How to remove nesting facet labels in facet_nested (ggh4x package) - r

Question

I try to get rid of the nesting (upper) facet label. So far I can only apply changes to all facet labels together and not just the grouping ones.
library(tidyverse)
library(ggh4x)
df <- as_tibble(iris) %>%
mutate(
Nester = if_else(Species == "setosa", "Short Leaves", "Long Leaves"),
Nester = factor(Nester)
) %>%
pivot_longer(!c(Species, Nester), names_to = "Measure", values_to = "Value")
ylim <- df %>%
split(.$Nester) %>%
map(., ~ range(.$Value))
df %>%
ggplot(aes(Measure, Value)) +
geom_boxplot() +
theme(axis.text.x = element_text(angle = 90, , vjust = .5)) +
facet_nested(~ Nester + Species, scales = "free_y", independent = "y") +
facetted_pos_scales(
y = list(
Species == "versicolor" ~ scale_y_continuous(limits = ylim[[1]]),
Species == "virginica" ~ scale_y_continuous(limits = ylim[[1]], guide = "none"),
Species == "setosa" ~ scale_y_continuous(limits = ylim[[2]])
)
)
Created on 2022-08-23 with reprex v2.0.2
I manually edited the plot above to show the desired plot:
I need a method that works with nested facets, as my actual data is far bigger and more complex.
EDIT: The solution needs to use nested facets.

Answer 1

Update on OP request: This is a solution using the color trick white on white is not visible:
From https://teunbrand.github.io/ggh4x/articles/Facets.html
my_strips <- strip_themed(
# Horizontal strips
background_x = elem_list_rect(fill = c("white", "dodgerblue")),
text_x = elem_list_text(colour = c("white", "white"),
face = c("bold", "bold")),
by_layer_x = TRUE,
text_y = elem_list_text(angle = c(0, 90)),
by_layer_y = FALSE
)
df %>%
pivot_longer(!c(Species, Nester), names_to = "Measure", values_to = "Value") %>%
ggplot(aes(Measure, Value)) +
geom_boxplot() +
facet_nested(~ Nester + Species , scales = "free_y", independent = "y",
strip = my_strips)+
theme(
axis.text.x=element_text(angle=90, hjust=0, vjust=0))
First answer:
We could use just facet_wrap for this. No need for facet_nested:
library(tidyverse)
p <- df %>%
pivot_longer(!c(Species, Nester), names_to = "Measure", values_to = "Value") %>%
ggplot(aes(Measure, Value)) +
geom_boxplot() +
theme(axis.text.x = element_text(angle = 90, , vjust = .5)) +
facet_wrap(.~Species)
p

Answer 2

This answer resembles #TarJae's answer, but uses element_blank()s instead. The result is pretty similar, but may be relevant if exporting graphics with a transparent background.
library(tidyverse)
library(ggh4x)
df <- as_tibble(iris) %>%
mutate(
Nester = if_else(Species == "setosa", "Short Leaves", "Long Leaves"),
Nester = factor(Nester)
) %>%
pivot_longer(!c(Species, Nester), names_to = "Measure", values_to = "Value")
ylim <- df %>%
split(.$Nester) %>%
map(., ~ range(.$Value))
strips <- strip_nested(
text_x = list(element_blank(), element_text()),
background_x = list(element_blank(), element_rect()),
by_layer_x = TRUE
)
df %>%
ggplot(aes(Measure, Value)) +
geom_boxplot() +
theme(axis.text.x = element_text(angle = 90, , vjust = .5)) +
facet_nested(
~ Nester + Species, scales = "free_y", independent = "y",
strip = strips
) +
facetted_pos_scales(
y = list(
Species == "versicolor" ~ scale_y_continuous(limits = ylim[[1]]),
Species == "virginica" ~ scale_y_continuous(limits = ylim[[1]], guide = "none"),
Species == "setosa" ~ scale_y_continuous(limits = ylim[[2]])
)
)
Created on 2022-08-23 by the reprex package (v2.0.1)