Abstract

Least squares regression is proposed as a moving-windows method for analysis of a series of spectra acquired as a function of external perturbation. The least squares moving-window (LSMW) method can be considered an extended form of the Savitzky–Golay differentiation for nonuniform perturbation spacing. LSMW is characterized in terms of moving-window size, perturbation spacing type, and intensity noise. Simulation results from LSMW are compared with results from other numerical differentiation methods, such as single-interval differentiation, autocorrelation moving-window, and perturbation correlation moving-window methods. It is demonstrated that this simple LSMW method can be useful for quantitative analysis of nonuniformly spaced spectral data with high frequency noise.

© 2017 The Author(s)

Spectral analysis for the generalized least squares phase-shifting algorithms with harmonic robustness

Sotero Ordones, Manuel Servin, Moises Padilla, Antonio Muñoz, Jorge L. Flores, and Ivan Choque
Opt. Lett. 44(9) 2358-2361 (2019)

Error analysis of the spectral shift for partial least squares models in Raman spectroscopy

Haiyi Bian and Jing Gao
Opt. Express 26(7) 8016-8027 (2018)

Nonlinear least squares analysis of atmospheric absorption spectra

E. Niple
Appl. Opt. 19(20) 3481-3490 (1980)

Computer Analysis of Resonance Profiles by the Method of Least Squares

David L. Ederer
Appl. Opt. 8(11) 2315-2325 (1969)

Least-squares analysis of the Mueller matrix

Michael Reimer and David Yevick
Opt. Lett. 31(16) 2399-2401 (2006)