Why not just change the names of the variables before invoking the call to
rq, and its formula?
Because I won't be able to do longer label e.g. 'Population Density Per Square
Mile'.
I actually found the solution by accident---
x<-rnorm(50)
x1<-rnorm(50)
y<-x+x1+rnorm(50) out<-rq(y~x+x1, tau=1:9/
Hi Roger,
Maybe I'm missing a clue. Here's an example:
x<-rnorm(50)
x1<-rnorm(50)
y<-x+x1+rnorm(50)
out<-rq(y~x+x1, tau=1:9/10)
plot(summary(out))
plot.out<-plot(summary(out))
#I change the variable names
dimnames(plot.out)<-list(c("intercept", "sex", "inc"))
Why not just change the names
Hi Roger,
Maybe I'm missing a clue. Here's an example:
x<-rnorm(50)
x1<-rnorm(50)
y<-x+x1+rnorm(50)
out<-rq(y~x+x1, tau=1:9/10)
plot(summary(out))
plot.out<-plot(summary(out))
#I change the variable names
dimnames(plot.out)<-list(c("intercept", "sex", "inc"))
#Look at the plot
#it's ONE si
It should be possible to say plot(..., xlab = "foo", ylab = "bar") )
if not please let me know.
Roger Koenker
rkoen...@illinois.edu
On Aug 22, 2011, at 9:00 PM, Kitty Lee wrote:
I have spent hours on this ---looked through the quantreg manual and
r-help site--- still couldn't figure ou
I have spent hours on this ---looked through the quantreg manual and r-help
site--- still couldn't figure out the answer. Can someone please help me on
this?
I plot the result from quantile regression and want to change the variable
labels:
temp<-rq(dep~inc+age50, data=newdata, tau=1:9/10)
5 matches
Mail list logo
