Given two tibbles, I want to assign one tibble to be a cell in the other tibble. However, the current approach consistently converts my sub-tibble to a different format, which leads to errors.
I know what I want to do is possible:
library(dplyr)
desired = tibble(
row = 1:4,
nested = list(
tibble(text = c("a","b","c")),
tibble(numbers = 1:4),
tibble(label = c("high", "med", "low"), value = c(10, 5, 0)),
NA
)
)
This code produces a tibble with two columns. The second column contains tibbles of different sizes and contents.
However, I can not construct this in incremental steps:
library(dplyr)
# outer tibble
df = tibble(row = 1:4)
# inner tibbles
t1 = tibble(text = c("a","b","c"))
t2 = tibble(numbers = 1:4)
t3 = tibble(label = c("high", "med", "low"), value = c(10, 5, 0))
# setup column of type tibble to assign into
df = df %>%
mutate(nested = purrr::map(row, ~{ tibble() }))
# assign (fails)
df$nested[1] = t1 # no error but df$nested[1] is of type character not tibble
df$nested[2] = t2 # no error but df$nested[2] is of type int not tibble
df$nested[3] = t3 # gives warning and only one column of t3 is assigned
The warning I receive is:
In df$nested[3] = t3 : number of items to replace is not a multiple of replacement length
The result of this approach looks like:
> df
# A tibble: 4 x 2
row nested
<int> <list>
1 1 <chr [3]>
2 2 <int [4]>
3 3 <chr [3]>
4 4 <tibble [0 × 0]>
How can I assign (inner) tibbles to individual cells in an existing (outer) tibble, so that I end up with neested tibbles?
Using [[ instead of [ you could do:
library(dplyr, w = FALSE)
df <- tibble(row = 1:4, nested = vector("list", 4))
df$nested[[1]] <- t1
df$nested[[2]] <- t2
df$nested[[3]] <- t3
df$nested[[4]] <- NA
df
#> # A tibble: 4 × 2
#> row nested
#> <int> <list>
#> 1 1 <tibble [3 × 1]>
#> 2 2 <tibble [4 × 1]>
#> 3 3 <tibble [3 × 2]>
#> 4 4 <lgl [1]>