Changeset 1100 for applications/bdmtoolbox
- Timestamp:
- 06/13/10 22:56:55 (14 years ago)
- Files:
-
- 1 modified
Legend:
- Unmodified
- Added
- Removed
-
applications/bdmtoolbox/tutorial/userguide/partfrg_estim.m
r1038 r1100 21 21 Apri.class = 'ARX'; 22 22 DSpri = DS; 23 DSpri.Data = DS.Data(:,1: 40);23 DSpri.Data = DS.Data(:,1:6); 24 24 % get decent prior -- estimate with data first 25 25 [Dum,post]=estimator(DSpri,{Apri}); … … 30 30 31 31 %%%%%% Random walk on frg - Dirichlet 32 walk.class = 'm Beta'; % random walk on coefficient phi32 walk.class = 'mDirich'; % random walk on coefficient phi 33 33 walk.rv = RV({'phi'},4); % 2D random walk - frg is the first element 34 walk.k = 0.001 *ones(1,4); % width of the random walk34 walk.k = 0.001; % width of the random walk 35 35 walk.betac = 0.1*ones(1,4); % stabilizing elememnt of random walk 36 36 … … 49 49 E.prior.alpha = 5*ones(1,4); % 50 50 E.prior.beta = ones(1,4); % 51 E.log_level = 'logbounds,logweights ,logmeans,logvars';51 E.log_level = 'logbounds,logweights'; 52 52 E.name = 'MPF'; 53 53