In strboul/mmy: Custom miscellaneous R functions

Bench with base R solutions

The base solution is adapted from: https://stackoverflow.com/a/5222365/ beat it up!

x <- sample(1e6)
microbenchmark::microbenchmark(
    cumsum(c(1L, diff(x) != 1L)),
    mmy::group_consecutive(x),
    check = "identical"
)

Testing the performance of early_exit option

microbenchmark::microbenchmark(
    with_early_exit = mmy::group_consecutive(x),
    without_early_exit = mmy::group_consecutive(x, .early_exit = FALSE),
    check = "identical"
)

Seems that early_exit does not affect performance in mixed samples. What about consecutive numbers?

y <- seq(1e6)
microbenchmark::microbenchmark(
    with_early_exit = mmy::group_consecutive(y),
    without_early_exit = mmy::group_consecutive(y, .early_exit = FALSE),
    check = "identical"
)

Seems that the early exit is slightly faster than the without early_exit but don't know whether the difference is significant.

strboul/mmy documentation built on Sept. 24, 2021, 12:08 p.m.