Combine output of two groups in generated quantities

Modeling
1

Hi! I used multilevel model to calculate parameters for gamma distribution in a semiparametric survival curve. The output I wanted to fit was time-to-event (infectious period) ~ gamma(shape, rate). I have right censored and observed period (_obs, _cens are assigned to parameters’ name). I managed to run it using two for loop for censored and observed data. now I wanted to combine them and extracted time-to-event of each group (define as isolated[I]).

data{
    int<lower=0> Total;
    int<lower=0> M; // data length observed
    int<lower=0> N; // data length censored
    int<lower=0> Niso; // total number of isolator
    real inf_obs[M];
    real inf_cens[N];
    int isol_obs[M];
    int isol_cens[N];
}

parameters{
    real<lower=0> alpha; // log recovery rate
    real<lower=0> beta;
    real<lower=0> zalpha_obs[M];
    real<lower=0> zbeta_obs[M];
     real<lower=0> zalpha_cens[N];
    real<lower=0> zbeta_cens[N];
    real<lower=0> logsigmaA; // variation
    real<lower=0> logsigmaB;
}

model{
    vector[M] alphaInf_obs;
    vector[M] betaInf_obs;
    vector[N] alphaInf_cens;
    vector[N] betaInf_cens;
    alpha ~ lognormal(0,0.05);
    beta ~ normal(0.01,0.05);
    zalpha_obs ~lognormal(0, 1);
    zbeta_obs ~ lognormal(0, 1);
    zalpha_cens ~lognormal(0, 1);
    zbeta_cens ~ lognormal(0, 1);
    logsigmaA ~normal(0,0.09);
    logsigmaB ~normal(0.01,0.01);
    for ( i in 1:M ) {  //M is observed
      alphaInf_obs[i]  = alpha + zalpha_obs[isol_obs[i]]*exp(logsigmaA);
      betaInf_obs[i] = exp(beta) + zbeta_obs[isol_obs[i]]*exp(logsigmaB);
      target +=  gamma_lpdf(inf_obs[i]| alphaInf_obs[i],betaInf_obs[i]/alphaInf_obs[i]);
    }
     for ( i in 1:N ) {  //N is observed
      alphaInf_cens[i]  = alpha + zalpha_cens[isol_cens[i]]*exp(logsigmaA);;
      betaInf_cens[i] = exp(beta) + zbeta_cens[isol_cens[i]]*exp(logsigmaB);
         target+= log(1-gamma_cdf(inf_cens[i],alphaInf_cens[i],betaInf_cens[i]/alphaInf_cens[i])); 
    }
}

generated quantities{
    vector[M] log_lik_obs;
    vector[N] log_lik_cens;
    vector[M] alphaInf_obs;
    vector[M] betaInf_obs;
    vector[N] alphaInf_cens;
    vector[N] betaInf_cens;
  for ( i in 1:M ) {  //M is observed
      alphaInf_obs[i]  = alpha + zalpha_obs[isol_obs[i]]*exp(logsigmaA) ;
      betaInf_obs[i] = exp(beta) + zbeta_obs[isol_obs[i]]*exp(logsigmaB) ;
      log_lik_obs[i] = gamma_lpdf(inf_obs[i]| alphaInf_obs[i],betaInf_obs[i]/alphaInf_obs[i]) ;
      }
 for ( i in 1:N ) {  //M is observed
      alphaInf_cens[i]  = alpha + zalpha_cens[isol_cens[i]]*exp(logsigmaA) ;
      betaInf_cens[i] = exp(beta) + zbeta_cens[isol_cens[i]]*exp(logsigmaB) ;
      log_lik_cens[i] = log(1-gamma_cdf(inf_cens[i],alphaInf_cens[i],betaInf_cens[i]/alphaInf_cens[i])); 
}
}

In generated quantities, I tried creating output of each group (isolator) by creating new variable and assigned group effect. but I got error since the beginning and i am also don’t know how to do it correctly.
Mainly, I want to extract result based on group that isolator as it assigned in function [isol_obs[I]] and also combine output of censored and observed together.

generated quantities{
  real alphaall[Niso]
     alphaOB = alpha + zalpha_obs*logsigmaA;
     betaOB= beta+zbeta_obs*logsigmaB;
     alphaC = alpha + zalpha_cens*logsigmaA;
     betaOB= beta+zbeta_cens*logsigmaB;

My data is simply infectious period (1-400 hours), censored status(0,1), and isolator groups (1-40) where multiple infectious period are in a isolator.
Thank you in advance for your help

2

I’m not exactly sure what you’re trying to do, but we can perhaps get you over the syntax issues if you can say what type of error you are getting.