backcor.h File Reference

Baseline estimation using iterative polynomial fitting with a non-quadratic cost function(Backcor algorithm). More...

#include <stdexcept>
#include <vector>

Go to the source code of this file.

Enumerations
enum class	sablib::BackcorFunc {   Huber , AHuber , TQuad , ATQuad ,   Indec , AIndec }
	Cost function types for the Backcor algorithm. More...

Functions
const std::vector< double >	sablib::BaselineBackcor (const std::vector< double > &y, const unsigned int polyorder, const BackcorFunc func=BackcorFunc::ATQuad, const double s=1, const double alpha=0.99 *0.5, const unsigned int loop=50, const double eps=1.0e-3)
	Performs baseline estimation using iterative polynomial fitting with a non-quadratic cost function (Backcor).

Detailed Description

Baseline estimation using iterative polynomial fitting with a non-quadratic cost function(Backcor algorithm).

Author

Izadori

References:

• Mazet, V.; Carteret, C.; Brie, D.; Idier, J.; Humbert, B. "Background removal from spectra by designing and minimising a non-quadratic cost function" Chemometrics and Intelligent Laboratory Systems, 2005, 76(2), 121-133.
• Liu, J.; Sun, J.; Huang, X.; Li, G.; Liu, B. "Goldindec: A Novel Algorithm for Raman Spectrum Baseline Correction" Appl. Spectrosc., 2015, 69(7), 834-842.

Definition in file backcor.h.

Enumeration Type Documentation

BackcorFunc

enum class sablib::BackcorFunc

strong

Cost function types for the Backcor algorithm.

These functions define how the fitting error is penalized during the iterative polynomial fitting process. Asymmetric versions are typically used for signals with positive peaks to ensure the baseline stays below the signal.

Enumerator
Huber	Huber function.
AHuber	Asymmetric Huber function.
TQuad	Truncated quadratic function.
ATQuad	Asymmetric truncated quadratic function.
Indec	Indec function.
AIndec	Asymmetric Indec function.

Definition at line 25 of file backcor.h.

Function Documentation

BaselineBackcor()

const std::vector< double > sablib::BaselineBackcor	(	const std::vector< double > &	y,
		const unsigned int	polyorder,
		const BackcorFunc	func = BackcorFunc::ATQuad,
		const double	s = 1,
		const double	alpha = 0.99 *0.5,
		const unsigned int	loop = 50,
		const double	eps = 1.0e-3 )

Performs baseline estimation using iterative polynomial fitting with a non-quadratic cost function (Backcor).

This algorithm fits a polynomial of order polyorder to the signal y. In each iteration, the signal is modified based on the chosen cost function func to progressively estimate the background.

Parameters

y	The input data vector (signal to be processed).
polyorder	The order of the polynomial to be fitted.
func	The type of cost function to use (default: ATQuad).
s	Threshold parameter for the cost function (default: 1.0).
alpha	Control parameter for the iterative update (default: 0.99 * 0.5). Should be in range [0, 1].
loop	The maximum number of iterations (default: 50).
eps	Convergence threshold for the relative change in the estimated baseline (default: 1.0e-3).

Returns

A vector of the same size as y containing the estimated baseline.

Exceptions

std::invalid_argument

One or more parameters are wrong.

Definition at line 20 of file backcor.cpp.