Boxplot color can't have a continuous variable when faceting

[JacobBumgarner]

Code

For some reason the boxplot layer isn't able to have a continuous variable (other than the one used to define the group, e.g., x in the plot below):

ggplot2::ggplot(
    mtcars,
    ggplot2::aes(x = factor(cyl), y = hp, color = hp)
) + 
    ggplot2::geom_point() + 
    ggplot2::geom_boxplot(outliers = FALSE) + 
    ggplot2::facet_wrap(gear ~ am)

Warning on plot

> ggplot2::ggplot(
+     mtcars,
+     ggplot2::aes(x = factor(cyl), y = hp, color = hp)
+ ) + 
+     ggplot2::geom_point() + 
+     ggplot2::geom_boxplot(outliers = FALSE) + 
+     ggplot2::facet_wrap(gear ~ am)
Warning: The following aesthetics were dropped during statistical transformation: colour.
ℹ This can happen when ggplot fails to infer the correct grouping structure in the data.
ℹ Did you forget to specify a `group` aesthetic or to convert a numerical variable into
  a factor?
Warning: The following aesthetics were dropped during statistical transformation: colour.
ℹ This can happen when ggplot fails to infer the correct grouping structure in the data.
ℹ Did you forget to specify a `group` aesthetic or to convert a numerical variable into
  a factor?
Warning: The following aesthetics were dropped during statistical transformation: colour.
ℹ This can happen when ggplot fails to infer the correct grouping structure in the data.
ℹ Did you forget to specify a `group` aesthetic or to convert a numerical variable into
  a factor?
Warning: The following aesthetics were dropped during statistical transformation: colour.
ℹ This can happen when ggplot fails to infer the correct grouping structure in the data.
ℹ Did you forget to specify a `group` aesthetic or to convert a numerical variable into
  a factor?

[Yunuuuu]

The issue arises because multiple colors are being assigned within each boxplot group, whereas each box should have a single, consistent color. To resolve this, you can summarize the color per group — for example, using the mean (or median) of hp for each group:

ggplot2::ggplot(
    mtcars,
    ggplot2::aes(x = factor(cyl), y = hp, color = hp)
) +
    ggplot2::geom_point() +
    ggplot2::geom_boxplot(
        aes(group = cyl, color = color),
        outliers = FALSE,
        data = ~ dplyr::mutate(.x, color = mean(hp), .by = cyl)
    ) +
    ggplot2::facet_wrap(gear ~ am)

[teunbrand]

This is indeed because you're giving instructions to give singular boxplots multiple colour values. The warning message indicates that there might be a conflict between the colour aesthetic and grouping structure. In your mind, what'd be the ideal outcome in this problem?

[JacobBumgarner]

I see - sorry for the confusion, and thanks for the explanation @Yunuuuu

@teunbrand I believe I was imagining the color value would automatically have taken an average of the grouped values. But after thinking about this, I realize how big of an assumption this is.

I think this was after a long night of coding, so I was assuming very poorly... please disregard!!