Merge pull request #56 from MChatzakis/feature/sing-dtw

Feature/sing dtw

Author: FrancescaDelGaudio

CMakeLists.txt

@@ -274,53 +274,53 @@ if(BUILD_PYTHON)
         message(STATUS "Subdirectory 'extern/pybind11' added.")
     endif()
 
-    # Build the Python module
-    pybind11_add_module(daisy pybinds/setup.cpp)
+    # Build the Python module as daisy._core
+    pybind11_add_module(daisy_core pybinds/setup.cpp)
+    set_target_properties(daisy_core PROPERTIES
+        OUTPUT_NAME "_core"
+        PREFIX ""  # no 'lib' prefix, produce _core.cpython-*.so
+        LIBRARY_OUTPUT_DIRECTORY "${PROJECT_SOURCE_DIR}/daisy"
+    )
     if(DEBUG_MSG)
-        message(STATUS "Python module 'daisy' created from pybinds/setup.cpp.")
+        message(STATUS "Python module 'daisy._core' created from pybinds/setup.cpp.")
     endif()
 
-    target_link_libraries(daisy PRIVATE dino_lib)
+    target_link_libraries(daisy_core PRIVATE dino_lib)
     if(DEBUG_MSG)
-        message(STATUS "Python module 'daisy' linked to 'dino_lib'.")
+        message(STATUS "Python module 'daisy._core' linked to 'dino_lib'.")
     endif()
 
     # Pass BUILD_SING_ENABLED definition to Python module
     if (BUILD_SING_AVAILABLE)
-        target_compile_definitions(daisy PRIVATE SING_CUDA_ENABLED=1)
+        target_compile_definitions(daisy_core PRIVATE SING_CUDA_ENABLED=1)
         if(DEBUG_MSG)
-            message(STATUS "Python module 'daisy' compiled with SING_CUDA_ENABLED=1.")
+            message(STATUS "Python module 'daisy._core' compiled with SING_CUDA_ENABLED=1.")
         endif()
     else()
-        target_compile_definitions(daisy PRIVATE SING_CUDA_ENABLED=0)
+        target_compile_definitions(daisy_core PRIVATE SING_CUDA_ENABLED=0)
         if(DEBUG_MSG)
-            message(STATUS "Python module 'daisy' compiled with SING_CUDA_ENABLED=0.")
+            message(STATUS "Python module 'daisy._core' compiled with SING_CUDA_ENABLED=0.")
         endif()
     endif()
 
+    # Link Python module 'daisy._core' with MPI when Odyssey is available
     if (BUILD_ODYSSEY_AVAILABLE)
         if(DEBUG_MSG)
-            message(STATUS "Linking Python module 'daisy' with MPI.")
+            message(STATUS "Linking Python module 'daisy._core' with MPI.")
         endif()
-        target_link_libraries(daisy PRIVATE ${MPI_CXX_LIBRARIES})
-        target_include_directories(daisy PRIVATE ${MPI_CXX_INCLUDE_DIRS})
+        target_link_libraries(daisy_core PRIVATE ${MPI_CXX_LIBRARIES})
+        target_include_directories(daisy_core PRIVATE ${MPI_CXX_INCLUDE_DIRS})
         if(DEBUG_MSG)
             message(DEBUG "MPI Libraries: ${MPI_CXX_LIBRARIES}")
             message(DEBUG "MPI Includes: ${MPI_CXX_INCLUDE_DIRS}")
         endif()
     else()
         if(DEBUG_MSG)
-            message(STATUS "BUILD_ODYSSEY_AVAILABLE is FALSE. Python module 'daisy' will NOT be linked with MPI.")
+            message(STATUS "BUILD_ODYSSEY_AVAILABLE is FALSE. Python module 'daisy._core' will NOT be linked with MPI.")
         endif()
-    endif()    
-
-    # Set output directory for the .so file (demos/ so that import daisy finds daisy.*.so when run from demos)
-    set_target_properties(daisy PROPERTIES
-        LIBRARY_OUTPUT_DIRECTORY "${CMAKE_SOURCE_DIR}/demos"
-    )
-    if(DEBUG_MSG)
-        message(STATUS "Python module 'daisy' output directory set to: ${CMAKE_SOURCE_DIR}/demos")
     endif()
+
+    # (Python module 'daisy._core' already configured to output into ${PROJECT_SOURCE_DIR}/daisy above)
 else()
     if(DEBUG_MSG)
         message(STATUS "BUILD_PYTHON is OFF. Skipping Python bindings.")

README.md

@@ -191,4 +191,3 @@ The logo of DaiSy was designed by [Eva Chamilaki](https://evachamilaki.github.io
 
 
 
-

demos/CMakeLists.txt

@@ -202,6 +202,28 @@ if(BUILD_DEMO)
         message(STATUS "Include directories added for demo_Messi_DTW.")
     endif()
 
+    # ////// SING DTW //////
+    if(BUILD_SING_AVAILABLE)
+        if(DEBUG_MSG)
+            message(STATUS "---")
+            message(STATUS "## Demo: Sing DTW")
+            message(STATUS "Attempting to add executable: demo_Sing_DTW")
+        endif()
+        add_executable(demo_Sing_DTW demo_Sing_DTW.cpp)
+        target_link_libraries(demo_Sing_DTW PRIVATE dino_lib commons_lib)
+        target_include_directories(demo_Sing_DTW PRIVATE
+            ${CMAKE_CURRENT_SOURCE_DIR}/../lib
+            ${CMAKE_CURRENT_SOURCE_DIR}/../commons
+        )
+        if(DEBUG_MSG)
+            message(STATUS "Executable demo_Sing_DTW added and configured (CUDA/Sing available).")
+        endif()
+    else()
+        if(DEBUG_MSG)
+            message(STATUS "BUILD_SING_AVAILABLE is FALSE. demo_Sing_DTW will NOT be built.")
+        endif()
+    endif()
+
 # ////// ODYSSEY L2Square //////
     if(DEBUG_MSG)
         message(STATUS "---")
@@ -325,25 +347,23 @@ if(BUILD_DEMO)
     # ////// SING L2Square //////
     if(DEBUG_MSG)
         message(STATUS "---")
-        message(STATUS "## Demo: Sing")
+        message(STATUS "## Demo: Sing L2Square")
     endif()
 
-    if(SING_CUDA_AVAILABLE)
+    if(BUILD_SING_AVAILABLE)
         if(DEBUG_MSG)
-            message(STATUS "SING_CUDA_AVAILABLE is TRUE. Building demo_Sing_L2Square.")
+            message(STATUS "BUILD_SING_AVAILABLE is TRUE. Building demo_Sing_L2Square.")
         endif()
 
         add_executable(demo_Sing_L2Square demo_Sing_L2Square.cpp)
-
-        target_link_libraries(demo_Sing_L2Square PRIVATE dino_lib commons_lib CUDA::cudart)
-
+        target_link_libraries(demo_Sing_L2Square PRIVATE dino_lib commons_lib)
         target_include_directories(demo_Sing_L2Square PRIVATE
             ${CMAKE_CURRENT_SOURCE_DIR}/../lib
             ${CMAKE_CURRENT_SOURCE_DIR}/../commons
         )
     else()
         if(DEBUG_MSG)
-            message(STATUS "SING_CUDA_AVAILABLE is FALSE. demo_Sing_L2Square will NOT be built.")
+            message(STATUS "BUILD_SING_AVAILABLE is FALSE. demo_Sing_L2Square will NOT be built.")
         endif()
     endif()
 

demos/demo_Messi_DTW.cpp

@@ -16,7 +16,7 @@ int main(){
     printf("Loaded %llu database points and %llu query points with dimension %llu\n", n_database, n_query, dim);
 
     daisy::Messi messi_search(daisy::DistanceType::DTW);
-    messi_search.setNumThreads(4);
+    messi_search.setNumThreads(1);
 
     int warp_window = std::max(1, static_cast<int>(dim * 0.1));
     messi_search.setWarpingWindow(warp_window);  

demos/demo_Messi_DTW.py

@@ -8,9 +8,9 @@
 
 def main():
 
-    n_database = 200000
-    dim = 96
-    n_query = 10
+    n_database = 840000
+    dim = 128
+    n_query = 20
     k = 5
 
     np.random.seed(100)   

demos/demo_Sing_DTW.cpp

@@ -0,0 +1,45 @@
+    #include "../commons/dataloaders.hpp"
+    #include "../lib/daisy.hpp"
+    #include <chrono>
+    #include <algorithm>
+
+    int main() {
+        daisy::idx_t n_database = 200000;
+        unsigned long long dim = 96;
+        unsigned long long n_query = 10;
+        daisy::idx_t k = 1; // current Sing DTW implementation supports only k = 1
+
+        float *database = loadRandomData(n_database, dim, 100, true);
+        float *query = loadRandomData(n_query, dim, 50, true);
+
+        daisy::Sing sing_search(daisy::DistanceType::DTW);
+        sing_search.setNumThreads(1);
+
+        // Sing::buildIndex expects a DataSource*, use InMemoryDataSource as in other components.
+        daisy::InMemoryDataSource data_source(database, n_database, dim);
+        printf("indexing\n");
+        sing_search.buildIndex(&data_source);
+        printf("finished indexing\n");
+
+        daisy::idx_t *I = new daisy::idx_t[n_query * k];
+        float *D = new float[n_query * k];
+        printf("querying\n");
+        sing_search.searchIndex(query, n_query, k, I, D);
+        printf("finished querying\n");
+
+        for (daisy::idx_t i = 0; i < n_query; i++) {
+            printf("Query %llu:\n", i);
+            printf("  Distances: ");
+            for (daisy::idx_t j = 0; j < k; j++) {
+                printf("%f ", D[i * k + j]);
+            }
+        }
+
+        delete[] database;
+        delete[] query;
+        delete[] I;
+        delete[] D;
+
+        return 0;
+    }
+

demos/demo_Sing_DTW.py

@@ -0,0 +1,52 @@
+"""
+Demo Sing DTW — build index from NumPy array and run DTW search.
+Requires Sing to be built with CUDA (SING_CUDA_ENABLED).
+Note: current Sing DTW implementation only supports k = 1.
+"""
+import sys
+import os
+import time
+import numpy as np
+
+sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), "..")))
+
+from daisy import DistanceType, Sing
+
+
+def main():
+    # Small-ish sizes so the demo runs quickly
+    n_database = 50000
+    dim = 128
+    n_query = 5
+    k = 1  # DTW implementation in Sing currently supports only k = 1
+
+    np.random.seed(100)
+    db = np.random.randn(n_database, dim).astype(np.float32)
+    np.random.seed(50)
+    query = np.random.randn(n_query, dim).astype(np.float32)
+    print(f"Loaded {n_database} database points and {n_query} query points with dimension {dim}")
+
+    sing_search = Sing(DistanceType.DTW)
+    # Optional: use multiple threads
+    sing_search.setNumThreads(4)
+
+    sing_search.buildIndex(db)
+    print(f"buildIndex (DTW) done")
+
+    print(f"Starting DTW search (n_query={n_query}, k={k})...")
+    I, D = sing_search.searchIndex(query, k)
+    print(f"Search done)")
+
+    print("\nPython Sing DTW Search Results:")
+    for i in range(n_query):
+        print(f"Query {i}:")
+        print("Distances:", D[i])
+        print("Indices:", I[i])
+        print()
+
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())
+

lib/algos/CMakeLists.txt

@@ -128,9 +128,9 @@ if(DEBUG_MSG)
 endif()
 
 # ////// SING //////
-if (SING_CUDA_AVAILABLE)
+if (BUILD_SING_AVAILABLE)
     if(DEBUG_MSG)
-        message(STATUS "SING_CUDA_AVAILABLE is TRUE. Configuring CUDA for SING components.")
+        message(STATUS "BUILD_SING_AVAILABLE is TRUE. Configuring CUDA for SING components.")
         message(STATUS "Adding CUDA Toolkit include directories to dino_lib...")
     endif()
 
@@ -171,7 +171,7 @@ if (SING_CUDA_AVAILABLE)
     endif()
 else()
     if(DEBUG_MSG)
-        message(STATUS "SING_CUDA_AVAILABLE is FALSE. CUDA configuration for SING skipped.")
+        message(STATUS "BUILD_SING_AVAILABLE is FALSE. CUDA configuration for SING skipped.")
     endif()
     target_compile_definitions(dino_lib PUBLIC SING_CUDA_ENABLED=0)
 endif()