pivot_longer 用于连续的多个观察

Question

我有一个数据集，其中包含 1 行内的多个观察值，我试图将其重塑为长格式。这是数据的样子

    df_have <- data.frame(group = c(1,2), p1_age = c(18,30), p1_sex = c("male", "male"),
                   p2_age = c(22,51), p2_sex = c("female", "male"))

  group p1_age p1_sex p2_age p2_sex
1     1     18   male     22 female
2     2     30   male     51   male

我想要重塑的人和变量会更多。我想重塑它，使它看起来像这样：

    df_want <- data.frame(person = c(1,2,1,2), group = c(1,1,2,2), age = c(18,22,30,51),
                   sex = c("male", "female", "male", "male"))
  person group age    sex
1      1     1  18   male
2      2     1  22 female
3      1     2  30   male
4      2     2  51   male

我尝试在 pivot_longer 步骤中使用“.value”来尝试生成我想要的内容，但我无法弄清楚。

Answer 1

我们可以使用pivot_longer 并指定names_pattern 来捕获_ 后面的数字（\\d+）和单词（\\w+）

library(dplyr)
library(tidyr)
df_have %>%
     pivot_longer(cols = -group, names_to = c("person", ".value"),
       names_pattern = ".(\\d+)_(\\w+)")%>% 
     mutate(person = as.integer(person))

-输出

# A tibble: 4 x 4
  group person   age sex   
  <dbl> <int>  <dbl> <chr> 
1     1 1         18 male  
2     1 2         22 female
3     2 1         30 male  
4     2 2         51 male  

Answer 2

library(tidyverse)
pivot_longer(df_have, -group, names_to = c("person", ".value"), names_sep = "_")

# A tibble: 4 x 4
  group person    age sex   
  <dbl>  <chr> <dbl> <chr> 
1     1  p1       18 male  
2     1  p2       22 female
3     2  p1       30 male  
4     2  p2       51 male