rpurrrfurrr

appending tibbles using purrr::map


Background

I wrote a function using map2 that aims to produce a dataframe or a list of dataframes. My problem is that this function produces individual tables that are not assigned to an object. My desired output is a dataframe which stores the output tibble for each iteration of the vectorised loop.

Reprex

df_coefs <- list()

df <- 
  tibble(names = c("john", "charlie", "lucy"), colours = c("green", "grey", "silver"))


funx <- function(name, colour) {
  
  coef <- runif(1, 0, 30)
  
  df_coefs[[i]] <- 
      tibble_row(names = name, 
                 colours = colour,
                 coefs = coef)
}

map2(.x = df$names,
     .y = df$colours,
     .f = funx)

Current output

[[1]]
# A tibble: 1 × 3
  names colours coefs
  <chr> <chr>   <dbl>
1 john  green    11.1

[[2]]
# A tibble: 1 × 3
  names   colours coefs
  <chr>   <chr>   <dbl>
1 charlie grey     3.73

[[3]]
# A tibble: 1 × 3
  names colours coefs
  <chr> <chr>   <dbl>
1 lucy  silver   29.4

Desired output

  names colours coefs
  <chr> <chr>   <dbl>
1 john  green    11.1
2 charlie grey   3.73
3 lucy  silver   29.4

Solution

  • The assignment with [i] seems to be a typo

    funx <- function(name, colour) {
      
      coef <- runif(1, 0, 30)
      
      
          tibble_row(names = name, 
                     colours = colour,
                     coefs = coef)
                     
    }
    

    Other than that, if we need to collapse the rows, use the suffix _dfr in map

    library(purrr)
    map2_dfr(.x = df$names,
          .y = df$colours,
          .f = funx)
    # A tibble: 3 × 3
      names   colours coefs
      <chr>   <chr>   <dbl>
    1 john    green    14.3
    2 charlie grey     25.8
    3 lucy    silver   13.1