Correlation plot between two variables with line and person r value in graph - seeking alternate example

Question

Would just like some clarity here and a different example if someone has one. Initially I wanted to use this example because it has the graph, the mean line, and the r value all presented in the graph: http://www.sthda.com/english/wiki/correlation-test-between-two-variables-in-r

However, I'm using r studio server and creating a shiny app. Library ggpubr will simply not install. I've tried several ways to get this library to install.

So, does anyone have an alternate example that might work?

Cheers ~!

Answer 1

How about this:

library(ggplot2)
data(mtcars)

r <- round(cor(mtcars$wt, mtcars$mpg), 2)
p <- cor.test(mtcars$wt, mtcars$mpg)$p.value
ggplot(mtcars, aes(y=wt, x=mpg)) + 
  geom_point() + 
  geom_smooth(method="lm", col="black") + 
  annotate("text", x=20, y=4.5, label=paste0("r = ", r), hjust=0) +
  annotate("text", x=20, y=4.25, label=paste0("p = ", round(p, 3)), hjust=0) +
  theme_classic() 
  

Answer 2

You could use the geom_smooth function from ggplot2 and implement the correlation coefficient aswell as the p-value as follows:

library(ggplot2)
my_data <- mtcars
cor_coefs <- cor.test(my_data$mpg, my_data$wt)
ggplot(data = my_data, aes(x = mpg, y = wt)) + 
  geom_point() +
  geom_smooth(method=lm , color="red", fill="#69b3a2", se=TRUE) +
  annotate("text", x = 30, y = 4, label = paste0("R: ", round(cor_coefs$estimate, 2))) +
  annotate("text", x = 30, y = 3.5, label = paste0("p-value: ", round(cor_coefs$p.value, 10)))

cor_coefs safes the correlation test and you can use it to get the desired values. With annotate from ggplot2 , you need to specify the x and y position of your text. You could implement that dynamically based on your needs (since you have not provided any data).