R help - Aug 2008 - Unable to send color palette through plot.Design to method="image"

I have been trying to specify a different color palette to the image  
method in plot.Design. My model has crossed two rcs() arguments and  
one two-level gender argument. The goal which appears to have been  
mostly achieved is to produce separate bivariate plots for men and women

The call to plot does produce a level plot but it appears only with  
the default color palette despite various efforts to supply a col  
argument. I even took a crack at hacking the plot.Design function,   
adding a col=col parameter to be passed in the function call to  
image(), but failed to get the desired effect:
  # else image(xseqn, y, zmat, xlab = xlab, ylab = laby , col = col)

library(Hmisc); library(Design)
lr.fit6 <- lrm(death ~  
rcs(BL_CHOLEST.A,c(180,220,280))*rcs(BL_HDL.A,c(40,55,70))*Sex, data =  
pref900)

#
str(pref900[,c("BL_HDL.A","BL_CHOLEST.A","death")])
$'data.frame':	910659 obs. of  3 variables:
  $ BL_HDL.A    : num  34 35 40 46 39 45 46 34 42 52 ...
  $ BL_CHOLEST.A: num  181 184 238 134 180 220 125 256 265 175 ...
  $ death       : logi  FALSE FALSE FALSE FALSE FALSE FALSE ...

plot(lr.fit6, BL_HDL.A=seq(25,125, length=50),BL_CHOLEST.A=seq(100,  
350, length=70), Sex="Male", col = rainbow(15),
method="image")

Respectfully;
David Winsemius, MD, MPH
Heritage Laboratories,
West Hartford, CT, USA

David Winsemius wrote:
> 
> I have been trying to specify a different color palette to the image 
> method in plot.Design. My model has crossed two rcs() arguments and one 
> two-level gender argument. The goal which appears to have been mostly 
> achieved is to produce separate bivariate plots for men and women
> 
> The call to plot does produce a level plot but it appears only with the 
> default color palette despite various efforts to supply a col argument. 
> I even took a crack at hacking the plot.Design function,  adding a 
> col=col parameter to be passed in the function call to image(), but 
> failed to get the desired effect:
I improved the source code to make use of col for 3-d plot types of 
image, contour, and persp.  I don't know why your hack didn't work.  The
new changes work for me, and the updates will be in the next release of 
Design to CRAN.  I will e-mail you a copy of the source code to use in 
the meantime.

Frank
>  # else image(xseqn, y, zmat, xlab = xlab, ylab = laby , col = col)
> 
> library(Hmisc); library(Design)
> lr.fit6 <- lrm(death ~ 
> rcs(BL_CHOLEST.A,c(180,220,280))*rcs(BL_HDL.A,c(40,55,70))*Sex, data = 
> pref900)
> 
> #
str(pref900[,c("BL_HDL.A","BL_CHOLEST.A","death")])
> $'data.frame':    910659 obs. of  3 variables:
>  $ BL_HDL.A    : num  34 35 40 46 39 45 46 34 42 52 ...
>  $ BL_CHOLEST.A: num  181 184 238 134 180 220 125 256 265 175 ...
>  $ death       : logi  FALSE FALSE FALSE FALSE FALSE FALSE ...
> 
> plot(lr.fit6, BL_HDL.A=seq(25,125, length=50),BL_CHOLEST.A=seq(100, 350, 
> length=70), Sex="Male", col = rainbow(15),
method="image")
> 
> Respectfully;
> David Winsemius, MD, MPH
> Heritage Laboratories,
> West Hartford, CT, USA
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide 
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
> 

-- 
Frank E Harrell Jr   Professor and Chair           School of Medicine
                      Department of Biostatistics   Vanderbilt University