r: Blank graph when plotting multiple lines on scatterplot

Question

My goal is to produce a graph showing the differences between regression lines using continuous vs categorical variables. I'm using is the "SleepStudy" dataset from Lock5Data , and I want to show the regression lines predicting GPA from ClassYear as either continuous or categorical. The code is below:

library(Lock5Data)
data("SleepStudy")
fit2 <- lm(GPA ~ factor(ClassYear), data = SleepStudy)
fit2_line <- aggregate(fit2$fitted.values ~ SleepStudy$ClassYear, FUN = mean)
colnames(fit2_line) <- c('ClassYear','GPA')

options(repr.plot.width=5, repr.plot.height=5)
library(ggplot2)
ggplot() +
    geom_line(data=fit2_line, aes(x=ClassYear, y=GPA)) + # Fit line, ClassYear factor
    geom_smooth(data=SleepStudy, method='lm', formula=GPA~ClassYear) + # Fit line, ClassYear continuous
    geom_point(data=SleepStudy, aes(x=ClassYear, y=GPA)) # Data points as dots

What is producing the blank graph? What am I missing here?

Answer 1

You have to define the data you are using for the geom_smooth in the ggplot() . This code works:

ggplot(data=SleepStudy, aes(y = GPA,x = ClassYear)) +
  geom_smooth(data=SleepStudy, method='lm', formula=y~x)+
  geom_line(data=fit2_line, aes(x=ClassYear, y=GPA)) + 
  geom_point(data=SleepStudy, aes(x=ClassYear, y=GPA))