calculate EstimateElementSize accroding to the stored vector size

add tests

Author: meiravgri

cmake/san.cmake

@@ -7,7 +7,7 @@ if (USE_ASAN OR USE_MSAN)
 	set(CMAKE_LINKER "${CMAKE_C_COMPILER}")
 
 	if (USE_ASAN)
-		set(CLANG_SAN_FLAGS "-fno-omit-frame-pointer -fsanitize=address")
+		set(CLANG_SAN_FLAGS "-fno-omit-frame-pointer -fsanitize=address -fsized-deallocation")
 
 	elseif (USE_MSAN)
 		set(CLANG_SAN_FLAGS "-fno-omit-frame-pointer -fsanitize=memory -fsanitize-memory-track-origins=2")

src/VecSim/index_factories/brute_force_factory.cpp

@@ -127,12 +127,9 @@ size_t EstimateInitialSize(const BFParams *params, bool is_normalized) {
     return est;
 }
 
-/*** ======================================================= */
-// TODO: replace params->dim * VecSimType_sizeof(params->type) with getStoredDataSize!!!
-/** ======================================================= */
-
 size_t EstimateElementSize(const BFParams *params) {
     // counting the vector size + idToLabel entry + LabelToIds entry (map reservation)
-    return params->dim * VecSimType_sizeof(params->type) + sizeof(labelType) + sizeof(void *);
+    return VecSimParams_GetStoredDataSize(params->type, params->dim, params->metric) +
+           sizeof(labelType) + sizeof(void *);
 }
 }; // namespace BruteForceFactory

src/VecSim/index_factories/hnsw_factory.cpp

@@ -126,7 +126,8 @@ size_t EstimateElementSize(const HNSWParams *params) {
     size_t elementGraphDataSize = sizeof(ElementGraphData) + sizeof(idType) * M * 2;
 
     size_t size_total_data_per_element =
-        elementGraphDataSize + params->dim * VecSimType_sizeof(params->type);
+        elementGraphDataSize +
+        VecSimParams_GetStoredDataSize(params->type, params->dim, params->metric);
 
     // when reserving space for new labels in the lookup hash table, each entry is a pointer to a
     // label node (bucket).

tests/unit/test_int8.cpp

@@ -301,6 +301,24 @@ TEST_F(INT8TieredTest, elementSizeEstimation) {
     EXPECT_NO_FATAL_FAILURE(element_size_test(hnsw_params));
 }
 
+// ======= Cosine size estimation (stored vectors include additional space for norm)
+TEST_F(INT8HNSWTest, elementSizeEstimation_Cosine) {
+    size_t M = 64;
+
+    HNSWParams params = {.dim = 4, .metric = VecSimMetric_Cosine, .M = M};
+    EXPECT_NO_FATAL_FAILURE(element_size_test(params));
+}
+
+TEST_F(INT8BruteForceTest, elementSizeEstimation_Cosine) {
+    BFParams params = {.dim = 4, .metric = VecSimMetric_Cosine};
+    EXPECT_NO_FATAL_FAILURE(element_size_test(params));
+}
+
+TEST_F(INT8TieredTest, elementSizeEstimation_Cosine) {
+    size_t M = 64;
+    HNSWParams hnsw_params = {.dim = 4, .metric = VecSimMetric_Cosine, .M = M};
+    EXPECT_NO_FATAL_FAILURE(element_size_test(hnsw_params));
+}
 /* ---------------------------- Functionality tests ---------------------------- */
 
 template <typename params_t>

tests/unit/test_uint8.cpp

@@ -300,6 +300,25 @@ TEST_F(UINT8TieredTest, elementSizeEstimation) {
     EXPECT_NO_FATAL_FAILURE(element_size_test(hnsw_params));
 }
 
+// ======= Cosine size estimation (stored vectors include additional space for norm)
+TEST_F(UINT8HNSWTest, elementSizeEstimation_Cosine) {
+    size_t M = 64;
+
+    HNSWParams params = {.dim = 4, .metric = VecSimMetric_Cosine, .M = M};
+    EXPECT_NO_FATAL_FAILURE(element_size_test(params));
+}
+
+TEST_F(UINT8BruteForceTest, elementSizeEstimation_Cosine) {
+    BFParams params = {.dim = 4, .metric = VecSimMetric_Cosine};
+    EXPECT_NO_FATAL_FAILURE(element_size_test(params));
+}
+
+TEST_F(UINT8TieredTest, elementSizeEstimation_Cosine) {
+    size_t M = 64;
+    HNSWParams hnsw_params = {.dim = 4, .metric = VecSimMetric_Cosine, .M = M};
+    EXPECT_NO_FATAL_FAILURE(element_size_test(hnsw_params));
+}
+
 /* ---------------------------- Functionality tests ---------------------------- */
 
 template <typename params_t>