[R] Difference MNP-package and rmnpGibbs from bayesm-package

[A. Martinovici]

Hi,

The first thing to check is if both functions use the same 'base' option - they usually don't.
On a more general note, the two approaches use a different data augmentation procedure, but this should not have a very large impact on the results provided enough draws and similar priors are used. If you want to go more into detail on what are the exact differences, the Imai and van Dyk (2005) paper is very useful.

Best regards,
Ana

	[[alternative HTML version deleted]]