计算多列中每组不同变量的数量？

Question

我有一个数据框，我想计算每组不同观察值的数量，不计算 NA 值。

以下是数据示例：

ID <-c("A", "A", "B", "B", "B", "C")
Act1 <- c("Football", "Swim", "Football", 'Basketball', "Swim", "Tennis")
Act2 <- c("Swim", "Football", "Tennis", 'Swim', "Football", "Swim")
Act3 <- c("NA", "Tennis", "NA", 'Football', "Tennis", "NA")
df <- data.frame(ID,Act1, Act2, Act3)

df

   ID       Act1     Act2     Act3
1  A   Football     Swim       NA
2  A       Swim Football   Tennis
3  B   Football   Tennis       NA
4  B Basketball     Swim Football
5  B       Swim Football   Tennis
6  C     Tennis     Swim       NA

正确答案应该是这样的...

因为 A 有三种不同的 活动（例如足球、游泳、网球），B 有四种（例如足球、游泳、网球、篮球），C 有两种（例如网球）和游泳）

我该怎么做？

Answer 1

假设空值实际上是 NA 值而不是字符串 "NA"，您可以使用包 dplyr 和 tidyr 来实现您的预期输出

library(dplyr)
library(tidyr)

df %>% 
  pivot_longer(-ID) %>% 
  filter(!is.na(value)) %>%   # if you have strings "NA" use   filter(value != "NA")   
  group_by(ID) %>%
  summarise(n = n_distinct(value))

# A tibble: 3 x 2
#   ID        n
#   <chr> <int>
# 1 A         3
# 2 B         4
# 3 C         2

计算多列中每组不同变量的数量？

Counting the number of different variables per group over multiple columns?

r

dataframe

dplyr

tidyr

summarize