Added support for univariate box plots with no x aesthetic.

[foo-bar-baz-qux]

Fixes #2110.

This now works:

ggplot(mtcars, aes(y = disp)) + geom_boxplot()

Which would essentially be the equivalent of this, without the mapped x-aesthetic:

library(ggplot2)
ggplot(mtcars, aes(x=0, y = disp)) + geom_boxplot()

Tried a few approaches, but this is the cleanest way I've found so far. I think it looks OK to have the x location be at 0 without an axis label; user can always map the x aesthetic if more fine-grained adjustments are required. Ideally, I think the univariate boxplot would look something like this, where the x-axis is essentially removed/treated like a discrete scale. However, I couldn't find a clean way of changing it from a continuous scale from within the geom/stat ggprotos without hard coding the exceptions into the layer building process.

library(ggplot2)
ggplot(mtcars, aes(y = disp, x = "all")) + geom_boxplot()

[hadley]

Could you please add a test for this too? (Maybe a visual test)

[hadley]

Are you interested in finishing this off? No problems if you don't have time; just let me know I'll get it over the finish line.

[foo-bar-baz-qux]

I'm haven't looked at the visual testing suite so might take some extra time to look through; perhaps you could add that test and finish this off? Thanks.

[hadley]

Looking at the code again, it feels like a visual test would be overkill.

[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/