Mon premier blog

Wavelet methods for time series analysis Andrew T. Walden, Donald B. Percival
Publisher: Cambridge University Press

CSSPM - Percival D.B., Walden A.T. This method advances Fourier analysis, where the basic shortcoming was that the Fourier spectrum contained only globally average information. They could be efficiently evaluated by passing γ through a series of filters (linear operators) obtaining at each step: i) wavelet coefficients for a given level, and ii) a downsampled signal to which the next round of evaluation is to be applied: