Do any of the R GUI assist with power calculations / analysis?

Question

I met with several people who are familiar with SPSS and would be reluctant to wade into R but might be encouraged to do so from the comfort of a GUI. Do any of the GUIs provide support for power calculations? I searched around for about 15 minutes and could not find anything to tell me that they could. Do you happen to know?

Answer 1

While it's targeting data mining tasks, you could have a look at Rattle (see the related R Journal paper). I'm not sure if it supports power calculations, but it is intended to be more of a user-friendly GUI.

Answer 2

Red-R wraps many packages in an easy to use GUI interface. If you're looking for a specific package that isn't included, email the devs and they'll be glad to add it, usually within a few days.

Answer 3

Red-R is a visual programming interface for R designed to bring the power of the R statistical environment to the general researcher or user. The goal of this project is to provide access to the massive library of packages in R without any programming expertise. The Red-R framework uses concepts of dataflow programming to make data the center of attention while hiding all the programming complexity. In visual programming, functional tools (called Widgets in Red-R) are linked together on a canvas to control the flow of data through the program. This framework allows novice users to quickly and easily build complex analysis pipelines.

There are a couple of videos demonstrating the applcation at Red-R Documentation

Some of the current functionality includes:

• Read/View Data
• Merge/Intersect/Filter
• Math/Apply
• Plotting
• Stats - Parametric
• Stats - Non-Parametric

Some advanced functionality:

• Bioconductor microarray analysis
• Survival analysis
• Spatial Stats
• SQLite
• ROCR – ROC Curves
• Neural Nets
• LME4

This is definitely a work in progress, but you can check out the application at Red-R.org.

We are looking for any feedback in improving the usability and expanding the functionality. We'd be happy to work with anyone for specific needs.

Anup

Answer 4

Not that I'm aware of, but it is fairly easy to add the functionality to a number of them. For example, below is a simple power analysis dialog for Deducer ( http://www.deducer.org/manual.html ) for the paired/one-sample/two-sample t-test. For more on making dialogs in Deducer, see http://www.deducer.org/pmwiki/pmwiki.php?n=Main.Development .

If they are coming from SPSS they should feel very comfortable working with Deducer.

dialog <- new(SimpleRDialog)
dialog$setSize(400L,600L)
dialog$setTitle("t-test power analysis")

#type of test
test<- new(ComboBoxWidget,"type of test",c("two.sample", "one.sample", "paired"))
test$setDefaultModel("two.sample")
addComponent(dialog, test,100,900,200, 100)

#Sample size
ss <- new(TextAreaWidget,"Sample size")
addComponent(dialog, ss,210,700,310, 300)

#sig
sig <- new(TextAreaWidget,"significance level")
sig$setDefaultModel("0.05")
addComponent(dialog, sig,320,700,420, 300)

#power
pow <- new(TextAreaWidget,"Power")
pow$setDefaultModel("0.80")
addComponent(dialog, pow,430,700,530, 300)

#effect size
eff <- new(TextAreaWidget,"Cohens D")
eff$setDefaultModel(".5")
addComponent(dialog, eff,540,700,640, 300)


#alternative
test<- new(ComboBoxWidget,"alternative",c("two.sided", "less","greater"))
test$setDefaultModel("two.sided")
addComponent(dialog, test,650,900,750, 100)


runDialog <- function(state){
    #print(state)
    cmd <- "require(pwr)\npwr.t.test("

    if(state[['Sample size']] == "")
        parameter <- "n=NULL"
    else
        parameter = paste("n=",state[['Sample size']],sep="")
    cmd <- paste(cmd,parameter);

    if(state[['significance level']] == "")
        parameter <- ",sig.level=NULL"
    else
        parameter = paste(",sig.level=",state[['significance level']],sep="")
    cmd <- paste(cmd,parameter);

    if(state[['Power']] == "")
        parameter <- ",power=NULL"
    else
        parameter = paste(",power=",state[['Power']],sep="")
    cmd <- paste(cmd,parameter);

    if(state[['Cohens D']] == "")
        parameter <- ",d=NULL"
    else
        parameter = paste(",d=",state[['Cohens D']],sep="")
    cmd <- paste(cmd,parameter);

    parameter = paste(",alternative='",state[['alternative']],"'",sep="")
    cmd <- paste(cmd,parameter);

    parameter = paste(",type='",state[['type of test']],"')",sep="")
    cmd <- paste(cmd,parameter);

    execute(cmd)
}

dialog$setRunFunction(toJava(runDialog))

deducer.addMenu("Power")
deducer.addMenuItem("t-test power",,"dialog$run()","Power")
if(.windowsGUI){
    winMenuAdd("Power")
    winMenuAddItem("Power", "t-test", "deducer('t-test power')")
}else if(.jgr){
    jgr.addMenu("Power")
    jgr.addMenuItem("Power", "t-test", "deducer('t-test power')")
}