I have submitted a revision of my quantile regression software for S(+)
to Statlib.  It may now be obtained by sending the message:

	send quantreg from S

to statlib@stat.cmu.edu.  The new version contains several new functions
which implement new forms of rank tests  for the linear model 
based on dual quantile regression process and are described in:

[1] Gutenbrunner, C. Jureckova, J. (1991). 
Regression quantile and regression rank score process in the 
linear model and derived statistics, Annals of Statistics, 20, 305-330.

[2] Gutenbrunner, C. Jureckova, J.  Koenker, R. and Portnoy, S. (1993). 
Tests of Linear Hypotheses  based on Regression Rank Scores,
Journal of Nonparametric Statistics,  2, 307-331.


The algorithm, like the original quantreg submission is based on a
modification of the Barrodale and Roberts l_1 regression algorithm
for details see:

[3] Koenker, R.W. and d'Orey (1994).  Remark on Alg. AS 229: Computing Dual
Regression Quantiles and Regression Rank Scores, Applied Statistics, 43, 410-414

Finally, inverting one of the  rank tests yields an interesting new way
to compute confidence intervals for quantile regression.  This is
described in:

[4] Koenker, R.W. (1994). Confidence Intervals for Regression Quantiles, in
P. Mandl and M. Huskova (eds.), Asymptotic Statistics, 349-359, Springer-Verlag,
New York.

The functions in quantreg do not incorporate recent work on computing
quantile smoothing splines or nonlinear quantile regression estimation,
but I hope statlib version of these functions will follow shortly.

Comments on any aspect of this submission are most welcome.


						Roger Koenker	
roger@ysidro.econ.uiuc.edu			Department of Economics
voice: 217-333-4558				University of Illinois
fax:   217-244-6678				Champaign, IL 61820