Practical Predictive Analytics

Nominal with quantitative variables

These comparisons are often examined using basic bar charts or, in the following example, using side-by-side box-plots. These boxplots give very clear comparisons.

The following code simulates two ad campaigns A, and B with two different outcomes:

set.seed(123)  rows=100 a <- data.frame(Sales=rnorm(rows,mean=75,sd=5),                 Treatment=factor(c("Campaign A"))) b <- data.frame(Sales=rnorm(rows,mean=80,sd=5),                 Treatment=factor(c("Campaign B"))) combined=rbind(a,b) #Boxplots which compare treatments  boxplot(Sales~Treatment,data=combined, main="Comparing Treatments",            xlab="Treatment", ylab="Sales")