Sort a bar plot based on two conditions in ggplot

Question

I would like to create a bar plot that is arranged into groups and within those groups, sorted from the *minimum to maximum value.

To recreate the example I will use the following data frame

df <- data.frame(
  stringsAsFactors = FALSE,
             Sites = c("Site 1","Site 2","Site 3",
                       "Site 4","Site 5","Site 6","Site 7","Site 8","Site 9",
                       "Site 10","Site 11"),
            Values = c(184.7955548,171.1466314,
                       245.5952181,188.3072784,259.9438698,210.3448318,
                       173.7977541,182.5497301,198.7985429,188.0458496,215.5709303),
            Groups = c(1, 1, 3, 3, 2, 3, 1, 3, 3, 2, 2))

For the plot I used:

df %>% arrange(Groups, Values) %>%
  mutate(name=factor(Groups, levels = Values)) %>%
  ggplot(aes(x = df$Sites, y = df$Values))+
  geom_bar(stat = "identity", fill = df$Groups)+
  scale_color_manual(values = c ('royalblue1', 'slategrey2', 'yellow1'))+
  ylab("Values")+
  xlab("")+
  theme(axis.text.x = element_text(angle = 90, hjust = 1))

that resulted into:

but what I was expecting is the following:

Any help would be appreciated

Answer 1

Couple of issues in the code -

• name=factor(Groups, levels = Values) gives all NA's. levels should be the value present in the data.
• We don't need $ in ggplot code. Also df$Sites does not have the factor levels that we need. The factor levels are added in the piped data and not in the original data.

library(dplyr)
library(ggplot2)

df %>% 
    arrange(Groups, Values) %>%
    mutate(Sites=factor(Sites, levels = Sites), 
           Groups = factor(Groups)) %>% 
    ggplot(aes(x = Sites, y = Values, fill = Groups)) +
    geom_bar(stat = "identity")+
    scale_fill_manual(values = c ('royalblue1', 'grey2', 'yellow1'))+
    ylab("Values")+
    xlab("")+
    theme(axis.text.x = element_text(angle = 90, hjust = 1))