A multivariate dataset, arranged as a matrix, can be interpreted as an example of two-way dataset. In this case, each value in the matrix is characterized by two factors (one factor is an object and the other factor is a property), so each cell contains the corresponding property of the corresponding object. Each row contains properties for the same object and each column the same property for various objects. For instance, chromatograms have to be recorded exactly in the same way (the same number of points, the same time sampling frequency, etc.) to fit all in one matrix.