这里有一种使用factor艾德方法的方法-
我们创建了两个新列n和n1，其中n1是每个number_petals中n值的sum。arrange按n1对数据进行排序，并根据其出现情况创建因子水平。

library(dplyr)
library(ggplot2)

iris_new %>% 
  group_by(Species, number_petals) %>%
  summarise(n=sum(Petal.Length, na.rm=TRUE), .groups = "drop") %>%
  group_by(number_petals) %>%
  mutate(n1 = sum(n)) %>%
  arrange(desc(n1)) %>%
  ungroup() %>%
  mutate(number_petals = factor(number_petals, unique(number_petals))) %>%
  ggplot(aes(x=number_petals,y=n,fill=Species)) +
  xlab("Number of Petals")+
  ylab("Total sum of petal lengths") +
  geom_col()

head(25)将选择前25行，而不考虑分组。如果要按组选择顶部行，请查看?slice_max或?slice。
要选择n（这里是5个）值，这里有一个使用连接的不同方法。

iris_new %>% 
  group_by(number_petals) %>%
  summarise(n=sum(Petal.Length, na.rm=TRUE), .groups = "drop") %>%
  slice_max(n, n = 5) %>%
  inner_join(iris_new %>% 
  group_by(Species, number_petals) %>%
  summarise(n1=sum(Petal.Length, na.rm=TRUE), .groups = "drop"), 
            by = 'number_petals') %>%
  arrange(desc(n)) %>%
  mutate(number_petals = factor(number_petals, unique(number_petals))) %>%
  ggplot(aes(x=number_petals,y=n1,fill=Species)) +
  xlab("Number of Petals")+
  ylab("Total sum of petal lengths") +
  geom_col()