Duplicated aesthetics on some documentations of Geom/Stat

[yutannihilation]

Probably due to the change of the specification of required_aes (#3506), some documentations of Geom/Stat now show both "x or y" and, "x" and "y" separately on their Aesthetics section. This seems a bit confusing, so it's nice if we can fix this.

For example, https://ggplot2.tidyverse.org/dev/reference/geom_linerange.html#aesthetics:

[carywreams]

Is this a matter of modifying R/utilities-help.r to remove any parameter found in x$requires_aes from x$aesthetics() prior to, or in conjunction with, the union() operation ?

from

rd_aesthetics_item <- function(x) {
  req <- x$required_aes
  req <- sub("|", "} \\emph{or} \\code{", req, fixed = TRUE)
  all <- union(req, sort(x$aesthetics()))

  ifelse(all %in% req,
    paste0("\\strong{\\code{", all, "}}"),
    paste0("\\code{", all, "}")
  )
}

to

rd_aesthetics_item <- function(x) {
  req <- x$required_aes
  req <- sub("|", "} \\emph{or} \\code{", req, fixed = TRUE)
  req_aes <- unlist(strsplit(x$required_aes, "|", fixed = TRUE))
  optional_aes <- setdiff(x$aesthetics(),req_aes)
  all <- union(req, sort(optional_aes))

  ifelse(all %in% req,
    paste0("\\strong{\\code{", all, "}}"),
    paste0("\\code{", all, "}")
  )
}

If I'm on the right track, let me know and I'll start figuring out how to do this.

[carywreams]

Seems unlist(strsplit(self$required_aes, "|", fixed = TRUE)) is used already to pull apart the "|" specs. So perhaps my previous question is answered (have edited above comment to include this approach).

I found that technique in R/stat-.r::aesthetics function definition, which appears to be where the individual required parameters are added into consideration (and thus duplicated for documentation) in R/utilities-help.r::rd_aesthetics_item(), as part of the union() call (see above comment).

While I'm tempted to suggest correcting it in R/stat-.r::aesthetics, some customers of that function may continue to expect required parameters (specified as "x|y" in required aesthetics) to appear as individual items when retrieving the list of aesthetics.

Changing the behavior in R/utilities-help.r::rd_aesthetics_item() to remove the duplicates may only affect documentation.

[carywreams]

Promising (at least on the documentation front)

diff --git a/man/geom_linerange.Rd b/man/geom_linerange.Rd
index 16fed3db..a0e073d1 100644
--- a/man/geom_linerange.Rd
+++ b/man/geom_linerange.Rd
@@ -134,12 +134,6 @@ This geom treats each axis differently and, thus, can thus have two orientations
 \item \code{group}
 \item \code{linetype}
 \item \code{size}
-\item \code{x}
-\item \code{xmax}
-\item \code{xmin}
-\item \code{y}
-\item \code{ymax}
-\item \code{ymin}
 }
 Learn more about setting these aesthetics in \code{vignette("ggplot2-specs")}.
 }

[carywreams]

appears to only affect these files:

#       modified:   geom_bar.Rd
#       modified:   geom_boxplot.Rd
#       modified:   geom_linerange.Rd
#       modified:   geom_ribbon.Rd

[carywreams]

testthat runs produce identical results on master branch and the branch with my candidate mod.

• on_i3618.txt
• on_master.txt

the pull request for #3506 looks to have had some build issues as well. Should I be concerned ?
As the results are the same, think I'm going to wrap this and submit a PR; will deal with any additional issues there.

[carywreams]

looking for a little help with automated code coverage checks in pr #3657

[clauswilke]

You can ignore them.

[carywreams]

.ignored

Thanks

[lock]

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