Consider the simple 1-factor model structure below:

R_it = a_i + beta_i*S_t + u_it

S_t = phi* S_t-1 + v_t

where R_it is a 5×1 vector. As you can see, the structure is simple but differs from the baseline factor models shown in statsmodels https://www.statsmodels.org/dev/examples/notebooks/generated/statespace_sarimax_stata.html

In particular, my state-space equation contains a constant. I know that I can use the simple command in Stata

sspace (u L.u, state noconstant)                                            ///
    (y_m1 u)(y_m3 u) (y_m6 u)                                   /// 
    (y_y1 u) (y_y2 u)   covstate(identity) 

How can I estimate such model in statsmodels by maximum likelihood?

Thanks!