如何在堆叠的条形图上叠加条形图？

Question

I am trying to overlay a bar plot over a stacked bar plot as you can see in the figure below. Although this is working well the dotted line type does not appear as such in the legend.

Dotted bar plot on top of stacked barplot

This is my data:

data = data.frame(treat=c("min","plus","min","plus"),
                  obs=c(8,7,121,98),
                  type=c("Tare","Tare","Net","Net"))

data2 = data.frame(treat=c("min","plus"),
                   obs=c(20,15),
                   type=c("Sugar","Sugar"))

And the code I use to make the graph:

ggplot(data = data, aes(x=treat, y=obs, fill = type)) + 
  theme_classic() + 
  geom_bar(stat="identity", col = "black", width = .75) +
  geom_bar(data = data2, stat="identity", fill = "transparent", col = "black", size = .8, linetype = "dotted", width = .75, show.legend = F) +
  scale_fill_manual(values = c("#F8766D", "#00BFC4", NA),
         limits = c("Tare", "Net", "Sugar"))

How do I make the second bar appear with dotted lines in the legend?

Answer 1

Is this what you're looking for?

library(tidyverse)
data = data.frame(treat=c("min","plus","min","plus"),
                  obs=c(8,7,121,98),
                  type=c("Tare","Tare","Net","Net"))

data2 = data.frame(treat=c("min","plus"),
                   obs=c(20,15),
                   type=c("Sugar","Sugar"))

ggplot(data = data, aes(x=treat, y=obs, fill = type)) + 
  theme_classic() + 
  geom_bar(stat="identity", col = "black", width = .75) +
  geom_bar(data = data2, stat="identity", 
           fill = "transparent", col = "black", size = .8, 
           aes(linetype = "dotted"), 
           width = .75, 
           show.legend = T) +
  scale_fill_manual(values = c("#F8766D", "#00BFC4", NA),
                    limits = c("Tare", "Net", "Sugar"),
                    name="type")+
  scale_linetype_manual(name="Sugar", values="dotted")+
  guides(fill = guide_legend(keywidth = 1, keyheight = 1),
         linetype=guide_legend(keywidth = 3, keyheight = 1))

^{Created on 2020-12-10 by the reprex package (v0.3.0)}