SAS files

Chapter 5 Regression Computations

Filename

Description

chex51.sas

Checks Example 5.1, showing the condition of a simple regression problem for changes in the response variable.

chex52.sas

Checks Example 5.2, showing the condition of a simple regression problem for changes in the explanatory variable.

brown.sas

Computes regression statistics for example from Brownlee, p 463, Example 13.1, which is used in demonstrations flyreg and sweep. Uses data in file brown463.dat.

 

 

Chapter 9 Maximum Likelihood and Nonlinear Regression

Filename

Description

conclk.sas

Computes results from simple nonlinear regression Example 9.6 (Bates & Watts, BOD 1) to be compared with demonstration conclk.

chex97.sas

Computes results for logistic regression in Example 9.7 (Cox) to be compared with results in the demonstration chex97.

chex98.sas

Computes results for logistic regression in Example 9.8 (Finney) to be compared with results in the demonstration chex98. Uses file finney.dat.

chex99.sas

Computes results for Poisson regression in Example 9.9 (Frome) to be compared with results in the demonstration chex99. Uses file frome.dat.

nllsqt0.sas

Computes results from polynomial regression test problem (Exercise 5.5) to be compared with results from demonstration nllsqt0.

nllsqt1.sas

Computes results from nonlinear regression on first test problem (Fuller) to be compared with the results from demonstration nllsqt1. Uses file fulnls.dat.

nllsqt2.sas

Computes results from nonlinear regression on second test problem (Gallant) to be compared with the results from demonstration nllsqt2. Uses file gallant.dat.

 

 

Chapter 13 Markov Chain Monte Carlo Methods

Filename

Description

chex131.sas

Mimics the computations in the demonstration chex131 (MCMC on variance components problem, Example 13.1), but does not reproduce the same output as the Fortran code since SAS uses different algorithms for generating normal and gamma variates.

chex133.sas

Mimics the computations in the demonstration chex133 (Metropolis within Gibbs on Pump Example 13.3), but does not reproduce the same output as the Fortran code since SAS uses different algorithms for generating gamma variates.

mh1.sas

Reproduces results from demonstration mh1 (Metropolis-Hastings on log-series posterior, Example 13.5 ), including means, autocorrelation, as well as batching and spectral density variance estimates.

mh2.sas

Reproduces results from demonstration mh2 (Metropolis-Hastings on log-series posterior, Example 13.5), including means, variances and spectral density variance estimates for doing the Geweke test.

mh3.sas

Reproduces results from demonstration mh3 (Metropolis-Hastings on log-series posterior, Example 13.5), including batch means and spectral density estimates, and Gelman-Rubin test results.