Made into class

gtsam/base/WeightedSampler.h

@@ -16,6 +16,9 @@
  * @date    May 2019
  **/
 
+#pragma once
+
+#include <cmath>
 #include <queue>
 #include <random>
 #include <stdexcept>
@@ -25,20 +28,36 @@
 namespace gtsam {
 /*
  * Fast sampling without replacement.
+ * Example usage:
+ *   std::mt19937 rng(42);
+ *   WeightedSampler<std::mt19937> sampler(&rng);
+ *   auto samples = sampler.sampleWithoutReplacement(5, weights);
  */
-template <class Engine>
+template <class Engine = std::mt19937>
-std::vector<size_t> sampleWithoutReplacement(Engine& rng, size_t s,
+class WeightedSampler {
+ private:
+  Engine* engine_;  // random number generation engine
+
+ public:
+  /**
+   * Construct from random number generation engine
+   * We only store a pointer to it.
+   */
+  explicit WeightedSampler(Engine* engine) : engine_(engine) {}
+
+  std::vector<size_t> sampleWithoutReplacement(size_t numSamples,
                                                std::vector<double> weights) {
-  // Implementation adapted from paper at
+    // Implementation adapted from code accompanying paper at
     // https://www.ethz.ch/content/dam/ethz/special-interest/baug/ivt/ivt-dam/vpl/reports/1101-1200/ab1141.pdf
     const size_t n = weights.size();
-  if (n < s) {
+    if (n < numSamples) {
-    throw std::runtime_error("s must be smaller than weights.size()");
+      throw std::runtime_error(
+          "numSamples must be smaller than weights.size()");
     }
 
-  // Return empty array if s==0
+    // Return empty array if numSamples==0
-  std::vector<size_t> result(s);
+    std::vector<size_t> result(numSamples);
-  if (s == 0) return result;
+    if (numSamples == 0) return result;
 
     // Step 1: The first m items of V are inserted into reservoir
     // Step 2: For each item v_i ∈ reservoir: Calculate a key k_i = u_i^(1/w),
@@ -47,10 +66,10 @@ std::vector<size_t> sampleWithoutReplacement(Engine& rng, size_t s,
     //  reservoir is a priority queue that pops the *maximum* elements)
     std::priority_queue<std::pair<double, size_t> > reservoir;
 
-  static const double kexp1 = exp(1.0);
+    static const double kexp1 = std::exp(1.0);
-  for (auto iprob = weights.begin(); iprob != weights.begin() + s; ++iprob) {
+    for (auto it = weights.begin(); it != weights.begin() + numSamples; ++it) {
-    double k_i = kexp1 / *iprob;
+      const double k_i = kexp1 / *it;
-    reservoir.push(std::make_pair(k_i, iprob - weights.begin() + 1));
+      reservoir.push(std::make_pair(k_i, it - weights.begin() + 1));
     }
 
     // Step 4: Repeat Steps 5–10 until the population is exhausted
@@ -60,37 +79,38 @@ std::vector<size_t> sampleWithoutReplacement(Engine& rng, size_t s,
       // Step 10: The new threshold T w is the new minimum key of reservoir
       const std::pair<double, size_t>& T_w = reservoir.top();
 
-    // Incrementing iprob is part of Step 7
+      // Incrementing it is part of Step 7
-    for (auto iprob = weights.begin() + s; iprob != weights.end(); ++iprob) {
+      for (auto it = weights.begin() + numSamples; it != weights.end(); ++it) {
         // Step 5: Let r = random(0, 1) and X_w = log(r) / log(T_w)
         // (Modification: Use e = -exp(1) instead of log(r))
-      double X_w = kexp1 / T_w.first;
+        const double X_w = kexp1 / T_w.first;
 
-      // Step 6: From the current item v_c skip items until item v_i, such that:
+        // Step 6: From the current item v_c skip items until item v_i, such
+        // that:
         double w = 0.0;
 
         // Step 7: w_c + w_{c+1} + ··· + w_{i−1} < X_w <= w_c + w_{c+1} + ··· +
         // w_{i−1} + w_i
-      for (; iprob != weights.end(); ++iprob) {
+        for (; it != weights.end(); ++it) {
-        w += *iprob;
+          w += *it;
           if (X_w <= w) break;
         }
 
         // Step 7: No such item, terminate
-      if (iprob == weights.end()) break;
+        if (it == weights.end()) break;
 
-      // Step 9: Let t_w = T_w^{w_i}, r_2 = random(t_w, 1) and v_i’s key: k_i =
+        // Step 9: Let t_w = T_w^{w_i}, r_2 = random(t_w, 1) and v_i’s key: k_i
-      // (r_2)^{1/w_i} (Mod: Let t_w = log(T_w) * {w_i}, e_2 =
+        // = (r_2)^{1/w_i} (Mod: Let t_w = log(T_w) * {w_i}, e_2 =
         // log(random(e^{t_w}, 1)) and v_i’s key: k_i = -e_2 / w_i)
-      double t_w = -T_w.first * *iprob;
+        const double t_w = -T_w.first * *it;
-      std::uniform_real_distribution<double> randomAngle(exp(t_w), 1.0);
+        std::uniform_real_distribution<double> randomAngle(std::exp(t_w), 1.0);
-      double e_2 = log(randomAngle(rng));
+        const double e_2 = std::log(randomAngle(*engine_));
-      double k_i = -e_2 / *iprob;
+        const double k_i = -e_2 / *it;
 
-      // Step 8: The item in reservoir with the minimum key is replaced by item
+        // Step 8: The item in reservoir with the minimum key is replaced by
-      // v_i
+        // item v_i
         reservoir.pop();
-      reservoir.push(std::make_pair(k_i, iprob - weights.begin() + 1));
+        reservoir.push(std::make_pair(k_i, it - weights.begin() + 1));
       }
     }
 
@@ -113,4 +133,5 @@ std::vector<size_t> sampleWithoutReplacement(Engine& rng, size_t s,
 
     return result;
   }
+};  // namespace gtsam
 }  // namespace gtsam

gtsam/base/tests/testWeightedSampler.cpp

@@ -27,8 +27,9 @@ using namespace gtsam;
 
 TEST(WeightedSampler, sampleWithoutReplacement) {
   vector<double> weights{1, 2, 3, 4, 3, 2, 1};
-  mt19937 rng(42);
+  std::mt19937 rng(42);
-  auto samples = sampleWithoutReplacement(rng, 5, weights);
+  WeightedSampler<std::mt19937> sampler(&rng);
+  auto samples = sampler.sampleWithoutReplacement(5, weights);
   EXPECT_LONGS_EQUAL(5, samples.size());
 }