Estimating unobservables

Hi everyone,
I want to estimate a DSGE model where I have two or more trends, for example, non-stationary unit root and investment specific technology trend. I give the actual GDP data, and the model returns me the GDP gap and the trend. But, here, is my question, how can I tell dynare that, for example, at 2015Q1 my GDP gap was 8% and do the filtering, given this value. Is it possible?
Thank You, in advance.

We have the filter_initial_state-block that may be helpful here. See the manual.

Thank You, a lot, for Your response professor, but in manual I can’t find anything about filter_initial_state block command. Can You provide more details about it?

From the manual:

Sorry, Professor, but when I am trying to search filter_initial_state in Dynare manual, it gives me no result. I can’t understand what’s the problem I am very sorry.
Screen Shot 2021-12-03 at 16.41.01

This is a feature of 4.7. See

Thank You Professor for Your reply. I read how the filter_initial_state works, but I guess it doesn’t give me the option to set a value within the period of filtering. I mean that, if for example, my data starts at 2004Q1 and I want to set the value of an unobservable at 2009Q1 to be 8. Is there an option to do this?