Using purrr (tidyverse) to map distance function across all columns of dataframe

Question

I have a distance function which takes in 2 (numeric) vectors and calculates the distance between them.

For a given dataframe ( mtcars_raw ) in the example below and a fixed input vector ( test_vec ) I would like to calculate the pairwise distances (ie apply the distance function) to each column and test_vec and return the vector of distances. The length of the vector should be the number of columns.

Please see the reproducible example:

library(datasets)

# The raw dataframe containing only numeric columns
mtcars_raw <- datasets::mtcars

# The distance function between 2 vectors (of the same length typically)
eucl_dist <- function(x, y){
    return(sqrt(sum((x-y)^2)))
}

# An example of a numeric vector to check the distance against each column
test_vec   <- rnorm(n = dim(mtcars_raw)[1], mean = 12, sd = 2)

# Manually for the first column, we would have:
dist_1 <- eucl_dist(x = test_vec, mtcars_raw[, 1])
dist_1
#> [1] 58.71256

# Manually for the second column, we would have:
dist_2 <- eucl_dist(x = test_vec, mtcars_raw[, 1])
dist_2
#> [1] 58.71256

# Would like dist_comb returned for all columns without having to manually do
# the combining
dist_comb <- c(dist_1, dist_2)
dist_comb
#> [1] 58.71256 58.71256

Could anyone please show the purrr (tidyverse) code to return vector on each column of mtcars_raw against test_vec?

Answer 1

Use map_dbl , which is a special case of map to loop through columns but explicitly return double type vector:

map_dbl(mtcars_raw[1:2], ~ eucl_dist(test_vec, .x))

#     mpg      cyl 
#58.06386 36.51686 

On all columns:

map_dbl(mtcars_raw, ~ eucl_dist(test_vec, .x))

#       mpg        cyl       disp         hp       drat         wt       qsec         vs         am       gear       carb 
#  58.06386   36.51686 1414.98943  850.71261   49.72837   51.74005   35.50658   67.25079   67.35504   49.34896   54.56577