Highest posterior density interval

Estimation ML/Bayesian estimation
1

Hi.

For calculation of the highest posterior density interval for example 95% , if we have a bimodal distribution we have two different intervals with two different values. How dynare report the values of HDI in these situations ?

In unimodal distributions we have only two different values for highest posterior density interval.

2

The percentiles of a distribution have the same definition regardless of the shape of the distribution. The difference rather is with their interpretation.