subtr_baseline: Automated background subtraction for spectral data

In wincowgerDEV/OpenSpecy: Analyze, Process, Identify, and Share Raman and (FT)IR Spectra

Automated background subtraction for spectral data

Description

This baseline correction routine iteratively finds the baseline of a spectrum using polynomial fitting methods or accepts a manual baseline.

Usage

subtr_baseline(x, ...)

## Default S3 method:
subtr_baseline(
  x,
  y,
  type = "polynomial",
  degree = 8,
  raw = FALSE,
  full = T,
  remove_peaks = T,
  refit_at_end = F,
  crop_boundaries = F,
  iterations = 10,
  peak_width_mult = 3,
  termination_diff = 0.05,
  degree_part = 2,
  bl_x = NULL,
  bl_y = NULL,
  make_rel = TRUE,
  ...
)

## S3 method for class 'OpenSpecy'
subtr_baseline(
  x,
  type = "polynomial",
  degree = 8,
  raw = FALSE,
  full = T,
  remove_peaks = T,
  refit_at_end = F,
  crop_boundaries = F,
  iterations = 10,
  peak_width_mult = 3,
  termination_diff = 0.05,
  degree_part = 2,
  baseline = list(wavenumber = NULL, spectra = NULL),
  make_rel = TRUE,
  ...
)

Arguments

x	a list object of class OpenSpecy or a vector of wavenumbers.
y	a vector of spectral intensities.
type	one of "polynomial" or "manual" depending on the desired baseline correction method.
degree	the degree of the full spectrum polynomial. Must be less than the number of unique points when raw is FALSE. Typically, a good fit can be found with an 8th order polynomial.
raw	if TRUE, use raw and not orthogonal polynomials.
full	logical, whether to use the full spectrum as in "imodpoly" or to partition as in "smodpoly".
remove_peaks	logical, whether to remove peak regions during first iteration.
refit_at_end	logical, whether to refit a polynomial to the end result (TRUE) or to use linear approximation.
crop_boundaries	logical, whether to smartly crop the boundaries to match the spectra based on peak proximity.
iterations	the number of iterations for automated baseline correction.
peak_width_mult	scaling factor for the width of peak detection regions.
termination_diff	scaling factor for the ratio of difference in residual standard deviation to terminate iterative fitting with.
degree_part	the degree of the polynomial for "smodpoly". Must be less than the number of unique points.
bl_x	a vector of wavenumbers for the baseline.
bl_y	a vector of spectral intensities for the baseline.
make_rel	logical; if TRUE, spectra are automatically normalized with make_rel().
baseline	an OpenSpecy object containing the baseline data to be subtracted (only for "manual").
...	further arguments passed to poly() or "smodpoly" parameters.

Details

This function supports two types of "polynomial" automated baseline correction with options. Default settings are closest to "imodpoly" for iterative polynomial fitting based on Zhao et al. (2007). Additionally options recommended by "smodpoly" for segmented iterative polynomial fitting with enhanced peak detection from the S-Modpoly algorithm (https://github.com/jackma123-rgb/S-Modpoly), and "manual" for applying a user-provided baseline.

Value

subtr_baseline() returns a data frame containing two columns named "wavenumber" and "intensity".

Author(s)

Win Cowger, Zacharias Steinmetz

References

Chen MS (2020). Michaelstchen/ModPolyFit. MATLAB. Retrieved from https://github.com/michaelstchen/modPolyFit (Original work published July 28, 2015)

Zhao J, Lui H, McLean DI, Zeng H (2007). “Automated Autofluorescence Background Subtraction Algorithm for Biomedical Raman Spectroscopy.” Applied Spectroscopy, 61(11), 1225–1232. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.1366/000370207782597003")}.

Jackma123 (2023). S-Modpoly: Segmented modified polynomial fitting for spectral baseline correction. GitHub Repository. Retrieved from https://github.com/jackma123-rgb/S-Modpoly.

See Also

poly(); smooth_intens()

Examples

data("raman_hdpe")

# Use polynomial
subtr_baseline(raman_hdpe, type = "polynomial", degree = 8)

subtr_baseline(raman_hdpe, type = "polynomial", iterations = 5)

# Use manual
bl <- raman_hdpe
bl$spectra$intensity <- bl$spectra$intensity / 2
subtr_baseline(raman_hdpe, type = "manual", baseline = bl)

wincowgerDEV/OpenSpecy documentation built on June 14, 2025, 5:57 a.m.