Free scales for y-axis not working for facet_nested (ggh4x)

I am trying to make scales for the y-axis free. But it looks like it is not working. I want adaptation to have its own scale and post adaptation also having its own scales. I am trying to make scales for the y-axis free. But it looks like it is not working. I want adaptation to have its own scale and post adaptation also having its own scales.

tgc <- structure(list(Group = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 
2L), .Label = c("Visible", "Remembered"), class = "factor"), 
    Condition = structure(c(1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 3L, 
    3L, 3L, 3L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 3L, 3L, 3L, 3L
    ), .Label = c("CEN", "IPS", "CTL"), class = "factor"), test = structure(c(1L, 
    1L, 2L, 2L, 1L, 1L, 2L, 2L, 1L, 1L, 2L, 2L, 1L, 1L, 2L, 2L, 
    1L, 1L, 2L, 2L, 1L, 1L, 2L, 2L), .Label = c("Pre-test", "Post-test"
    ), class = "factor"), Session = structure(c(1L, 2L, 1L, 2L, 
    1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 2L, 1L, 
    2L, 1L, 2L, 1L, 2L), .Label = c("Adaptation", "Post-adaptation"
    ), class = "factor"), N = c(12, 12, 12, 12, 12, 12, 12, 12, 
    12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 
    12), EE = c(2.134379625, 0.333942625, 1.742841125, 0.317361916666667, 
    2.84197270833333, 0.307057416666667, 2.403112375, 0.281202, 
    3.49590529166667, 0.305657666666667, 2.85211466666667, 0.3131155, 
    1.44857545833333, 0.269328166666667, 1.740270875, 0.243361833333333, 
    2.10702266666667, 0.286209125, 2.145855125, 0.305474083333333, 
    1.60016616666667, 0.281528625, 1.94182179166667, 0.294655916666667
    ), sd = c(0.727246182828044, 0.0816168443914292, 0.549168068103643, 
    0.0894916121701392, 1.14554677132408, 0.0958562360654162, 
    1.06827971273128, 0.0953131237162305, 1.18204258551111, 0.0896670491921828, 
    1.32864473484909, 0.109865886496798, 0.605344957514288, 0.0815454655757737, 
    0.833908172662699, 0.0798994165789182, 1.11582277105041, 
    0.0976064300150272, 0.667812406644538, 0.142929179817685, 
    0.686043669971901, 0.109794818975944, 1.39509308576833, 0.161854932615856
    ), se = c(0.209937889711449, 0.0235607535398997, 0.158531165974993, 
    0.0258340031883217, 0.330690868396632, 0.0276713118479362, 
    0.308385789857611, 0.0275145288174349, 0.341226302469221, 
    0.0258846474942731, 0.383546697661249, 0.0317155495718416, 
    0.174748037086728, 0.0235401482506832, 0.240728553983119, 
    0.0230649748349663, 0.322110288616933, 0.0281765493219072, 
    0.192780836372198, 0.0412601002213964, 0.198043748767058, 
    0.0316950341456936, 0.402728684306467, 0.0467234944577166
    ), ci = c(0.462070179795855, 0.0518568689018959, 0.348924743722983, 
    0.0568602576432562, 0.727845693918804, 0.0609041467375754, 
    0.678752547059741, 0.0605590696140879, 0.751034027967696, 
    0.0569717250090983, 0.844180589754564, 0.069805453951774, 
    0.384617836383033, 0.0518115169661108, 0.529839974927164, 
    0.0507656673296478, 0.708959965158704, 0.0620161669201078, 
    0.424307760005262, 0.0908128682911871, 0.435891352085212, 
    0.0697602998032695, 0.886399857701764, 0.102837717929058)), row.names = c(NA, 
-24L), class = "data.frame")

 library(ggh4x)

 p <- ggplot(tgc, aes(x = Condition, y = EE), fill = test) +
   geom_errorbar(aes(ymin=EE-se, ymax=EE+se, group = test), position = position_dodge(0.5), width=.1) +
   geom_bar(aes(fill = test), stat = "identity", width = 0.5, color = "black", position='dodge') + ylim(0,4) + theme_bw() + theme(
     axis.text.x = element_text(size = 12,face="bold"),#, angle = 10, hjust = .5, vjust = .5),
     axis.text.y = element_text(size = 12, face = "bold"),
     axis.title.y = element_text(vjust= 1.8, size = 16),
     axis.title.x = element_text(vjust= -0.5, size = 16),
     axis.title = element_text(face = "bold")) + xlab("Workspace") + ylab("EE (cm)") + theme(legend.position="top") +
   scale_fill_manual(values = c("grey80", "grey20")) + facet_nested(. ~ Session + Group, scales = "free_y") + theme(aspect.ratio = 6/4)
 
 p + guides(fill=guide_legend(title="Test:")) + theme(legend.text=element_text(size=14),legend.title=element_text(size=14) ) +
   theme(strip.text = element_text(face="bold", size=12))

Solution

It is not directly an answer to that question, but an alternative would be to create the panels separately. You can do this semi-programmatically by creating a list first. The biggest disadvantage that I can see in this way is that patchwork still cannot merge / combine x-/y- axis titles like it does with legends, so you would need to create each plot separately which would in turn defeat the point of this approach...

library(tidyverse)
library(ggh4x)
library(patchwork)

ls_p <- 
  tgc %>%
  split(., .$Session) %>%
  map(function(x){
## I've tried to de-clutter your code. 
ggplot(x, aes(x = Condition, y = EE, fill = test)) +
  geom_errorbar(aes(ymin = EE - se, ymax = EE + se, group = test), position = position_dodge(0.5), width = .1) +
  geom_col(width = 0.5, color = "black", position = "dodge") +
  labs(x= "Workspace", y = "EE (cm)") +
  scale_fill_manual("Test:", values = c("grey80", "grey20")) +
  facet_nested(. ~ Session + Group, scales = "free_y") +
  ## I am using cowplot::theme_minimal because I think it's a well designed theme
  cowplot::theme_minimal_hgrid() 
})

wrap_plots(ls_p) + plot_layout(guides = "collect") & 
  theme(legend.position = "top")