分组条形图，R中的ggplot2，需要一次显示一个观察

Question

I have a dataset that looks like the following:

df <- data.frame(Name=rep(c('Sarah', 'Casey', 'Mary', 'Tom'), 3), 
                 Scale=rep(c('Scale1', 'Scale2', 'Scale3'), 4), 
                 Score=sample(1:7, 12, replace=T))

I am trying to create a barchat in ggplot2 that currently looks like this:

ggplot(df, aes(x=Name, y=Score, fill=Scale)) + geom_bar(stat='identity', position='dodge') +
  coord_flip() +
  scale_y_continuous(breaks=seq(0, 7, 1), limits = c(0, 7)) +  
  scale_x_discrete() +    
  scale_fill_manual(values=c('#253494', '#2c7fb8', '#000000')) + 
  theme(panel.background = element_blank(),
        legend.position = 'right',  
        axis.line = element_line(),  
        axis.title = element_blank(),    
        axis.text = element_text(size=10)) 

However, I only want to show one observation (one Name) at a time. Is this possible to do without creating a ton of separate datasets, one for each person? I would like the end result to look like the example below, where I can just iterate through the names to produce a separate plot for each, or some similar process.

# Trying to avoid creating separate datasets, but for the sake of the example:
df2 <- data.frame(Name=rep(c('Sarah'), 3), 
                 Scale=c('Scale1', 'Scale2', 'Scale3'), 
                 Score=sample(1:7, 3, replace=T))

ggplot(df2, aes(x=Name, y=Score, fill=Scale)) + geom_bar(stat='identity', position='dodge') +
  coord_flip() +
  scale_y_continuous(breaks=seq(0, 7, 1), limits = c(0, 7)) +  
  scale_x_discrete() +    
  scale_fill_manual(values=c('#253494', '#2c7fb8', '#000000')) + 
  theme(panel.background = element_blank(),
        legend.position = 'right',  
        axis.line = element_line(),  
        axis.title = element_blank(),    
        axis.text = element_text(size=10)) 

Answer 1

Since your data is already tidy ie. in long format, you can use facet_wrap as suggested and set the scales as "free" thus creating facets with your different Name groups.

df %>% ggplot(aes(y = Score, x = Name)) +
  geom_bar(stat = "identity", aes(colour = Scale, fill = Scale),
           position = "dodge") +
  coord_flip() + 
  facet_wrap(~Name, scales = "free")

You can get rid of the facet labels or the axis labels depending which you prefer.

EDIT: in response to comment. You can use the same data frame to create seperate plots by just piping a filter in at the start, hence,

df %>% 
  filter(Name == "Sarah") %>%
  ggplot(aes(y = Score, x = Name)) +
  geom_bar(stat = "identity", aes(colour = Scale, fill = Scale),
           position = "dodge") +
  coord_flip()

Since you are using Rmarkdown you could throw a for loop around that to plot all the names

for(i in c("Sarah", "Casey", "Mary", "Tom")){
  df %>% 
    filter(Name == i) %>%
    ggplot(aes(y = Score, x = Name)) +
    geom_bar(stat = "identity", aes(colour = Scale, fill = Scale),
             position = "dodge") +
    coord_flip()
}

If you want to arrange all these into a group you can use ggpubr::ggarrange to place all the plots into the same object.

Answer 2

• facet_grid(.~Name)

Maybe somehow implement this, it'll plot them all, but should do so in individual plots.