In this paper, we present a decomposition method of multivariate functions. This method shows that any multivariate function f on [0, 1]d is a finite sum of the form ∑jФjψj, where each Фj can be extended to a smooth periodic function, each ψj is an algebraic polynomial, and each Фjψj is a product of separated variable type and its smoothness is same as f. Since any smooth periodic function can be approximated well by trigonometric polynomials, using our decomposition method, we find that any smooth multivariate function on [0, 1]d can be approximated well by a combination of algebraic polynomials and trigonometric polynomials. Meanwhile, we give a precise estimate of the approximation error.
When one applies the wavelet transform to analyze finite-length time series, discontinuities at the data boundaries will distort its wavelet power spectrum in some regions which are defined as a wavelength-dependent cone of influence (COI). In the COI, significance tests are unreliable. At the same time, as many time series are short and noisy, the COI is a serious limitation in wavelet analysis of time series. In this paper, we will give a method to reduce boundary effects and discover significant frequencies in the COI. After that, we will apply our method to analyze Greenland winter temperature and Baltic sea ice. The new method makes use of line removal and odd extension of the time series. This causes the derivative of the series to be continuous (unlike the case for other padding methods). This will give the most reasonable padding methodology if the time series being analyzed has red noise characteristics.