Instructions for installation (Unix)
------------------------------------

This is the S-plus library `sm' for nonparametric 
estimation and inference using kernel smoothing methods,
and related scripts.

Please read the file `Licence' before proceeding,

Execute `SMunpack' to expand the *.shar files and then execute `Ins'.
(If you don't have a Fortran compiler, execute `Ins nofort'.) 
The `Example' section below shows steps in more detail. 
Optional step: follow the instructions of the `Colours' section.


System requirements:
--------------------
S-Plus 3.3 or later version (S-Plus 3.2 works in most cases)
Fortran-77 compiler (not for the no-Fortran version)


Example:
-------
<as S-Plus owner>
cd `Splus SHOME`/library
mkdir sm
cd sm
uncompress -c <somewhere>/sm.tar.Z | tar xvf -
SMunpack
Ins
rm routines.f *.q *.shar SMunpack Ins

You can also install as a private library as any user, when sm must be
invoked by

library(sm, lib.loc="<path_to_library">)


Scripts
-------
There is a set of scripts available to run the examples discussed in
the book by Bowman & Azzalini (1997).  A complete list of the scripts
can be obtained by typing
	sm.script()
in S-Plus, after the sm library has been installed.  Individual
scripts can be executed by typing
	sm.script(scriptname)
The scripts are text files which can be found in the directory scripts
under the sm library directory.


Colours:
--------
If you use a colour monitor and a colour device (e.g. motif),
the effectiveness of some of the pictures may be affected by the
colour settings.  The following settings were used during much of the
development of the sm library and the associated scripts.

  background          : white
  lines,text,polygons : black red green blue yellow cyan magenta
  images              : red 20 cyan

Running motif(xcm.SM.motif) or openlook(xcm.SM.openlook) will set this
colour scheme.

Updated: 1 Sept 1997
