Reorganize benchmarks/dbscan\

* Rename benchmarks/{dbscan => clustering}

* Reorganize clustering benchmarks

* Remove data converter

* Instantiate data loader only once

* benchmarks/clustering => benchmarks/cluster

* Add missing copyright header

* Maybe fix?

* Address review comments

* Fix else-after-return check

benchmarks/CMakeLists.txt

@@ -30,7 +30,7 @@ else()
 ENDIF()
 
 add_subdirectory(brute_force_vs_bvh)
-add_subdirectory(dbscan)
+add_subdirectory(cluster)
 add_subdirectory(execution_space_instances)
 if(NOT WIN32)
   # FIXME: for now, skip the benchmarks using Google benchmark

benchmarks/cluster/CMakeLists.txt

@@ -0,0 +1,23 @@
+add_library(cluster_benchmark_helpers
+  data.cpp
+  print_timers.cpp
+)
+target_link_libraries(cluster_benchmark_helpers PRIVATE ArborX::ArborX)
+
+set(input_file "input.txt")
+configure_file(${CMAKE_CURRENT_SOURCE_DIR}/${input_file} ${CMAKE_CURRENT_BINARY_DIR}/${input_file} COPYONLY)
+
+add_executable(ArborX_Benchmark_DBSCAN.exe dbscan.cpp)
+target_include_directories(ArborX_Benchmark_DBSCAN.exe PRIVATE ${CMAKE_CURRENT_SOURCE_DIR})
+target_link_libraries(ArborX_Benchmark_DBSCAN.exe ArborX::ArborX Boost::program_options cluster_benchmark_helpers)
+add_test(NAME ArborX_Benchmark_DBSCAN COMMAND ArborX_Benchmark_DBSCAN.exe --filename=${input_file} --eps=1.4 --verify)
+
+add_executable(ArborX_Benchmark_MST.exe mst.cpp)
+target_include_directories(ArborX_Benchmark_MST.exe PRIVATE ${CMAKE_CURRENT_SOURCE_DIR})
+target_link_libraries(ArborX_Benchmark_MST.exe ArborX::ArborX Boost::program_options cluster_benchmark_helpers)
+add_test(NAME ArborX_Benchmark_HDBSCAN COMMAND ArborX_Benchmark_HDBSCAN.exe --filename=${input_file})
+
+add_executable(ArborX_Benchmark_HDBSCAN.exe hdbscan.cpp)
+target_include_directories(ArborX_Benchmark_HDBSCAN.exe PRIVATE ${CMAKE_CURRENT_SOURCE_DIR})
+target_link_libraries(ArborX_Benchmark_HDBSCAN.exe ArborX::ArborX Boost::program_options cluster_benchmark_helpers)
+add_test(NAME ArborX_Benchmark_MST COMMAND ArborX_Benchmark_MST.exe --filename=${input_file})

benchmarks/cluster/data.cpp

@@ -0,0 +1,62 @@
+/****************************************************************************
+ * Copyright (c) 2025, ArborX authors                                       *
+ * All rights reserved.                                                     *
+ *                                                                          *
+ * This file is part of the ArborX library. ArborX is                       *
+ * distributed under a BSD 3-clause license. For the licensing terms see    *
+ * the LICENSE file in the top-level directory.                             *
+ *                                                                          *
+ * SPDX-License-Identifier: BSD-3-Clause                                    *
+ ****************************************************************************/
+#include "data.hpp"
+
+#include <Kokkos_Core.hpp>
+
+#include <fstream>
+#include <string>
+
+#include "data_timpl.hpp"
+
+namespace ArborXBenchmark
+{
+
+// Explicit instantiations
+using MemorySpace = typename Kokkos::DefaultExecutionSpace::memory_space;
+#define INSTANTIATE_LOADER(DIM)                                                \
+  template Kokkos::View<ArborX::Point<DIM> *, MemorySpace>                     \
+  loadData<DIM, MemorySpace>(ArborXBenchmark::Parameters const &);
+INSTANTIATE_LOADER(2)
+INSTANTIATE_LOADER(3)
+INSTANTIATE_LOADER(4)
+INSTANTIATE_LOADER(5)
+INSTANTIATE_LOADER(6)
+#undef INSTANTIATE_LOADER
+
+int getDataDimension(std::string const &filename, bool binary)
+{
+  std::ifstream input;
+  if (!binary)
+    input.open(filename);
+  else
+    input.open(filename, std::ifstream::binary);
+  if (!input.good())
+    throw std::runtime_error("Error reading file \"" + filename + "\"");
+
+  int num_points;
+  int dim;
+  if (!binary)
+  {
+    input >> num_points;
+    input >> dim;
+  }
+  else
+  {
+    input.read(reinterpret_cast<char *>(&num_points), sizeof(int));
+    input.read(reinterpret_cast<char *>(&dim), sizeof(int));
+  }
+  input.close();
+
+  return dim;
+}
+
+} // namespace ArborXBenchmark

benchmarks/cluster/data.hpp

@@ -0,0 +1,31 @@
+/****************************************************************************
+ * Copyright (c) 2025, ArborX authors                                       *
+ * All rights reserved.                                                     *
+ *                                                                          *
+ * This file is part of the ArborX library. ArborX is                       *
+ * distributed under a BSD 3-clause license. For the licensing terms see    *
+ * the LICENSE file in the top-level directory.                             *
+ *                                                                          *
+ * SPDX-License-Identifier: BSD-3-Clause                                    *
+ ****************************************************************************/
+#ifndef ARBORX_BENCHMARK_DATA_HPP
+#define ARBORX_BENCHMARK_DATA_HPP
+
+#include <ArborX_Point.hpp>
+
+#include <Kokkos_Core.hpp>
+
+#include "parameters.hpp"
+
+namespace ArborXBenchmark
+{
+
+int getDataDimension(std::string const &filename, bool binary);
+
+template <int DIM, typename MemorySpace>
+Kokkos::View<ArborX::Point<DIM> *, MemorySpace>
+loadData(ArborXBenchmark::Parameters const &params);
+
+} // namespace ArborXBenchmark
+
+#endif

benchmarks/dbscan/data.hpp → benchmarks/cluster/data_timpl.hpp

@@ -9,16 +9,27 @@
  * SPDX-License-Identifier: BSD-3-Clause                                    *
  ****************************************************************************/
 
-#ifndef DATA_HPP
-#define DATA_HPP
+#ifndef ARBORX_BENCHMARK_DATA_TIMPL_HPP
+#define ARBORX_BENCHMARK_DATA_TIMPL_HPP
 
 #include <ArborX_Point.hpp>
 #include <misc/ArborX_Exception.hpp>
 
+#include <Kokkos_Core.hpp>
+
+#include <cassert>
+#include <fstream>
 #include <iostream>
+#include <iterator>
 #include <random>
 #include <vector>
 
+#include "data.hpp"
+#include "parameters.hpp"
+
+namespace ArborXBenchmark
+{
+
 using ArborX::Point;
 
 template <int DIM>
@@ -298,4 +309,42 @@ std::vector<Point<DIM>> GanTao(int n, bool variable_density = false,
   return points;
 }
 
+template <typename... P, typename T>
+auto vec2view(std::vector<T> const &in, std::string const &label = "")
+{
+  Kokkos::View<T *, P...> out(
+      Kokkos::view_alloc(label, Kokkos::WithoutInitializing), in.size());
+  Kokkos::deep_copy(out, Kokkos::View<T const *, Kokkos::HostSpace,
+                                      Kokkos::MemoryTraits<Kokkos::Unmanaged>>{
+                             in.data(), in.size()});
+  return out;
+}
+
+template <int DIM, typename MemorySpace>
+Kokkos::View<ArborX::Point<DIM> *, MemorySpace>
+loadData(ArborXBenchmark::Parameters const &params)
+{
+  if (!params.filename.empty())
+  {
+    // Read in data
+    printf("filename          : %s [%s, max_pts = %d]\n",
+           params.filename.c_str(), (params.binary ? "binary" : "text"),
+           params.max_num_points);
+    printf("samples           : %d\n", params.num_samples);
+    return vec2view<MemorySpace>(loadData<DIM>(params.filename, params.binary,
+                                               params.max_num_points,
+                                               params.num_samples),
+                                 "Benchmark::primitives");
+  }
+
+  // Generate data
+  int dim = params.dim;
+  printf("generator         : n = %d, dim = %d, density = %s\n", params.n, dim,
+         (params.variable_density ? "variable" : "constant"));
+  return vec2view<MemorySpace>(GanTao<DIM>(params.n, params.variable_density),
+                               "Benchmark::primitives");
+}
+
+} // namespace ArborXBenchmark
+
 #endif