DPMSV2024

fse_results.mat: full sample estimation results file using 2022-12 (2022-12-ccmm16-r).
2022-12-ccmm16-r.txt: full sample estimation dataset.
2022-12-r.txt: Fred-MD 2022-12 with 99 variables, the wide dataset.
BVAR_DP_CSV_fred_008.m: the main file for running forecast excercises, to run the file, 1) modify the file to load the dataset or 2) use the genqforecast file to generate fcst file and then run the fcst file, it will load data and this main file.
construct_prior_A_woa0.m: construct the parameter matrix A without the global intercept vector.
crpsDraws.m: calculate CRPS.
fredMD16-2021-04-nvn.csv: data file with 16 variables generated by CCMM without variable name labels.
genqforecast_2021_04.m: used to generate the fcst file using Fred-MD 2021-04 dataset.
genqforecast_dmpsv99.m: used to generate the fcst file using Fred-MD 2022-12 dataset.
nigSV.m: the class file for multivariate normal inverse gamma with stochastic volatility.
oneDPM.m: used to generate the initialization of cluster assignments for DPM.
randmn.m: used to generate samples from a multinomial distribution.
sample_h.m: used to generate samples the common stochastic volatility h.
full sample estimation: this folder contains all files to run the full sample estimation, run BVAR_DPM_CSV_FE.m.
AppenB 2021_04: this folder contains all fcst result files using 2021-04 to generate the tables.
2022_12: this folder contains all fcst result files using 2022-12, 455 dpmsv 01 to 04 are results using 99 variables.