A question about OSR command and policy frontier

Dear all,
I am now using OSR to calculate the simple optimal rule and plot the policy frontier.
But I find that all the papers on the topic of policy frontier have given a limitation on variance of the key variable,e.g., in Iacoviello(2005), "I constrain interest rate volatility by imposing an upper bound 25 percent larger than the estimated standard deviation…"
So how should I put a variance constraint in OSR?

Anyone knows that?
Great thanks.

That is not easy to impose. Dynare allows you to impose bounds on the parameters, not on the endogenous objects. In the exampe posted at [Policy frontier) you would probably need to check whether


is above a particular value and then set the output to NaN.

Thanks, I will try.
And could you tell me how to impose constraint on the value of the parameters?

See the manual of the unstable version/4.5 on osr_params_bounds

Thanks a lot!
So is there a guide for how to setup dynare 4.5?
and can it co-exist with 4.4.3?

The unstable version is available at dynare.org/download/dynare-unstable/
You can select the Dynare version you want to use by providing Matlab with the path to the Matlab subfolder of the respective Dynare version.

Thanks a lot .
So if I simultaneously add both 4.43 and 4.5 unstable into the matlab path, which version will matlab choose to run ?
And how could I choose to run one of these two versions if I wanna to manually, if both of them are in my matlab path?

Always only add the one you want to your path. For example, when you use 4.4.3, use something like


in Matlab to select 4.4.3 (of course with your path)