08/07/2021

Motivation

One issue I have come across during the Data Analysis phase is looking for correlations between variables.

Functions such as ggpairs() in the GGally package are useful, but plots are often small and hard to read.

The following slide illustrates the difficulty of reading such outputs.

ggpairs example

library(GGally)
ggpairs(mtcars)

Solution

Following this issue, one can find themselves typing many commands to view correlations between variables.

A simple solution to this is an App which can produce a single chart based on the users selection of two variables.

This App includes the plot with a linear regression line with confidence interval.

How to use the App

  • Choose two factors of interest from the two selection boxes
  • Investigate whether the two variables may be strongly correlated or not