Petals and plots à la Tufte

🌸🌸🌸

Hi,

In the mornings, I’m in the fields; in the afternoons, I’m back to stats. And since I have to stare at lots of correlation coefficients, I thought I’d recommend three really good videos:

1. on correlation which isn’t causation

2. on degrees of freedom, here and here

I also need to quickly plot some variables against each other from time to time. If you like the idea of minimalistic plots in the spirit of Edward Tufte, here are just a handful of lines to achieve that (inspiration comes from this chapter). And since everything around is blossoming, let’s use the in-built iris dataset.

### Load the libraries ###
library(ggplot2)
library(ggthemes)

### Draft the plot ###

summary(iris)

ggplot(iris, aes(Sepal.Length, Sepal.Width)) + 
  geom_point() + 
  theme_tufte() +
  xlab("Sepal length") + 
  ylab("Sepal width")

The summary() function gives us a quick overview of the number and type of columns, so we can check whether sepal width (the sepal is part of the flower head) correlates with sepal length. Here it how it looks like:

Enjoy the break!

Aga

PS: Here is the survey in which you can tell me what R topic you find particularly confusing and why you want to learn it, so that we can shape this space together!