Using purrr and map to loop over deeper layers of a list?

Question

I'm trying to see how to extract values from the nth level of a nested list. For example...if these were all on the second layer, I could just use map() like so:

> testlist1 <- 
+   list(
+     list('veggie' = 'apple', 'food' = 'bacon'),
+     list('veggie' = 'banana', 'food' = 'burger'),
+     list('veggie' = 'tomato', 'food' = 'sausage'),
+     list('veggie' = 'pickle', 'food' = 'chicken'),
+     list('veggie' = 'chestnut', 'food' = 'salmon'),
+     list('veggie' = 'seaweed', 'food' = 'tuna')
+   )
> 
> testlist1 %>% map('veggie')
[[1]]
[1] "apple"

[[2]]
[1] "banana"

[[3]]
[1] "tomato"

[[4]]
[1] "pickle"

[[5]]
[1] "chestnut"

[[6]]
[1] "seaweed"

But when they are deeper, I imagine there must be some map() method of looping to the layers? Example:

testlist2 <- list(
  list(
    list(
      list('veggie' = 'apple', 'food' = 'bacon')
    ), 
    list(
      list('veggie' = 'banana', 'food' = 'burger')
    )
  ),
  list(
    list(
      list('veggie' = 'tomato', 'food' = 'sausage')
    ), 
    list(
      list('veggie' = 'pickle', 'food' = 'chicken')
    )
  ),
  list(
    list(
      list('veggie' = 'chestnut', 'food' = 'salmon')
    ), 
    list(
      list('veggie' = 'seaweed', 'food' = 'tuna')
    )
  )
)

Are you looking for something like map_depth(testlist2, 3, "veggie") %>% unlist() in this case? — MrFlick
– MrFlick, Commented Nov 18, 2019 at 3:47

pgcudahy · Accepted Answer · 2019-11-18 14:38:27Z

As @MrFlick suggested, map_depth can help here

library(tidyverse)

testlist2 <- list(
  list(
    list(
      list('veggie' = 'apple', 'food' = 'bacon')
    ), 
    list(
      list('veggie' = 'banana', 'food' = 'burger')
    )
  ),
  list(
    list(
      list('veggie' = 'tomato', 'food' = 'sausage')
    ), 
    list(
      list('veggie' = 'pickle', 'food' = 'chicken')
    )
  ),
  list(
    list(
      list('veggie' = 'chestnut', 'food' = 'salmon')
    ), 
    list(
      list('veggie' = 'seaweed', 'food' = 'tuna')
    )
  )
)

Use the first argument to set the desired list depth

testlist2 %>% map_depth(3, 'veggie')
#> [[1]]
#> [[1]][[1]]
#> [[1]][[1]][[1]]
#> [1] "apple"
#> 
#> 
#> [[1]][[2]]
#> [[1]][[2]][[1]]
#> [1] "banana"
#> 
#> 
#> 
#> [[2]]
#> [[2]][[1]]
#> [[2]][[1]][[1]]
#> [1] "tomato"
#> 
#> 
#> [[2]][[2]]
#> [[2]][[2]][[1]]
#> [1] "pickle"
#> 
#> 
#> 
#> [[3]]
#> [[3]][[1]]
#> [[3]][[1]][[1]]
#> [1] "chestnut"
#> 
#> 
#> [[3]][[2]]
#> [[3]][[2]][[1]]
#> [1] "seaweed"

Or use negative numbers to count up from the lowest level

testlist2 %>% map_depth(-2, 'veggie')
#> [[1]]
#> [[1]][[1]]
#> [[1]][[1]][[1]]
#> [1] "apple"
#> 
#> 
#> [[1]][[2]]
#> [[1]][[2]][[1]]
#> [1] "banana"
#> 
#> 
#> 
#> [[2]]
#> [[2]][[1]]
#> [[2]][[1]][[1]]
#> [1] "tomato"
#> 
#> 
#> [[2]][[2]]
#> [[2]][[2]][[1]]
#> [1] "pickle"
#> 
#> 
#> 
#> [[3]]
#> [[3]][[1]]
#> [[3]][[1]][[1]]
#> [1] "chestnut"
#> 
#> 
#> [[3]][[2]]
#> [[3]][[2]][[1]]
#> [1] "seaweed"

Created on 2019-11-18 by the reprex package (v0.3.0)

Ronak Shah · Accepted Answer · 2019-11-18 04:12:36Z

1

How about unlisting the data and select the elements with the required name ?

temp <- unlist(testlist2)
unname(temp[names(temp) == "veggie"])
#[1] "apple"    "banana"   "tomato"   "pickle"   "chestnut" "seaweed"

This will work even when data is multiple layers deep.

answered Nov 18, 2019 at 4:12

Ronak Shah

391k20 gold badges173 silver badges237 bronze badges

Collectives™ on Stack Overflow

Using purrr and map to loop over deeper layers of a list?

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related