correct_rt: Correct retention time

In ethanbass/chromatographR: Chromatographic Data Analysis Toolset

Correct retention time

Description

Aligns chromatograms using one of two algorithms, according to the value of alg: either parametric time warping, as implemented in ptw, or variable penalty dynamic time warping, as implemented in VPdtw. The init.coef and n.traces arguments apply only to ptw warping, while penalty and maxshift apply only to vpdtw warping.

Usage

correct_rt(
  chrom_list,
  lambdas,
  models = NULL,
  reference = "best",
  alg = c("ptw", "vpdtw"),
  what = c("corrected.values", "models"),
  init.coef = c(0, 1, 0),
  n.traces = NULL,
  n.zeros = 0,
  scale = FALSE,
  trwdth = 200,
  plot_it = FALSE,
  penalty = 5,
  maxshift = 50,
  verbose = getOption("verbose"),
  show_progress = NULL,
  cl = 2,
  ...
)

Arguments

chrom_list	List of chromatograms in matrix format.
lambdas	Select wavelengths to use by name.
models	List of models to warp by. The models provided here (if any) must match the algorithm selected in alg.
reference	Index of the sample that is to be considered the reference sample.
alg	algorithm to use: parametric time warping (ptw) or variable penalty dynamic time warping (vpdtw).
what	What to return: either the 'corrected.values' (useful for visual inspection) or the warping 'models' (for further programmatic use).
init.coef	Starting values for the optimization.
n.traces	Number of traces to use.
n.zeros	Number of zeros to add.
scale	Logical. If true, scale chromatograms before warping.
trwdth	width of the triangle in the WCC criterion.
plot_it	Logical. Whether to plot alignment.
penalty	The divisor used to calculate the penalty for VPdtw. The warping penalty is calculated by dividing the dilation by this number. Thus, a higher number will produce a lower penalty and be more permissive, while a lower number will produce a higher penalty and allow less warping. Defaults to 5.
maxshift	Integer. Maximum allowable shift for VPdtw. Defaults to 50.
verbose	Whether to print verbose output.
show_progress	Logical. Whether to show progress bar. Defaults to TRUE if pbapply is installed. Currently works only for ptw alignments.
cl	Argument to pblapply or mclapply. Either an integer specifying the number of clusters to use for parallel processing or a cluster object created by makeCluster. Defaults to 2. On Windows integer values will be ignored.
...	Optional arguments for the ptw function. The only argument that cannot be changed is warp.type: this is always equal to "global".

Value

A list of warping models or a list of warped absorbance profiles, depending on the value of the what argument.

Note

Adapted from correctRT function in the alsace package by Ron Wehrens.

Author(s)

Ethan Bass

References

• Clifford, D., Stone, G., Montoliu, I., Rezzi, S., Martin, F. P., Guy, P., Bruce, S., & Kochhar, S. 2009. Alignment using variable penalty dynamic time warping. Analytical chemistry, 81(3):1000-1007. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.1021/ac802041e")}.

• Clifford, D., & Stone, G. 2012. Variable Penalty Dynamic Time Warping Code for Aligning Mass Spectrometry Chromatograms in R. Journal of Statistical Software, 47(8):1-17. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.18637/jss.v047.i08")}.

• Eilers, P.H.C. 2004. Parametric Time Warping. Anal. Chem., 76:404-411. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.1021/ac034800e")}.

• Wehrens, R., Bloemberg, T.G., and Eilers P.H.C. 2015. Fast parametric time warping of peak lists. Bioinformatics, 31:3063-3065. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.1093/bioinformatics/btv299")}.

• Wehrens, R., Carvalho, E., Fraser, P.D. 2015. Metabolite profiling in LC–DAD using multivariate curve resolution: the alsace package for R. Metabolomics, 11:143-154. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.1007/s11306-014-0683-5")}.

See Also

ptw, correct_peaks, VPdtw

Examples

data(Sa_pr)
warping.models <- correct_rt(Sa_pr, what = "models", lambdas=c(210))
warp <- correct_rt(chrom_list = Sa_pr, models = warping.models)

ethanbass/chromatographR documentation built on April 17, 2025, 10:55 a.m.