| plot.bctrans {alr3} | R Documentation |
Extract, view, and plot useful summaries of a ‘bctrans’ object for transformations to normality.
## S3 method for class 'bctrans': plot(x,y=x$Y, lambda=coef(x),family=x$family,plot=pairs,...) ## S3 method for class 'bctrans': summary (object, digits=4, ...) ## S3 method for class 'bctrans': coef(object,...) ## S3 method for class 'bctrans': vcov(object,...)
x |
An object created by bctrans. |
y |
If not NULL, add this variable to the scatterplot matrix |
object |
An object created by bctrans. |
lambda |
A vector of transformation parameters. |
family |
A family of transformations, probably "box.cox" or
"yeo.johnson" |
plot |
Name of the plotting function to view the results, usually
pairs or similar. |
digits |
Accuracy of summary output for ‘bctrans’ objects. |
... |
Arguments passed to the plotting method. |
The plot method is used for its side effect of producing a scatterplot matrix.
The helper summary
summarizes a bctrans object. This helper calls
lrt.bctrans, but the latter can be
called directly. The helpers coef and vcov
return, respectively, the estimated value of lambda, and an estimate of
its covariance matrix (based on a profile log-likelihood).
Sanford Weisberg, sandy@stat.umn.edu
Weisberg, S. (2005) Applied Linear Regression, third edition. Wiley.
bctrans, lrt.bctrans,
pairs,
inv.res.plot
data(highway) b <- highway[,c(8,1,2,10,5)] # select interesting columns summary(ans <- bctrans(b,family="yeo.johnson")) # zeros ==> use yeo.johnson b$Sigs <- (round(b$Sigs*b$Len)+1)/b$Len # redefine so no zeroes summary(ans <- bctrans(b)) # fit with box.cox b$Sigs <- (round(b$Sigs*b$Len)+1)/b$Len # redefine so no zeroes summary(ans <- bctrans(b)) # fit with box.cox lrt.bctrans(ans,lrt=list(c(0,-1,1,0))) plot(ans,family="power") # plot, but use ordinary powers b <- cbind(b,powtran(ans)) # add transformed variables to data frame