Renamed quantize_binary to binary_quantize

CHANGELOG.md

@@ -3,9 +3,9 @@
 - Added `halfvec` type
 - Added `sparsevec` type
 - Added support for `bit` vectors to HNSW
+- Added `binary_quantize` function
 - Added `hamming_distance` function
 - Added `jaccard_distance` function
-- Added `quantize_binary` function
 - Added `subvector` function
 - Updated comparison operators to support vectors with different dimensions
 

README.md

@@ -482,20 +482,20 @@ Also supports Jaccard distance (`<%>`)
 Use expression indexing for binary quantization
 
 ```sql
-CREATE INDEX ON items USING hnsw ((quantize_binary(embedding)::bit(3)) bit_hamming_ops);
+CREATE INDEX ON items USING hnsw ((binary_quantize(embedding)::bit(3)) bit_hamming_ops);
 ```
 
 Get the nearest neighbors by Hamming distance
 
 ```sql
-SELECT * FROM items ORDER BY quantize_binary(embedding)::bit(3) <~> quantize_binary('[1,-2,3]') LIMIT 5;
+SELECT * FROM items ORDER BY binary_quantize(embedding)::bit(3) <~> binary_quantize('[1,-2,3]') LIMIT 5;
 ```
 
 Re-rank by the original vectors for better recall
 
 ```sql
 SELECT * FROM (
-    SELECT * FROM items ORDER BY quantize_binary(embedding)::bit(3) <~> quantize_binary('[1,-2,3]') LIMIT 20
+    SELECT * FROM items ORDER BY binary_quantize(embedding)::bit(3) <~> binary_quantize('[1,-2,3]') LIMIT 20
 ) ORDER BY embedding <=> '[1,-2,3]' LIMIT 5;
 ```
 
@@ -858,11 +858,11 @@ Operator | Description | Added
 
 Function | Description | Added
 --- | --- | ---
+binary_quantize(vector) → bit | binary quantize | unreleased
 cosine_distance(vector, vector) → double precision | cosine distance |
 inner_product(vector, vector) → double precision | inner product |
 l1_distance(vector, vector) → double precision | taxicab distance | 0.5.0
 l2_distance(vector, vector) → double precision | Euclidean distance |
-quantize_binary(vector) → bit | quantize | unreleased
 subvector(vector, integer, integer) → vector | subvector | unreleased
 vector_dims(vector) → integer | number of dimensions |
 vector_norm(vector) → double precision | Euclidean norm |
@@ -893,12 +893,12 @@ Operator | Description | Added
 
 Function | Description | Added
 --- | --- | ---
+binary_quantize(halfvec) → bit | binary quantize | unreleased
 cosine_distance(halfvec, halfvec) → double precision | cosine distance | unreleased
 inner_product(halfvec, halfvec) → double precision | inner product | unreleased
 l1_distance(halfvec, halfvec) → double precision | taxicab distance | unreleased
 l2_distance(halfvec, halfvec) → double precision | Euclidean distance | unreleased
 l2_norm(halfvec) → double precision | Euclidean norm | unreleased
-quantize_binary(halfvec) → bit | quantize | unreleased
 subvector(halfvec, integer, integer) → halfvec | subvector | unreleased
 vector_dims(halfvec) → integer | number of dimensions | unreleased
 

sql/vector--0.6.2--0.7.0.sql

@@ -1,7 +1,7 @@
 -- complain if script is sourced in psql, rather than via CREATE EXTENSION
 \echo Use "ALTER EXTENSION vector UPDATE TO '0.7.0'" to load this file. \quit
 
-CREATE FUNCTION quantize_binary(vector) RETURNS bit
+CREATE FUNCTION binary_quantize(vector) RETURNS bit
 	AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
 
 CREATE FUNCTION subvector(vector, int, int) RETURNS vector
@@ -77,8 +77,8 @@ CREATE FUNCTION vector_dims(halfvec) RETURNS integer
 CREATE FUNCTION l2_norm(halfvec) RETURNS float8
 	AS 'MODULE_PATHNAME', 'halfvec_l2_norm' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
 
-CREATE FUNCTION quantize_binary(halfvec) RETURNS bit
-	AS 'MODULE_PATHNAME', 'halfvec_quantize_binary' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
+CREATE FUNCTION binary_quantize(halfvec) RETURNS bit
+	AS 'MODULE_PATHNAME', 'halfvec_binary_quantize' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
 
 CREATE FUNCTION subvector(halfvec, int, int) RETURNS halfvec
 	AS 'MODULE_PATHNAME', 'halfvec_subvector' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;

sql/vector.sql

@@ -49,7 +49,7 @@ CREATE FUNCTION vector_dims(vector) RETURNS integer
 CREATE FUNCTION vector_norm(vector) RETURNS float8
 	AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
 
-CREATE FUNCTION quantize_binary(vector) RETURNS bit
+CREATE FUNCTION binary_quantize(vector) RETURNS bit
 	AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
 
 CREATE FUNCTION subvector(vector, int, int) RETURNS vector
@@ -370,8 +370,8 @@ CREATE FUNCTION vector_dims(halfvec) RETURNS integer
 CREATE FUNCTION l2_norm(halfvec) RETURNS float8
 	AS 'MODULE_PATHNAME', 'halfvec_l2_norm' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
 
-CREATE FUNCTION quantize_binary(halfvec) RETURNS bit
-	AS 'MODULE_PATHNAME', 'halfvec_quantize_binary' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
+CREATE FUNCTION binary_quantize(halfvec) RETURNS bit
+	AS 'MODULE_PATHNAME', 'halfvec_binary_quantize' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
 
 CREATE FUNCTION subvector(halfvec, int, int) RETURNS halfvec
 	AS 'MODULE_PATHNAME', 'halfvec_subvector' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;

src/halfvec.c

@@ -869,9 +869,9 @@ halfvec_mul(PG_FUNCTION_ARGS)
 /*
  * Quantize a half vector
  */
-PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_quantize_binary);
+PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_binary_quantize);
 Datum
-halfvec_quantize_binary(PG_FUNCTION_ARGS)
+halfvec_binary_quantize(PG_FUNCTION_ARGS)
 {
 	HalfVector *a = PG_GETARG_HALFVEC_P(0);
 	half	   *ax = a->x;

src/vector.c

@@ -873,9 +873,9 @@ vector_mul(PG_FUNCTION_ARGS)
 /*
  * Quantize a vector
  */
-PGDLLEXPORT PG_FUNCTION_INFO_V1(quantize_binary);
+PGDLLEXPORT PG_FUNCTION_INFO_V1(binary_quantize);
 Datum
-quantize_binary(PG_FUNCTION_ARGS)
+binary_quantize(PG_FUNCTION_ARGS)
 {
 	Vector	   *a = PG_GETARG_VECTOR_P(0);
 	float	   *ax = a->x;

test/expected/halfvec_functions.out

@@ -278,14 +278,14 @@ SELECT l1_distance('[0,0]'::halfvec, '[0,1]');
 
 SELECT l1_distance('[1,2]'::halfvec, '[3]');
 ERROR:  different halfvec dimensions 2 and 1
-SELECT quantize_binary('[1,0,-1]'::halfvec);
- quantize_binary 
+SELECT binary_quantize('[1,0,-1]'::halfvec);
+ binary_quantize 
 -----------------
  100
 (1 row)
 
-SELECT quantize_binary('[0,0.1,-0.2,-0.3,0.4,0.5,0.6,-0.7,0.8,-0.9,1]'::halfvec);
- quantize_binary 
+SELECT binary_quantize('[0,0.1,-0.2,-0.3,0.4,0.5,0.6,-0.7,0.8,-0.9,1]'::halfvec);
+ binary_quantize 
 -----------------
  01001110101
 (1 row)

test/expected/vector_functions.out

@@ -272,14 +272,14 @@ SELECT l1_distance('[3e38]'::vector, '[-3e38]');
     Infinity
 (1 row)
 
-SELECT quantize_binary('[1,0,-1]'::vector);
- quantize_binary 
+SELECT binary_quantize('[1,0,-1]'::vector);
+ binary_quantize 
 -----------------
  100
 (1 row)
 
-SELECT quantize_binary('[0,0.1,-0.2,-0.3,0.4,0.5,0.6,-0.7,0.8,-0.9,1]'::vector);
- quantize_binary 
+SELECT binary_quantize('[0,0.1,-0.2,-0.3,0.4,0.5,0.6,-0.7,0.8,-0.9,1]'::vector);
+ binary_quantize 
 -----------------
  01001110101
 (1 row)

test/sql/halfvec_functions.sql

@@ -59,8 +59,8 @@ SELECT l1_distance('[0,0]'::halfvec, '[3,4]');
 SELECT l1_distance('[0,0]'::halfvec, '[0,1]');
 SELECT l1_distance('[1,2]'::halfvec, '[3]');
 
-SELECT quantize_binary('[1,0,-1]'::halfvec);
-SELECT quantize_binary('[0,0.1,-0.2,-0.3,0.4,0.5,0.6,-0.7,0.8,-0.9,1]'::halfvec);
+SELECT binary_quantize('[1,0,-1]'::halfvec);
+SELECT binary_quantize('[0,0.1,-0.2,-0.3,0.4,0.5,0.6,-0.7,0.8,-0.9,1]'::halfvec);
 
 SELECT subvector('[1,2,3,4,5]'::halfvec, 1, 3);
 SELECT subvector('[1,2,3,4,5]'::halfvec, 3, 2);

test/sql/vector_functions.sql

@@ -58,8 +58,8 @@ SELECT l1_distance('[0,0]'::vector, '[0,1]');
 SELECT l1_distance('[1,2]'::vector, '[3]');
 SELECT l1_distance('[3e38]'::vector, '[-3e38]');
 
-SELECT quantize_binary('[1,0,-1]'::vector);
-SELECT quantize_binary('[0,0.1,-0.2,-0.3,0.4,0.5,0.6,-0.7,0.8,-0.9,1]'::vector);
+SELECT binary_quantize('[1,0,-1]'::vector);
+SELECT binary_quantize('[0,0.1,-0.2,-0.3,0.4,0.5,0.6,-0.7,0.8,-0.9,1]'::vector);
 
 SELECT subvector('[1,2,3,4,5]'::vector, 1, 3);
 SELECT subvector('[1,2,3,4,5]'::vector, 3, 2);