BayesSUR 2.3-0
- Make arma::invas an alternative toarma::inv_sympdin the case of a matrix is singular or not
positive definite
- Fixed a bug about HRR model introduced in v2.2-0 by bringing back
the control of a_omegaandb_omegain functionBayesSUR()
- Fixed a bug in function SUR_Chain::stepWGibbs()for
extracting a submatrix
- Simplified the calculation in function
SUR_Chain::logPGamma
BayesSUR 2.2-1
- Fixed a bug about SUR_Chain.cppintroduced in version
2.2-0
- Fixed argument mrfG index issue in function
BayesSUR()
BayesSUR 2.2-0
- Fixed a bug about the use of temperature parameter in
HRR_Chain.cpp, also minor update inSUR_Chain.cpp
- Fixed random effects’ sampler HRR_Chain::stepW0()
- Added argument beta.typein functionplotEstimator()to plot MPM coefficient estimates
- Fixed argument mrfGindex issue in functionBayesSUR()
BayesSUR 2.1-7
- Increased the threshold of predictor dimension to 100000 (previously
5000) when pre-computing XtX
- Minor changes of the parameter updates of the HRR models, especially
in function step()
BayesSUR 2.1-6
- Fixed issues of knitr VignetteBuilder and README
BayesSUR 2.1-5
- Updated citation, added README and one more vignette
BayesSUR 2.1-4
- Fixed a documentation issue by adding
@alias BayesSUR-package
BayesSUR 2.1-3
- Updated C++11 specification to current default of C++17
- Cleaned R scripts
BayesSUR 2.1-2
- Fixed gcc-UBSAN error reference binding to null pointerin SUR_Chain.cpp:3818
BayesSUR 2.1-1
- Fixed R session aborted fatal error
BayesSUR 2.1-0
- Fixed an issue with omp_set_num_threadsthat did not
work
BayesSUR 2.0-1
BayesSUR 2.0-0
- Major update, introducing random effects for mandatory variables
without variable selection.