Vous êtes sur la page 1sur 4

HOUSING DATA ANALYSIS

Graphic Analysis
Boxplot: It looks like there are seven outliers in the price set

Using the uvapairs code, I developed the following chart

It looks like the size and price of the house have a roughly linear relationship, with the highest pearson coeff amongst the variables

Principle Components Analysis


I first found the principal components for the variables Price, Rooms, Baths, and Size. We leave out the age variable since it is categorical.
house.pca <- princomp(house[,c("Rooms","Baths","Size","Price")],cor=T) > par(mfrow = c(1,1)) > biplot(house.pca)

Notably, Price is roughly othoganal with rooms, while price, size, and baths are all in the same general direction.

Size comprises .559 of the first component, followed by price at .51.


house.pca$loadings[,1] Rooms Baths Size Price -0.4253316 -0.4910334 -0.5589513 -0.5153180

Vous aimerez peut-être aussi