plotXRinvreg {XReg} | R Documentation |
plotXRinvreg
Returns a plot of the inverse regression function, as a matrix plot. Rows of matrix are variables and columns correspond to miniumum terms.
This representation of the regression function probably only works well as a graphical display for up to about 5 unique variables and 5 terms.
plotXRinvreg(x,xrout,shading=F,quant=F,both=F,qlev=c(.2,.4,.6,.8), spec.ylim=F,mat.ylim=NULL,level.set=NA)
x |
matrix of predictors |
xrout |
output from XR regression function XReg |
shading |
if shading=T shade the directions of the rules in panel associated with sequence of rules x:f(x)>=q |
quant |
if quant=T, plot rules in terms of quantiles or ranked predictions. Note, models will not appear linear univarite. |
both |
if both=T, plot in linear form, but put quantiles on the top of plot |
qlev |
if both=T, these are the labels for the quantiles put on the top of the plot |
spec.ylim |
if spec.ylim=F, do not specify the limits of the y axis of the plot. Note, in this case the same variable different plots could have different axes. If spec.ylim=T, then the user must provide the limits y-axis for each of the variables. |
mat.ylim |
This gives the y-axis limits (min,max) for each of the variables, it is a 2xp matrix, where each column corresponds to a predictor variable. It is only used if spec.ylim=T. |
level.set |
If this is set to a numeric value, a highlighted line will be added to the plot to indicate associated rule |
Michael LeBlanc mleblanc@fhcrc.org
## # See XReg