[R] truncated regressor

[Prof Brian D Ripley]

On Sun, 14 May 2000, Troels Ring wrote:

> Dear friends. I have the problem of assessing the importance of bleeding 
> times censored at 20 minutes for predicting blood loss incurred after a 
> liver biopsy. How would I use the knowledge that the censored values were 
> 20 minutes or more ?

This is a (partially) missing-value problem.  Are values > 20 mins 
always censored?  If so I would group the times and use an ordered factor
as a regressor.  If you really want a linear effect on, say `bt' you can
use

... + bt + I(bt==20) + ...

which allows a separate value of bt==20 over and above a linear model.

If your times are not all censored, you could build a model for the
missing data and use multiple imputation.  If you need, that, more
details of your problem, please.

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._