Miguel Juárez and Mark Steel, "Non-Gaussian Dynamic Bayesian Modelling for
Panel Data", Journal of Applied Econometrics, Vol. 25, No. 7, 2010, pp.
1128-1154.
The three data sets in matlab-data.zip are MATLAB mat files. The same data
are also provided as ASCII files (in DOS format) in the file ascii-data.zip.
Unix/Linux users should use "unzip -a" for this zip file only.
In all cases, the data are ordered such that the first T_1 observations
correspond to the first time period, etc. The data on GDP growth pertain to
29 countries, of which we select 25 in the paper (compare the `OECD_names'
file with Appendix B of the paper).
Hirano.mat was kindly provided by K. Hirano, from Hirano (2002),
Econometrica, 70, 781-799. These data are also contained in the files:
Hirano_cg.txt
Hirano_hs.txt
Hirano_hsd.txt
Hirano_tot.txt
OECD.mat can be downloaded from the Penn World Tables, V.6.1. These data are
also contained in the files:
OECD_growth.txt
OECD_open.txt
OECD_gdp.txt
OECD_names.txt
OECD_save.txt
Tiao.mat is taken from Liu, M.C. and Tiao. G.C. (1980), Journal of
Econometrics, 13, 305-325. These data are also contained in the file
tiao.txt.
All code below is for use in a MATLAB environment:
Code and readme files for the pooled Skew-t model are in PSkt.zip
Code and readme files for the UN-pooled Skew-t model are in USkt.zip
Code and readme files for the bridge sampler are in bsPSkt.zip
MCMCAppendix.pdf describes briefly the MCMC scheme of the above samplers and
contains some results on convergence of the chains.