address comments

Author: aamijar

cpp/include/cuvs/neighbors/cagra.hpp

@@ -471,8 +471,8 @@ struct CUVS_EXPORT index : cuvs::neighbors::index {
   /** Total length of the index (number of vectors). */
   [[nodiscard]] constexpr inline auto size() const noexcept -> IdxT
   {
+    if (dataset_fd_.has_value() || graph_fd_.has_value()) { return n_rows_; }
     auto data_rows = dataset_.n_rows();
-    if (dataset_fd_.has_value()) { return n_rows_; }
     return data_rows > 0 ? data_rows : graph_view_.extent(0);
   }
 
@@ -484,7 +484,14 @@ struct CUVS_EXPORT index : cuvs::neighbors::index {
   /** Graph degree */
   [[nodiscard]] constexpr inline auto graph_degree() const noexcept -> uint32_t
   {
-    return dataset_fd_.has_value() ? graph_degree_ : graph_view_.extent(1);
+    return graph_fd_.has_value() ? graph_degree_ : graph_view_.extent(1);
+  }
+
+  /** Number of rows represented by the graph. */
+  [[nodiscard]] constexpr inline auto graph_size() const noexcept -> IdxT
+  {
+    return graph_fd_.has_value() ? static_cast<IdxT>(n_rows_)
+                                 : static_cast<IdxT>(graph_view_.extent(0));
   }
 
   /** Non-owning dataset binding stored by the index. */

cpp/src/neighbors/cagra.cuh

@@ -5,15 +5,16 @@
 
 #pragma once
 
+#include "detail/ann_utils.cuh"
 #include "detail/cagra/add_nodes.cuh"
 #include "detail/cagra/cagra_build.cuh"
 #include "detail/cagra/cagra_merge.cuh"
 #include "detail/cagra/cagra_search.cuh"
 #include "detail/cagra/graph_core.cuh"
 
-#include "detail/ann_utils.cuh"
 #include <raft/core/device_mdspan.hpp>
 #include <raft/core/host_device_accessor.hpp>
+#include <raft/core/logger.hpp>
 #include <raft/core/mdspan.hpp>
 #include <raft/core/resources.hpp>
 #include <raft/linalg/norm.cuh>
@@ -22,12 +23,11 @@
 #include <cuvs/core/bitset.hpp>
 #include <cuvs/distance/distance.hpp>
 #include <cuvs/neighbors/cagra.hpp>
-
 #include <cuvs/neighbors/common.hpp>
+
 #include <rmm/cuda_stream_view.hpp>
 
 #include <algorithm>
-#include <memory>
 #include <optional>
 #include <type_traits>
 
@@ -573,11 +573,16 @@ auto update_dataset(raft::resources const& res,
                     index<T, IdxT, SrcDatasetViewT>&& cagra_index,
                     DstDatasetViewT dataset) -> index<T, IdxT, DstDatasetViewT>
 {
-  if constexpr (!std::is_same_v<SrcDatasetViewT, DstDatasetViewT>) {
-    RAFT_EXPECTS(dataset.n_rows() == static_cast<int64_t>(cagra_index.size()),
-                 "The new dataset row count must match the source dataset row count");
-    RAFT_EXPECTS(cagra_index.dim() == 0 || dataset.dim() == cagra_index.dim(),
-                 "The new dataset dimension must match the source dataset dimension");
+  auto const graph_rows = static_cast<int64_t>(cagra_index.graph_size());
+  if (dataset.n_rows() != graph_rows) {
+    RAFT_LOG_WARN("The new dataset row count (%ld) does not match the graph row count (%ld)",
+                  static_cast<long>(dataset.n_rows()),
+                  static_cast<long>(graph_rows));
+  }
+  if (dataset.dim() != cagra_index.dim()) {
+    RAFT_LOG_WARN("The new dataset dimension (%u) does not match the index dimension (%u)",
+                  static_cast<unsigned>(dataset.dim()),
+                  static_cast<unsigned>(cagra_index.dim()));
   }
 
   index<T, IdxT, DstDatasetViewT> new_index(res, std::move(cagra_index), dataset);

cpp/src/neighbors/detail/cagra/add_nodes.cuh

@@ -408,8 +408,8 @@ void extend_core(raft::resources const& handle,
     cuvs::neighbors::cagra::add_graph_nodes<T, IdxT>(
       handle, extended_strided, index, updated_graph.view(), params);
 
-    index = cuvs::neighbors::cagra::update_dataset(handle, std::move(index), extended_dataset);
     index.update_graph(handle, raft::make_const_mdspan(updated_graph.view()));
+    index = cuvs::neighbors::cagra::update_dataset(handle, std::move(index), extended_dataset);
   }
 }
 }  // namespace cuvs::neighbors::cagra