How to change baseline patient in Predict function in rms package in R? I am doing a time dependent Cox model using cph function in rms package. I use Predict and plot.Predict to plot the hazard ratio on y axis and a continuous covariate (e.g. LDL cholesterol) on X axis for 3 levels of a treatment. I get 3 curves for 3 treatment across the range of my continous covariate LDL. I use the R code below: 
dd<- with (mydata, datadist( age, sex, LDL, Treatment))
options(datadist='dd')
S<-Surv(mydata$tstart, mydata$tstop, mydata$followUpTime)
fit <- cph( S ~ rcs(age,3) + sex + rcs(LDL,4) + Treatment, x=T, y=T, data=mydata)
plot(Predict(fit, LDL, Treatment, fun='exp'))

The hazard ratio I get on Y axis comes from applying "exp" on "log relative hazard" or Xbeta produced by Predict function. This hazard ratio is the hazard relative to the "baseline" patient - a theoretical patient with all covariates set to 0. 
My problem is that I want to make a plot where I compare hazard to a pre-defined patient (e.g. 50 year-old, Male, LDL=80) and not the baseline. I mean I simply want to replace the theoretical baseline patient (all covariates=0) with a patient I define.
I would appreciate any help.
Thanks
 A: Type ?contrast.rms and see the examples that involve plotting.
A: Thank you very much Dr. Harrell! I am wondering if the following code would be the answer to my question above? I would appreciate a confirmation from you or other experts:
library(rms)
dd <- datadist(mydata)
options(datadist='dd')
S <- with(mydata, Surv(start, tstop, followUpTime))
fit <- cph( S ~ rcs(age,3) + sex + rcs(LDL,4) + Treatment, x=T, y=T, data=mydata)

# A sequence of LDLs across LDL range in data by 5 units 
LDLs <- seq(10, 300, by=5) 
#Reference is a patient with Treatment=1 and LDL=80, and the rest of covariates at  
# their medians or default references for categorical variables as performed in cph fit 
# (e.g. age= median of age [50 years], sex= default ref. [Male]).
w1 <- contrast( fit, list(Treatment=1, LDL=LDLs), list(Treatment=1, LDL=80) )
w2 <- contrast( fit, list(Treatment=2, LDL=LDLs), list(Treatment=1, LDL=80))
w3 <- contrast( fit, list(Treatment=3, LDL=LDLs), list(Treatment=1, LDL=80))

#Plot separately for each treatment. (used exp() to get HR=Hazard Ratio) :
xYplot(Cbind(exp(Contrast), exp(Lower), exp(Upper)) ~ LDLs, data=w1, ylab='HR, Treatment1 – Reference Patient')
xYplot(Cbind(exp(Contrast), exp(Lower), exp(Upper)) ~ LDLs, data=w2, ylab='HR, Treatment2 – Reference Patient')
xYplot(Cbind(exp(Contrast), exp(Lower), exp(Upper)) ~ LDLs, data=w3, ylab='HR, Treatment3 – Reference Patient')

#To make a plot with all 3 treatment curves, I made a data-frame by extracting
# Contrast, Upper and Lower using w1$Contrast, w2$Contrast etc. and using rbind. 
# Then I could use ggplot or matplot to get a plot all in one. Not shown here.

#--------------A second shorter solution? ----------------
w.all <- contrast( fit, list(Treatment=c(1,2,3), LDL=LDLs), list(Treatment=1, LDL=80), conf.type='simultaneous' )
xYplot(Cbind(exp(Contrast), exp(Lower), exp(Upper)) ~ LDLs, data=w.all, ylab='HR, Treatment – Reference Patient')
#This one gave a very similar plot as above but one needs to find a way to label and color by Treatment….

