opt-class {TIMP} | R Documentation |
Class "opt" stores options for fitting and plotting applicable to all model types
See kinopt-class
, specopt-class
and
massopt-class
for
the specification of fitting/plotting options that are specific to the
class type.
Objects can be created by calls of the form new("opt", ...)
or
opt(...)
.
algorithm
:"character"
that defaults to
algorithm="nls"
, so that the function nls
is used to
optimize nonlinear parameters under least squares criteria.
Other options are
nls.lm
:nls.lm
optim
:optim
; if this option is used then it MUST be used in
conjunction with nnls=TRUE
. Currently, it must also be
used with stderrclp=FALSE
. nnls
:"logical"
that defaults to
FALSE
. If nnls=TRUE
,
constrain the conditionally linear parameters
to nonnegativity via a nonnegative least squares algorithm as
implemented via the function nnls
from the package by the same
name.writecon
:"logical"
that defaults to
FALSE
; if true then
concentrations are written to a txt file; row labels are x
writespec
:"logical"
that defaults to
FALSE
; if TRUE
then
spectra are written to a txt file; row labels are x2
writenormspec
:"logical"
that
defaults to FALSE
; if TRUE
then normalized spectra are
written to a txt file; row labels are x2
writefit
:"logical"
that defaults to FALSE
; if TRUE
then fit
is written to a txt file; row and column labels are x
and
x2
writeclperr
:"logical"
that defaults to
FALSE; if true then
the error bars for clp are written to a txt file. This option is only
sensible with stderrclp=TRUE
.output
:"character"
that defaults to "ps"
, which means that
plots written to file are postscript. Alternatively, specify
output = "pdf"
, and plots are written as pdf filesaddfilename
:"logical"
that, for each data file, tries to add the
filename to plots associated with output for that data. residplot
:"logical"
defaults to
FALSE
; if TRUE
generate a plot of residuals in a
separate window.adddataimage
:"logical"
defaults to
FALSE
; if TRUE
adding imageplot of data in summary plot.plot
:"logical"
that defaults to
TRUE
; if FALSE
then do not write output in the form of
plots and other windows to the screen. divdrel
:"logical"
that defaults to
FALSE
; if TRUE
, plot traces and concentration
profiles divided by the dataset scaling parameters where they apply; this
allows for the fit of datasets having different intensities on the same
scale.plotkinspec
:"logical"
that defaults
to FALSE
; if TRUE
, generates a separate plot of the spectra
associated with the components that are not a part of a coherent
artifact/scatter model. superimpose
:"vector"
containing
dataset indices whose results should be superimposed in plotsxlab
:"character"
containing
label for x-axis, e.g., "nanoseconds"
or "picoseconds"
ylab
:"character"
containing label
for y-axis, e.g., "wavelength"
title
:"character"
containing title
to write at the top of plots.makeps
:"character"
containing
prefix to plot files written to postscript;
if present postscript will be written. Note that this string is also
used as the preffix of txt output files linrange
:"numeric"
giving linear
range of time axis for plotting; time will be plotted linearly from
-linrange to linrange and plotted on a logarithmic (base 10) axis elsewheresummaryplotrow
:"numeric"
giving
number of rows in summary plot; defaults to 4
summaryplotcol
:"numeric"
giving
number of columns in summary plot; defaults to 4
iter
:"numeric"
giving
number of iterations to optimize model parameters; if
nls=FALSE
so that the Levenberg-Marquardt algorithm is
applied, then iter
is interpretted as the
maximum number of residual function evaluations (see the help
page of the function nls.lm
for details)paropt
:"list"
of graphical parameters in format par(...)
to apply to plots. stderrclp
:"logical"
that defaults
to FALSE
; if TRUE
, estimates of the standard error of
conditionally linear parameters are madeaddest
:"vector"
containing
character strings of which parameter estimates should be added to the
summary plot, e.g., addest = c("kinpar", "irfpar")
"logical"
that defaults to
FALSE
; if TRUE
, add standard error estimates to the clp
a plot generated with kinspecest=TRUE
or
plotkinspec=TRUE
. This option can only be
used if the estimates were generated during fitting via the option
stderrclp=TRUE
"vector"
that defaults to
vector()
; if changed, it should specify the desired x-limits of
the plot of clp"vector"
that defaults to
vector()
; if changed, it should specify the desired y-limits of
the plot of clp. In the case of plotting the results of FLIM image
analysis, ylimspec
can be used to determine the range used in the
image plot of lifetimes. "vector"
that defaults to
vector()
; if changed, the first value should specify a vector to
add to the y-limits of the plot of clp"logical"
that defaults to
FALSE
; if TRUE
, then the line-type for clp is the same
for all datasets
"logical"
that defaults to
FALSE
; if TRUE
, use spline instead of lines between
the points representing estimated clp"logical"
that defaults to
TRUE
; if TRUE
, add points representing the actual estimates
for clp to plots of the curves respresenting smoothed clp"numeric"
that defaults to
50
; represents the number of segments used in a spline-based
representation of clp"logical"
that defaults
to FALSE
; determines whether a B-spline based representation of
clp is used (when specinterpol=TRUE
) or a piecewise polynomial
representation"logical"
that determines whether
clp are normalized in plots"logical"
that defaults to
FALSE
; if TRUE
, a spline-based representation of clp
is written to ASCII files"character"
that defaults to
"default"
and determines the algorithm used by nls
, if
nls
is used in optimization. See help(nls)
for other
possibilities, such as "port"
, which is more stable with
respect to starting values but requires more time. "numeric"
if given, sets the line
type of the fit in plots of the fit/data; see lty
in
help(par)
for options. "numeric"
if given, sets the line
type of the data in plots of the fit/data; see lty
in
help(par)
for options. "vector"
if given, sets the color
of the fit corresponding to each dataset
in plots of the fit/data; see col
in
help(par)
for options. If given
length(colfit)
must be equal to
the number of datasets in the analysis"vector"
if given, sets the color
of the data for each dataset
in plots of the fit/data; see col
in
help(par)
for options. If given,
length(coldata)
must be equal to
the number of datasets in the analysis
Katharine M. Mullen, Ivo H. M. van Stokkum