Is it possible to solve part of a system in a model analytically, and part of it numerically (using fsolve?) Can you quote some examples please? thanks
You mean for the steady state?
Yes, for the steady state
Yes it is.
The technique is the same as in https://github.com/JohannesPfeifer/DSGE_mod/tree/master/Born_Pfeifer_2018/Welfare
which made use of the get_Calvo_theta function.
A better example would be https://github.com/JohannesPfeifer/DSGE_mod/blob/master/Gali_2010/Gali_2010_calib_target_steadystate.m