How can I change the width of jitter using visreg and ggplot2 for plotting my data?

Question

I am using the R packages visreg() and ggplot2() to display the relationship between length and acclimation temperature. The temperatures are 23 and 28 degrees, so I would like to add jitter to the points to make sure the points do not all appear on the same line. However, when I use the jitter = TRUE option in visreg(), I find the jitter too wide.

Here is what I tried:

library(ggplot2)
library(visreg)

p1 <- visreg(m,"a_temp",by="ploisex", overlay=TRUE, jitter = TRUE, gg=TRUE) + 
  xlab("Acclimation temperature (°C)") +
  ylab("Log final length") + ylim(2.8,3.4) +
  scale_fill_manual(name = "Ploidy level/sex", values = c("2n f" = "blue", "2n m" = "lightblue", "3n m" = "red")) + 
  scale_colour_manual(name = "Ploidy level/sex", values = c("2n f" = "blue", "2n m" = "lightblue", "3n m" = "red")) +
  theme_bw()

p1

Which gives the following plot: plot1

I have tried to change the width of the jitter by using geom_jitter, which is does, but somehow the points are now plotted double (I did not change the color of the new points to make it more easy to see that the points are plotted double).

Here is my code for the second plot:

p2 <- visreg(m,"a_temp",by="ploisex", overlay=TRUE, gg=TRUE) + 
  geom_jitter(width = .1) +
  #geom_jitter(width = .1, aes(color=ploisex)) +
  xlab("Acclimation temperature (°C)") +
  ylab("Log final length") + ylim(2.8,3.4) +
  scale_fill_manual(name = "Ploidy level/sex", values = c("2n f" = "blue", "2n m" = "lightblue", "3n m" = "red")) + 
  scale_colour_manual(name = "Ploidy level/sex", values = c("2n f" = "blue", "2n m" = "lightblue", "3n m" = "red")) + 
  theme_bw()

p2

Which gives the following plot: plot2

When I combine jitter = TRUE and geom_jitter the width of the jitter is not changed at all, but the points are still plotted double:

p3 <- visreg(m,"a_temp",by="ploisex", overlay=TRUE, jitter = TRUE, gg=TRUE) + 
  geom_jitter(width = .1) +
  #geom_jitter(width = .1, aes(color=ploisex)) +
  xlab("Acclimation temperature (°C)") +
  ylab("Log final length") + ylim(2.8,3.4) +
  scale_fill_manual(name = "Ploidy level/sex", values = c("2n f" = "blue", "2n m" = "lightblue", "3n m" = "red")) + 
  scale_colour_manual(name = "Ploidy level/sex", values = c("2n f" = "blue", "2n m" = "lightblue", "3n m" = "red")) + 
  theme_bw()

p3

plot3

I would like to know how I can change the width of the jitter, without plotting the points double.

By the way I am not a programmer, so I hope I asked this question correctly. Your help is much appreciated!

Answer 1

You could set the amount of jitter via points = list(position = position_jitter(width = ...)).

Using the default example from visage and adding the unjittered data via a second geom_point:

library(visreg)
library(ggplot2)

fit <- lm(Ozone ~ Solar.R + Wind + Temp, data=airquality)
visreg(fit, "Wind", gg=TRUE, points = list(position = position_jitter(width = 20))) +
  geom_point(aes(color = "Unjittered Points"))