Nonlinear Regression Dialog Functions

DESCRIPTION:

These functions are used by the Nonlinear Regression dialog. menuNls calls tabSummary.nls and tabPredict.nls if summary and prediction results are requested.

USAGE:

menuNls(formula, data=sys.parent(), start=<<see below>>, trace=F, 
        maxiter=50, tolerance=0.001, minscale=0.001, plinear.p=F, 
        print.short.p=F, print.long.p=T, save.name=NULL, 
        save.fit.p=F, save.resid.p=F, newdata=NULL, 
        predobj.name=NULL, predict.p=F, ci.p=F, se.p=F, 
        conf.level=0.95) 
tabSummary.nls(nlsobj, print.short.p=T, print.long.p=T, 
        save.name=NULL, save.fit.p=F, save.resid.p=F) 
tabPredict.nls(nlsobj, newdata=NULL, save.name, predict.p=F, 
        ci.p=F, se.p=F, conf.level=0.95) 

REQUIRED ARGUMENTS:

formula
a formula which specifies the nonlinear regression model.
nlsobj
an object that inherits from class nls.

OPTIONAL ARGUMENTS:

data
a data frame in which to do the computations. In addition to the usual data variables, the data frame may contain parameters (set, typically, by using the assignment form of parameters or param) that establish initial values for the model parameters.
start
a list or numerical vector. Although it is optional, use of start is recommended for unambiguous specification of the parameters. If start is omitted, the assumption is that any names occurring in formula that are not variables in the data frame are parameters. The list form of start allows the individual parameter names to refer to subsets of the parameters of arbitrary length. If a numeric starting vector is supplied, the named parameters must each be of length 1. In the case of partially linear models ( plinear.p = T), only the nonlinear parameters should be supplied.
trace
if TRUE, details of the iterations are printed.
maxiter
the maximum number of iterations during fitting.
tolerance
the tolerance for the convergence criterion in the algorithm. This is a relative offset criterion that measures the numerical imprecision in the parameter estimates compared to the statistical variability. Smaller values of this will require more iterations while larger values will result in convergence being declared earlier.
minscale
the minimum factor by which to shrink the default step size in an attempt to decrease the sum of squares.
plinear.p
if TRUE, the Golub-Pereyra algorithm for partially linear least-squares models is used.
print.short.p
if TRUE, a short summary of the nonlinear model is printed. This output is from the function print.nls.
print.long.p
if TRUE, a long summary of the nonlinear model is printed. This output is from the function summary.nls.
save.name
a character string for the name of the data frame to save the fit and residuals in. If data frame with this name already exists in database 1 and it has the appropriate number of rows then the saved values will be appended to the data frame. If the object already exist in database 1 and it is not a data frame or it does not have the appropriate number of rows then a new name is created by appending a number to save.name and the results are saved in the data frame with the new name.
save.fit.p
if TRUE, the fitted values from the regression are saved in the data frame save.name.
save.resid.p
f TRUE, the working residuals are saved in the data frame save.name. The working residuals are the response minus the fitted value.
newdata
a data frame to use for computing predictions. It must contain the same names as the terms in the right side of the formula for the model. If missing, the predictions for the original data are computed.
predobj.name
a character string for the name of the data frame to save the predictions, standard errors and confidence intervals in. If data frame with this name already exists in database 1 and it has the appropriate number of rows then the values will be appended to the data frame. If the object already exist in database 1 and it is not a data frame or it does not have the appropriate number of rows then a new name is created by appending a number to predobj.name and the values are saved in data frame with the new name.
predict.p
if TRUE, the predicted values are saved in predobj.name.
ci.p
if TRUE, lower and upper confidence limits will be stored in the predobj.name The column names will be "xx % L.C.L." and "xx% U.C.L" where xx is the value specified in conf.level. These confidence limits are for the mean response and are computed as the prediction plus/minus t-value * standard error.
se.p
if TRUE, the pointwise standard errors for the predictions will be stored in predobj.name.
conf.level
the confidence level to use when computing confidence intervals.

VALUE:

invisibly returns an object of class nls. See the nls.object help file for details.

SIDE EFFECTS:

Printed output will be displayed if requested. Plots will be drawn if requested. The objects save.name and predobj.name will be created or appended to if fitted values, residuals or predictions are saved.

SEE ALSO:

, , , .