BayesLogit: Logistic Regression
The BayesLogit package does posterior simulation for binomial and
multinomial logistic regression using the Polya-Gamma latent variable
technique. This method is fully automatic, exact, and fast. A routine to
efficiently sample from the Polya-Gamma class of distributions is included.
Version: |
0.6 |
Depends: |
R (≥ 2.14.0) |
Published: |
2016-10-20 |
Author: |
Nicholas G. Polson, James G. Scott, and Jesse Windle |
Maintainer: |
Jesse Windle <jesse.windle at gmail.com> |
License: |
GPL (≥ 3) |
NeedsCompilation: |
yes |
Citation: |
BayesLogit citation info |
CRAN checks: |
BayesLogit results |
Downloads:
Reverse dependencies:
Linking:
Please use the canonical form
https://CRAN.R-project.org/package=BayesLogit
to link to this page.