RFgui: Graphical User Interface For Fitting Covariance Models And Variograms

Description

This is a nice instructive graphical tool useful in particular for teaching classes

Usage

RFgui(data, x, y, same.algorithm = TRUE, ev, bin = NULL, xcov, ycov, sim_only1dim=FALSE, wait = 0, ...)

Arguments

data

see RFempiricalvariogram. If data is given, the empirical variogram is shown

a sequence of the locations of the simulated process; if not given, x is determined by data and if data is not given by default values

a sequence of numbers if a simulation on $R^d$ is performed. Default is y = x; see x for details.

same.algorithm

Force the picture being simulated with the same algorithm so that the pictures are always directly comparable. The disadvantage is that some models are simulated only (very) approximatively.

instead of the data, the emprirical variogram itself might be passed

bin

only considered if data is given. See RFempiricalvariogram for details.

xcov

sequence of the locations where the covariance function is plotted

ycov

Only for anisotropic models. sequence of the locations where the covariance function is also plotted

sim_only1dim

Logical. The argument determines whether a process should be simulated on the line or on the plane

wait

integer. See details.

...

further options and control arguments for the simulation that are passed to and processed by RFoptions.

Value

If wait < 0 the the function returns NULL else it returns the last chosen RMmodel.If wait < 0, a side effect effect of RFgui is the creation of the variable RFgui.model on .GlobalEnv.

Details

If wait is negative the xterm does not wait for the tkltk-window to be finished. Further the variable RFgui.model is created in the environment .GlobalEnv and contains the currently chosen variable in the gui. RFgui always return NULL.

If wait is non-negative the xterm waits for the tkltk-window to be finished. RFgui returns invisibly the last chosen model (or NULL if no model has been chosen). RFgui idles a lot when wait=0. It idles less for higher values by sleeping about wait microseconds. Of course the handling in the tkltk window get slower also. Reasonable values for wait are within [0,1000].

same.alg = TRUE is equivalent to setting circulant.trials=1, circulant.simu_method = "RPcirculant", circulant.force=TRUE, circulant.mmin=-2.

Examples

Run this code

RFoptions(seed=0) ## *ANY* simulation will have the random seed 0; set
##                   RFoptions(seed=NA) to make them all random again
RFgui()