Daniel F. Ahelegbey, Monica Billio, and Roberto Casarin,
"Bayesian Graphical Models for Structural Vector Autoregressive
Processes", Journal of Applied Econometrics, Vol. 31, No. 2, 2016,
pp. 357-386.

All code and data files are in the zip file abc-codes-and-data.zip.
All the datasets and lists of variables are tab-delimited text files
in DOS format. Unix/Linux users should use "unzip -a".

All data files, lists of variables, and the script file for data
generation are available in the data folder:

There are two data files (Macro_Data_1.txt and Macro_Data_2.txt), each
consisting of quarterly observations from 1959Q1 to 2008Q4
corresponding to the datasets used for the macroeconomic application
in Section 5 and the robustness check in the appendix of the paper. 

Macro_Data_1.txt (40KB) contains 20 variables used in Section 5 of
the paper.

Macro_Data_2.txt (42KB) contains 21 variables used for the Robustness
Check of the paper.

Macro_Variables_1.txt (2KB) and Macro_Variables_2.txt (2KB) list the
variables of the two data files. Each line provides the Short ID,
Mnemonic, Transformation Code, and Description of one of the variables,
as described in Table IV in Appendix.

There is another data file consisting of monthly observations from
1/31/2001 to 8/30/2013 corresponding to the dataset used for the
financial application in Section 6 of the paper. 

Stoxx600_Data.txt (20KB) contains the 19 Super-sectors and 152
observations used for the Financial Application in Section 6 of the
paper.

The list of variables of the financial application are reported in
Stoxx_Variables.txt (1KB).

SIMVAR.m (2KB) is the Matlab script used to generate the data for
the simulation experiment in Section 4. 


The following program files are contained in the function folder:

SAMPLE_BGMAR_DAG.m (2KB) contains the Matlab script for sampling the
temporal dependence between explanatory variables and responses.

SAMPLE_BGMIN_DAG.m (3KB) contains the Matlab script for sampling the
cross-sectional dependence among response variables.

PROC_DATA.m (2KB) contains the Matlab script for procession the data

LOG_SCORE.m (1KB) contains the Matlab script for computing the
logarithmic score of the graph

CONVERGENCE.m (4KB) contains the Matlab script for convergence
diagnostics

Estimate_BIC.m (1KB) contains the Matlab script for estimating the lag
using BIC

transform.m (3BK) contains the Matlab script for transforming the data
for stationarity


The following files are contained in the figure folder:

Draw_Adj.m (2KB) contains the Matlab script for graphical
representation of the connectivity matrix

heatmap.m (23KB) contains the Matlab script for displaying a matrix as
a heatmap image

rotateticklabel.m (2KB) contains the Matlab script for rotating tick
labels


Other Scripts (in the main folder):

Mxample.m (1KB) -- simulation experiment

Finance_exp.m (1KB) -- application to finance

Macro_exp.m (1KB) -- application to macro