Calculate percentage of occurrence of a factor of the total amount of occurrences of a group factor

I have a sample dataset:

Species <-c("Bass", "Bass", "Bass", "Bass", "Bass", "Bass","Bass","Bass","Bass")
FishID <- c("a1", "a1", "a1", "a2", "a2", "a3","a3","a3","a3")
Prey <- c("Amphipoden", "Mysis", "Polychaeten", "Amphipoden", "Mysis", "Amphipoden","Mysis","Polychaeten","Mollusca")

df <- data.frame(Species, FishID, Prey)

For having Bass as a predator, there are 3 unique individual Basses as different FishID: a1, a2 and a3. I would like to calculate the absolute percentage of occurrence of a prey species per FishID (individual Bass).

So in this case:
Amphipods occurs 3 times, so 100% in the stomachs of Bass (found in all three of the individuals), for Mysis idem. For polychaete however, is found only two times in the stomach of Bass: so this would be then 66,6%. And Moluscs are only found one time, so 33,3 %

As an end result, I am looking for something like this:

Species <-c("Bass", "Bass", "Bass", "Bass")
Prey <- c("Amphipoden", "Mysis", "Polychaeten", "Mollusca")
Percentage <- c(100, 100, 66,6, 33,3)
df2 <- data.frame(Species,Prey, Percentage)

I tried this:

df %>%
  group_by(Species,Prey) %>% 
  summarise(n = n()) %>%
  mutate(percent = n / sum(n) * 100)

But it isn't giving me hat I want.

Anty help is welcome.

Thank you in advance!

You just have to change a little point: Instead of dividing by sum(n), you have to divide by length(unique(FishID)) in order to get the correct number of individual FishID. Also note that the last element of FishID has to be a3, not A3.

library(dplyr)

FishID <- c("a1", "a1", "a1", "a2", "a2", "a3","a3","a3","a3")

df %>%
    summarise(n = n(), .by = Prey) %>%
    mutate(percent = n / length(unique(FishID)) * 100)

         Prey n   percent
1  Amphipoden 3 100.00000
2       Mysis 3 100.00000
3 Polychaeten 2  66.66667
4    Mollusca 1  33.33333

</details>



# 答案2
**得分**: 1

```R
library(tidyverse)
df |&gt;
  # 只计算每个种类/鱼ID一次的猎物（我假设它们总是一起的）
  distinct(Species, FishID, Prey) |&gt;
  mutate(count = 1) |&gt;
  # 为每个鱼ID完成缺失的组合
  complete(nesting(Species, FishID), Prey, fill = list(count = 0)) |&gt; 
  summarize(percent = sum(count) / n(), .by = c(Species, Prey))


# 一个数据表：4 × 3
  种类    猎物        百分比
1 鲈鱼    介形虫    1    
2 鲈鱼    软体动物  0.333
3 鲈鱼    沙蚕    1    
4 鲈鱼    多毛类  0.667

library(tidyverse)
df |>
  # only count Prey once per Species/FishID (which I presume always go together)
  distinct(Species, FishID, Prey) |>
  mutate(count = 1) |>
  # complete with missing combinations for each FishID
  complete(nesting(Species, FishID), Prey, fill = list(count = 0)) |> 
  summarize(percent = sum(count) / n(), .by = c(Species, Prey))


# A tibble: 4 × 3
  Species Prey        percent
  <chr>   <chr>         <dbl>
1 Bass    Amphipoden    1    
2 Bass    Mollusca      0.333
3 Bass    Mysis         1    
4 Bass    Polychaeten   0.667