A principal

component analysis (called Principal Compone nt Analysis, PCA) - one of the statistical methods of factor anal ysis. Data set consisting of N observations, each of which includ es k variables can be interpreted as a cloud of N points in k-dim ensional space. The purpose of PCA is a coordinate system rotation to maximize the variance in the first instance of the first coordinate, then the variance of the second coordinate, and so on. Thus transformed values of coordinates are called the loadings generated by the factors (principal components).