Fix formatting of bag imputation on factors

[topepo]

There is a bug (reported to maintainer today) in ipred::predict.classbag() where the factor levels of the predictions may not be correct. This results in a reverse dependency breakage.

This PR

1. temporarily converts the outcome to a factor if a character column is to be imputed. The original data is untouched.
2. casts the imputation predictions into a format consistent with raw data (factor, character, or numeric)
3. no longer casts the data to be a factor (to be consistent with prep(rec, strings_as_factors = FALSE))

I also made changes to a $%&# Rmd file that is always changing because of cli::rule() widths and random step id's.

[juliasilge]

In workflows, we handle the printing width like this:

options(cli.width = 70, width = 70, cli.unicode = FALSE)

I think this is a really good thing to make sure we have so this is more consistent. 👍

I agree that the random IDs are frustrating, but if you look at how this documentation ends up, we don't ever explain about what id does. It seems like a bad idea (confusing to folks who land here) to use it in the documentation as it exists overall now.

We could add more explanation, like what we have here, but this page is already getting really long-winded and detailed.

• Do we prefer to add this discussion here so we can use id? (I don't think I would want to add it here on its own merits TBH. Although now that I look, I can find it on TMwR but maybe not anywhere in the recipes docs?)
• Do we want to get used to not staging this chunk? This is what I typically do when I run devtools::document(); I either don't stage this whole file or if I need something from this file, I don't stage that chunk.

[DavisVaughan]

I fixed the random step id's here by setting the seed. We shouldn't have to worry about that anymore
#794

[topepo]

Torsten fixed the ipred bug within like 15 and the new version is on CRAN

[github-actions]

This pull request has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex https://reprex.tidyverse.org) and link to this issue.