Contour analysis of Hungarian folk music in a multidimensional metric-space

A collection of folk tunes of a Hungarian ethnic group was investigated using a mathematical model, relating multi-dimensional points to melodies. The distribution of the points was studied using eigenvector analysis of the correlation matrix. A scalar measure of musical distance was formulated, in accord with the Euclidean norm, in the basis of the most important eigenvectors. The analysis of the distances between the points indicated cluster formation in agreement with the results of classical musicology. Visual representation was accomplished by projecting the multi-dimensional space to a two dimensional plane of variable position.