= Estimation over sub-samples, structural change =

== Computing log-posterior over sub-samples ==

Using information provided by estimation_info_

New dsge_likelihood.m function:
 1. Initialize kalman_algo
 1. Check boundaries for all estimated parameters
 1. Initialize M_.Sigma_e, M_.H and M_.params for 1st sub-sample (new separate function)
 1. call dsge_likelihood_sub()
 1. For each sub-sample
      i. Update M_.Sigma_e, M_.H and M_.params
      i. call dsge_likelihood_sub()

New dsge_likelihood_sub.m 
 1. Compute T, R and steady_state (using dynare_resolve())
 1. Demean/detrend observations (new separate function)
 1. Initialize Kalman filter (Pstar, Pinf) (new separate function) '''[first subsample only]'''
 1. Run diffuse filter if necessary. '''[first subsample only]'''
 1. Run Kalman filter

New dsge_log_posterior function:
 1. call (new) dsge_likelihood.m
 1. add log-prior density

New dsge_smoother.m function:
 1. Initialize kalman_algo
 1. Check boundaries for all estimated parameters
 1. Initialize M_.Sigma_e, M_.H and M_.params for 1st sub-sample (new separate function)
 1. Compute T, R and steady_state (using dynare_resolve())
 1. Demean/detrend observations (new separate function)
 1. Initialize Kalman filter (Pstar, Pinf) (new separate function)
 1. Run diffuse filter if necessary. Report error if doesn't converge in first sub-sample.
 1. Run Kalman filter (make separate function) on 1st sub-sample
 1. For each sub-sample
      i. Update M_.Sigma_e, M_.H and M_.params
      i. Compute T, R and steady_state
      i. Demean/detrend observations
      i. Run Kalman filter
 1. For each sub-sample backwards:
      i. Update M_.Sigma_e, M_.H and M_.params
      i. Compute T, R and steady_state
      i. Demean/detrend observations
      i. Run Kalman smoother (make separate function)
 1. If necessary call diffuse smoother