Update .assert_package()?

[ddsjoberg]

Should we update to use rlang::is_installed() and rlang::check_installed() internally? I think it'll improve the user experience when interactive.

• rlang::is_installed() doesn't interact with the user. It simply returns TRUE or FALSE depending on whether the packages are installed.

• In interactive sessions, rlang::check_installed() asks the user whether to install missing packages. If the user accepts, the packages are installed with pak::pkg_install() if available, or utils::install.packages() otherwise. If the session is non interactive or if the user chooses not to install the packages, the current evaluation is aborted.

[larmarange]

Seems good to me. Could you propose a PR?

[ddsjoberg]

Seems good to me. Could you propose a PR?

Will do!

[ddsjoberg]

hey hey @larmarange !

I think the issue is pretty strange on R 3.5. The error is stemming from a call to reformulate() in {lme4}, and the error is unused argument (env = environment(f)). The env= argument wasn't added until 3.6.

You can see in the table below, the function was changed in 3.6 and the env arg was added.

I looked in the package's GH repo and they don't use a prefix stats::reformulate() when they reference the function (https://github.com/lme4/lme4/search?q=reformulate). So what I think is happening is that the package is installing properly on R 3.5, it's using the old base R of reformulate that doesn't have an env argument.

I think lme4 likely isn't preforming the tests to ensure there are no errors on the older versions of R it claims to support. (they also had a release of a new version last week) What do you think the next best steps are?

[larmarange]

I guess we have two options:

• decide that now broom.helpers requires 3.6 as a minimum (but it will also impact gtsummary). Some how, we are now version 4.1
• test R version in examples and test and do not run examples/tests requiring lme4 if R < 3.6

[ddsjoberg]

I filed an issue on the lme4 repo lme4/lme4#664 . I vote would to just leave it as is for now with the erroring R Cmd Check. Hopefully lme4 will update either the call to reformulate(env=) or update the R versions that are supported, either of which will fix our errors upon the next release.

[larmarange]

OK in that case I propose to merge the PR but to keep this issue open. Is it fine with you?

[ddsjoberg]

that sounds good to me! i think they'll update soon enough. i may turn off my 3.5 checks for a while until they do

[ddsjoberg]

hey hey @larmarange ! The lme4 package had a new release 11 days ago, and should solve our R CMD Check problems!
https://cran.r-project.org/web/packages/lme4/index.html

[ddsjoberg]

FYI, I think I am about 3 to 4 weeks out from a gtsummary release (https://github.com/ddsjoberg/gtsummary/milestone/14). Is there anything we could add to broom.helpers in the interim? Would you be ok with a broom.helpers release in a few weeks?

[larmarange]

Is there anything we could add to broom.helpers in the interim?

I do not see anything right now

Would you be ok with a broom.helpers release in a few weeks?

No problem