prt_sem.m

计量工具箱

fprintf(fid,'\n');
fprintf(fid,'Bayesian spatial autoregressive Tobit model \n');
if (nflag == 1)
fprintf(fid,'Dependent Variable = %16s \n',vnames(1,:));
end;
fprintf(fid,'sigma^2            = %9.4f \n',sige);
if results.rdraw == 0
fprintf(fid,'r-value            = %6d   \n',results.r);
else
fprintf(fid,'mean of rdraws     = %9.4f \n',mean(results.rdraw));
fprintf(fid,'gam(m,k) prior     = %6d,%6d \n',results.m,results.k);
end; 
fprintf(fid,'Nobs, Nvars        = %6d,%6d \n',results.nobs,results.nvar);
fprintf(fid,'# censored values  = %6d \n',results.nobsc);
fprintf(fid,'ndraws,nomit       = %6d,%6d \n',results.ndraw,results.nomit);
fprintf(fid,'time in secs       = %9.4f   \n',results.time);
fprintf(fid,'min and max rho    = %9.4f,%9.4f \n',results.rmin,results.rmax);
fprintf(fid,'***************************************************************\n');

vstring = 'Variable';
bstring = 'Prior Mean';
tstring = 'Std Deviation';

tmp = [results.bmean results.bstd];

cnames = strvcat(bstring,tstring);
rnames = vstring;
for i=1:nvar
rnames = strvcat(rnames,Vname(i+1,:));
end;

pin.fmt = '%16.6f';
pin.fid = fid;
pin.cnames = cnames;
pin.rnames = rnames;

mprint(tmp,pin);
fprintf(fid,'***************************************************************\n');
fprintf(fid,'      Posterior Estimates \n');

 if strcmp(results.tflag,'tstat')
% now print coefficient estimates, t-statistics and probabilities
tout = norm_prb(results.tstat); % find asymptotic z (normal) probabilities
      
tmp = [bout results.tstat tout];  % matrix to be printed
% column labels for printing results
bstring = 'Coefficient'; tstring = 'Asymptot t-stat'; pstring = 'z-probability';
cnames = strvcat(bstring,tstring,pstring);
in.cnames = cnames;
in.rnames = Vname;
in.fmt = '%16.6f';
in.fid = fid;
mprint(tmp,in);
 else % use p-levels for Bayesian results
tmp = [bout bstd tout];  % matrix to be printed
% column labels for printing results
bstring = 'Coefficient'; tstring = 'Std Deviation'; pstring = 'p-level';
cnames = strvcat(bstring,tstring,pstring);
in.cnames = cnames;
in.rnames = Vname;
in.fmt = '%16.6f';
in.fid = fid;
mprint(tmp,in);
end;

return;

% <=================== end of sart_g case


case {'sarp_g','sarp_gc'} % <=================== Gibbs spatial autoregressive Probit model

nobs = results.nobs;
nvar = results.nvar;
  
% find posterior means
tmp1 = mean(results.bdraw);
pout = mean(results.pdraw);
bout = [tmp1'
        pout];

sige = mean(results.sdraw);
tmp1 = std(results.bdraw);
tmp2 = std(results.pdraw);
bstd = [tmp1'
        tmp2];

if strcmp(results.tflag,'tstat')
 tstat = bout./bstd;
 % find t-stat marginal probabilities
 tout = tdis_prb(tstat,results.nobs);
 results.tstat = bout./bstd; % trick for printing below
else % find plevels
 draws = [results.bdraw results.pdraw];
 for i=1:results.nvar+1;
 if bout(i,1) > 0
 cnt = find(draws(:,i) > 0);
 tout(i,1) = 1 - (length(cnt)/(results.ndraw-results.nomit));
 else
 cnt = find(draws(:,i) < 0);
 tout(i,1) = 1 - (length(cnt)/(results.ndraw-results.nomit));
 end; % end of if - else
 end; % end of for loop
end; 


fprintf(fid,'\n');
fprintf(fid,'Bayesian spatial autoregressive Probit model \n');
if (nflag == 1)
fprintf(fid,'Dependent Variable = %16s \n',vnames(1,:));
end;
fprintf(fid,'sigma^2         = %9.4f \n',sige);
if results.rdraw == 0
fprintf(fid,'r-value         = %6d   \n',results.r);
else
fprintf(fid,'mean of rdraws  = %9.4f \n',mean(results.rdraw));
fprintf(fid,'gam(m,k) prior  = %6d,%6d \n',results.m,results.k);
end; 
fprintf(fid,'Nobs, Nvars     = %6d,%6d \n',results.nobs,results.nvar);
fprintf(fid,'# 0, 1 y-values = %6d,%6d \n',results.zip,nobs-results.zip);
fprintf(fid,'ndraws,nomit    = %6d,%6d \n',results.ndraw,results.nomit);
fprintf(fid,'time in secs    = %9.4f   \n',results.time);
fprintf(fid,'min and max rho = %9.4f,%9.4f \n',results.rmin,results.rmax);
fprintf(fid,'***************************************************************\n');


 if strcmp(results.tflag,'tstat')
% now print coefficient estimates, t-statistics and probabilities
tout = norm_prb(results.tstat); % find asymptotic z (normal) probabilities
      
tmp = [bout results.tstat tout];  % matrix to be printed
% column labels for printing results
bstring = 'Coefficient'; tstring = 'Asymptot t-stat'; pstring = 'z-probability';
cnames = strvcat(bstring,tstring,pstring);
in.cnames = cnames;
in.rnames = Vname;
in.fmt = '%16.6f';
in.fid = fid;
mprint(tmp,in);
 else % use p-levels for Bayesian results
tmp = [bout bstd tout];  % matrix to be printed
% column labels for printing results
bstring = 'Coefficient'; tstring = 'Std Deviation'; pstring = 'p-level';
cnames = strvcat(bstring,tstring,pstring);
in.cnames = cnames;
in.rnames = Vname;
in.fmt = '%16.6f';
in.fid = fid;
mprint(tmp,in);
end;

return;

% <=================== end of sarp_g case


case {'far'} % <=================== first-order autoregressive model

nvar = 1;
nobs = results.nobs;


% special handling of vnames
Vname = 'Variable';
% add spatial rho parameter name
    Vname = strvcat(Vname,'rho');

fprintf(fid,'\n');
fprintf(fid,'First-order spatial autoregressive model Estimates \n');
if (nflag == 1)
fprintf(fid,'Dependent Variable = %16s \n',vnames(1,:));
end;
fprintf(fid,'R-squared          = %9.4f \n',results.rsqr);
fprintf(fid,'sigma^2            = %9.4f \n',results.sige);
fprintf(fid,'log-likelihood     = %16.8g \n',results.lik);
fprintf(fid,'Nobs, Nvars        = %6d,%3d \n',results.nobs,results.nvar);
fprintf(fid,'# of iterations    = %6d   \n',results.iter);

% print timing information
fprintf(fid,'total time in secs = %9.4f \n',results.time);
fprintf(fid,'time for optimiz   = %9.4f \n',results.time4);
if results.time1 ~= 0
fprintf(fid,'time for lndet     = %9.4f \n',results.time1);
end;
if results.time2 ~= 0
fprintf(fid,'time for eigs      = %9.4f \n',results.time2);
end;
if results.time3 ~= 0
fprintf(fid,'time for t-stat    = %9.4f \n',results.time3);
end;
% print min and max rho used
fprintf(fid,'min and max rho    = %8.4f,%8.4f \n',results.rmin,results.rmax);

if results.lflag == 0
fprintf(fid,'No lndet approximation used \n');
end;
% put in information regarding Pace and Barry approximations
if results.lflag == 1
fprintf(fid,'Pace and Barry, 1999 MC lndet approximation used \n');
fprintf(fid,'order for MC appr  = %6d  \n',results.order);
fprintf(fid,'iter  for MC appr  = %6d  \n',results.liter);
end;
if results.lflag == 2
fprintf(fid,'Pace and Barry, 1998 spline lndet approximation used \n');
end;

fprintf(fid,'***************************************************************\n');

bout = results.rho;

% <=================== end of far case

case {'far_g','far_gc'} % <=================== Gibbs first-order autoregressive model

nobs = results.nobs;
nvar = 1;

% special handling of vnames
Vname = 'Variable';
% add spatial rho parameter name
    Vname = strvcat(Vname,'rho');
    
% find posterior means
bout = mean(results.pdraw);
sige = mean(results.sdraw);
bstd = std(results.pdraw);

if strcmp(results.tflag,'tstat')
 tstat = bout./bstd;
 % find t-stat marginal probabilities
 tout = tdis_prb(tstat,results.nobs);
 results.tstat = bout./bstd; % trick for printing below
else % find plevels
 draws = [results.pdraw];
 if bout > 0
 cnt = find(draws > 0);
 tout = 1 - (length(cnt)/(results.ndraw-results.nomit));
 else
 cnt = find(draws < 0);
 tout = 1 - (length(cnt)/(results.ndraw-results.nomit));
 end; % end of if - else
end; 


y = results.y;
e = y - results.yhat;
sigu = e'*e;
ym = y - mean(y);
rsqr1 = sigu;
rsqr2 = ym'*ym;
rsqr = 1.0 - rsqr1/rsqr2; % conventional r-squared

fprintf(fid,'\n');
fprintf(fid,'Bayesian First-order spatial autoregressive model \n');
if (nflag == 1)
fprintf(fid,'Dependent Variable = %16s \n',vnames(1,:));
end;
fprintf(fid,'R-squared          = %9.4f   \n',rsqr);
fprintf(fid,'sigma^2            = %9.4f   \n',sige);
if results.rdraw ~= 0
fprintf(fid,'rmean              = %9.4f \n',mean(results.rdraw));
else
fprintf(fid,'r-value            = %6d  \n',results.r);
end;
fprintf(fid,'Nobs, Nvars        = %6d,%3d \n',results.nobs,nvar);
fprintf(fid,'ndraws,nomit       = %6d,%6d \n',results.ndraw,results.nomit);
fprintf(fid,'min and max rho    = %9.4f,%9.4f \n',results.rmin,results.rmax);

% print timing information
fprintf(fid,'total time in secs = %9.4f \n',results.time);
if results.time1 ~= 0
fprintf(fid,'time for lndet     = %9.4f \n',results.time1);
end;
if results.time2 ~= 0
fprintf(fid,'time for eigs      = %9.4f \n',results.time2);
end;
if results.time3 ~= 0
fprintf(fid,'time for sampling  = %9.4f \n',results.time3);
end;

if results.lflag == 0
fprintf(fid,'No lndet approximation used \n');
end;
% put in information regarding Pace and Barry approximations
if results.lflag == 1
fprintf(fid,'Pace and Barry, 1999 MC lndet approximation used \n');
fprintf(fid,'order for MC appr  = %6d  \n',results.order);
fprintf(fid,'iter  for MC appr  = %6d  \n',results.liter);
end;
if results.lflag == 2
fprintf(fid,'Pace and Barry, 1998 spline lndet approximation used \n');
end;

fprintf(fid,'***************************************************************\n');


 if strcmp(results.tflag,'tstat')
% now print coefficient estimates, t-statistics and probabilities
tout = norm_prb(results.tstat); % find asymptotic z (normal) probabilities
      
tmp = [bout results.tstat tout];  % matrix to be printed
% column labels for printing results
bstring = 'Coefficient'; tstring = 'Asymptot t-stat'; pstring = 'z-probability';
cnames = strvcat(bstring,tstring,pstring);
in.cnames = cnames;
in.rnames = Vname;
in.fmt = '%16.6f';
in.fid = fid;
mprint(tmp,in);
 else % use p-levels for Bayesian results
tmp = [bout bstd tout];  % matrix to be printed
% column labels for printing results
bstring = 'Coefficient'; tstring = 'Std Deviation'; pstring = 'p-level';
cnames = strvcat(bstring,tstring,pstring);
in.cnames = cnames;
in.rnames = Vname;
in.fmt = '%16.6f';
in.fid = fid;
mprint(tmp,in);
end;

return;

% <=================== end of far_g case

case {'sac'} % <=================== general spatial model

nobs = results.nobs;
nvar = results.nvar;

% special handling of vnames
if ( nflag == 0) %  no variable names supplied, make some up
Vname = 'Variable';
 for i=1:nvar
    tmp = ['variable ',num2str(i)];
    Vname = strvcat(Vname,tmp);
 end;
% add spatial rho, lambda parameter name
    Vname = strvcat(Vname,'rho');
    Vname = strvcat(Vname,'lambda');

elseif (nflag == 1) % the user supplied variable names
Vname = 'Variable';
[tst_n nsize] = size(vnames);
 if tst_n ~= nvar+1
 error('Wrong # of variable names in prt_sem -- check vnames argument');
 end;
 for i=1:nvar
    Vname = strvcat(Vname,vnames(i+1,:));
 end;
% add spatial rho, lambda parameter name
    Vname = strvcat(Vname,'rho');
    Vname = strvcat(Vname,'lambda');

end; % end of nflag issue


fprintf(fid,'\n');
fprintf(fid,'General Spatial Model Estimates \n');
if (nflag == 1)
fprintf(fid,'Dependent Variable = %16s \n',vnames(1,:));
end;
fprintf(fid,'R-squared          = %9.4f \n',results.rsqr);
fprintf(fid,'Rbar-squared       = %9.4f \n',results.rbar);
fprintf(fid,'sigma^2            = %9.4f \n',results.sige);
fprintf(fid,'log-likelihood     = %16.8g \n',results.lik);
fprintf(fid,'Nobs, Nvars        = %6d,%6d \n',results.nobs,results.nvar);
fprintf(fid,'# iterations       = %6d \n',results.iter);
% print timing information
fprintf(fid,'total time in secs = %9.4f \n',results.time);
fprintf(fid,'time for optimiz   = %9.4f \n',results.time4);
if results.time1 ~= 0
fprintf(fid,'time for lndet     = %9.4f \n',results.time1);
end;
if results.time2 ~= 0
fprintf(fid,'time for eigs      = %9.4f \n',results.time2);
end;
if results.time3 ~= 0
fprintf(fid,'time for t-stat    = %9.4f \n',results.time3);
end;
fprintf(fid,'***************************************************************\n');

bout = [results.beta
        results.rho
        results.lam];
        

% <=================== end of sac case

case {'sac_g'} % <=================== Gibbs general spatial model

nobs = results.nobs;
nvar = results.nvar;

% handling of vnames
Vname = 'Variable';
 for i=1:nvar
    tmp = ['variable ',num2str(i)];
    Vname = strvcat(Vname,tmp);
 end;
% add spatial rho parameter name
    Vname = strvcat(Vname,'rho');
% add spatial rho parameter name
    Vname = strvcat(Vname,'lambda');

if (nflag == 1) % the user supplied variable names
Vname = 'Variable';
[tst_n nsize] = size(vnames);
 if tst_n ~= nvar+1
 fprintf(fid,'Wrong # of variable names in prt_sem -- check vnames argument \n');
 fprintf(fid,'will use generic variable names \n');
 nflag = 0;
 else,
 for i=1:nvar
    Vname = strvcat(Vname,vnames(i+1,:));
 end;
% add spatial rho parameter name
    Vname = strvcat(Vname,'rho');
    Vname = strvcat(Vname,'lambda');
 end; % end of if-else
end; % end of nflag issue

  
% find posterior means
tmp1 = mean(results.bdraw);
pout = mean(results.pdraw);
lout = mean(results.ldraw);
bout = [tmp1'
        pout
        lout];

y = results.y;
sige = mean(results.sdraw);