updating precision in round2()

[ddsjoberg]

What changes are proposed in this pull request?

• Updating the levels of precision used in round2(), which is used in the background for every rounded/formatted number in a gtsummary table. (Bug Report: round2 is unreliable #1494)

If there is an GitHub issue associated with this pull request, please provide link.
closes #1494
closes #1472

---

Reviewer Checklist (if item does not apply, mark is as complete)

• Ensure all package dependencies are installed by running renv::install()
• PR branch has pulled the most recent updates from master branch. Ensure the pull request branch and your local version match and both have the latest updates from the master branch.
• If an update was made to tbl_summary(), was the same change implemented for tbl_svysummary()?
• If a new function was added, function included in _pkgdown.yml
• If a bug was fixed, a unit test was added for the bug check
• Run pkgdown::build_site(). Check the R console for errors, and review the rendered website.
• Code coverage is suitable for any new functions/features. Review coverage with withr::with_envvar(new = c("NOT_CRAN" = "true"), covr::report()). Begin in a fresh R session without any packages loaded.
• R CMD Check runs without errors, warnings, and notes
• usethis::use_spell_check() runs with no spelling errors in documentation

When the branch is ready to be merged into master:

• Update NEWS.md with the changes from this pull request under the heading "# gtsummary (development version)". If there is an issue associated with the pull request, reference it in parentheses at the end update (see NEWS.md for examples).
• Increment the version number using usethis::use_version(which = "dev")
• Run codemetar::write_codemeta()
• Run usethis::use_spell_check() again
• Approve Pull Request
• Merge the PR. Please use "Squash and merge".

[nikostr]

Thanks for looking into this! Won't this implementation simply shift where problems occur? Running

number <- 10^seq(1,15)
(number + sqrt(.Machine$double.eps)) == number
#> [1] FALSE FALSE FALSE FALSE FALSE FALSE FALSE FALSE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE  TRUE
round2(2.5e10, digits=-10)
#> [1] 2e+10

and since sqrt(.Machine$double.eps) is about 1.5e-8 any numbers in that range will start behaving badly.

Very frustrating that this isn't something that just works in R!

[ddsjoberg]

@nikostr oh yes, i was just in the middle of testing out a few things. I was hoping to land on a solution that had fewer computations....but as you noted, the previous code wasn't perfect.

I've tested the janitor/stackoverflow solution too, and it's better, but has some issue too round2(0.499999999, digits = 0) = 1 ¯_(ツ)_/¯