Software and Datasets
R packages
-
BMA:Bayesian model averaging (BMA) for linear regression, generalized linear
models and survival analysis
-
ensembleBMA: Probabilistic forecasting using Bayesian Model Averaging
-
ProbForecastGOP: Probabilistic Weather Field Forecast using the GOP method
-
covRobust: Robust covariance matrix estimation using nearest neighbor cleaning.
-
mcgibbsit: Warnes and Raftery's MCGibbsit MCMC diagnostic for multiple chains.
By G. Warnes, 2004.
Bioconductor R packages
For the Bioconductor packages here,
go to the Bioconductor home page and
click on "Developmental Packages".
- spotSegmentation: Spot segmentation in microarrays via model-based clustering
- rama: Robust analysis of microarrays (Bioconductor R package). By R. Gottardo, 2004.
- bridge: Bayesian robust inference for differential gene expression. By R. Gottardo, 2005.
S functions
- ace.logit.
S function to calculate and plot the ACE transformations for logistic regression
(adapted from linear regression).
Adrian E. Raftery, raftery@stat.washington.edu. [23/Jan/95] (4k)
- ar.robust.
S-function to compute robust estimates of partial autocorrelation for
AR processes (upto order 6).
This function can also be used to obtain posterior probabilities for
the competing AR models.
Nhu Le (nhu@epi.terryfox.ubc.ca). [1/Mar/95] (99 k)
- bicreg.
Bayesian model selection and accounting for model uncertainty in linear regression
models, using the BIC approximation.
Contributed by "Adrian E. Raftery" (raftery@stat.washington.edu)
[17/Oct/94][23/Jan/95] (24 kbytes).
- bic.logit.
S function to do
logistic regression with Bayesian model selection and accounting for model
uncertainty, using the BIC approximation.
Contributed by "Adrian E. Raftery"
(raftery@stat.washington.edu) [29/Nov/94] (30 kbytes)
- bic.surv
S-Plus function to implement Bayesian Model Averaging for Cox Proportional
Hazards Models, using the BIC Approximation. Includes a summary function.
Contributed by Chris T. Volinsky (volinsky@thelma.stat.washington.edu)
[1/Dec/95] (9 kbytes)
- bma.
Bayesian model averaging (BMA) to account for model uncertainty in linear
regression models.
Contributed by Jennifer Hoeting (jah@lamar.stat.colostate.edu)
[3/Apr/95] (24kbytes)
- fracdiff.
Maximum likelihood estimation of the parameters of a fractionally-differenced
ARIMA (p,d,q) model (Haslett and Raftery, Applied Statistics 1989). Fortran
source code in a shar archive. See also fracdiff in the general collection of StatLib.
Submitted by Chris Fraley, (fraley@stat.washington.edu) [24/Mar/91]. [12/Aug/91]
[9/Oct/91] [4/Nov/91] [13/Nov/91] [24/March/92] [15/Sept/92] [22/Feb/93]
[11/Mar/94] [4/May/94] (135 kbytes)
Note: This function is now available in Splus under the name
arima.fracdiff, and several other related functions are also included in
Splus (versions 3.1 and later).
- gibbsit.
Determines the number of iterations required to achieve a specified accuracy level
in Markov chain Monte Carlo. An S translation of the Fortran function
gibbsit.
Submitted by Steven Lewis (slewis@stat.washington.edu) [1/Mar/95]
[4/Oct/95](7 kbytes)
- glib.
Bayesian generalized linear modeling, model comparison and accounting for model
uncertainty. An S-PLUS function.
Adrian Raftery, (raftery@stat.washington.edu).
[17/May/93][18/May/93] (29 Kbytes).
- principal.curve.
Functions for fitting principal curves. Implements the bias correction method
of Banfield and Raftery (JASA, 1992).
Created by Trevor Hastie
(trevor@research.att.com) [7/Aug/93] (15 kbytes)
- wlb.
S code to perform a weighted likelihood bootstrap.
Michael Newton (newton@stat.wisc.edu) [17/Dec/92] (3k)
Fortran programs
- baysn.
Bayesian synthesis method for bowhead whale population dynamics model
(Raftery, Givens, Zeh 1995 JASA). Fortran. Also see bergs.
Submitted by Geof
Givens (geof@space.stat.ColoState.EDU) [1/Mar/95] (266 kbytes)
- bergs.
Reweighting sensitivity analysis for Bayesian synthesis bowhead whale pop.
assessment (Raftery, Givens, Zeh 1995 JASA). Fortran. See baysyn.
Submitted by
Geof Givens (geof@space.stat.ColoState.EDU) [1/Mar/95] (2.46 Mbytes)
- eha.
Discrete time event history analysis, written for World Fertility Survey data, but
nearly all of the subroutines are more generally applicable.
Steven M. Lewis (slewis@stat.washington.edu) [1/Mar/95] (347 k)
- fracdiff.
Maximum likelihood estimation of the parameters of a fractionally-differenced
ARIMA (p,d,q) model (Haslett and Raftery, Applied Statistics 1989). Fortran
source code in a shar archive.
See also fracdiff in the S collection of StatLib (see link above).
Submitted by Chris Fraley, (fraley@stat.washington.edu) [7/Oct/91] [24/Mar/91].
[4/Nov/91] [15/Sept/92] [22/Feb/93] [11/Mar/94] [4/May/94](220 kbytes)
- gibbsit.
Determines the number of iterations required to achieve a specified accuracy level
in the Gibbs sampler. A Fortran program.
(Adrian Raftery,
raftery@stat.washington.edu) [13/Sept/91] [1/Mar/95] (114 kbytes)
- laplacpg.
Fortran program used for Bayes factors in hierarchical models:
calculates the Laplace approximation to the marginal likelihood
for the Poisson-Gamma (negative binomial) regression model.
Susan Rosenkranz (slr@biostat.washington.edu) [29/Mar/94] (252 kbytes)
- mtd.
Higher order Markov chains: Estimating the MTD (Markov Transition
Distribution) model of Raftery (JRSS B, 1985) and
Raftery and Tavare (Applied Statistics, 1994) [calls NAG].
"Adrian E. Raftery"
(raftery@stat.washington.edu) [11/Mar/94][25/Jul/94] (32kbytes)
Data sets available
Click here
to return to Adrian Raftery's home page.
Click here
to return to the Statistics Dept. home page.
Click here
to return to the University of Washington home page.
Last updated May 4, 2005