Learn R Programming

pmg (version 0.9-43)

pmg-dynamic: "Dynamic" widgets for pmg

Description

We call a widget "dynamic" if it updates itself immediately when an event occurs, such as a drag and drop, or a change in some value. The dynamic widgets documented here, are meant to provide quick, easy (but limited) access to R's modeling functions, R's significance tests, and R's lattice functions

Usage

dModelsDialog() dTestsDialog() dLatticeExplorer(container = NULL, ...)

Arguments

container
A container to attach the object to
...
Currently ignored

Value

Although there are methods for dModelsDialog, these widgets aren't meant to be interacted with from the command line.

Details

For each "dynamic" widget, the variables can be specified by drag and drop, or by editing the widget. The bold-face areas of each widget can be edited by clicking on them or by dropping values. If the drop value comes from a column of an idf instance, then when that column is edited, the dynamic widget is updated. Such variables can not be edited or changed. Other variables may, such as writing powers, or applying functions.

The "dynamic" widgets are meant for easy exploration, but not for saving of actions.

The dModelsDialog shows an interface to lm, aov, and rlm. The user can only specifiy formulas of the type y ~ 1 + x1 + x2 + ... + xn. Dropping a value on "response" changes the response. Dropping a value on the right side of the ~ adds the term (using +). If the terms are edited by clicking, the values are split on the + sign.

For each model fit, a drop list allows one to generate several of the diagnostic plots.

The dTestsDialog offers an interface to most of the tests in the stats package of class ctests. (The chisq.test is not implemented yet.) Not only can variables be dropped, but one can also change, as appropriate, the choice of the null, the alternative, etc. Again, the bold-face terms may be edited by clicking on them.

The ilatticeexplorer function creates a dynamic graphing widget based on lattice graphics. Up to three variables (only 2 for univariate graphs) may be dropped on the widget. The order is for univariate graphs: ~x then ~x | y. And for bivariate graphs x, x ~ y, x ~ y | z. The panel functions add to the plots of dots by, typically, incorporating some trend line.

Examples

Run this code
## Not run: 
# dTestsDialog()
# ## End(Not run)

Run the code above in your browser using DataLab