DotPlot using split.by condition - avg expression bar disappears

[yonlada-nawilai]

I would like to create DotPlot using split.by function. My problem is I couldn't manage to 1) apply my custom palette to the split.by dotplot 2) the average expression legend bar disappears when creating split.by dotplot

So, I would like to create this kind of plot (found this from random paper)

However, when I try running my code:

plot3 <- DotPlot(subset_newlabel2, features = matrix_genes, dot.scale = 6, split.by = 'Condition', cols = c('darkred','darkred','darkred'))

plot3 + theme(
  axis.text = element_text(colour = "black"),
  axis.line = element_line(linewidth = 1),
  axis.text.x = element_text(size = 12, angle = 45, hjust = 1),
  axis.text.y = element_text(size = 12, hjust = 1))& 
  coord_flip()

This is what I got

As you can see, the avg expression bar disappears. Also, I would like to add my custom color palette to the plot as I usually did

plot2 <- DotPlot(subset_newlabel2, features = matrix_genes, dot.scale = 6)
plot2 + theme(
  axis.text = element_text(colour = "black"),
  axis.line = element_line(linewidth = 1),
  axis.text.x = element_text(size = 12, angle = 45, hjust = 1),
  axis.text.y = element_text(size = 12, hjust = 1)) + scale_x_discrete(labels = function(x) toupper(x))+
  scale_colour_gradient2(
    low = "darkred",
    mid = "lightyellow",
    high = "darkblue",  #this palette that I want to apply to split.by dotplot
    midpoint = 0
  ) + coord_flip()

I try to paste the same coding to the plot3 -

scale_colour_gradient2(
    low = "darkred",
    mid = "lightyellow",
    high = "darkblue"

but this error occurs

Error in `train_continuous()`:
! Discrete value supplied to a continuous scale
Run `rlang::last_trace()` to see where the error occurred.

---

So, I am wondering if anyone knows how to solve these two issues?