wlsbaseline
Purpose
Weighted least squares baseline function.
Synopsis
[bldata,wts] = wlsbaseline(data,baseline,options)
[bldata,wts] = wlsbaseline(data,order,options)
Description
Subtracts a baseline (or other signal) from a spectrum with
the constraint that residuals below zero be weighted more heavily than those
above zero. This achieves a robust "non-negaitve" residual fit when
residuals of significant amplitude (e.g. signals on a background) are present.
Inputs are data
the spectral data, baseline
the reference spectrum/spectra to use for baseline OR an integer value
representing the order of polynomial baselining to use and options an optional options
structure.
Outputs are the baselined data bldata and the weightings wts indicating the amount of baseline which
was removed from each spectrum in data. (i.e. bldata = data - wts*baseline)
Polynomial baseline Option: If a positive scalar value is
given instead of the input baseline,
then a polynomial baseline of that order will be used. In this mode, any row of
the output wts can be
used with the polyval function to obtain the baseline removed from the
corresponding row of data.
Options
plots : [{'none'} | 'debug' | 'intermediate' |
'final'] governs plots
weightmode
: [ {1} | 2 ] flag indicating which weighting mode to use.
Mode 1 = Power method.
Negative residuals are weighted up by the power of 10.^(option.negw). All
residuals are then raised to the power of (option.power)
Mode 2 = T squared
method. Negative residuals are weighted up by the extent to which the surpass
an estimate of the noise limit and the approximate t-limit defined by
(option.tsqlim)
trbflag : [ 'bottom' | 'top' ] baseline to top or bottom of
data
negw : {1} deweighting scale of negative values
(10^negw) (used only for weightmode = 1),
power : {2} exponential amplification of residuals
(used only for weightmode = 1),
tsqlim : [0.99] t-test confidence limit for significant
negative residuals which need to be up-weighted. (used only for weightmode =
2),
nonneg : ['no'|{'yes'}] flag to force non-negative baseline
weighting, most often used when "real" spectra are used for
baslineing and they should not be "flipped" by a negative weighting.
Using nonneg = 'yes', WLSBASELINE an be used as a partial CLS prediction to
estimate the concentration of a species when not all species' pure component
spectra are known,
delta : [1e-4] change-of-fit convergence
criterion,
maxiter : [100] maximum iterations allowed per spectrum,
maxtime : [600] maximum time (in seconds) permitted for
baselining of all data.
Examples
To swap 4 BYTES in a 32 bit number:
See Also
baseline, baselinew, mscorr, savgol
