How can I store a vector or list in a single tibble cell?

Question

For each row in a tibble, a function call to quantile() will generate multiple values. How can I put each in a different column?

I've tried putting the whole set in a temporary column, but that doesn't seem to work.

tibble(category = LETTERS[1:10]) %>% 
  rowwise() %>% 
  # put everything in a temporary "range" column
  mutate(range = quantile(rnorm(count), c(0.025, 0.975), names = FALSE) ) %>% 
  # put each value in a column
  mutate(lo = range[[1]], hi = range[[2]]) %>% 
  ungroup() %>% 
  select(-range)

Answer 1

you can use an approach which relies on purrr::map . For each row you compute the quantiles and save the results into a tibble using ´enframe`. Then you unnest the result:

library(tidyverse)
df <- tibble(count = c(100, 1000, 10000))

set.seed(1)
df %>%
  mutate(q = map(count, ~quantile(rnorm(.x), c(0.025, 0.975)) %>% 
                   enframe %>% 
                   spread(name, value))) %>%
  unnest(q)

# A tibble: 3 x 3
  count `2.5%` `97.5%`
  <dbl>  <dbl>   <dbl>
1   100  -1.67    1.80
2  1000  -2.13    2.01
3 10000  -1.99    1.97