Draw from the prior distribution of the original parameters of the DSGE model. By default, YADA suggest the number of draws to be equal to the number determined by the maximum number of posterior draws to use for prediction option in the posterior sampling frame on the Options tab. The user can change this number within a range from 100 to 50,000 before drawing random numbers from the prior distribution of the parameters.
For models with a system prior, prior sampling will be conducted with either a random-walk Metropolis algorithm with a normal proposal density or with a Metropolis-Hastings algorithm using the marginal prior distributions as proposals.
The prior draws are stored in a mat-file in a sub-directory of the base output directory called priordraws. The draws can then be used by the functions where the prior distribution can be used.
This function is also available on the toolbar.
Icon: |
Additional Information
• | More detailed information about the prior distributions supported by YADA for DSGE models is provided in Section 4.2 of the YADA Manual. |
Page url: http://www.texlips.net/yada/help/index.html?prior_sampling.htm