How do you change the x-axis tick labels on plot_model() and format the lines on the graph to be closer together?

Question

I am following the example here: https://strengejacke.github.io/sjPlot/articles/plot_interactions.html

I copy the following:

library(sjPlot)
library(sjmisc)
library(ggplot2)
data(efc)
theme_set(theme_sjplot())

# make categorical
efc$c161sex <- to_factor(efc$c161sex)

# fit model with interaction
fit <- lm(neg_c_7 ~ c12hour + barthtot * c161sex, data = efc)

And then I do the following:

plot_model(fit, type = "pred", terms = c("c161sex", "barthtot [0, 100]"))

Why does the x-axis ticks show up as "male" and "female" when their values are 1 and 2 in the data frame?

Is there a way to replace the x-axis ticks when using plot_model ? I've been adding scale_x_discrete(breaks=c("1","2"), labels=c("male", "female")) in other datasets and it makes my x-axis ticks disappear altogether.

And is it possible to have the lines closer to each other rather than so far apart?

When I use my own dataset (I'm unable to post the whole dataset), I get something that looks like this:

Replication code:

data = read.table(text="
X   Z   Y
0   1   1
1   0   1
1   0   1
0   1   1
1   1   0
1   0   0
0   0   0
1   1   1
1   1   1
0   0   1
1   0   1
1   1   1
0   1   0
1   0   0
1   0   0
0   1   1
1   1   1
1   0   1
0   0   1
1   1   1
1   1   0
0   0   0
1   0   0
1   1   1
0   1   1
1   0   1
1   0   1
0   1   1
1   1   0
1   0   0", header=TRUE)

data$X <- as.factor(data$X)
data$Z <- as.factor(data$Z)

h <- lm_robust(Y ~ X*Z, data = data)

plot_model(h, type = "pred", terms = c("X", "Z")) +
  xlab("Fun Level") + ylab("") + theme_bw() + 
  theme(legend.title=element_blank()) +
  scale_x_discrete(limits = c("Not Fun", "Fun"))

Answer 1

Use scale_x_discrete(limits = c("1", "2")) ; this will also bring the two categories closer together.

library(sjPlot)
#> Learn more about sjPlot with 'browseVignettes("sjPlot")'.
library(sjmisc)
library(ggplot2)
data(efc)
theme_set(theme_sjplot())

# make categorical
efc$c161sex <- to_factor(efc$c161sex)

# fit model with interaction
fit <- lm(neg_c_7 ~ c12hour + barthtot * c161sex, data = efc)

plot_model(fit, type = "pred", terms = c("c161sex", "barthtot [0, 100]")) +
  scale_x_discrete(limits = c("1", "2"))
#> Scale for x is already present.
#> Adding another scale for x, which will replace the existing scale.

---

NEW EDIT (after additional data/info)

Somehow you should indicate, if the original post is being substantially changed. In this new addition you use a function lm_robust which is from a package estimate`.

Key here is that you must define factor levels of data$X : levels(data$X) <- list('Not Fun' = 0, "Fun" = 1) . Then you can together with scale_x_discrete() change the axis.ticks.

Within plot_model I have used axis.lim = list(c(.6,2.4), c(-.2,1.5) to tweak the x-axis, so that they come closer together.

library(sjPlot)
#> Learn more about sjPlot with 'browseVignettes("sjPlot")'.
library(sjmisc)
library(ggplot2)

library(estimatr)
data = read.table(text="
X   Z   Y
0   1   1
1   0   1
1   0   1
0   1   1
1   1   0
1   0   0
0   0   0
1   1   1
1   1   1
0   0   1
1   0   1
1   1   1
0   1   0
1   0   0
1   0   0
0   1   1
1   1   1
1   0   1
0   0   1
1   1   1
1   1   0
0   0   0
1   0   0
1   1   1
0   1   1
1   0   1
1   0   1
0   1   1
1   1   0
1   0   0", header=TRUE)

data$X <- factor(data$X)
levels(data$X) <- list('Not Fun' = 0, "Fun" = 1)
data$Z <- factor(data$Z)

h <- lm_robust(Y ~ X + Z, data = data)

plot_model(h, type ='eff', terms = c("X", "Z"),
           axis.lim = list(c(.6,2.4), c(-.2,1.5))
           )+
  scale_x_discrete(limits = c("0" = "Not Fun", "1" = "Fun"))