This directory contains S/Splus software to
implement relative distribution methods.
These methods are described in the book
Relative Distribution Methods in the Social Sciences, by
Mark S. Handcock and Martina Morris,
Springer-Verlag, Inc., New York, 1999
ISBN 0387987789 .
The software can be used freely for non-commercial
purposes. You may modify and distribute the code for
NON-commercial purposes, as long as this
statement
and the above contact information is included.
While this software has now been developed and used for
several years by the authors and others, a new user may well
experience some problems or bugs. Please report these by email
(to either
handcock@stat.ucla.edu
or
morrism@u.washington.edu),
and we will work with you to resolve the problem.
To use the files:
Double click on the files to save it into the UNIX or PC sub-directory in which you use Splus.
Start Splus, and source the files e.g.: % Splus
> source("reldistplus.s")
or, if you downland the data as well:
> source("install.RDB")
The software:
General relative distribution software
- reldist: Software
to estimate and graph the relative density, CDF and
related functions. Also functions to estimate summary
measures and their uncertainties. This contains
only functions directly related to the relative
distribution.
of "reldist" slightly modified to work in R.
- reldistaux:
Auxillary programs required to reconstruct the
figures and numbers in the book. These programs
estimate and graph densities, and related summary
measures. This requires the above reldist
software to be effective.
Documentation
- Short documentation and installation instructions
- Longer documentation for the general software (in PDF form)
-
install.RDB: S/Splus code to install
both the software and data (i.e.
reldist,
reldistaux,
and all the
data
files) once they have been downloaded. Remember to
unzip the data files (if it is not done automatically
by your browser). This program is not necessary,
but helpful.
Software to create the figures and tables in
the book. Remember to unzip the archive for each
chapter (if it is not done automatically by your
browser). Figure X.Y is created by file "FigRDBX.Y".
- Figures from Chapter 2
- Figures from Chapter 3
- Figures from Chapter 4
- Figures from Chapter 6
- Figures from Chapter 8
- Figures from Chapter 9
- Figures from Chapter 11
- Figures from Chapter 12
- Figures from Chapter 13
Some of the figures require the use of ancillary
Splus libraries that are linked here:
- logspline
Only necessary if logspline density estimation option is chosen.
Used for Figure 2.1, 9.3, 13.1 and 13.2 in the book.
- S functions to, fully and automatically, estimate an unknown
density based on possibly grouped or censored data and
obtain corresponding probabilities, quantiles and
random samples. Charles Kooperberg (clk@fhcrc.org).
[24/Mar/97] (182 kbytes)
- Modapplstat
Only necessary if the Sheather-Jones bandwidth for
univariate density estimation is chosen.
Used for Figures 2.1 and 9.2 in the book.
- Software, scripts, and data from the book
Modern Applied Statistics with S-Plus by W.N. Venables & B.D. Ripley.
There are unix archives and a Windows version. This material is not
available via e-mail. Mirrored from Brian Ripley's collection
at Oxford. (ripley@stats.ox.ac.uk).
- quantreg
Only necessary if quantile regression is used.
Used for Figures 13.3 and 13.4 in the book.
- Routines to compute quantile regressions including
analogues of trimmed means for the linear model.
Updated version contains new functions
implementing new forms of rank tests for linear models
based on the dual quantile regression process.
(roger@ysidro.econ.uiuc.edu) [1/Jun/95](48 kbytes)
- nlsd
Only necessary if logspline density estimation with smoothing
parameter selection is chosen.
Used only for Figure 9.3 in the book.
- A version of the "logspline" S functions that
incorporated case weights. Written by Charles
Kooperberg (clk@fhcrc.org) and based on ideas
in Section 3 and 4 of "The use of polynomial
splines and their tensor products in extended
linear modeling," by Charles J. Stone, Mark Hansen,
Charles Kooperberg, and Young K. Truong, Annals of
Statistics, 25 (1997), 1371-1470. The code was
adapted from "Logspline density estimation under
censoring and truncation," by Ja-Yong Koo, Charles
Kooperberg and Jinho Park, Scandinavian Journal of
Statistics, 26 (1999), no. 1, 87-105.
For additional information see http://lynx.fhcrc.org/~clk.
Description of data sources, and further information about the data sets,
can be found in the book.