Bayesian and Likelihood Analysis of Dynamic Linear Models
Provides routines for Maximum likelihood,
Kalman filtering and smoothing, and Bayesian
analysis of Normal linear State Space models, also known as
Dynamic Linear Models.
News
Changes to package dlm
Version 1.1-0, 2010-02-15
- In functions dlmLL and dlmFilter, whenever the observation
variance V is found to be numerically singular, an 'epsilon'
is added to it (with a warning) in order to make the SVD-based
algorithms work.