Reproducing "An Introduction to State Space Time Series Analysis" using Stan

• Original Author: github.com/sinhrks
• Extended/Modified: github.com/dantonnoriega

Trying to reproduce the examples introduced in "An Introduction to State Space Time Series Analysis" using Stan.

Data

SOURCE: http://www.ssfpack.com/CKbook.html

• data/
  • logUKpetrolprice.txt
  • NorwayFinland.txt
  • UKdriversKSI.txt
  • UKinflation.txt
  • UKfrontrearseatKSI.txt

R scripts

All the R scripts are in R/.

All the original files were edited to be more concise. In many cases, I switch from using ggplot2 to base R graphics.

Each file is tied to a headline figure. Most files will also carry forward and plot any proceeding plots between one headline figure and the next e.g. fig04_02.R creates figures 4.2 - 4.5, up to the next headline file fig04_06.R. This is not always the case, particularly in Chapter 8 where consecutive plots jump between different datasets. Instead, in Chapter 8, I sort of pick and choose the figures where I thought there was the most to be learned from attempted replication.

Figures that I added are labeled with "*(NEW)**" in the list below.

I'll add more over time. Feel free to contribute as well.

1. Introduction
   • fig01_01.R: Linear regression
2. The local level model
   • fig02_01.R: Deterministic level
   • fig02_03.R: Stochastic level
   • fig02_05.R: The local level model and Norwegian fatalities
3. The local linear trend model
   • fig03_01.R: Stochastic level and slope
   • fig03_04.R: Stochastic level and deterministic slope
   • fig03_05.R: The local linear trend model and Finnish fatalities
4. The local level model with seasonal
   • fig04_02.R: Deterministic level and seasonal
   • fig04_06.R: Stochastic level and seasonal
   • fig04_10.R: The local level and seasonal model and UK inflation
5. The local level model with explanatory variable
   • fig05_01.R: Deterministic level and explanatory variable
   • fig05_04.R: Stochastic level and explanatory variable
6. The local level model with intervention variable
   • fig06_01.R: Deterministic level and intervention variable
   • fig06_04.R: Stochastic level and intervention variable
7. The UK seat belt and inflation models
   • fig07_01.R: Deterministic level and seasonal
   • fig07_02.R: Stochastic level and seasonal
   • fig07_05.R: Stochastic level and deterministic seasonal (NEW)
   • fig07_07.R: The UK inflation model
8. General treatment of univariate state space models
   • fig08_02.R: Confidence Intervals (NEW)
   • fig08_05.R: Filtering and Prediction (NEW)
   • fig08_08.R: Standardised one-step prediction errors of model (NEW)
   • fig08_14.R: Filtered trend, and five-year forecasts for Finnish fatalities (NEW)
9. Multivariate time series analysis
10. State space and Box–Jenkins methods for time series analysis

IMPORTANT Some models output different results from textbook and R's {dlm} package.