emeanframe$realmaxmean<-realvalueemean[,2]
emeanframe$realminmean<-realvalueemean[,3]

type_of_night_levels <- c(
  "UnDisturbed",
  "Disturbed_Feed",
  "Disturbed_Foot",
  "Disturbed_Temp",
  "Recovery"
)

emeanframe2 <- 
  emeanframe |>
  mutate(emean_Treatment = factor(emean_Treatment, levels = type_of_night_levels))

emeanframe2 |>
  ggplot(aes(x = emean_Sleep_State,
             y = realmean,
             fill = emean_Treatment)) +
  geom_col(position = 'dodge',
           color = 'black') +
  geom_errorbar(
    aes(ymin = realminmean,
        ymax = realmaxmean),
    color = 'black',
    width = 0.2,
    position = position_dodge(width = 0.9),
    linewidth = 1
  ) +
  scale_x_discrete(limits = c('WAKE', 'SWS', 'REM')) +
  labs(x = 'Sleep State',
       y = 'Proportion of Behaviour',
       color = "Type of Night",
       size = 14) +
  ylim(0, 1) +
  guides(fill = guide_legend(title = 'Type of Night')) +
  theme(
    axis.text.x = element_text(size = 12),
    axis.text.y = element_text(size = 12),
    axis.title.x = element_text(size = 14),
    axis.title.y = element_text(size = 14)
  ) +
  scale_fill_manual(values = c('pink', '#F8766D', 'red', '#00BA38', '#619CFF'))

How do I reorder a single column within ggplot2?

1 Answers1