R: Preprocessing Algorithm for Quantile Regression
rq.fit.pfn
R Documentation
Preprocessing Algorithm for Quantile Regression
Description
A preprocessing algorithm for the Frisch Newton algorithm
for quantile regression. This is one possible method for rq().
Usage
rq.fit.pfn(x, y, tau=0.5, Mm.factor=0.8, max.bad.fixup=3, eps=1e-06)
Arguments
x
design matrix usually supplied via rq()
y
response vector usually supplied via rq()
tau
quantile of interest
Mm.factor
constant to determine sub sample size m
max.bad.fixup
number of allowed mispredicted signs of residuals
eps
convergence tolerance
Details
Preprocessing algorithm to reduce the effective sample size for QR
problems with (plausibly) iid samples. The preprocessing relies
on subsampling of the original data, so situations in which the
observations are not plausibly iid, are likely to cause problems.
The tolerance eps may be relaxed somewhat.
Value
Returns an object of type rq
Author(s)
Roger Koenker <rkoenker@uiuc.edu>
References
Portnoy and Koenker, Statistical Science, (1997) 279-300