Some notes on Principal Component Analysis

Principal Component Analysis (PCA) is an ordination method that reduces the dimensionality of multivariate data by creating few new key explanatory variables called principal components (PCs).
Each PC accounts for as much variance in the data as possible, provided that all the PAs are uncorrelated: therefore all PCs are independent and orthogonal.
It is possible to order the PCs according to the amount of total variation they explain, as well as to determine the relative contribution of each of the original variables to each PA.

Click here for a very good, interactive explanation of the idea behind PCA.

A practical example follows using the software R on the “iris” dataset: Continue reading