Fig. 2

Principal component analysis (PCA), functional principal component analysis (FPCA) and wavelet-based principal component analysis (WPCA). Panel a – Principal components (PCs) resulting from a PCA on raw data; Panel b – Functional principal components (FPCs) resulting from a FPCA using Fourier basis functions and three different smoothing parameters; Panel c – Functional principal components (FPCs) resulting from a FPCA using B-splines basis functions and three different smoothing parameters. Panel d– Wavelet principal components (WPCs) resulting from a WPCA using the Haar mother wavelet and three different shrinkage rules; Panel e – Wavelet principal components (WPCs) resulting from a WPCA using the Daubechies extremal phase mother wavelet and three different shrinkage rules; Panel f – Wavelet principal components (WPCs) resulting from a WPCA using the Daubechies least asymmetric mother wavelet and three different shrinkage rules