Help creating a tibble that has a column made of tibbles?

[ignacio]

I want to create a tibble that has a column that is made of tibbles.

In particular, I want to do what the code below does but from i in 1:n_rows. What is the best way of doing that?

library(dplyr)
library(tidyr)
n_rows <- 4000
n_cols <- 1000

# my data looks something like this
A <- matrix(data = rnorm(n=n_rows*n_cols), nrow = n_rows)
betas <- rnorm(n = n_rows)
# I want the first row of my tibble to look like this

i <- 1

my_data <- tibble(i=i,
                  beta = betas[i],
                  data = tibble(a = A[i,],
                                b = sample(c(1,2), size = n_cols, replace = T))) %>% 
  group_by(i, beta) %>% 
  nest()
my_data
#> # A tibble: 1 x 3
#>       i  beta data                
#>   <dbl> <dbl> <list>              
#> 1     1 -2.26 <tibble [1,000 × 1]>

What is the tidy way of doing this?

^{Created on 2019-05-24 by the reprex package (v0.2.1)}

[von_olaf]

you re already doing it right, little fella. what do you want to do exactly?

[ignacio]

My code only generates the the first row of the tibble. How would you do a tibble that has rows for in 1:4K ?

[von_olaf]

is this what you want young padawan?

tibble(idx = seq(1, n_rows)) %>%
  mutate(mytib = map(idx , ~tibble(i=.x,
                beta = betas[.x],
                data = tibble(a = A[.x,],
                              b = sample(c(1,2), size = n_cols, 
                                         replace = T)))))

# A tibble: 4,000 x 2
     idx mytib               
   <int> <list>              
 1     1 <tibble [1,000 x 3]>
 2     2 <tibble [1,000 x 3]>
 3     3 <tibble [1,000 x 3]>