1168 Blackfield Way
Mountain View, CA 94040

3 March 1992


These scientific subroutines are a part of my personal lisp library.
Although they were written to fulfill my needs at the time, they were
written to be general so I or somebody else could use them for other
tasks later.  This grab-bag of routines has served me well, and you
should find them useful.

They are not, however, polished routines.  You will see pieces of
test code here and there, and many pages of code are really notes
about what didn't work or what needs to be done.

The routines have also suffered as several versions of lisp have come
and gone.  Many of the routines were written a long time ago for
MACLisp on a PDP-10.  Then came ZetaLisp, and now there is Common
Lisp.  The routines are mostly Common Lisp, but my Lisp Machine has
many backward compatibility features, so some minor changes may be
required.

There is, of course, no warranty about the fitness or the accuracy
of these routines.  Although I have tried to be careful, I have
fixed too many routines that I once thought were right.

Except for the routines below, you may use these routines in your own
work and pass them on to your colleagues.  You may incorporate
individual modules into an application that you sell.  You may
distribute the source (and even charge a nominal fee) as long
as the copyright notices are not removed.

I cannot give away some of these routines because they are not
mine, but rather transliterations of someone else's Fortran or C code
to lisp.  In particular, some or all the routines in

	regress.lisp    come from Bevington

	fmfp.lisp       come from Kuester & Mize
	marq.lisp        

	beta.lisp	come from Press
	gamma.lisp
				
These authors published their code, and I expect they meant that
people should use it.  If you own a copy of the book, then there
should be no problem with you using the code.

You can ftp a copy of these routines (pub/clmath.tar) from
the Internet host ftp.ai.mit.edu.

If you discover bugs, please let me know so the fixes can
be made to the source.

Sincerely,
Gerald Roylance

---------

Philip R. Bevington,
  Data Reduction and Error Analysis for the Physical Sciences,
  McGraw-Hill, New York, 1969.

James Kuester and Joe Mize,
  Optimization Techniques with Fortran,
  McGraw-Hill, New York, 1973.

William H. Press, Brian P. Flannery, Saul A. Teukolsky, William T. Vetterling
  Numerical Recipes in C, The Art of Scientific Computing
  Cambridge University Press, 1988
