Skip to content
Prev 173513 / 398506 Next

help with predict and plotting confidence intervals

On Mar 12, 2009, at 11:45 AM, David Winsemius wrote:

            
At this point I should not have accepted your starting point. A better  
starting point would be to use the wt.mod model:

pp <- predict(wt.mod, int = "p", newdata= list(logA=seq(0,6,  
length.out = 24)) )

# Followed by:

plot( seq(0,6, length.out = 24),  pp[ ,"fit"] )
  lines(seq(0,6, length.out = 24),  pp[ ,"lwr"], lty=2)
  lines(seq(0,6, length.out = 24),  pp[ ,"upr"], lty=2)
David Winsemius, MD
Heritage Laboratories
West Hartford, CT