Rstudio: Adjusting bar width of grouped bar plot in facets

Question

I have a Data set which I am trying to plot as grouped columns in facets, with the columns in equal size.

Data set:

Day=rep(1:6, times=15)

Colour=c("Yellow", "Yellow","Yellow","Yellow","Yellow","Yellow", "Red", "Red","Red","Red","Red","Red", "Green","Green","Green","Green","Green","Green", "Yellow", "Yellow","Yellow","Yellow","Yellow","Yellow", "Red", "Red","Red","Red","Red","Red", "Green","Green","Green","Green","Green","Green", "Blue","Blue","Blue","Blue","Blue","Blue", "Purple","Purple","Purple","Purple","Purple","Purple", "Yellow", "Yellow","Yellow","Yellow","Yellow","Yellow", "Red", "Red","Red","Red","Red","Red", "Green","Green","Green","Green","Green","Green", "Yellow", "Yellow","Yellow","Yellow","Yellow","Yellow", "Red", "Red","Red","Red","Red","Red", "Green","Green","Green","Green","Green","Green", "Yellow", "Yellow","Yellow","Yellow","Yellow","Yellow")

Values=rep(c(9,8,7,6,5,8,7,6,5,4,7,6,5,4,3), times=6)

Fruit=rep(c("CApple", "Banana", "ABlueberry","Mango", "Melon", "Pear"), times = c(18,18,12,18,18,6))

Data <-data.frame(Day, Fruit, Colour, Values) %>%
mutate(unten=Values-0.2, oben=Values+0.2)

My code is:

ForPlot <-  ggplot(Data, aes(Colour, Values), fill=Colour)

design <- theme(strip.background = element_blank(),
                 panel.background = element_blank(),  
                 panel.grid.major = element_blank(),
                 strip.placement = "outside",
                 panel.border=element_blank(),
                 #     axis.line = element_line(colour = "black"),
                 axis.line.x = element_line(colour = "white"),
                 axis.ticks.x=element_blank(),
                 legend.title = element_blank(),
                 # axis.text=element_text(size=8),
                 axis.text.x=element_blank(),
                 axis.title=element_text(size=10),
                 strip.text = element_text(size = 10)) 

  ForPlot+
  geom_hline(yintercept = c(0, 2.5, 5, 7.5, 10), colour= "lightgrey" )+
  geom_col(aes(fill=Colour), 
           position=position_dodge2(preserve="single"))+
  ggtitle("Fruity Colours")+
  ylab("Values") + xlab("Day") +
  facet_wrap(~Fruit,  scales= "free", ncol=3)+
  scale_y_continuous()+
  scale_fill_discrete()+
  geom_text(aes(label = Day), position=position_dodge2(width= 0.9), 
            y=-0.2, size=2.4) +
  geom_errorbar(aes(ymin=pmax(0,unten), ymax=oben), 
                position=position_dodge2())+
  design+
  geom_segment(aes(x=0.4,xend=Inf, y=0, yend=0), color="black")+
  geom_segment(aes(x=0.4,xend=0.4, y=0, yend=Inf), colour="black")

Currently, the bars in the different facets all have different width. I have tried to adjust the problem using "width=" but without success (bars only shrink relative to their current size or I loose my Day labels and errorbars).

Help is appreciated!

The `space` argument (available in `facet_grid()` but not `facet_wrap()`) might help here: `facet_grid(~Fruit, scales= "free", space = "free")`. Is that more what you want? The bars are the same width but, of course, the facets themselves become different sizes. — aosmith, Jul 31 '20 at 17:00

Duck · Accepted Answer · 2020-07-31T19:33:40.107

Pointing to the useful comment of @aosmith this can help:

ForPlot+
  geom_hline(yintercept = c(0, 2.5, 5, 7.5, 10), colour= "lightgrey" )+
  geom_col(aes(fill=Colour), 
           position=position_dodge2(preserve="single"))+
  ggtitle("Fruity Colours")+
  ylab("Values") + xlab("Day") +
  facet_grid(.~Fruit,  scales= "free", space = 'free')+
  scale_y_continuous()+
  scale_fill_discrete()+
  geom_text(aes(label = Day), position=position_dodge2(width= 0.9), 
            y=-0.2, size=2.4) +
  geom_errorbar(aes(ymin=pmax(0,unten), ymax=oben), 
                position=position_dodge2())+
  design+
  geom_segment(aes(x=0.4,xend=Inf, y=0, yend=0), color="black")+
  geom_segment(aes(x=0.4,xend=0.4, y=0, yend=Inf), colour="black")

Update: You can get more space placing legend on top with theme(legend.position = 'top') I used your new data:

Update 2: With the new data, this is the closer I get about what you want (Just be careful about the colors in scale_fill_manual()), try this code:

library(tidyverse)
library(patchwork)
#Create a split
Fruit <- unique(Data$Fruit)
Key <- sort(rep(c(1,2),length(Fruit)/2))
DKeys <- data.frame(Fruit,Key)

#Join data for splits
Data2 <- Data %>% left_join(DKeys)

#Now split
List <- split(Data2,Data2$Key)

#Plot function
myplot <- function(x)
{
  ForPlot <-  ggplot(x, aes(Colour, Values), fill=Colour)
  
  design <- theme(strip.background = element_blank(),
                  panel.background = element_blank(),  
                  panel.grid.major = element_blank(),
                  strip.placement = "outside",
                  panel.border=element_blank(),
                  #     axis.line = element_line(colour = "black"),
                  axis.line.x = element_line(colour = "white"),
                  axis.ticks.x=element_blank(),
                  legend.title = element_blank(),
                  # axis.text=element_text(size=8),
                  axis.text.x=element_blank(),
                  axis.title=element_text(size=10),
                  strip.text = element_text(size = 10),
                  legend.position = 'top') 
  
  
  ForPlot <- ForPlot+
    geom_hline(yintercept = c(0, 2.5, 5, 7.5, 10), colour= "lightgrey" )+
    geom_col(aes(fill=Colour), 
             position=position_dodge2(preserve="single"))+
    ylab("Values") + xlab("Day") +
    facet_grid(.~Fruit,  scales= "free", space = 'free')+
    scale_y_continuous()+
    # scale_fill_discrete()+
    geom_text(aes(label = Day), position=position_dodge2(width= 0.9), 
              y=-0.2, size=2.4) +
    geom_errorbar(aes(ymin=pmax(0,unten), ymax=oben), 
                  position=position_dodge2())+
    design+
    geom_segment(aes(x=0.4,xend=Inf, y=0, yend=0), color="black")+
    geom_segment(aes(x=0.4,xend=0.4, y=0, yend=Inf), colour="black")+
    scale_fill_manual(values=c(Blue='Blue',Green='Green',Purple='Purple',Red='Red',Yellow='Yellow'))
  return(ForPlot)
}

#Apply to plots
List2 <- lapply(List,myplot)
#Format 
List2[2:length(List2)] <- lapply(List2[2:length(List2)], function(x) {x<-x+theme(legend.position = 'none')})
names(List2) <- paste0('G',1:length(List))
list2env(List2,envir = .GlobalEnv)
#Arrange plots
G1/G2+ plot_annotation(title = 'Fruity Colours')

Hi, thanks for your quick reply. I apologize, my data set wasn't aqurate enough (it is adjusted now). I have enough data that it has to shows up in two rows while faceting on only one variable. Is there still a way to get this working using facet grid? — BeccaLi, Jul 31 '20 at 17:29
@BeccaLi You can add `+theme(legend.position = 'top')` at the end. It will place the legend on top and I checked on the plot and you have more space. Is that valid for what you want? — Duck, Jul 31 '20 at 17:43
Sadly no, this is just an example data set. The actual Data still has too many bars to be placed next to each other (15 groups with 6 bars each, the groups are distributed to 6 facets: 2,3,3,3,3,1 ) — BeccaLi, Jul 31 '20 at 17:54
@BeccaLi Understood. I could suggest you try splitting in two different plots and then you can merge with patchwork. If you could add to the example data all the groups you mention I could do for you. — Duck, Jul 31 '20 at 17:59
Thanks a lot for your help! I adjusted the sample set again. — BeccaLi, Jul 31 '20 at 18:24
@BeccaLi I have added an update. It is the closer I get about what you want! I hope this can help! — Duck, Jul 31 '20 at 19:35

Rstudio: Adjusting bar width of grouped bar plot in facets

1 Answers1