How to make a scatterplot with grouping variables?

Question

Say I have a dataset that looks like what is shown in the picture, and I want to make a scatterplot whose x-axis presents the results during the A test phase and y-axis presents the results during the B test phase. The results is a continous variable. There should be four data points pinpointed in the graph, representing the four subjects. How am I supposed to do it with ggplot in R?

ggplot(data, aes(x = results, y = results)) + geom_point()

This is what I tried, but it does not give me what I wanted.

Please don’t use images of data as they cannot be used without a lot of unnecessary effort. [For multiple reasons](//meta.stackoverflow.com/q/285551). Best if your question is [reproducible0]https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example. — Peter, Jun 26 '23 at 08:17

score 0 · Accepted Answer · answered Jun 26 '23 at 08:30

You'll need the test_phase data in wide format in this case as the values provide the x, y coordinates of the plot. This can be achieved using tidyr::pivot_wider. After that it all down to formatting with ggplot.

# Some dummy data

set.seed(12)
df1 <- data.frame(subject = rep(1:4, 2),
                  test_phase = rep(c("A", "B"), each = 4),
                  results = sample(1000:6000, 8))

# packages for data wrangling and plotting
library(tidyr)
library(ggplot2)

df1 |> 
  pivot_wider(names_from = test_phase, values_from = results) |> 
  ggplot(aes(A, B, colour = factor(subject))) +
  geom_point() +
  labs(colour = "Subject")

^{Created on 2023-06-26 with reprex v2.0.2}

How to make a scatterplot with grouping variables?

1 Answers1