Update with pp changes

Author: WardBrian

BPA/Ch.06/M0.stan

@@ -6,7 +6,7 @@ data {
 transformed data {
   array[M] int<lower=0> s; // Totals in each row
   int<lower=0> C; // Size of observed data set
-
+  
   C = 0;
   for (i in 1 : M) {
     s[i] = sum(y[i]);
@@ -23,20 +23,19 @@ model {
   // Priors are imlicitly defined;
   //  omega ~ uniform(0, 1);
   //  p ~ uniform(0, 1);
-
+  
   // Likelihood
   for (i in 1 : M) {
     if (s[i] > 0) {
       // z[i] == 1
       target += bernoulli_lpmf(1 | omega) + binomial_lpmf(s[i] | T, p);
-    }
-    // s[i] == 0
-    else {
-      target += log_sum_exp(bernoulli_lpmf(1 | omega) // z[i] == 1
+    } else // s[i] == 0
+    {
+      target += log_sum_exp(bernoulli_lpmf(1 | omega)
+                            // z[i] == 1
                             + binomial_lpmf(0 | T, p),
                             bernoulli_lpmf(0 | omega) // z[i] == 0
                             );
-
     }
   }
 }

BPA/Ch.06/Mb.stan

@@ -6,7 +6,7 @@ data {
 transformed data {
   array[M] int<lower=0> s; // Totals in each row
   int<lower=0> C; // Size of observed data set
-
+  
   C = 0;
   for (i in 1 : M) {
     s[i] = sum(y[i]);
@@ -24,11 +24,11 @@ parameters {
 }
 transformed parameters {
   array[M] vector<lower=0, upper=1>[T] p_eff;
-
+  
   for (i in 1 : M) {
     // First occasion
     p_eff[i, 1] = p;
-
+    
     // All subsequent occasions
     for (j in 2 : T) {
       p_eff[i, j] = (1 - y[i, j - 1]) * p + y[i, j - 1] * c;
@@ -39,16 +39,16 @@ model {
   // Priors are implicitly define;
   //  omega ~ uniform(0, 1);
   //  p ~ uniform(0, 1);
-
+  
   // Likelihood
   for (i in 1 : M) {
     if (s[i] > 0) {
       // z[i] == 1
       target += bernoulli_lpmf(1 | omega) + bernoulli_lpmf(y[i] | p_eff[i]);
-    }
-    // s[i] == 0
-    else {
-      target += log_sum_exp(bernoulli_lpmf(1 | omega) // z[i] == 1
+    } else // s[i] == 0
+    {
+      target += log_sum_exp(bernoulli_lpmf(1 | omega)
+                            // z[i] == 1
                             + bernoulli_lpmf(y[i] | p_eff[i]),
                             bernoulli_lpmf(0 | omega) // z[i] == 0
                             );

BPA/Ch.06/Mh.stan

@@ -5,7 +5,7 @@ data {
 }
 transformed data {
   int<lower=0> C; // Size of observed data set
-
+  
   C = 0;
   for (i in 1 : M) {
     if (y[i] > 0) {
@@ -32,17 +32,17 @@ model {
   // In case a weakly informative is used
   //  sigma ~ normal(2.5, 1.25);
   eps_raw ~ normal(0, 1);
-
+  
   // Likelihood
   for (i in 1 : M) {
     if (y[i] > 0) {
       // z[i] == 1
       target += bernoulli_lpmf(1 | omega)
                 + binomial_logit_lpmf(y[i] | T, eps[i]);
-    }
-    // y[i] == 0
-    else {
-      target += log_sum_exp(bernoulli_lpmf(1 | omega) // z[i] == 1
+    } else // y[i] == 0
+    {
+      target += log_sum_exp(bernoulli_lpmf(1 | omega)
+                            // z[i] == 1
                             + binomial_logit_lpmf(0 | T, eps[i]),
                             bernoulli_lpmf(0 | omega) // z[i] == 0
                             );
@@ -52,7 +52,7 @@ model {
 generated quantities {
   array[M] int<lower=0, upper=1> z;
   int<lower=C> N;
-
+  
   for (i in 1 : M) {
     if (y[i] > 0) {
       // animal was detected at least once

BPA/Ch.06/Mt.stan

@@ -6,7 +6,7 @@ data {
 transformed data {
   array[M] int<lower=0> s; // Totals in each row
   int<lower=0> C; // Size of observed data set
-
+  
   C = 0;
   for (i in 1 : M) {
     s[i] = sum(y[i]);
@@ -23,16 +23,16 @@ model {
   // Priors are implicitly defined.
   //  omega ~ uniform(0, 1);
   //  p ~ uniform(0, 1);
-
+  
   // Likelihood
   for (i in 1 : M) {
     if (s[i] > 0) {
       // z[i] == 1
       target += bernoulli_lpmf(1 | omega) + bernoulli_lpmf(y[i] | p);
-    }
-    // s[i] == 0
-    else {
-      target += log_sum_exp(bernoulli_lpmf(1 | omega) // z[i] == 1
+    } else // s[i] == 0
+    {
+      target += log_sum_exp(bernoulli_lpmf(1 | omega)
+                            // z[i] == 1
                             + bernoulli_lpmf(y[i] | p),
                             bernoulli_lpmf(0 | omega) // z[i] == 0
                             );

BPA/Ch.06/MtX.stan

@@ -8,7 +8,7 @@ data {
 }
 transformed data {
   array[M] int<lower=0> s; // Totals in each row
-
+  
   for (i in 1 : M) {
     s[i] = sum(y[i]);
   }
@@ -24,7 +24,7 @@ parameters {
 transformed parameters {
   vector[T] alpha = logit(mean_p);
   matrix[M, T] logit_p;
-
+  
   for (i in 1 : C) {
     logit_p[i] = alpha' + beta * bsize[i];
   }
@@ -39,24 +39,24 @@ model {
   beta ~ normal(0, 10);
   mu_size ~ normal(0, 10);
   //  sd_size ~ uniform(0, prior_sd_upper);   // Provide upper bound as data
-
+  
   // Likelihood
   for (i in 1 : C) {
     bsize[i] ~ normal(mu_size, sd_size) T[-6, 6];
   }
   for (i in (C + 1) : M) {
     bsize_mis[i - C] ~ normal(mu_size, sd_size) T[-6, 6];
   }
-
+  
   for (i in 1 : M) {
     if (s[i] > 0) {
       // z[i] == 1
       target += bernoulli_lpmf(1 | omega)
                 + bernoulli_logit_lpmf(y[i] | logit_p[i]);
-    }
-    // s[i] == 0
-    else {
-      target += log_sum_exp(bernoulli_lpmf(1 | omega) // z[i] == 1
+    } else // s[i] == 0
+    {
+      target += log_sum_exp(bernoulli_lpmf(1 | omega)
+                            // z[i] == 1
                             + bernoulli_logit_lpmf(y[i] | logit_p[i]),
                             bernoulli_lpmf(0 | omega) // z[i] == 0
                             );
@@ -67,7 +67,7 @@ generated quantities {
   matrix<lower=0, upper=1>[M, T] p = inv_logit(logit_p);
   array[M] int<lower=0, upper=1> z;
   int<lower=C> N;
-
+  
   for (i in 1 : M) {
     if (s[i] > 0) {
       // species was detected at least once

BPA/Ch.06/Mtbh.stan

@@ -6,7 +6,7 @@ data {
 transformed data {
   array[M] int<lower=0> s; // Detection times for each species
   int<lower=0> C; // Number of observed species
-
+  
   C = 0;
   for (i in 1 : M) {
     s[i] = sum(y[i]);
@@ -28,10 +28,10 @@ transformed parameters {
   vector[M] eps = sigma * eps_raw;
   vector[T] alpha = logit(mean_p);
   matrix[M, T] logit_p;
-
+  
   // First occasion: no term for recapture (gamma)
   logit_p[ : , 1] = alpha[1] + eps;
-
+  
   // All subsequent occasions: includes recapture term (gamma)
   for (i in 1 : M) {
     for (j in 2 : T) {
@@ -49,17 +49,17 @@ model {
   // In case a weakly informative prior is used
   //  sigma ~ normal(1.5, 0.75);
   eps_raw ~ normal(0, 1);
-
+  
   // Likelihood
   for (i in 1 : M) {
     if (s[i] > 0) {
       // z[i] == 1
       target += bernoulli_lpmf(1 | omega)
                 + bernoulli_logit_lpmf(y[i] | logit_p[i]);
-    }
-    // s[i] == 0
-    else {
-      target += log_sum_exp(bernoulli_lpmf(1 | omega) // z[i] == 1
+    } else // s[i] == 0
+    {
+      target += log_sum_exp(bernoulli_lpmf(1 | omega)
+                            // z[i] == 1
                             + bernoulli_logit_lpmf(y[i] | logit_p[i]),
                             bernoulli_lpmf(0 | omega) // z[i] == 0
                             );
@@ -70,7 +70,7 @@ generated quantities {
   matrix<lower=0, upper=1>[M, T] p = inv_logit(logit_p);
   array[M] int<lower=0, upper=1> z;
   int<lower=C> N;
-
+  
   for (i in 1 : M) {
     if (s[i] > 0) {
       // animal was detected at least once

BPA/Ch.06/Mth.stan

@@ -6,7 +6,7 @@ data {
 transformed data {
   array[M] int<lower=0, upper=T> s; // Totals in each row
   int<lower=0, upper=M> C; // Size of observed data set
-
+  
   C = 0;
   for (i in 1 : M) {
     s[i] = sum(y[i]);
@@ -27,7 +27,7 @@ transformed parameters {
   vector[M] eps = sigma * eps_raw; // Random effects
   array[T] real mean_lp = logit(mean_p);
   matrix[M, T] logit_p;
-
+  
   for (j in 1 : T) {
     logit_p[ : , j] = mean_lp[j] + eps;
   }
@@ -40,17 +40,17 @@ model {
   // In case a weakly informative prior is used
   //  sigma ~ normal(2.5, 1.25);
   eps_raw ~ normal(0, 1);
-
+  
   // Likelihood
   for (i in 1 : M) {
     if (s[i] > 0) {
       // z[i] == 1
       target += bernoulli_lpmf(1 | omega)
                 + bernoulli_logit_lpmf(y[i] | logit_p[i]);
-    }
-    // s[i] == 0
-    else {
-      target += log_sum_exp(bernoulli_lpmf(1 | omega) // z[i] == 1
+    } else // s[i] == 0
+    {
+      target += log_sum_exp(bernoulli_lpmf(1 | omega)
+                            // z[i] == 1
                             + bernoulli_logit_lpmf(y[i] | logit_p[i]),
                             bernoulli_lpmf(0 | omega) // z[i] == 0
                             );
@@ -61,7 +61,7 @@ generated quantities {
   matrix<lower=0, upper=1>[M, T] p = inv_logit(logit_p);
   array[M] int<lower=0, upper=1> z;
   int<lower=C> N;
-
+  
   for (i in 1 : M) {
     if (s[i] > 0) {
       // animal was detected at least once

BPA/Ch.08/mr_ss.stan

@@ -106,9 +106,8 @@ model {
       if (last[i] > first[i]) {
         // Recovered
         pr = cell_prob(n_occasions, s[i], r[i], first[i], last[i] - 1);
-      }
-      // Never Recovered
-      else {
+      } else // Never Recovered
+      {
         pr = cell_prob(n_occasions, s[i], r[i], first[i], n_occasions);
       }
       1 ~ bernoulli(pr);

bugs_examples/vol1/equiv/equiv.stan

@@ -12,7 +12,7 @@ transformed data {
   array[N, P] int T;
   for (n in 1 : N) {
     for (p in 1 : P) {
-      T[n, p] = (group[n] * (2 * p - 3) + 3) / 2;
+      T[n, p] = (group[n] * (2 * p - 3) + 3) %/% 2;
     }
   }
 }

bugs_examples/vol2/pines/pines.stan

@@ -8,7 +8,7 @@ transformed data {
   vector[N] y_std;
   vector[N] x_std;
   vector[N] z_std;
-
+  
   y_std = (y - mean(y)) / sd(y);
   x_std = (x - mean(x)) / sd(x);
   z_std = (z - mean(z)) / sd(z);
@@ -24,11 +24,11 @@ parameters {
 transformed parameters {
   vector<lower=0>[2] sigma;
   vector[2] log_py;
-
+  
   for (i in 1 : 2) {
     sigma[i] = 1 / sqrt(tau[i]);
   }
-
+  
   log_py[1] = log(lambda) + log(0.9995)
               + normal_lpdf(y_std | alpha + beta * x_std, sigma[1])
               + normal_lpdf(alpha | 0, sqrt(1e6))
@@ -38,7 +38,7 @@ transformed parameters {
               + normal_lpdf(gamma | 0, sqrt(1 / 400.0))
               + normal_lpdf(delta | 1, sqrt(1 / 400.0))
               + gamma_lpdf(tau[2] | 46, 4.5);
-
+  
   log_py[2] = log(lambda) + log1m(0.0005)
               + normal_lpdf(y_std | gamma + delta * z_std, sigma[2])
               + normal_lpdf(gamma | 0, sqrt(1e6))
@@ -59,10 +59,10 @@ generated quantities {
 /*
  * p(lambda) = 1
  *
- * p(y|lambda,zeta1,zeta2)
+ * p(y|lambda,zeta1,zeta2) 
  *   = lambda * p1(y|zeta1) + (1 - lambda) * p2(y|zeta2)
  *
- * log p(y|lambda,zeta1,zeta2)
+ * log p(y|lambda,zeta1,zeta2) 
  *   = log_sum_exp(log(lambda) + log(p1(y|zeta1)),
  *                 log1m(lambda) + log(p2(y|zeta2));
  *