View summary information about the posterior draws of the DSGE model parameters under adaptive learning for up to 10 parallel chains. As in the case of the DSGE model under rational expectation, the results are divided into the following categories:
• | Model information: The data includes the files used by the estimation routine. |
• | Sample and data information: The data contains the selected sample, the length of the training sample, the estimation sample, names of variables, etc. |
• | Posterior draws information: Presents the number of posterior draws, the number of burn-in draws, the acceptance ratio of the random-walk Metropolis algorithm, the choice of inverse Hessian estimator, the scale factors for the proposal density and the initial value. |
• | Marginal likelihood estimation: If you have chosen to estimate the marginal likelihood in the DSGE posterior sampling frame on the Settings tab, then the full sample estimates a presented here. |
• | Parameters: The mean, median, and mode of the marginal posterior distribution of the original parameters are presented along with related parameter data. Specifically, the standard deviation based on assumed independent draws and autocorrelated draws (Newey and West, 1987, corrected), relative numerical efficiency (RNE), inefficiency factor (IF), numerical standard error of the posterior mean, as well as prior distribution information. Finally, quantile values are presented based on the percentiles specified in the data construction file. An example is provided in Table 2. |
Page url: http://www.texlips.net/yada/help/index.html?posterior_sampling_summary3.htm