Elliptical

Copyright elliptical © 2005 Francisco José A. Cysneiros

A library of  S-PLUS  and R functions (and data) for analysis in symmetrical linear and nonlinear regression models with unrestricted and restricted in the parameters by Francisco José A. Cysneiros  

The S-PLUS library Elliptical. They can be obtained from here. Instructions how to install the library are given here.

elliptical v.2.0

Univariate elliptical inference for linear and nonlinear regression models.
 

Details of the theorical are given in:

Cysneiros, F.J.A. and (2004). Métodos Restritos e Validação de Modelos Simétricos de Regressão. (pdf) P.h.d. Thesis, IME-University of Sao Paulo.

Cysneiros, F.J.A. and Paula, G.A. (2005). Restricted methods in symmetrical linear regression models. (pdf) Computational Statistics and Data Analysis.

The library also contains some data sets. All of the major functions and the data sets are fully documented with help files. The reference below should be cited, when the library elliptical had been used.

Cysneiros, F.J.A. and Paula, G. A. (2003). One-Sided Tests in Univariate Elliptical Linear in Elliptical Linear Regression Models. In: Proceedings of the 18th International Workshop on Statistical Modelling, Verbeke,G., Molenberghs,G., Aerts, A., and Fieuws, S. (Eds). Leuven: Katholieke Universiteit Leuven, pp.103-108.

Where to download the library

Please click on the appropriate link for the section and version you require.

  S  elliptical   written in S-PLUS 2000  professional Release 3

  S  elliptical   written in S-PLUS 6.x 
 

  R elliptical   written in R 2.0.1

How to install the library

The single library sections are zip files. Expand the archives and follow then the instructions in the INSTALL file to complete the installation of the library. If all library sections are to be installed, it is best to create a main directory elliptical which will contain the library sections. Copy the Zip files to the corresponding directories before unpacking them.

You should then follow the instructions above for each library section. The README file for the whole library. Save it in the main directory ELLIPTICAL. Once this is done, and once inside S-PLUS or R in your usual working directory, the functions and data are accessed by typing  command “source”

Examples

         Two examples are used to explain library and can be founded here. The first example consist in linear  regression model with Student-t errors. In the literature, this example was very studied for example Lange et al. (1989). The fitted model was given by

Ammonia_i=alpha*intercept +beta*air_i + nu*water_i + theta*acid_i + e_i   i=1,...,21

e_i ~ t(0,Ф,4)

        The second example,  is founded in Ratkowsky(1983) where was propused a  nonlinear regression model is given by

Log(Y_i) = -Log(alpha + beta*X_i+gamma*X²_i) +  e_i                 i=1,...,42

e_i ~ N(0,Ф)    

Browlee, K.A (1965) Statistical Theory and Methodology in Science and Engineering    (2nd ed). New York: John Wiley.

Lange K.L, Little, J.A. and Taylor, J.M.G. (1989) Robust Statistical Modeling Using the  t Distribution. JASA, Vol 84, 881-896.

 Ratkowsky, D.A. (1983) Nonlinear regression modeling, a unifield practical approach. New York: Marcel Dekker.

This library is free software for academic use; you can redistribute it and/or modify it under permissions of the authors and the terms of the GNU General Public License as published by the Free Software Foundation, either version 2 of the License, or (at your option) any later version. This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

Please send any comments, suggestions or errors reported to F.J.A. Cysneiros.

 DE        UFPE

      Last updated on February 20, 2005