R Graphical Manual

Browse All

Last data update: 2014.03.03

R: Change the Scales of a quickPredict Object for an Interaction...

changeScale

R Documentation

Change the Scales of a quickPredict Object for an Interaction Term

Description

Designed to transform results of quickPredict obtained on interaction terms from the transformed scale (on which the variables are approximately uniformly distributed) onto the "native", linear scale.

Usage

changeScale(obj, xFct, yFct)

Arguments

`obj`	a `quickPredict` object.
`xFct`	a function to be applied on the `xx` element of `obj`. This function should be the `qFct` attribute of the function, returned by `mkM2U`, used to transform the variable from the "native" to the "uniform" scale.
`yFct`	a function to be applied on the `yy` element of `obj`. This function should be the `qFct` attribute of the function, returned by `mkM2U`, used to transform the variable from the "native" to the "uniform" scale.

Value

A quickPredict object.

Author(s)

Christophe Pouzat christophe.pouzat@gmail.com

Examples

## Not run: 
data(e060824spont)
DFA <- subset(mkGLMdf(e060824spont,0.004,0,59),neuron==1)
DFA <- within(DFA,i1 <- isi(DFA,lag=1))
DFA <- DFA[complete.cases(DFA),]
m2u1 <- mkM2U(DFA,"lN.1",0,29)
m2ui <- mkM2U(DFA,"i1",0,29,maxiter=200)
DFA <- within(DFA,e1t <- m2u1(lN.1))
DFA <- within(DFA,i1t <- m2ui(i1))
with(DFA,plot(ecdf(e1t[time>29]),pch="."))
abline(a=0,b=1,col=2,lty=2)
with(DFA,plot(ecdf(i1t[time>29]),pch="."))
abline(a=0,b=1,col=2,lty=2)
m1.fit <- gssanova(event~e1t*i1t, data=subset(DFA,time>29), family="binomial", seed=20061001)
inter.pred <- m1.fit %qp% "e1t:i1t"
contour(inter.pred,what="mean",nlevels=10,col=2,lwd=2)
contour(inter.pred,what="sd",nlevels=5,col=1,lwd=1,lty=2,add=TRUE)
inter.predN <- changeScale(inter.pred,attr(m2u1,"qFct"),attr(m2ui,"qFct"))
contour(inter.predN,what="mean",nlevels=5,col=2,lwd=1)
contour(inter.predN,what="sd",nlevels=3,col=1,lwd=1,lty=2,add=TRUE)

## End(Not run)