stepan
/
dynare
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

More efficient code using recursive approach to compute the mean and covariance matrix. 

The posterior mode, obtained from the metropolis-hastings, and the inverse of the estimated covariance matrix are saved in <NAME_OF_THE_MOD_FILE>_mh_mode.mat.

git-svn-id: https://www.dynare.org/svn/dynare/dynare_v4@1285 ac1d8469-bf42-47a9-8791-bf33cf982152
time-shift
adjemian 2007-05-19 16:54:39 +00:00
parent 243484b387
commit e1385d2cd5
1 changed files with 37 additions and 30 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

67
matlab/compute_mh_covariance_matrix.m
View File

@ -1,12 +1,13 @@
function covar = compute_mh_covariance_matrix()
% Estimation of the posterior covariance matrix.
function [m0,s0] = compute_mh_covariance_matrix()
% Estimation of the posterior covariance matrix and expectation.
%
% INPUTS
% None.
%
% OUTPUTS
% o covar [double] p*p matrix, posterior covariance of the estimated
% parameters (computed from previous metropolis hastings).
% OUTPUTS
% o m0 [double] (n*1) vector, posterior expectation of the parameters.
% o s0 [double] (n*n) matrix, posterior covariance of the parameters
% (computed from previous metropolis hastings).
%
%
% ALGORITHM
@ -18,43 +19,49 @@ function covar = compute_mh_covariance_matrix()
%
% part of DYNARE, copyright S. Adjemian, M. Juillard (2006)
% Gnu Public License.
global M_ options_ estim_params_ oo_
global M_ options_ estim_params_
npar = estim_params_.np+estim_params_.nvn+estim_params_.ncx+estim_params_.ncn+estim_params_.nvx;
n = estim_params_.np + ...
estim_params_.nvn+ ...
estim_params_.ncx+ ...
estim_params_.ncn+ ...
estim_params_.nvx;
nblck = options_.mh_nblck;
MhDirectoryName = CheckPath('metropolis');
load([ MhDirectoryName '/' M_.fname '_mh_history'])
FirstMhFile = record.KeepedDraws.FirstMhFile;
FirstLine = record.KeepedDraws.FirstLine; ifil = FirstLine;
FirstLine = record.KeepedDraws.FirstLine;
TotalNumberOfMhFiles = sum(record.MhDraws(:,2));
TotalNumberOfMhDraws = sum(record.MhDraws(:,1));
MAX_nruns = ceil(options_.MaxNumberOfBytes/(npar+2)/8);
TODROP = floor(options_.mh_drop*TotalNumberOfMhDraws);
MU = zeros(1,npar);
SIGMA = zeros(npar,npar);
load([ MhDirectoryName '/' M_.fname '_mh' int2str(1) '_blck' int2str(1)],'x2','logpo2');
params = x2(1,:); oldlogpo2 = logpo2(1);
for blck = 2:nblck
load([ MhDirectoryName '/' M_.fname '_mh' int2str(1) '_blck' int2str(blck)],'x2','logpo2');
if logpo2(1)>oldlogpo2
oldlogpo2 = logpo2(1);
params = x2(1,:);
end
end
fprintf('MH: I''m computing the posterior mean... ');
offset = 0;
m0 = zeros(n,1); s0 = zeros(n,n);
for n = FirstMhFile:TotalNumberOfMhFiles
for b = 1:nblck
load([ MhDirectoryName '/' M_.fname '_mh' int2str(n) '_blck' int2str(b)],'x2','logpo2');
MU = MU + sum(x2(ifil:end,:));
[tmp,idx] = max(logpo2);
if tmp>oldlogpo2
oldlogpo2 = tmp;
params = x2(idx,:);
end
[m0,s0,offset] = recursive_moments(m0,s0,x2(FirstLine,:),offset);
end
ifil = 1;
FirstLine = 1;
end
MU = MU/((TotalNumberOfMhDraws-TODROP)*nblck);
MU1 = repmat(MU,MAX_nruns,1);
fprintf(' Done!\n');
fprintf('MH: I''m computing the posterior covariance matrix... ');
ifil = FirstLine;
for n = FirstMhFile:TotalNumberOfMhFiles
for b = 1:nblck
load([MhDirectoryName '/' M_.fname '_mh' int2str(n) '_blck' int2str(b)],'x2');
x = x2(ifil:end,:)-MU1(1:size(x2(ifil:end,:),1),:);
SIGMA = SIGMA + x'*x;
end
ifil = 1;
end
SIGMA = SIGMA/((TotalNumberOfMhDraws-TODROP)*nblck);%<=== Variance of the parameters (ok!)
xparam1 = params';
hh = inv(s0);
fval = oldlogpo2;
save([M_.fname 'mh_mode'],'xparam1','hh','fval');