r - 如何将具有不同数据类型的嵌套列表转换为数据框？

Question

我有一个包含不同类型的嵌套元素的列表

library(tidyverse)

# list to be transfomed into a dataframe/tibble
mylist <- 
  list(
    lois = list(
      hair = list(color = "orange", form = "flat"),
      sex = "female"
    )
  )

# structure
str(mylist)
#> List of 1
#>  $ lois:List of 2
#>   ..$ hair:List of 2
#>   .. ..$ color: chr "orange"
#>   .. ..$ form : chr "flat"
#>   ..$ sex : chr "female"

目标是将此列表转换为数据框/tibble。期望的输出是

# A tibble: 3 x 4
  name  value_id attribute text  
  <chr> <chr>    <chr>     <chr> 
1 lois  hair     color     orange
2 lois  hair     form      flat  
3 lois  sex      NA        female

我尝试了 tidyr::unnest_longer()。但是很难取消嵌套，因为 value 列中有不同的类型:

# unnest_longer does not work like this
mylist %>% 
  enframe() %>% 
  unnest_longer(col = value) %>% 
  unnest_longer(col = value)
#> Error: Can't combine `..1$value` <list> and `..2$value` <character>.

^{由 reprex package 创建于 2021-01-13 (v0.3.0)}

这个问题有什么好的解决方案？

Answer 1

一种选择是使用rrapply 库:

rrapply(mylist, how = "melt")

    L1   L2    L3  value
1 lois hair color orange
2 lois hair  form   flat
3 lois  sex  <NA> female