具有多因素条件的行和

Question

我试图对具有不同条件的行求和，但我在这部分遇到了困难。

这是我的数据库的一个类似示例：

fl<-c("rose", "sunflower", "rose", "sunflower", "lily", "sunflower", "rose", "sunflower")
an<-c("sunbird", "beetle", "beetle", "bee", "bee", "bee", "fly", "bat")
gr<-c("bird", "insect", "insect", "insect", "insect", "insect", "insect", "mammal")
n<-c(2, 3, 7, 15, 23, 48,11, 3)
df<-data.frame(fl, an, gr, n)

我想要的是用相同类型的访问者计算来自相同花种的n(gr列)，甚至在相关列( an )中有两个条目:例如，我想要的最终产品是一个表，给我一个表，其中有3种不同的os动物(蜜蜂、甲虫和蝙蝠)访问向日葵，66次昆虫访问(所有动物的总和标记为‘’昆虫‘’)和3次来自哺乳动物。

我不知道我是否恰当地表达了自己的意思，但我会感谢任何形式的帮助

Answer 1

library(tidyverse)

fl<-c("rose", "sunflower", "rose", "sunflower", "lily", "sunflower", "rose", "sunflower")
an<-c("sunbird", "beetle", "beetle", "bee", "bee", "bee", "fly", "bat")
gr<-c("bird", "insect", "insect", "insect", "insect", "insect", "insect", "mammal")
n<-c(2, 3, 7, 15, 23, 48,11, 3)
df<-data.frame(fl, an, gr, n, stringsAsFactors = FALSE)

df %>%
    group_by(fl, an) %>%
    summarise(howMany = sum(n))

Answer 2

我相信@georgery right..being是tidyverse的粉丝，我更喜欢他的方法而不是@arkun。

以下是对他的解决方案的略微改进：

library(dplyr)
fl<-c("rose", "sunflower", "rose", "sunflower", "lily", "sunflower", "rose", "sunflower")
an<-c("sunbird", "beetle", "beetle", "bee", "bee", "bee", "fly", "bat")
gr<-c("bird", "insect", "insect", "insect", "insect", "insect", "insect", "mammal")
n<-c(2, 3, 7, 15, 23, 48,11, 3)
df<-data.frame(fl, an, gr, n, stringsAsFactors = FALSE)

df %>% 
  group_by(fl, gr) %>%
  summarise(category = sum(n), typewithincategory=length(unique(an)))

结果将是一个tibble

  fl        gr     category typewithincategory
  <chr>     <chr>     <dbl>              <int>
1 lily      insect       23                  1
2 rose      bird          2                  1
3 rose      insect       18                  2
4 sunflower insect       66                  2
5 sunflower mammal        3                  1