Improve knn search performance (#138)

tzaeschke · Mar 8, 2023 · 731a21e · 731a21e
1 parent 19a5da2
commit 731a21e
Show file tree

Hide file tree

Showing 25 changed files with 1,085 additions and 101 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -15,6 +15,8 @@ and this project adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.
 - Changed `bpt_vectot` to use `std::destroy` i.o. default dstr. [#132](https://github.com/tzaeschke/phtree-cpp/pull/132)
 - Moved B+trees into own namespace. [#131](https://github.com/tzaeschke/phtree-cpp/pull/131)
 - Moved some stuff in `common` into `nemaspace detail`. [#129](https://github.com/tzaeschke/phtree-cpp/issues/129)
+- Improved kNN search implementation. This also deprecates the post-increment iterator.
+  [#118](https://github.com/tzaeschke/phtree-cpp/issues/118)
 
 ### Fixed
 - Replaced deprecated `<assert.h>` imports with `<cassert>`. [#134](https://github.com/tzaeschke/phtree-cpp/pull/134)

diff --git a/benchmark/BUILD b/benchmark/BUILD
@@ -248,6 +248,21 @@ cc_binary(
     ],
 )
 
+cc_binary(
+    name = "knn_mm_d_benchmark",
+    testonly = True,
+    srcs = [
+        "knn_mm_d_benchmark.cc",
+    ],
+    linkstatic = True,
+    deps = [
+        ":benchmark",
+        "//:phtree",
+        "@gbenchmark//:benchmark",
+        "@spdlog",
+    ],
+)
+
 cc_binary(
     name = "query_benchmark",
     testonly = True,

diff --git a/benchmark/bpt_insert_benchmark.cc b/benchmark/bpt_insert_benchmark.cc
@@ -1,5 +1,5 @@
 /*
-* Copyright 2022-2023 Tilmann Zäschke
+ * Copyright 2022-2023 Tilmann Zäschke
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -30,6 +30,7 @@ const int GLOBAL_MAX = 10000;
 enum Scenario {
     MAP,
     MULTIMAP,
+    MULTIMAP2,
     HASH_MAP,
     STD_MAP,
     STD_MULTIMAP,
@@ -118,7 +119,22 @@ void IndexBenchmark<DIM, TYPE>::SetupWorld(benchmark::State& state) {
 template <size_t DIM, Scenario TYPE>
 void IndexBenchmark<DIM, TYPE>::Insert(benchmark::State& state, Index& tree) {
     switch (TYPE) {
-    default: {
+    case MAP: {
+        for (size_t i = 0; i < num_entities_; ++i) {
+            tree.emplace(points_[i][0], (payload_t)i);
+        }
+        break;
+    }
+    case MULTIMAP:
+    case MULTIMAP2:
+    case STD_MULTIMAP: {
+        for (size_t i = 0; i < num_entities_; ++i) {
+            tree.emplace(points_[i][0], (payload_t)i);
+        }
+        break;
+    }
+    case HASH_MAP:
+    case STD_MAP: {
         for (size_t i = 0; i < num_entities_; ++i) {
             tree.emplace(points_[i][0], (payload_t)i);
         }

diff --git a/benchmark/hd_erase_d_benchmark.cc b/benchmark/hd_erase_d_benchmark.cc
@@ -34,7 +34,7 @@ using payload_t = std::uint32_t;
 template <dimension_t DIM>
 class IndexBenchmark {
   public:
-    IndexBenchmark(benchmark::State& state);
+    explicit IndexBenchmark(benchmark::State& state);
     void Benchmark(benchmark::State& state);
 
   private:

diff --git a/benchmark/knn_d_benchmark.cc b/benchmark/knn_d_benchmark.cc
@@ -86,8 +86,6 @@ void IndexBenchmark<DIM>::SetupWorld(benchmark::State& state) {
         tree_.emplace(points_[i], (int)i);
     }
 
-    state.counters["total_result_count"] = benchmark::Counter(0);
-    state.counters["total_query_count"] = benchmark::Counter(0);
     state.counters["query_rate"] = benchmark::Counter(0, benchmark::Counter::kIsRate);
     state.counters["result_rate"] = benchmark::Counter(0, benchmark::Counter::kIsRate);
     state.counters["avg_result_count"] = benchmark::Counter(0, benchmark::Counter::kAvgIterations);
@@ -104,8 +102,6 @@ void IndexBenchmark<DIM>::QueryWorld(benchmark::State& state, PhPointD<DIM>& cen
         ++n;
     }
 
-    state.counters["total_query_count"] += 1;
-    state.counters["total_result_count"] += n;
     state.counters["query_rate"] += 1;
     state.counters["result_rate"] += n;
     state.counters["avg_result_count"] += n;
@@ -140,6 +136,12 @@ BENCHMARK_CAPTURE(PhTree3D, KNN_CU_10_of_10K, TestGenerator::CUBE, 10000, 10)
 BENCHMARK_CAPTURE(PhTree3D, KNN_CU_10_of_1M, TestGenerator::CUBE, 1000000, 10)
     ->Unit(benchmark::kMillisecond);
 
+BENCHMARK_CAPTURE(PhTree3D, KNN_CU_100_of_10K, TestGenerator::CUBE, 10000, 100)
+    ->Unit(benchmark::kMillisecond);
+
+BENCHMARK_CAPTURE(PhTree3D, KNN_CU_100_of_1M, TestGenerator::CUBE, 1000000, 100)
+    ->Unit(benchmark::kMillisecond);
+
 // index type, scenario name, data_type, num_entities, query_result_size
 // PhTree 3D CLUSTER
 BENCHMARK_CAPTURE(PhTree3D, KNN_CL_1_of_10K, TestGenerator::CLUSTER, 10000, 1)
@@ -154,4 +156,10 @@ BENCHMARK_CAPTURE(PhTree3D, KNN_CL_10_of_10K, TestGenerator::CLUSTER, 10000, 10)
 BENCHMARK_CAPTURE(PhTree3D, KNN_CL_10_of_1M, TestGenerator::CLUSTER, 1000000, 10)
     ->Unit(benchmark::kMillisecond);
 
+BENCHMARK_CAPTURE(PhTree3D, KNN_CL_100_of_10K, TestGenerator::CLUSTER, 10000, 100)
+    ->Unit(benchmark::kMillisecond);
+
+BENCHMARK_CAPTURE(PhTree3D, KNN_CL_100_of_1M, TestGenerator::CLUSTER, 1000000, 100)
+    ->Unit(benchmark::kMillisecond);
+
 BENCHMARK_MAIN();