A way to retrieve statespace DynamicFactor(MQ) Factor Predictions

[NMRobert]

When I estimate statsmodels.tsa.statespace.dynamic_factor_mq models and forecast out-of-sample, it would be very useful to be able to retrieve the forecasts for the factors as well as the observation variables. I've attempted to get at the underlying internal data by doing something like this:

fitted_model = dfm.fit(...)
res = fitted_model.get_forecast()
res.prediction_results.predicted_state

but the shape of the predicted_state numpy matrix doesn't quite make sense to me: I was expecting it to have dimension (n_obs x num_factors), but it seems to have dimension (n_obs x (n_end+n_factors*2)). Is there a way to accomplish this? Thank you.