void MBC_MCMC_loop(ERGMM_MCMC_Model *model, ERGMM_MCMC_Priors *prior,
ERGMM_MCMC_MCMCState *cur, ERGMM_MCMC_MCMCSettings *setting, ERGMM_MCMC_ROutput *outlists)
{
unsigned int pos=0;
unsigned int iter, total_iters = setting->sample_size*setting->interval;
// Rprintf("Started MCMC loop.\n");
/* Note that indexing here starts with 1.
It can be thought of as follows:
At the end of the updates, we have made iter complete MCMC updates. */
for(iter=1;iter<=total_iters;iter++){
R_CheckUserInterrupt(); // So that CTRL-C can interrupt the run.
ERGMM_MCMC_CV_up(model,prior,cur);
ERGMM_MCMC_logp_Z(model, cur->state);
// If we have a new MLE (actually, the highest likelihood encountered to date), store it.
if( cur->state->lpZ > outlists->lpZ[0] ) MBC_MCMC_store_iteration(0,model,cur->state,setting,outlists);
if( cur->state->lpZ + cur->state->lpLV > outlists->lpZ[1] + outlists->lpLV[1] ) MBC_MCMC_store_iteration(1,model,cur->state,setting,outlists);
/* every interval save the results */
if((iter % setting->interval) == 0){
pos = (iter/setting->interval-1)+MBC_OUTLISTS_RESERVE;
// Store current iteration.
MBC_MCMC_store_iteration(pos, model, cur->state, setting, outlists);
}
} // end main MCMC loop
return;
}
void ERGMM_MCMC_loop(ERGMM_MCMC_Model *model, ERGMM_MCMC_Priors *prior,
ERGMM_MCMC_MCMCState *cur, ERGMM_MCMC_MCMCSettings *setting, ERGMM_MCMC_ROutput *outlists)
{
unsigned int n_accept_z=0, n_accept_b=0, pos=0;
unsigned int iter, total_iters = setting->sample_size*setting->interval;
/* Note that indexing here starts with 1.
It can be thought of as follows:
At the end of the updates, we have made iter complete MCMC updates. */
for(iter=1;iter<=total_iters;iter++){
R_CheckUserInterrupt(); // So that CTRL-C can interrupt the run.
if(model->latent || cur->state->sender || cur->state->receiver)
n_accept_z += ERGMM_MCMC_Z_RE_up(model, prior, cur, setting);
if(model->latent){
// Update cluster parameters (they are separated from data by Z, so full conditional sampling).
// Note that they are also updated in coef_up_scl_tr_Z_shift_RE.
if(model->clusters>0)
ERGMM_MCMC_CV_up(model,prior,cur);
else
ERGMM_MCMC_LV_up(model,prior,cur);
}
/* Update coef given this new value of Z and conditioned on everything else.
Also propose to scale Z and shift random effects.
*/
if( ERGMM_MCMC_coef_up_scl_Z_shift_RE(model,prior,cur,setting) ){
n_accept_b++;
}
if(cur->state->sender || cur->state->receiver){
ERGMM_MCMC_REV_up(model,prior,cur);
}
// If we have a new MLE (actually, the highest likelihood encountered to date), store it.
if( cur->state->llk > GET_DEFAULT(outlists->llk,0,0) ) ERGMM_MCMC_store_iteration(0,model,cur->state,setting,outlists);
// If we have a new posterior mode (actually, the highest joint density of all variables but K observed to date), store it.
if( cur->state->llk + cur->state->lpZ + cur->state->lpLV +
cur->state->lpcoef + cur->state->lpRE + cur->state->lpREV + cur->state->lpdispersion >
GET_DEFAULT(outlists->llk,1,0) + GET_DEFAULT(outlists->lpZ,1,0) + GET_DEFAULT(outlists->lpLV,1,0) +
GET_DEFAULT(outlists->lpcoef,1,0) + GET_DEFAULT(outlists->lpRE,1,0) + GET_DEFAULT(outlists->lpREV,1,0) +
GET_DEFAULT(outlists->lpdispersion,1,0) )
ERGMM_MCMC_store_iteration(1,model,cur->state,setting,outlists);
/* every interval save the results */
if((iter % setting->interval) == 0){
pos = (iter/setting->interval-1)+ERGMM_OUTLISTS_RESERVE;
// Store current iteration.
ERGMM_MCMC_store_iteration(pos, model, cur->state, setting, outlists);
// Acceptance rates.
outlists->coef_rate[pos] = (double) ((double)n_accept_b)/((double)setting->interval);
if(outlists->Z_rate_move){
outlists->Z_rate_move[pos] = (double) ((double)n_accept_z)/((double)setting->interval*model->verts);
}
n_accept_z=0;
n_accept_b=0;
}
} // end main MCMC loop
return;
}
