#This is an archive of software and test problems for the paper
An Interior Point Algorithm for Nonlinear Quantile Regression
by Roger Koenker and BJ Park
This version was prepared March 31 1994 in Prague


Problems and Questions regarding this software should be directed to
	roger@ysidro.econ.uiuc.edu

The archive is in tar.Z format and can be obtained by clicking
<a href="nlrq.tar.Z">here<\a><br>
This software may be redistributed for any non-commercial purpose.

The archive contains the following files

data.s		S+ files which should be installed with source("data.s")
		These files contain the S+ source for the function
		nlrq() which contains the algorithm, all lower level
		functions it calls, and the problem specific code for
		all of the test problems used in the paper.
nlrq.d		S+ documentation file for the function nlrq, which
		should be installed in some appropriate .Help directory
		i.e. cp nlrq.d .Data/.Help/nlrq
test.rq.1.s	An S+ source file which runs the complete suite of test
		problems using the k=1 setting (Meketon steps per
		iteration =1) invoking this in S+ (source("test.rq.1.s")
		produces a file test.rq.1.oo which may be compared with
		the .o files below)
test.rq.2.s	An S+ source file which runs the complete suite of test
		problems using the k=2 setting (Meketon steps per
		iteration =2) invoking this in S+ (source("test.rq.2.s")
		produces a file test.rq.2.oo which may be compared with
		the .o files below)
test.rq.1.o	An S+ source output file for the complete suite of test
		problems using the k=1 setting (Meketon steps per
		iteration =1)
test.rq.2.o	An S+ source output file for the complete suite of test
		problems using the k=2 setting (Meketon steps per
		iteration =2)
text		directory containing postscript files of various pieces
		of the paper see readme file there for details.