[R] weights in quantile regression
Chris Wilcox
c.wilcox at uq.edu.au
Thu Jul 18 20:54:38 CEST 2002
I am just starting to use R, converting over from SAS. I want to estimate
the upper quartile in a scatterplot of demographic rates against rainfall in
a small mammal population. The data looks like this:

* * < point in question
*
* *
* *
* * *
* * * * * *
_______________________________________
Based on a paper by Scharf et al. in Ecology (vol 79, p448) I am trying to
use the qreg package in R to estimate the upper quartile in the data. My
problem is that I have relatively few data points (~40) which are mostly
distributed at the low end of the x axis, and an outlier in the y direction
at the high end of the x axis.
I would like to find a way either by iteratively weighting the observations
or by bootstrapping the regression to quantify the relationship in the mass
of the data. I looked in the manual for qreg, and while it allows for
weights, there is no guidance for routines to calculate them. I also
checked the help archives and on the web, and while I found a bit on the
web, I didnt see much that was specific to my problem. It seems like this
would be a general issue when working with naturally occurring phenomena
(like rainfall) as there will often be few data points at the extremes.
Any suggestions would be most welcome.
Thanks,
Chris
.......................................
rhelp mailing list  Read http://www.ci.tuwien.ac.at/~hornik/R/RFAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: rhelprequest at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
More information about the Rhelp
mailing list