| kin-class {TIMP} | R Documentation |
kin is the class for kinetic models;
an object
of class "kin" is initialized if
mod_type = "kin" is an
argument of initModel.
All objects of class kin are sub-classes of
class dat; see documentation for dat
for a description of
these slots.
See dat-class for an
example of the initialization of a
kin object via the initModel function.
Objects can be created by calls of the form new("kin", ...) or
kin(...).
specpar:"list" parameters for
spectral constraintsseqmod:"logical" that is TRUE if a sequential model is to be applied
and FALSE otherwiseirf:"logical" that is TRUE is an IRF is modeled and FALSE otherwisemirf:"logical" that is TRUE if a measured IRF is modeled and FALSE
otherwisemeasured_irf:"vector" containing a measured IRFconvalg:"numeric" 1-3 determining the numerical convolution algorithm
used in the case of modeling a measured IRF; if 3 then supply a
reference lifetime in the slot reftau.reftau:"numeric" containing a
reference lifetime to be used when convalg=3irffun:"character" describing the
function to use to describe the IRF, by default "gaus"irfpar:"vector" of IRF parameters; for the common Gaussian IRF this
vector is ordered c(location, width)dispmu:"logical" that is TRUE if dispersion of the parameter for IRF
location is to be modeled and FALSE otherwisedispmufun:"character" describing the functional form of the
dispersion of the IRF location parameter; if equal to "discrete" then the
IRF location is shifted per element of x2 and parmu should have the same
length as x2. defaults to a polynomial descriptionparmu:"list" of starting values for the dispersion model for the
IRF locationdisptau:"logical" that is TRUE if dispersion of the parameter for
IRF width is to be modeled and FALSE otherwisedisptaufun:"character" describing the functional form of the
dispersion of the IRF width parameter; if equal to "discrete" then the
IRF width is parameterized per element of x2 and partau should have the same
length as x2. defaults to a polynomial descriptionpartau:"vector" of starting values for the dispersion model for the
IRF FWHM fullk:"logical" that is TRUE if the data are to be modeled using a
compartmental model defined in a K matrix and FALSE otherwisekmat:"array" containing the K matrix descriptive of a compartmental
modeljvec:"vector" containing the J vector descriptive of the inputs to a
compartmental modelncolc:"vector" describing the number of columns of the C matrix for
each clp in x2kinscal:"vector" of starting values for branching parameters in a
compartmental modelkmatfit:"array" of fitted values for a compartmental modelcohspec:"list" describing the model for
coherent artifact/scatter component(s) containing the element type
and optionally the element numdatasets. The element
type can be set as follows:
"irf":type="irf", the coherent artifact/scatter has the
time profile of the IRF."freeirfdisp":type="freeirfdisp", the coherent
artifact/scatter has a
Gaussian time profile whose location and width are parameterized in the
vector coh. "irfmulti":type="irfmulti" the time profile of
the IRF is used for
the coherent artifact/scatter model, but the IRF parameters are taken per
dataset (for the multidataset case), and the integer argument
numdatasets must be equal to the
number of datasets modeled."seq":type="seq" a sequential exponential decay
model is applied, whose starting value are contained in an additional
list element start. This
often models oscillating behavior well, where the number of
oscillations is the number of parameter starting values given in
start. The starting values after optimization will be found in
the slot coh of the object of class theta corresponding
to each dataset modeled. "mix":type="mix" if type="mix" a sequential exponential decay
model is applied along with a model that follows the time profile of the IRF;
the coherent artifact/scatter is then a linear superposition of these two
models; see the above description of seq for how to supply the
starting values. coh:"vector" of starting values for the parameterization of a
coherent artifactwavedep:"logical" describing whether the kinetic model is dependent on
x2 index (i.e., whether there is clp-dependence)lambdac:"numeric" for the center wavelength to be used in a polynomial
description of x2-dependence amplitudes:"vector"
that may be used to multiply the concentrations by a square diagonal
matrix with the number of columns that the concentration matrix has;
the diagonal is given in amplitudes and these values will be
treated as parameters to be optimized.
streak:"logical"
that defaults to FALSE; if streak=TRUE then
the period of the laser is expected via streakT.streakT:"numeric"
the period of the laser; this will be used to add a
backsweep term to the concentration matrix and should be
set in conjunction streak=TRUE.doublegaus:"logical"
that defaults to FALSE and determines whether a double
Gaussian should be used to model the IRF. If doublegaus=TRUE
then irfpar should contain four numeric values corresponding to
the location (mean) of the IRF, the FWHM of the first Gaussian, the
FWHM of the second Gaussian, and the relative amplitude of the
second Gaussian, respectively.
Class dat-class, directly.
Katharine M. Mullen, Ivo H. M. van Stokkum