Note that the early examples (pages 82 to 93) are actually quite complicated and are designed to show what you can do with various structural time series models. The simpler ones start with harveyp218.rpf and generally work up to the level of that first set.
Harvey, Forecasting, structural time series,...
Harvey, Forecasting, structural time series,...
The attached zip has the worked examples from Harvey's Forecasting, structural time series and the Kalman filter, Cambridge Univ. Press, 1989. Most of these are applications of what Harvey calls structural time series models, which additively decompose a series into fundamental components like the trend and seasonal, using state space representations. With only a few exceptions, these use the DLM instruction. There's some overlap with the Durbin and Koopman book. Some applications which are unique to this are the three examples for dynamic models for count data, and harveyp447, which does a three variable state space model with seasonal with "homogeneous" dynamics.
Note that the early examples (pages 82 to 93) are actually quite complicated and are designed to show what you can do with various structural time series models. The simpler ones start with harveyp218.rpf and generally work up to the level of that first set.
Note that the early examples (pages 82 to 93) are actually quite complicated and are designed to show what you can do with various structural time series models. The simpler ones start with harveyp218.rpf and generally work up to the level of that first set.
Last bumped by TomDoan on Thu Jun 28, 2012 10:51 am.