R, ggplot: Change line type between 2 x axis factors

Question

I am trying to draw a plot with geom_line, and I want to change line type (solid to dashed) from a point to another point as in the picture.

my data frame look like this:

'data.frame':   33 obs. of  7 variables:
 $ pta.simnum      : int  2 2 3 3 5 5 8 8 15 15 ...
 $ dose            : num  100 100 200 200 400 400 700 700 1400 1400 ...
 $ group1          : chr  "arc" "nonarc" "arc" "nonarc" ...
 $ group2          : chr  "ei6" "ei6" "ei6" "ei6" ...
 $ pta.target      : num  0.25 0.25 0.5 0.5 1 1 2 2 4 4 ...
 $ pta.prop.success: num  0.925 0.935 0.925 0.935 0.925 0.935 0.901 0.904 0.901 0.904 ...
 $ toxic           : chr  "nontoxic" "nontoxic" "nontoxic" "nontoxic" ...

The idea is I want to illustrate non toxic dose and toxic dose in two types of line from low to high MIC of bacteria.

And this is my code:

graph.pta <- ggplot(data = pta.40, aes(x = as.factor(pta.target), y =  as.numeric(dose), group = group2, color = group2)) + 
  geom_point() + geom_line(data = filter(pta.40, toxic == "nontoxic"), size = 1,aes(linetype = "dashed")) + 
  geom_line(data = filter(pta.40, toxic == "toxic"), size = 1,aes(linetype = "solid")) +
  facet_wrap(~group1) 
graph.pta

and the result as shown in picture with these warning lines:

geom_path: Each group consists of only one observation. Do you need to adjust the group aesthetic?
geom_path: Each group consists of only one observation. Do you need to adjust the group aesthetic?

I want to change line type from y value for each group according to the picture below however, the dashed line did not appear.

I suppose the result is not as expected since x axis is in factor, but the point I want to change to dashed line lies between two factors.

Is there any ways to plot this figure with my idea?

Thank you.

PS: I added part of my data as below:

structure(list(pta.simnum = c(2L, 2L, 3L, 3L, 5L, 5L, 8L, 8L, 
15L, 15L, 29L, 29L, 57L, 56L, 112L, 110L, 2L, 3L, 2L, 4L, 3L, 
6L, 5L, 10L, 8L, 19L, 14L, 36L, 26L, 71L, 50L, 141L, 98L), dose = c(100, 
100, 200, 200, 400, 400, 700, 700, 1400, 1400, 2800, 2800, 5600, 
5500, 11100, 10900, 100, 200, 100, 300, 200, 500, 400, 900, 700, 
1800, 1300, 3500, 2500, 7000, 4900, 14000, 9700), group1 = c("arc", 
"nonarc", "arc", "nonarc", "arc", "nonarc", "arc", "nonarc", 
"arc", "nonarc", "arc", "nonarc", "arc", "nonarc", "arc", "nonarc", 
"arc", "arc", "nonarc", "arc", "nonarc", "arc", "nonarc", "arc", 
"nonarc", "arc", "nonarc", "arc", "nonarc", "arc", "nonarc", 
"arc", "nonarc"), group2 = c("ei6", "ei6", "ei6", "ei6", "ei6", 
"ei6", "ei6", "ei6", "ei6", "ei6", "ei6", "ei6", "ei6", "ei6", 
"ei6", "ei6", "ei8", "ei8", "ei8", "ei8", "ei8", "ei8", "ei8", 
"ei8", "ei8", "ei8", "ei8", "ei8", "ei8", "ei8", "ei8", "ei8", 
"ei8"), pta.target = c(0.25, 0.25, 0.5, 0.5, 1, 1, 2, 2, 4, 4, 
8, 8, 16, 16, 32, 32, 0.125, 0.25, 0.25, 0.5, 0.5, 1, 1, 2, 2, 
4, 4, 8, 8, 16, 16, 32, 32), pta.prop.success = c(0.925, 0.935, 
0.925, 0.935, 0.925, 0.935, 0.901, 0.904, 0.901, 0.904, 0.901, 
0.904, 0.901, 0.902, 0.9, 0.9, 0.958, 0.958, 0.933, 0.92, 0.933, 
0.909, 0.933, 0.901, 0.926, 0.901, 0.915, 0.9, 0.91, 0.9, 0.908, 
0.9, 0.904), toxic = c("nontoxic", "nontoxic", "nontoxic", "nontoxic", 
"nontoxic", "nontoxic", "nontoxic", "nontoxic", "nontoxic", "nontoxic", 
"nontoxic", "nontoxic", "nontoxic", "nontoxic", "toxic", "toxic", 
"nontoxic", "nontoxic", "nontoxic", "nontoxic", "nontoxic", "nontoxic", 
"nontoxic", "nontoxic", "nontoxic", "nontoxic", "nontoxic", "nontoxic", 
"nontoxic", "nontoxic", "nontoxic", "toxic", "toxic")), row.names = c(NA, 
33L), class = "data.frame")

Answer 1

It's rather old question, but hopefully still helpful to you or others on the internet to be answered.

To correctly get the dashed and solid lines portrayed, you need to stack them correctly. In ggplot geom_layers stack on top on the previous.

Therefore, we start with the dashed lines.

p1= ggplot()+ 
geom_line(aes(as.factor(pta.target), as.numeric(dose), 
group=interaction(group1, group2)),size=1,linetype="dashed", data=pta.simnum)+
facet_wrap(~group1)

The last step is lay on top the lines that match you criteria. You want the solid lines to stop after pta target of 16. You can make a seperate dataframe or filter within a ggplot command as follows.

p1+ geom_line(aes(as.factor(pta.target), as.numeric(dose),
col=group2, group=interaction(group1, group2)), size=1, 
data=filter(pta.simnum, pta.target<16))

Let me know if this is what you meant

Answer 2

Try this as an option, you have only 1 point in some cases so lines would not work. You can try with points based on conditions to color them:

library(ggplot2)
library(dplyr)
#Code
df <- pta.40 %>%
  mutate(Line=ifelse(dose>=6400 & group2=='ei6' & group1=='arc',1,
                     ifelse(dose>=7400 & group2=='ei6' & group1=='nonarc',2,
                            ifelse(dose>=7900 & group2=='ei8' & group1=='arc',3,
                                   ifelse(dose>=8200 & group2=='ei8' & group1=='nonarc',4,5))))) %>%
  mutate(pta.target=factor(pta.target,levels = sort(unique(pta.target)),
                           ordered = T))

#Plot
ggplot(data = df, aes(x = pta.target,
                      y =  as.numeric(dose),
                      group = group2,
                      color = interaction(factor(Line),group2),
                      linetype=factor(Line))) +
  geom_point() +
  facet_wrap(~group1)+
  labs(color='Var')

Output: