Address reviewer comments.

cxx/distributions/BUILD

@@ -4,7 +4,6 @@ cc_library(
     name = "distributions",
     visibility = ["//:__subpackages__"],
     deps = [
-        ":adapter",
         ":base",
         ":beta_bernoulli",
         ":bigram",
@@ -14,15 +13,6 @@ cc_library(
     ],
 )
 
-cc_library(
-    name = "adapter",
-    hdrs = ["adapter.hh"],
-    visibility = ["//:__subpackages__"],
-    deps = [
-        ":base",
-    ],
-)
-
 cc_library(
     name = "base",
     hdrs = ["base.hh"],
@@ -87,17 +77,6 @@ cc_library(
     ],
 )
 
-cc_test(
-    name = "adapter_test",
-    srcs = ["adapter_test.cc"],
-    deps = [
-        ":adapter",
-        ":normal",
-        "@boost//:algorithm",
-        "@boost//:test",
-    ],
-)
-
 cc_test(
     name = "beta_bernoulli_test",
     srcs = ["beta_bernoulli_test.cc"],

cxx/distributions/base.hh

@@ -3,7 +3,7 @@
 template <typename T>
 class Distribution {
   // Abstract base class for probability distributions in HIRM.
-  // New distribution subclasses need to be added to 
+  // New distribution subclasses need to be added to
   // `util_distribution_variant` to be used in the (H)IRM models.
  public:
   typedef T SampleType;
@@ -39,8 +39,5 @@ class Distribution {
   // e^logp_score() under those hyperparameters.
   virtual void transition_hyperparameters() = 0;
 
-  // Return a copy of the distribution with no observed data.
-  virtual Distribution<T>* prior() const = 0;
-
   virtual ~Distribution() = default;
 };

cxx/distributions/beta_bernoulli.cc

@@ -61,7 +61,3 @@ void BetaBernoulli::transition_hyperparameters() {
   alpha = hypers[i].first;
   beta = hypers[i].second;
 }
-
-BetaBernoulli* BetaBernoulli::prior() const {
-  return new BetaBernoulli (prng);
-}

cxx/distributions/beta_bernoulli.hh

@@ -7,6 +7,8 @@
 #include "distributions/base.hh"
 #include "util_math.hh"
 
+// TODO(thomaswc, emilyaf): Change BetaBernoulli to use bool instead of
+// double.
 class BetaBernoulli : public Distribution<double> {
  public:
   double alpha = 1;  // hyperparameter
@@ -35,6 +37,4 @@ class BetaBernoulli : public Distribution<double> {
   double sample();
 
   void transition_hyperparameters();
-
-  BetaBernoulli* prior() const;
 };

cxx/distributions/bigram.cc

@@ -118,7 +118,3 @@ void Bigram::transition_hyperparameters() {
   int i = sample_from_logps(logps, prng);
   set_alpha(alphas[i]);
 }
-
-Bigram* Bigram::prior() const {
-  return new Bigram (prng);
-}

cxx/distributions/bigram.hh

@@ -47,6 +47,4 @@ class Bigram : public Distribution<std::string> {
   void set_alpha(double alphat);
 
   void transition_hyperparameters();
-
-  Bigram* prior() const;
 };

cxx/distributions/crp.cc

@@ -112,7 +112,3 @@ void CRP::transition_alpha() {
   int idx = log_choice(logps, prng);
   this->alpha = grid[idx];
 }
-
-CRP* CRP::prior() const {
-  return new CRP(prng);
-}

cxx/distributions/crp.hh

@@ -6,14 +6,13 @@
 #include <unordered_map>
 #include <unordered_set>
 
-
 typedef int T_item;
 
 // TODO(emilyaf): Make this a distribution subclass.
 class CRP {
  public:
   double alpha = 1.;  // concentration parameter
-  int N = 0;         // number of customers
+  int N = 0;          // number of customers
   std::unordered_map<int, std::unordered_set<T_item>>
       tables;  // map from table id to set of customers
   std::unordered_map<T_item, int> assignments;  // map from customer to table id
@@ -36,6 +35,4 @@ class CRP {
   std::unordered_map<int, double> tables_weights_gibbs(int table) const;
 
   void transition_alpha();
-
-  CRP* prior() const;
 };

cxx/distributions/dirichlet_categorical.cc

@@ -63,7 +63,3 @@ void DirichletCategorical::transition_hyperparameters() {
   int i = sample_from_logps(logps, prng);
   alpha = alphas[i];
 }
-
-DirichletCategorical* DirichletCategorical::prior() const {
-  return new DirichletCategorical(prng, counts.size());
-}

cxx/distributions/dirichlet_categorical.hh

@@ -33,6 +33,4 @@ class DirichletCategorical : public Distribution<double> {
   double sample();
 
   void transition_hyperparameters();
-
-  DirichletCategorical* prior() const;
 };

cxx/distributions/normal.cc

@@ -105,7 +105,3 @@ void Normal::transition_hyperparameters() {
   m = std::get<2>(hypers[i]);
   s = std::get<3>(hypers[i]);
 }
-
-Normal* Normal::prior() const {
-  return new Normal(prng);
-}

cxx/distributions/normal.hh

@@ -60,8 +60,6 @@ class Normal : public Distribution<double> {
 
   void transition_hyperparameters();
 
-  Normal* prior() const;
-
   // Disable copying.
   Normal& operator=(const Normal&) = delete;
   Normal(const Normal&) = delete;

cxx/hirm.hh

@@ -11,7 +11,6 @@
 #include "relation.hh"
 #include "util_distribution_variant.hh"
 
-
 // Map from names to T_relation's.
 typedef std::map<std::string, T_relation> T_schema;
 
@@ -212,8 +211,11 @@ class IRM {
         }
         auto v = std::get<
             typename std::remove_reference_t<decltype(*rel)>::ValueType>(value);
-        return rel->clusters.contains(z) ? rel->clusters.at(z)->logp(v) 
-                                         : rel->cluster_prior->logp(v);
+        auto prior =
+            std::get<typename std::remove_reference_t<decltype(*rel)>::DType*>(
+                cluster_prior_from_spec(rel->dist_spec, prng));
+        return rel->clusters.contains(z) ? rel->clusters.at(z)->logp(v)
+                                         : prior->logp(v);
       };
       for (const auto& [r, items, value] : observations) {
         auto g = std::bind(f_logp, std::placeholders::_1, items, value);
@@ -252,7 +254,8 @@ class IRM {
       domain_to_relations.at(d).insert(name);
       doms.push_back(domains.at(d));
     }
-    relations[name] = relation_from_spec(name, relation.distribution_spec, doms, prng);
+    relations[name] =
+        relation_from_spec(name, relation.distribution_spec, doms, prng);
     schema[name] = relation;
   }
 
@@ -300,7 +303,7 @@ class HIRM {
     }
   }
 
-  void incorporate(const std::string& r, const T_items& items, 
+  void incorporate(const std::string& r, const T_items& items,
                    const ObservationVariant& value) {
     IRM* irm = relation_to_irm(r);
     irm->incorporate(r, items, value);

cxx/relation.hh

@@ -44,8 +44,8 @@ class Relation {
   const std::string name;
   // Relation spec.
   T_relation trel;
-  // Distribution prior over the relation's codomain.
-  const DistributionType* cluster_prior;
+  // Distribution spec over the relation's codomain.
+  const DistributionSpec dist_spec;
   // list of domain pointers
   const std::vector<Domain*> domains;
   // map from cluster multi-index to Distribution pointer
@@ -63,13 +63,10 @@ class Relation {
 
   Relation(const std::string& name, const DistributionSpec& dist_spec,
            const std::vector<Domain*>& domains, std::mt19937* prng)
-      : name(name), domains(domains) {
+      : name(name), dist_spec(dist_spec), domains(domains) {
     assert(!domains.empty());
     assert(!name.empty());
     this->prng = prng;
-    DistributionVariant cluster_prior_var =
-        cluster_prior_from_spec(dist_spec, prng);
-    cluster_prior = std::get<DistributionType*>(cluster_prior_var);
     for (const Domain* const d : domains) {
       this->data_r[d->name] =
           std::unordered_map<T_item, std::unordered_set<T_items, H_items>>();
@@ -104,7 +101,8 @@ class Relation {
       //      Cannot use clusters[z] because BetaBernoulli
       //      does not have a default constructor, whereas operator[]
       //      calls default constructor when the key does not exist.
-      clusters[z] = cluster_prior->prior();
+      clusters[z] =
+          std::get<DistributionType*>(cluster_prior_from_spec(dist_spec, prng));
     }
     clusters.at(z)->incorporate(value);
   }
@@ -243,8 +241,9 @@ class Relation {
     T_items z =
         get_cluster_assignment_gibbs(items_list[0], domain, item, table);
 
-    DistributionType* cluster =
-        clusters.contains(z) ? clusters.at(z) : cluster_prior->prior();
+    DistributionType* prior =
+        std::get<DistributionType*>(cluster_prior_from_spec(dist_spec, prng));
+    DistributionType* cluster = clusters.contains(z) ? clusters.at(z) : prior;
     double logp0 = cluster->logp_score();
     for (const T_items& items : items_list) {
       // assert(z == get_cluster_assignment_gibbs(items, domain, item, table));
@@ -326,8 +325,10 @@ class Relation {
         z.push_back(zi);
         logp_w += wi;
       }
-      double logp_z = clusters.contains(z) ? clusters.at(z)->logp(value)
-                                           : cluster_prior->logp(value);
+      DistributionType* prior =
+          std::get<DistributionType*>(cluster_prior_from_spec(dist_spec, prng));
+      DistributionType* cluster = clusters.contains(z) ? clusters.at(z) : prior;
+      double logp_z = cluster->logp(value);
       double logp_zw = logp_z + logp_w;
       logps.push_back(logp_zw);
     }
@@ -360,7 +361,8 @@ class Relation {
       T_items z_new = get_cluster_assignment_gibbs(items, domain, item, table);
       if (!clusters.contains(z_new)) {
         // Move to fresh cluster.
-        clusters[z_new] = cluster_prior->prior();
+        clusters[z_new] = std::get<DistributionType*>(
+            cluster_prior_from_spec(dist_spec, prng));
         clusters.at(z_new)->incorporate(x);
       } else {
         // Move to existing cluster.