R Statistics Cookbook

How to do it...

In the following example, we will load a dataset containing house prices and several variables. We will direct our attention towards the first assumption we have previously defined (that the residuals should be homoscedastic: in other words, have the same variance):

1. We'll start by loading our dataset and estimating our model:

data = read.csv("./people_shopping.csv") model = lm(sales ~ people_in + discount,data=data) 

2. We can direct our attention towards the first plot. It is evident that the variability gets larger as the fitted values increase. This is a violation of our homoscedasticity assumption:

plot(model) 

The following screenshot shows the Residuals vs Fitted. The variance is non-constant:

3. We can use another ...