Ordered bar plot with multiple groupings-CodePudding

Example data are .

The correct figure should be grouped by Class and then otherwise the bars should be in descending order - essentially exactly the same as example 1 above.

CodePudding user response：

ggplot automatically order alphabetically. If you want a different order, you can order the levels yourself.

df$Family <- factor(df$Family, levels = c("Muridae","Mustelidae",....))

I would also advise you to use facets to group the Classes: ggplot(...) facet_wrap("~Animal.group")

# randomize the order
df$Family <- factor(cdfiao$Family,
                      levels = sample(unique(df$Family)))


ggplot(df, aes(x = Family, y = n, fill = Animal.group))  
   facet_wrap("~Animal.group" ,,scales = "free_y", ncol =1)  
   geom_bar(stat = "identity") 
   scale_y_continuous(breaks = c(0, 250, 500, 750, 1000, 1250))  
   theme_classic()  
   labs(x = "Family", y = "Number", fill = "Class")  
   theme(axis.title.x = element_text(face = "bold"),
       axis.title.y = element_text(face = "bold"),
       axis.text.x = element_text(angle = 45, hjust = 1),
       axis.text.y = element_text(),
       legend.title = element_text(face = "bold"))

CodePudding user response：

The problem ended up being in the dataset itself - the dataset was at the level of n.bp.sp$Common.name, by grouping the dataset at the level of n.bp.sp$Family one of @fra previous answers then works fine.

Code below:

n.bp.f2 <- n.bp.sp %>% 
  group_by(Animal.group, Family) %>%
  summarise(n = sum(n))

n.bp.f2 %>%
  arrange(Animal.group, desc(n)) %>%
  mutate(Family = fct_inorder(Family)) %>%
  ggplot()   
  aes(x = Family, y = n, fill = Animal.group)   
  geom_bar(stat = "identity") 
  scale_y_continuous(breaks = c(0, 250, 500, 750, 1000, 1250))  
  theme_classic() 
  labs(x = "Family", y = "Number", fill = "Class")  
  theme(axis.title.x = element_text(face = "bold"),
        axis.title.y = element_text(face = "bold"),
        axis.text.x = element_text(angle = 45, hjust = 1),
        axis.text.y = element_text(),
        legend.title = element_text(face = "bold"))