Why does "scale_x_discrete" screw up the zoom on my plot? - r

Question

I'm trying to plot my data in R and trying to manually relabel the x axis keeps creating weird extra space on either side of the plot (see pictures).
Here's my code:
ACL_data_frame <- data.frame(ACLdata, Pre_PHQ_Score, Post_PHQ_Score)
colnames(ACL_data_frame)[c(328:329)] <- c("PHQ.1", "PHQ.2")
ACL_data_frame_long <- reshape(ACL_data_frame, direction="long", varying=328:329, sep=".")
ACL_data_frame_long$Condition <- factor(ACL_data_frame_long$Condition)
ggplot(data = ACL_data_frame_long, aes(x = time, y = PHQ, linetype = Condition)) +
stat_summary(fun=mean, geom="line", size=1.25) +
labs(x="Time", y="Depression (PHQ-9)") +
scale_x_discrete(limits=c("1","2"),
labels=c("Pre", "Post")) +
scale_linetype_discrete(name="Condition",
breaks=c("1", "2"),
labels=c("Control", "Intervention"))
The relevant columns in my data basically look like this (but longer):
time
PHQ
Condition
1
0.6666667
1
1
1.1111111
2
2
0.7777778
2
2
1.3333333
1
Does anyone know how to get rid of the weird space on either side of the plot (see pics)? I tried xlim but that gives an error. Removing the scale_x_discrete also fixes it, but then I can't figure out another way to label the x axis correctly (i.e., there are only 2 timepoints so having labels for Time 1.25, Time 1.5, etc doesn't make sense, they need to just be labeled "Pre" and "Post")
Plot using scale_x_discrete, with correct x axis labels but incorrect margins
Plot without scale_x_discrete, with correct margins but incorrect x axis labels

Answer 1

As your time column is a numeric you get a continuous scale by default. Therefore, to set the breaks and labels use scale_x_continuous(breaks = 1:2, labels = c("Pre", "Post")). Adding scale_x_discrete will switch to a discrete scale which by default adds an expansion of .6 to the lower and the upper end of the scale. In contrast, scale_x_continuous adds an expansion of 5% of the data range to the lower and the upper end of the scale.
library(ggplot2)
ggplot(data = ACL_data_frame_long, aes(x = time, y = PHQ, linetype = Condition)) +
stat_summary(fun = mean, geom = "line", size = 1.25) +
labs(x = "Time", y = "Depression (PHQ-9)") +
scale_x_continuous(breaks = 1:2, labels = c("Pre", "Post")) +
scale_linetype_discrete(
name = "Condition",
breaks = c("1", "2"),
labels = c("Control", "Intervention")
)