Added random_vector function

CHANGELOG.md

@@ -1,3 +1,7 @@
+## 0.4.1 (unreleased)
+
+- Added `random_vector` function
+
 ## 0.4.0 (2023-01-11)
 
 If upgrading with Postgres < 13, see [this note](https://github.com/pgvector/pgvector#040).

README.md

@@ -29,7 +29,7 @@ Then load it in databases where you want to use it
 CREATE EXTENSION vector;
 ```
 
-You can also install it with [Docker](#docker), [Homebrew](#homebrew), or [PGXN](#pgxn)
+You can also install it with [Docker](#docker), [Homebrew](#homebrew), [PGXN](#pgxn), or [conda-forge](#conda-forge)
 
 ## Getting Started
 
@@ -184,26 +184,26 @@ inner_product(vector, vector) → double precision | inner product
 l2_distance(vector, vector) → double precision | Euclidean distance
 vector_dims(vector) → integer | number of dimensions
 vector_norm(vector) → double precision | Euclidean norm
+random_vector(integer) → vector | random vector [unreleased]
 
 ### Aggregate Functions
 
-Function | Description | Partial Mode
+Function | Description
---- | --- | ---
+--- | ---
-avg(vector) → vector | arithmetic mean | Yes
+avg(vector) → vector | arithmetic mean
 
 ## Libraries
 
-Libraries that use pgvector:
+Language | Libraries
-
+--- | ---
-- [pgvector-python](https://github.com/pgvector/pgvector-python) (Python)
+Python | [pgvector-python](https://github.com/pgvector/pgvector-python)
-- [Neighbor](https://github.com/ankane/neighbor) (Ruby)
+Ruby | [Neighbor](https://github.com/ankane/neighbor), [pgvector-ruby](https://github.com/pgvector/pgvector-ruby)
-- [pgvector-ruby](https://github.com/pgvector/pgvector-ruby) (Ruby)
+Node | [pgvector-node](https://github.com/pgvector/pgvector-node)
-- [pgvector-node](https://github.com/pgvector/pgvector-node) (Node.js)
+Go | [pgvector-go](https://github.com/pgvector/pgvector-go)
-- [pgvector-go](https://github.com/pgvector/pgvector-go) (Go)
+PHP | [pgvector-php](https://github.com/pgvector/pgvector-php)
-- [pgvector-php](https://github.com/pgvector/pgvector-php) (PHP)
+Rust | [pgvector-rust](https://github.com/pgvector/pgvector-rust)
-- [pgvector-rust](https://github.com/pgvector/pgvector-rust) (Rust)
+C++ | [pgvector-cpp](https://github.com/pgvector/pgvector-cpp)
-- [pgvector-cpp](https://github.com/pgvector/pgvector-cpp) (C++)
+Elixir | [pgvector-elixir](https://github.com/pgvector/pgvector-elixir)
-- [pgvector-elixir](https://github.com/pgvector/pgvector-elixir) (Elixir)
 
 ## Frequently Asked Questions
 
@@ -258,12 +258,22 @@ Install from the [PostgreSQL Extension Network](https://pgxn.org/dist/vector) wi
 pgxn install vector
 ```
 
+### conda-forge
+
+Install from [conda-forge](https://anaconda.org/conda-forge/pgvector) with:
+
+```sh
+conda install -c conda-forge pgvector
+```
+
+This method is [community-maintained](https://github.com/conda-forge/pgvector-feedstock) by [@mmcauliffe](https://github.com/mmcauliffe)
+
 ## Hosted Postgres
 
 Some Postgres providers only support specific extensions. To request a new extension:
 
 - Amazon RDS - follow the instructions on [this page](https://aws.amazon.com/rds/postgresql/faqs/)
-- Google Cloud SQL - follow the instructions on [this page](https://cloud.google.com/sql/docs/postgres/extensions#requesting-support-for-a-new-extension)
+- Google Cloud SQL - vote or comment on [this page](https://issuetracker.google.com/issues/265172065)
 - DigitalOcean Managed Databases - vote or comment on [this page](https://ideas.digitalocean.com/app-framework-services/p/pgvector-extension-for-postgresql)
 - Azure Database for PostgreSQL - vote or comment on [this page](https://feedback.azure.com/d365community/idea/7b423322-6189-ed11-a81b-000d3ae49307)
 
@@ -279,7 +289,7 @@ ALTER EXTENSION vector UPDATE;
 
 ### 0.4.0
 
-For Postgres < 13, remove this line from `sql/vector--0.3.2--0.4.0.sql`:
+If upgrading with Postgres < 13, remove this line from `sql/vector--0.3.2--0.4.0.sql`:
 
 ```sql
 ALTER TYPE vector SET (STORAGE = extended);

sql/vector--0.4.0--0.4.1.sql

@@ -0,0 +1,5 @@
+-- complain if script is sourced in psql, rather than via CREATE EXTENSION
+\echo Use "ALTER EXTENSION vector UPDATE TO '0.4.1'" to load this file. \quit
+
+CREATE FUNCTION random_vector(integer) RETURNS vector
+	AS 'MODULE_PATHNAME' LANGUAGE C VOLATILE STRICT PARALLEL SAFE;

sql/vector.sql

@@ -52,6 +52,9 @@ CREATE FUNCTION vector_add(vector, vector) RETURNS vector
 CREATE FUNCTION vector_sub(vector, vector) RETURNS vector
 	AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
 
+CREATE FUNCTION random_vector(integer) RETURNS vector
+	AS 'MODULE_PATHNAME' LANGUAGE C VOLATILE STRICT PARALLEL SAFE;
+
 -- private functions
 
 CREATE FUNCTION vector_lt(vector, vector) RETURNS bool

src/ivfflat.h

@@ -10,15 +10,10 @@
 #include "access/generic_xlog.h"
 #include "access/reloptions.h"
 #include "nodes/execnodes.h"
-#include "port.h"				/* for strtof() and random() */
 #include "utils/sampling.h"
 #include "utils/tuplesort.h"
 #include "vector.h"
 
-#if PG_VERSION_NUM >= 150000
-#include "common/pg_prng.h"
-#endif
-
 #ifdef IVFFLAT_BENCH
 #include "portability/instr_time.h"
 #endif
@@ -68,14 +63,6 @@
 #define IvfflatBench(name, code) (code)
 #endif
 
-#if PG_VERSION_NUM >= 150000
-#define RandomDouble() pg_prng_double(&pg_global_prng_state)
-#define RandomInt() pg_prng_uint32(&pg_global_prng_state)
-#else
-#define RandomDouble() (((double) random()) / MAX_RANDOM_VALUE)
-#define RandomInt() random()
-#endif
-
 /* Variables */
 extern int	ivfflat_probes;
 

src/vector.c

@@ -950,3 +950,22 @@ vector_avg(PG_FUNCTION_ARGS)
 
 	PG_RETURN_POINTER(result);
 }
+
+/*
+ * Generate a random vector
+ */
+PGDLLEXPORT PG_FUNCTION_INFO_V1(random_vector);
+Datum
+random_vector(PG_FUNCTION_ARGS)
+{
+	int32		dim = PG_GETARG_INT32(0);
+	Vector	   *result;
+
+	CheckDim(dim);
+
+	result = InitVector(dim);
+	for (int i = 0; i < dim; i++)
+		result->x[i] = RandomDouble();
+
+	PG_RETURN_POINTER(result);
+}

src/vector.h

@@ -3,6 +3,12 @@
 
 #include "postgres.h"
 
+#include "port.h"				/* for strtof() and random() */
+
+#if PG_VERSION_NUM >= 150000
+#include "common/pg_prng.h"
+#endif
+
 #define VECTOR_MAX_DIM 16000
 
 #define VECTOR_SIZE(_dim)		(offsetof(Vector, x) + sizeof(float)*(_dim))
@@ -10,6 +16,14 @@
 #define PG_GETARG_VECTOR_P(x)	DatumGetVector(PG_GETARG_DATUM(x))
 #define PG_RETURN_VECTOR_P(x)	PG_RETURN_POINTER(x)
 
+#if PG_VERSION_NUM >= 150000
+#define RandomDouble() pg_prng_double(&pg_global_prng_state)
+#define RandomInt() pg_prng_uint32(&pg_global_prng_state)
+#else
+#define RandomDouble() (((double) random()) / MAX_RANDOM_VALUE)
+#define RandomInt() random()
+#endif
+
 typedef struct Vector
 {
 	int32		vl_len_;		/* varlena header (do not touch directly!) */

test/t/001_wal.pl

@@ -44,11 +44,6 @@ sub test_index_replay
 	return;
 }
 
-# Use ARRAY[random(), random(), random(), ...] over
-# SELECT array_agg(random()) FROM generate_series(1, $dim)
-# to generate different values for each row
-my $array_sql = join(",", ('random()') x $dim);
-
 # Initialize primary node
 $node_primary = get_new_node('primary');
 $node_primary->init(allows_streaming => 1);
@@ -75,7 +70,7 @@ $node_replica->start;
 $node_primary->safe_psql("postgres", "CREATE EXTENSION vector;");
 $node_primary->safe_psql("postgres", "CREATE TABLE tst (i int4, v vector($dim));");
 $node_primary->safe_psql("postgres",
-	"INSERT INTO tst SELECT i % 10, ARRAY[$array_sql] FROM generate_series(1, 100000) i;"
+	"INSERT INTO tst SELECT i % 10, random_vector($dim) FROM generate_series(1, 100000) i;"
 );
 $node_primary->safe_psql("postgres", "CREATE INDEX ON tst USING ivfflat (v);");
 
@@ -91,7 +86,7 @@ for my $i (1 .. 10)
 	test_index_replay("vacuum $i");
 	my ($start, $end) = (100001 + ($i - 1) * 10000, 100000 + $i * 10000);
 	$node_primary->safe_psql("postgres",
-		"INSERT INTO tst SELECT i % 10, ARRAY[$array_sql] FROM generate_series($start, $end) i;"
+		"INSERT INTO tst SELECT i % 10, random_vector($dim) FROM generate_series($start, $end) i;"
 	);
 	test_index_replay("insert $i");
 }

test/t/003_recall.pl

@@ -46,7 +46,7 @@ $node->start;
 $node->safe_psql("postgres", "CREATE EXTENSION vector;");
 $node->safe_psql("postgres", "CREATE TABLE tst (i int4, v vector(3));");
 $node->safe_psql("postgres",
-	"INSERT INTO tst SELECT i, ARRAY[random(), random(), random()] FROM generate_series(1, 100000) i;"
+	"INSERT INTO tst SELECT i, random_vector(3) FROM generate_series(1, 100000) i;"
 );
 
 # Generate queries

test/t/005_query_recall.pl

@@ -13,7 +13,7 @@ $node->start;
 $node->safe_psql("postgres", "CREATE EXTENSION vector;");
 $node->safe_psql("postgres", "CREATE TABLE tst (i int4 primary key, v vector(3));");
 $node->safe_psql("postgres",
-	"INSERT INTO tst SELECT i, ARRAY[random(), random(), random()] FROM generate_series(1, 100000) i;"
+	"INSERT INTO tst SELECT i, random_vector(3) FROM generate_series(1, 100000) i;"
 );
 
 # Check each index type

test/t/006_lists.pl

@@ -13,7 +13,7 @@ $node->start;
 $node->safe_psql("postgres", "CREATE EXTENSION vector;");
 $node->safe_psql("postgres", "CREATE TABLE tst (v vector(3));");
 $node->safe_psql("postgres",
-	"INSERT INTO tst SELECT ARRAY[random(), random(), random()] FROM generate_series(1, 100000) i;"
+	"INSERT INTO tst SELECT random_vector(3) FROM generate_series(1, 100000) i;"
 );
 
 $node->safe_psql("postgres", "CREATE INDEX lists50 ON tst USING ivfflat (v) WITH (lists = 50);");

test/t/007_inserts.pl

@@ -6,8 +6,6 @@ use Test::More tests => 5;
 
 my $dim = 768;
 
-my $array_sql = join(",", ('random()') x $dim);
-
 # Initialize node
 my $node = get_new_node('node');
 $node->init;
@@ -17,7 +15,7 @@ $node->start;
 $node->safe_psql("postgres", "CREATE EXTENSION vector;");
 $node->safe_psql("postgres", "CREATE TABLE tst (v vector($dim));");
 $node->safe_psql("postgres",
-	"INSERT INTO tst SELECT ARRAY[$array_sql] FROM generate_series(1, 10000) i;"
+	"INSERT INTO tst SELECT random_vector($dim) FROM generate_series(1, 10000) i;"
 );
 $node->safe_psql("postgres", "CREATE INDEX ON tst USING ivfflat (v);");
 
@@ -28,7 +26,7 @@ $node->pgbench(
 	[qr{^$}],
 	"concurrent INSERTs",
 	{
-		"007_inserts" => "INSERT INTO tst SELECT ARRAY[$array_sql] FROM generate_series(1, 10) i;"
+		"007_inserts" => "INSERT INTO tst SELECT random_vector($dim) FROM generate_series(1, 10) i;"
 	}
 );
 

test/t/009_storage.pl

@@ -17,7 +17,7 @@ $node->safe_psql("postgres", "CREATE TABLE tst (v1 vector(1024), v2 vector(1024)
 
 # Test insert succeeds
 $node->safe_psql("postgres",
-	"INSERT INTO tst SELECT array_agg(n), array_agg(n), array_agg(n) FROM generate_series(1, $dim) n"
+	"INSERT INTO tst SELECT random_vector($dim), random_vector($dim), random_vector($dim)"
 );
 
 # Change storage to PLAIN
@@ -27,6 +27,6 @@ $node->safe_psql("postgres", "ALTER TABLE tst ALTER COLUMN v3 SET STORAGE PLAIN"
 
 # Test insert fails
 my ($ret, $stdout, $stderr) = $node->psql("postgres",
-	"INSERT INTO tst SELECT array_agg(n), array_agg(n), array_agg(n) FROM generate_series(1, $dim) n"
+	"INSERT INTO tst SELECT random_vector($dim), random_vector($dim), random_vector($dim)"
 );
 like($stderr, qr/row is too big/);