Which is the most efficient way to generate m dataframes, where each one has a variable that includes random values, and append them?
Here's an example:
df <- data.frame(id = 1:10, var = sample(1:500), 10, replace=TRUE) id var 1 65 2 123 3 42 4 16 5 463 6 129 7 367 8 99 9 489 10 63 If m = 2, two dataframes should be generated and appended, having:
id var 1 65 2 123 3 42 4 16 5 463 6 129 7 367 8 99 9 489 10 63 1 321 2 410 3 78 4 166 5 320 6 478 7 231 8 100 9 105 10 206 Put the dataframe to be generated in a function
fun <- function() { df <- data.frame(id = 1:10, var = sample(1:500, 10, replace=TRUE)) df } Then there are multiple ways to call this function m times and bind.
Base R using replicate
m <- 2 do.call(rbind, replicate(m, fun(), simplify = FALSE)) Base R using lapply
do.call(rbind, lapply(seq_len(m),\(x) fun())) purrr::map_df
purrr::map_df(seq_len(m), ~fun()) The most efficient way to create a data.frame with those characteristics for an arbitrary m is to avoid appending by creating a single data.frame to begin with:
df <- data.frame(id = 1:10, var = sample(500, 10*m, 1)) 