ssfortran.Regression

class ssfortran.Regression(exog, random_walk=False, series=0, at_observations=False)

Regression effects \(x_t' \beta_t\) (DK §3.2.5, §3.6).

The coefficients are diffuse states, fixed or random walks.

Parameters:

exogarray_like, shape (n,) or (n, k)

Regressors; also interventions (step, pulse or slope dummies).

random_walkbool or sequence of bool

Make all, or the flagged, coefficients random walks with estimated variances (DK eq. 3.15).

seriesint

The series (0-based) the effects enter.

at_observationsbool

Act on the observations rather than the signals.