I am new to Dynare, and am wondering if there is a simple way to code the expectation of function of a random variable when the underlying distribution is non-Gaussian. For example, I would like to evaluate expressions such as

integral [x**(1/(1-rho))] dF(x),

where F is the cdf of the Pareto distribution. I would like to use expressions like this in the model statements of a DSGE model.

I am also interested in evaluating functions of a random variable subject to a cutoff, such as

integral{x<x^} [x**(1/(1-rho))] dF(x).