faiss 0.6.2 → 0.6.4

data/vendor/faiss/faiss/IndexIVFSpectralHash.cpp

@@ -28,7 +28,6 @@
 // NOLINTNEXTLINE(facebook-hte-InlineHeader)
 // NOLINTNEXTLINE(facebook-hte-InlineHeader)
 #include <faiss/impl/binary_hamming/IndexIVFSpectralHash_impl.h>
-#include <faiss/utils/hamming_distance/hamming_computer-generic.h>
 #undef THE_SIMD_LEVEL
 
 namespace faiss {
@@ -88,7 +87,7 @@ void IndexIVFSpectralHash::train_encoder(
     if (!vt->is_trained) {
         vt->train(n, x);
     }
-    FAISS_THROW_IF_NOT(!by_residual);
+    FAISS_THROW_IF_MSG(by_residual, "by_residual not supported for this index");
 
     if (threshold_type == Thresh_global) {
         // nothing to do
@@ -185,7 +184,7 @@ void IndexIVFSpectralHash::encode_vectors(
         uint8_t* codes,
         bool include_listnos) const {
     FAISS_THROW_IF_NOT(is_trained);
-    FAISS_THROW_IF_NOT(!by_residual);
+    FAISS_THROW_IF_MSG(by_residual, "by_residual not supported for this index");
     float freq = 2.0 / period;
     size_t coarse_size = include_listnos ? coarse_code_size() : 0;
 
@@ -222,7 +221,7 @@ InvertedListScanner* IndexIVFSpectralHash::get_InvertedListScanner(
         bool store_pairs,
         const IDSelector* sel,
         const IVFSearchParameters*) const {
-    FAISS_THROW_IF_NOT(!sel);
+    FAISS_THROW_IF_MSG(sel, "id selector not supported for this index");
     return with_simd_level([&]<SIMDLevel SL>() {
         return make_spectral_hash_scanner_fixSL<SL>(
                 code_size, this, store_pairs);
@@ -254,8 +253,11 @@ void IndexIVFSpectralHash::replace_vt(IndexPreTransform* encoder, bool own) {
     auto sub_index = dynamic_cast<IndexLSH*>(encoder->index);
     FAISS_THROW_IF_NOT_MSG(sub_index, "final index should be LSH");
     FAISS_THROW_IF_NOT(sub_index->nbits == nbit);
-    FAISS_THROW_IF_NOT(!sub_index->rotate_data);
-    FAISS_THROW_IF_NOT(!sub_index->train_thresholds);
+    FAISS_THROW_IF_MSG(
+            sub_index->rotate_data, "LSH sub-index must not rotate data");
+    FAISS_THROW_IF_MSG(
+            sub_index->train_thresholds,
+            "LSH sub-index thresholds must already be trained");
     replace_vt(encoder->chain[0], own);
 }
 

data/vendor/faiss/faiss/IndexLSH.cpp

@@ -119,8 +119,7 @@ void IndexLSH::search(
         float* distances,
         idx_t* labels,
         const SearchParameters* params) const {
-    FAISS_THROW_IF_NOT_MSG(
-            !params, "search params not supported for this index");
+    FAISS_THROW_IF_MSG(params, "search params not supported for this index");
     FAISS_THROW_IF_NOT(k > 0);
     FAISS_THROW_IF_NOT(is_trained);
     const float* xt = apply_preprocess(n, x);
@@ -151,7 +150,7 @@ void IndexLSH::transfer_thresholds(LinearTransform* vt) {
         vt->b.resize(nbits, 0);
         vt->have_bias = true;
     }
-    FAISS_THROW_IF_NOT(!vt->b.empty());
+    FAISS_THROW_IF_MSG(vt->b.empty(), "bias vector must not be empty");
     for (int i = 0; i < nbits; i++) {
         vt->b[i] -= thresholds[i];
     }

data/vendor/faiss/faiss/IndexLattice.cpp

@@ -26,6 +26,11 @@ IndexLattice::IndexLattice(idx_t d_in, int nsq_in, int scale_nbit_in, int r2)
     lattice_nbit = 0;
     while (!(((uint64_t)1 << lattice_nbit) >= zn_sphere_codec.nv)) {
         lattice_nbit++;
+        FAISS_THROW_IF_NOT_FMT(
+                lattice_nbit < 64,
+                "IndexLattice: nv=%zu too large, lattice code would exceed "
+                "63 bits (likely corrupt r2/dsq)",
+                (size_t)zn_sphere_codec.nv);
     }
 
     int total_nbit = (lattice_nbit + scale_nbit_in) * nsq_in;

data/vendor/faiss/faiss/IndexNNDescent.cpp

@@ -101,8 +101,7 @@ void IndexNNDescent::search(
         float* distances,
         idx_t* labels,
         const SearchParameters* params) const {
-    FAISS_THROW_IF_NOT_MSG(
-            !params, "search params not supported for this index");
+    FAISS_THROW_IF_MSG(params, "search params not supported for this index");
     FAISS_THROW_IF_NOT_MSG(
             storage,
             "Please use IndexNNDescentFlat (or variants) "
@@ -183,11 +182,19 @@ void IndexNNDescent::add(idx_t n, const float* x) {
 
 void IndexNNDescent::reset() {
     nndescent.reset();
+    FAISS_THROW_IF_NOT_MSG(
+            storage,
+            "Please use IndexNNDescentFlat (or variants) "
+            "instead of IndexNNDescent directly");
     storage->reset();
     ntotal = 0;
 }
 
 void IndexNNDescent::reconstruct(idx_t key, float* recons) const {
+    FAISS_THROW_IF_NOT_MSG(
+            storage,
+            "Please use IndexNNDescentFlat (or variants) "
+            "instead of IndexNNDescent directly");
     storage->reconstruct(key, recons);
 }
 

data/vendor/faiss/faiss/IndexNSG.cpp

@@ -62,8 +62,7 @@ void IndexNSG::search(
         float* distances,
         idx_t* labels,
         const SearchParameters* params) const {
-    FAISS_THROW_IF_NOT_MSG(
-            !params, "search params not supported for this index");
+    FAISS_THROW_IF_MSG(params, "search params not supported for this index");
     FAISS_THROW_IF_NOT_MSG(
             storage,
             "Please use IndexNSGFlat (or variants) instead of IndexNSG directly");
@@ -245,12 +244,18 @@ void IndexNSG::add(idx_t n, const float* x) {
 
 void IndexNSG::reset() {
     nsg.reset();
+    FAISS_THROW_IF_NOT_MSG(
+            storage,
+            "Please use IndexNSGFlat (or variants) instead of IndexNSG directly");
     storage->reset();
     ntotal = 0;
     is_built = false;
 }
 
 void IndexNSG::reconstruct(idx_t key, float* recons) const {
+    FAISS_THROW_IF_NOT_MSG(
+            storage,
+            "Please use IndexNSGFlat (or variants) instead of IndexNSG directly");
     storage->reconstruct(key, recons);
 }
 

data/vendor/faiss/faiss/IndexPQ.cpp

@@ -107,7 +107,7 @@ void IndexPQ::search(
     if (iparams) {
         params = dynamic_cast<const SearchParametersPQ*>(iparams);
         FAISS_THROW_IF_NOT_MSG(params, "invalid search params");
-        FAISS_THROW_IF_NOT_MSG(!params->sel, "selector not supported");
+        FAISS_THROW_IF_MSG(params->sel, "selector not supported");
         param_search_type = params->search_type;
     }
 
@@ -459,7 +459,7 @@ struct SortedArray {
 
     void init(const T* x_2) {
         this->x = x_2;
-        FAISS_THROW_IF_NOT(!perm.empty());
+        FAISS_THROW_IF_MSG(perm.empty(), "permutation array must not be empty");
         for (int n = 0; n < N; n++) {
             perm[n] = n;
         }
@@ -543,7 +543,7 @@ struct SemiSortedArray {
 
     void init(const T* x_2) {
         this->x = x_2;
-        FAISS_THROW_IF_NOT(!perm.empty());
+        FAISS_THROW_IF_MSG(perm.empty(), "permutation array must not be empty");
         for (int n = 0; n < N; n++) {
             perm[n] = n;
         }
@@ -665,7 +665,7 @@ struct MinSumK {
 
     void mark_seen(int64_t i) {
         if (use_seen) {
-            FAISS_THROW_IF_NOT(!seen.empty());
+            FAISS_THROW_IF_MSG(seen.empty(), "seen bitmap must not be empty");
             seen[i >> 3] |= 1 << (i & 7);
         }
     }
@@ -795,8 +795,7 @@ void MultiIndexQuantizer::search(
         float* distances,
         idx_t* labels,
         const SearchParameters* params) const {
-    FAISS_THROW_IF_NOT_MSG(
-            !params, "search params not supported for this index");
+    FAISS_THROW_IF_MSG(params, "search params not supported for this index");
     if (n == 0) {
         return;
     }
@@ -943,8 +942,7 @@ void MultiIndexQuantizer2::search(
         float* distances,
         idx_t* labels,
         const SearchParameters* params) const {
-    FAISS_THROW_IF_NOT_MSG(
-            !params, "search params not supported for this index");
+    FAISS_THROW_IF_MSG(params, "search params not supported for this index");
 
     if (n == 0) {
         return;

data/vendor/faiss/faiss/IndexPreTransform.cpp

@@ -229,6 +229,7 @@ size_t IndexPreTransform::remove_ids(const IDSelector& sel) {
 }
 
 void IndexPreTransform::reconstruct(idx_t key, float* recons) const {
+    FAISS_THROW_IF_NOT_MSG(index, "IndexPreTransform: null sub-index");
     float* x = chain.empty() ? recons : new float[index->d];
     std::unique_ptr<float[]> del(recons == x ? nullptr : x);
     // Initial reconstruction
@@ -239,6 +240,7 @@ void IndexPreTransform::reconstruct(idx_t key, float* recons) const {
 }
 
 void IndexPreTransform::reconstruct_n(idx_t i0, idx_t ni, float* recons) const {
+    FAISS_THROW_IF_NOT_MSG(index, "IndexPreTransform: null sub-index");
     float* x = chain.empty() ? recons : new float[ni * index->d];
     std::unique_ptr<float[]> del(recons == x ? nullptr : x);
     // Initial reconstruction
@@ -345,6 +347,19 @@ struct PreTransformDistanceComputer : DistanceComputer {
     float operator()(idx_t i) override {
         return (*sub_dc)(i);
     }
+
+    void distances_batch_4(
+            const idx_t idx0,
+            const idx_t idx1,
+            const idx_t idx2,
+            const idx_t idx3,
+            float& dis0,
+            float& dis1,
+            float& dis2,
+            float& dis3) override {
+        sub_dc->distances_batch_4(
+                idx0, idx1, idx2, idx3, dis0, dis1, dis2, dis3);
+    }
 };
 
 } // anonymous namespace

data/vendor/faiss/faiss/IndexRaBitQ.cpp

@@ -122,8 +122,8 @@ struct Run_search_with_dc_res {
                     // RaBitQuantizer.cpp for details.
                     auto* dc = dynamic_cast<RaBitQDistanceComputer*>(
                             dc_base.get());
-                    FAISS_THROW_IF_NOT_MSG(
-                            dc != nullptr,
+                    FAISS_THROW_IF_MSG(
+                            dc == nullptr,
                             "Failed to cast to RaBitQDistanceComputer for two-stage search");
 
                     bool is_similarity =

data/vendor/faiss/faiss/IndexRaBitQFastScan.cpp

@@ -512,8 +512,7 @@ void IndexRaBitQFastScan::search(
         float* distances,
         idx_t* labels,
         const SearchParameters* params) const {
-    FAISS_THROW_IF_NOT_MSG(
-            !params, "search params not supported for this index");
+    FAISS_THROW_IF_MSG(params, "search params not supported for this index");
 
     // Create query factors array on stack - memory managed by caller
     std::vector<rabitq_utils::QueryFactorsData> query_factors_storage(n);

data/vendor/faiss/faiss/IndexRaBitQFastScan.h

@@ -209,8 +209,12 @@ struct RaBitQHeapHandler
                 : 0;
 
         const size_t block_idx = base_db_idx / rabitq_index->bbs;
+        // aux_base points at this 32-lane sub-block's factors: the bbs block
+        // base plus the loop-invariant intra-block offset
+        // ((base_db_idx % bbs)). Only the per-element i term varies below.
         const uint8_t* aux_base = rabitq_index->codes.get() +
-                block_idx * full_block_size + packed_block_size;
+                block_idx * full_block_size + packed_block_size +
+                (base_db_idx % rabitq_index->bbs) * storage_size;
 
         for (size_t i = 0; i < max_vectors; i++) {
             const size_t db_idx = base_db_idx + i;

data/vendor/faiss/faiss/IndexRefine.cpp

@@ -156,7 +156,17 @@ void IndexRefine::range_search(
     SearchParameters* base_index_params =
             (params != nullptr) ? params->base_index_params : nullptr;
 
-    base_index->range_search(n, x, radius, result, base_index_params);
+    const float kf = (params != nullptr) ? params->k_factor : this->k_factor;
+
+    const bool is_similarity = is_similarity_metric(metric_type);
+
+    // Scale the base_index search radius by k_factor. Results are filtered to
+    // the exact radius below, so this only affects recall, not correctness; it
+    // is up to the user to pick a k_factor that widens the base search for
+    // their metric.
+    const float base_radius = radius * kf;
+
+    base_index->range_search(n, x, base_radius, result, base_index_params);
 
 #pragma omp parallel if (n > 1)
     {
@@ -177,6 +187,25 @@ void IndexRefine::range_search(
             }
         }
     }
+
+    // Exact filtering done here, as base index's approximations
+    // can fall outside the specified exact radius.
+    const std::vector<size_t> prev_lims(result->lims, result->lims + n + 1);
+    size_t wp = 0;
+    for (idx_t i = 0; i < n; i++) {
+        for (size_t j = prev_lims[i]; j < prev_lims[i + 1]; j++) {
+            const float dis = result->distances[j];
+            const bool within = is_similarity
+                    ? CMin<float, idx_t>::cmp(radius, dis)
+                    : CMax<float, idx_t>::cmp(radius, dis);
+            if (within) {
+                result->labels[wp] = result->labels[j];
+                result->distances[wp] = result->distances[j];
+                wp++;
+            }
+        }
+        result->lims[i + 1] = wp;
+    }
 }
 
 void IndexRefine::reconstruct(idx_t key, float* recons) const {

data/vendor/faiss/faiss/IndexReplicas.cpp

@@ -127,8 +127,7 @@ void IndexReplicasTemplate<IndexT>::search(
         distance_t* distances,
         idx_t* labels,
         const SearchParameters* params) const {
-    FAISS_THROW_IF_NOT_MSG(
-            !params, "search params not supported for this index");
+    FAISS_THROW_IF_MSG(params, "search params not supported for this index");
     FAISS_THROW_IF_NOT(k > 0);
     FAISS_THROW_IF_NOT_MSG(this->count() > 0, "no replicas in index");
 

data/vendor/faiss/faiss/IndexShards.cpp

@@ -143,8 +143,8 @@ void IndexShardsTemplate<IndexT>::add_with_ids(
             "request them to be shifted");
 
     if (successive_ids) {
-        FAISS_THROW_IF_NOT_MSG(
-                !xids,
+        FAISS_THROW_IF_MSG(
+                xids,
                 "It makes no sense to pass in ids and "
                 "request them to be shifted");
         FAISS_THROW_IF_NOT_MSG(
@@ -243,8 +243,8 @@ void IndexShardsTemplate<IndexT>::search(
 
     this->runOnIndex(fn);
 
-    if (this->metric_type == METRIC_L2) {
-        merge_knn_results<idx_t, CMin<distance_t, int>>(
+    if (is_similarity_metric(this->metric_type)) {
+        merge_knn_results<idx_t, CMax<distance_t, int>>(
                 n,
                 k,
                 nshard,
@@ -253,7 +253,7 @@ void IndexShardsTemplate<IndexT>::search(
                 distances,
                 labels);
     } else {
-        merge_knn_results<idx_t, CMax<distance_t, int>>(
+        merge_knn_results<idx_t, CMin<distance_t, int>>(
                 n,
                 k,
                 nshard,

data/vendor/faiss/faiss/IndexShardsIVF.cpp

@@ -13,6 +13,7 @@
 #include <cstdio>
 #include <functional>
 
+#include <faiss/MetricType.h>
 #include <faiss/impl/FaissAssert.h>
 #include <faiss/utils/Heap.h>
 #include <faiss/utils/WorkerThread.h>
@@ -106,8 +107,8 @@ void IndexShardsIVF::add_with_ids(
             "request them to be shifted");
 
     if (successive_ids) {
-        FAISS_THROW_IF_NOT_MSG(
-                !xids,
+        FAISS_THROW_IF_MSG(
+                xids,
                 "It makes no sense to pass in ids and "
                 "request them to be shifted");
         FAISS_THROW_IF_NOT_MSG(
@@ -227,8 +228,8 @@ void IndexShardsIVF::search(
 
     this->runOnIndex(fn);
 
-    if (this->metric_type == METRIC_L2) {
-        merge_knn_results<idx_t, CMin<distance_t, int>>(
+    if (is_similarity_metric(metric_type)) {
+        merge_knn_results<idx_t, CMax<distance_t, int>>(
                 n,
                 k,
                 nshard,
@@ -237,7 +238,7 @@ void IndexShardsIVF::search(
                 distances,
                 labels);
     } else {
-        merge_knn_results<idx_t, CMax<distance_t, int>>(
+        merge_knn_results<idx_t, CMin<distance_t, int>>(
                 n,
                 k,
                 nshard,

data/vendor/faiss/faiss/MetaIndexes.cpp

@@ -63,8 +63,7 @@ void IndexSplitVectors::search(
         float* distances,
         idx_t* labels,
         const SearchParameters* params) const {
-    FAISS_THROW_IF_NOT_MSG(
-            !params, "search params not supported for this index");
+    FAISS_THROW_IF_MSG(params, "search params not supported for this index");
     FAISS_THROW_IF_NOT_MSG(k == 1, "search implemented only for k=1");
     FAISS_THROW_IF_NOT_MSG(
             sum_d == d, "not enough indexes compared to # dimensions");
@@ -187,8 +186,7 @@ void IndexRandom::search(
         float* distances,
         idx_t* labels,
         const SearchParameters* params) const {
-    FAISS_THROW_IF_NOT_MSG(
-            !params, "search params not supported for this index");
+    FAISS_THROW_IF_MSG(params, "search params not supported for this index");
     FAISS_THROW_IF_NOT(k <= ntotal);
 #pragma omp parallel for if (n > 1000)
     for (idx_t i = 0; i < n; i++) {