-------------------------------------------------------------------------
This directory contains source code and executables for linear
regression software developed by:

	M. G. Akritas	
	G. J. Babu
	M. A. Bershady
	T. Bird
	E. D. Feigelson	
	T. Isobe

as described in Akritas, Bershady and Bird "Linear Regressions for
Astronomical Data with Measurement Errors and Intrinsic Scatter,"
1995, submitted to ApJ, and references therein.

If you find problems or bugs with this code, please contact me
(Bershady), as listed below.
-------------------------------------------------------------------------
        Matthew A. Bershady
        Pennsylvania State University           voice: (814) 865-0150
        Dept. of Astronomy & Astrophysics       email: mab@astro.psu.edu
        525 Davey Lab                           fax:   (814) 863-3399
        University Park, PA 16802
---------------------------------------------------------------------------
Modification History
	Apr  2 1995	MAB	initial Release
	Oct 18 1995	MAB	modify for public distribution
---------------------------------------------------------------------------

Makefile		unix executable for compilation
Notes			software notes
README.install		instructions for installation <<< READ THIS NEXT!

bin/			executables - PUT THIS IN YOUR PATH!

preprint.ps		postscript of preprint where this methodology
			is developed: "Linear Regressions for Astronomical 
			Data with Measurement Errors and Intrinsic Scatter,"
			Akritas, Bershady and Bird 1995, submitted to ApJ

src/			source code

* Contents of src/	(this lists the contents of Notes)

bces_regress.f          linear regions OLS(Y|X), OLS(X|Y), bisector, and
			orthogonal for model with heteroscedastic errors
			in X and Y, and homescedastic intrinsic scatter. 
			Variances calculated analytically and with bootstrap 
			resampling.

wls_regress.f           weighted OLS for model with heteroscedastic errors
			and homescedastic intrinsic scatter in Y only.
			Variances calculated analytically and with bootstrap 
			resampling.

slopes.f                linear regressions OLS(Y|X), OLS(X|Y), bisector,
			orthogonal, reduced major axis, and mean-OLS 
			regressions. Variances calculated analytically and with 
			bootstrap and jacknife resampling.

sixlin.f		subroutine called by slopes and wls_regress to
			calulate linear regressions

bootsp.f		bootstrap subroutine for slopes.f
bootspbec.f		bootstrap subroutine for bces_regress.f
bootspe.f		bootstrap subroutine for wls_regress.f

ran3.f			random number generator for bootstrap subroutines
			from Numerical Recipes
