Round a numeric vector to any number, rounded up by default (`round_up = TRUE`

).

## Details

This function's precision is limited to 15 significant digits in order to account for the limits of R's floating point representation.

A computationally more efficient alternative would be the unexported `scales:::round_any()`

which drives `scales::number()`

– with the drawback that it lacks
control of rounding up exact remainders of `accuracy / 2`

, i.e. it *always* rounds *off*.

## See also

Other statistical computing / numeric functions:
`safe_max()`

,
`safe_min()`

,
`safe_seq_len()`

,
`stat_mode()`

## Examples

```
vals = c(0.025, 0.1, 0.1999, 0.099999, 0.49, 0.55, 0.5, 0.9, 1)
vals |> pal::round_to(to = 0.05)
#> [1] 0.05 0.10 0.20 0.10 0.50 0.55 0.50 0.90 1.00
vals |> pal::round_to(to = 0.05,
round_up = FALSE)
#> [1] 0.00 0.10 0.20 0.10 0.50 0.55 0.50 0.90 1.00
```