Improved distance functions

Author: ankane

lib/pgvector/ecto/query.ex

@@ -9,7 +9,7 @@ if Code.ensure_loaded?(Ecto) do
     """
     defmacro l2_distance(column, value) do
       quote do
-        fragment("(? <-> ?)", unquote(column), ^Pgvector.Ecto.Utils.to_sql(unquote(value)))
+        fragment("(? <-> ?)", unquote(column), unquote(value))
       end
     end
 
@@ -18,7 +18,7 @@ if Code.ensure_loaded?(Ecto) do
     """
     defmacro max_inner_product(column, value) do
       quote do
-        fragment("(? <#> ?)", unquote(column), ^Pgvector.Ecto.Utils.to_sql(unquote(value)))
+        fragment("(? <#> ?)", unquote(column), unquote(value))
       end
     end
 
@@ -27,7 +27,7 @@ if Code.ensure_loaded?(Ecto) do
     """
     defmacro cosine_distance(column, value) do
       quote do
-        fragment("(? <=> ?)", unquote(column), ^Pgvector.Ecto.Utils.to_sql(unquote(value)))
+        fragment("(? <=> ?)", unquote(column), unquote(value))
       end
     end
 
@@ -36,7 +36,7 @@ if Code.ensure_loaded?(Ecto) do
     """
     defmacro l1_distance(column, value) do
       quote do
-        fragment("(? <+> ?)", unquote(column), ^Pgvector.Ecto.Utils.to_sql(unquote(value)))
+        fragment("(? <+> ?)", unquote(column), unquote(value))
       end
     end
 

lib/pgvector/ecto/utils.ex

@@ -30,54 +30,55 @@ defmodule EctoTest do
   end
 
   test "vector l2 distance" do
-    items = Repo.all(from i in Item, order_by: l2_distance(i.embedding, [1, 1, 1]), limit: 5)
+    # TODO restore support for list
+    items = Repo.all(from i in Item, order_by: l2_distance(i.embedding, ^Pgvector.new([1, 1, 1])), limit: 5)
     assert Enum.map(items, fn v -> v.id end) == [1, 3, 2]
     assert Enum.map(items, fn v -> v.embedding |> Pgvector.to_list() end) == [[1.0, 1.0, 1.0], [1.0, 1.0, 2.0], [2.0, 2.0, 3.0]]
   end
 
   test "vector max inner product" do
-    items = Repo.all(from i in Item, order_by: max_inner_product(i.embedding, [1, 1, 1]), limit: 5)
+    items = Repo.all(from i in Item, order_by: max_inner_product(i.embedding, ^Pgvector.new([1, 1, 1])), limit: 5)
     assert Enum.map(items, fn v -> v.id end) == [2, 3, 1]
   end
 
   test "vector cosine distance" do
-    items = Repo.all(from i in Item, order_by: cosine_distance(i.embedding, [1, 1, 1]), limit: 5)
+    items = Repo.all(from i in Item, order_by: cosine_distance(i.embedding, ^Pgvector.new([1, 1, 1])), limit: 5)
     assert Enum.map(items, fn v -> v.id end) == [1, 2, 3]
   end
 
   test "vector cosine similarity" do
-    items = Repo.all(from i in Item, order_by: (1 - cosine_distance(i.embedding, [1, 1, 1])), limit: 5)
+    items = Repo.all(from i in Item, order_by: (1 - cosine_distance(i.embedding, ^Pgvector.new([1, 1, 1]))), limit: 5)
     assert Enum.map(items, fn v -> v.id end) == [3, 2, 1]
   end
 
   test "vector l1 distance" do
-    items = Repo.all(from i in Item, order_by: l1_distance(i.embedding, [1, 1, 1]), limit: 5)
+    items = Repo.all(from i in Item, order_by: l1_distance(i.embedding, ^Pgvector.new([1, 1, 1])), limit: 5)
     assert Enum.map(items, fn v -> v.id end) == [1, 3, 2]
   end
 
   test "halfvec l2 distance" do
-    items = Repo.all(from i in Item, order_by: l2_distance(i.half_embedding, Pgvector.HalfVector.new([1, 1, 1])), limit: 5)
+    items = Repo.all(from i in Item, order_by: l2_distance(i.half_embedding, ^Pgvector.HalfVector.new([1, 1, 1])), limit: 5)
     assert Enum.map(items, fn v -> v.id end) == [1, 3, 2]
     assert Enum.map(items, fn v -> v.half_embedding |> Pgvector.to_list() end) == [[1.0, 1.0, 1.0], [1.0, 1.0, 2.0], [2.0, 2.0, 3.0]]
   end
 
   test "halfvec max inner product" do
-    items = Repo.all(from i in Item, order_by: max_inner_product(i.half_embedding, Pgvector.HalfVector.new([1, 1, 1])), limit: 5)
+    items = Repo.all(from i in Item, order_by: max_inner_product(i.half_embedding, ^Pgvector.HalfVector.new([1, 1, 1])), limit: 5)
     assert Enum.map(items, fn v -> v.id end) == [2, 3, 1]
   end
 
   test "halfvec cosine distance" do
-    items = Repo.all(from i in Item, order_by: cosine_distance(i.half_embedding, Pgvector.HalfVector.new([1, 1, 1])), limit: 5)
+    items = Repo.all(from i in Item, order_by: cosine_distance(i.half_embedding, ^Pgvector.HalfVector.new([1, 1, 1])), limit: 5)
     assert Enum.map(items, fn v -> v.id end) == [1, 2, 3]
   end
 
   test "halfvec cosine similarity" do
-    items = Repo.all(from i in Item, order_by: (1 - cosine_distance(i.half_embedding, Pgvector.HalfVector.new([1, 1, 1]))), limit: 5)
+    items = Repo.all(from i in Item, order_by: (1 - cosine_distance(i.half_embedding, ^Pgvector.HalfVector.new([1, 1, 1]))), limit: 5)
     assert Enum.map(items, fn v -> v.id end) == [3, 2, 1]
   end
 
   test "halfvec l1 distance" do
-    items = Repo.all(from i in Item, order_by: l1_distance(i.half_embedding, Pgvector.HalfVector.new([1, 1, 1])), limit: 5)
+    items = Repo.all(from i in Item, order_by: l1_distance(i.half_embedding, ^Pgvector.HalfVector.new([1, 1, 1])), limit: 5)
     assert Enum.map(items, fn v -> v.id end) == [1, 3, 2]
   end
 
@@ -92,28 +93,28 @@ defmodule EctoTest do
   end
 
   test "sparsevec l2 distance" do
-    items = Repo.all(from i in Item, order_by: l2_distance(i.sparse_embedding, Pgvector.SparseVector.new([1, 1, 1])), limit: 5)
+    items = Repo.all(from i in Item, order_by: l2_distance(i.sparse_embedding, ^Pgvector.SparseVector.new([1, 1, 1])), limit: 5)
     assert Enum.map(items, fn v -> v.id end) == [1, 3, 2]
     assert Enum.map(items, fn v -> v.sparse_embedding |> Pgvector.to_list() end) == [[1.0, 1.0, 1.0], [1.0, 1.0, 2.0], [2.0, 2.0, 3.0]]
   end
 
   test "sparsevec max inner product" do
-    items = Repo.all(from i in Item, order_by: max_inner_product(i.sparse_embedding, Pgvector.SparseVector.new([1, 1, 1])), limit: 5)
+    items = Repo.all(from i in Item, order_by: max_inner_product(i.sparse_embedding, ^Pgvector.SparseVector.new([1, 1, 1])), limit: 5)
     assert Enum.map(items, fn v -> v.id end) == [2, 3, 1]
   end
 
   test "sparsevec cosine distance" do
-    items = Repo.all(from i in Item, order_by: cosine_distance(i.sparse_embedding, Pgvector.SparseVector.new([1, 1, 1])), limit: 5)
+    items = Repo.all(from i in Item, order_by: cosine_distance(i.sparse_embedding, ^Pgvector.SparseVector.new([1, 1, 1])), limit: 5)
     assert Enum.map(items, fn v -> v.id end) == [1, 2, 3]
   end
 
   test "sparsevec cosine similarity" do
-    items = Repo.all(from i in Item, order_by: (1 - cosine_distance(i.sparse_embedding, Pgvector.SparseVector.new([1, 1, 1]))), limit: 5)
+    items = Repo.all(from i in Item, order_by: (1 - cosine_distance(i.sparse_embedding, ^Pgvector.SparseVector.new([1, 1, 1]))), limit: 5)
     assert Enum.map(items, fn v -> v.id end) == [3, 2, 1]
   end
 
   test "sparsevec l1 distance" do
-    items = Repo.all(from i in Item, order_by: l1_distance(i.sparse_embedding, Pgvector.SparseVector.new([1, 1, 1])), limit: 5)
+    items = Repo.all(from i in Item, order_by: l1_distance(i.sparse_embedding, ^Pgvector.SparseVector.new([1, 1, 1])), limit: 5)
     assert Enum.map(items, fn v -> v.id end) == [1, 3, 2]
   end