Furrr

The combination of map and future (parallel processing)

future::plan()

specifies how future():s are resolved, e.g. sequentially or in parallel.

plan(sequential) # (1)
plan(multisession, workers = 3) # (2)

1. sequential: Resolves futures sequentially in the current R process
2. multisession: Resolves futures asynchronously (in parallel) in separate R sessions running in the background on the same machine

implementation examples

df <- data.frame(
  x = c("apple", "banana", "cherry"),
  pattern = c("p", "n", "h"),
  replacement = c("x", "f", "q"),
  stringsAsFactors = FALSE
)

future_pmap(df, gsub)
future_pmap_chr(df, gsub)

close open connections

if (!inherits(plan(), "sequential")) plan(sequential)