mirror of
https://github.com/pgvector/pgvector.git
synced 2026-07-22 12:07:34 +08:00
Compare commits
71 Commits
| Author | SHA1 | Date | |
|---|---|---|---|
|
|
421116c8cd | ||
|
|
3623c64d82 | ||
|
|
18be1fda24 | ||
|
|
e10a5c90df | ||
|
|
3b0976b725 | ||
|
|
724bde5020 | ||
|
|
261ba4a0d9 | ||
|
|
4293bc439e | ||
|
|
044f0c6441 | ||
|
|
22efa8ec22 | ||
|
|
0a2803130f | ||
|
|
6105a01881 | ||
|
|
9925b92d73 | ||
|
|
e49bae9dc3 | ||
|
|
eea2c44fae | ||
|
|
87ac108bf7 | ||
|
|
97cf990e0f | ||
|
|
55dc735e1a | ||
|
|
be4e9a9df2 | ||
|
|
d5e8fc96a5 | ||
|
|
6d2af6d3f9 | ||
|
|
a6ab5d07c0 | ||
|
|
aa77346103 | ||
|
|
b0da2d95d9 | ||
|
|
3fb05eb847 | ||
|
|
b738ffecc1 | ||
|
|
7117513532 | ||
|
|
85d877d540 | ||
|
|
05fb382031 | ||
|
|
8e1853fbf3 | ||
|
|
f9d68a061a | ||
|
|
4f8ab574c9 | ||
|
|
a15806196e | ||
|
|
5c9429a0f8 | ||
|
|
4b44d6e745 | ||
|
|
16ca608f42 | ||
|
|
8dde14a736 | ||
|
|
d74d3065bc | ||
|
|
a1b80faa67 | ||
|
|
4af5a127e0 | ||
|
|
d02d71a398 | ||
|
|
2aca04b8de | ||
|
|
e47984e616 | ||
|
|
da7d3959a3 | ||
|
|
dadbbc3758 | ||
|
|
6af0a43d62 | ||
|
|
ffcb90d094 | ||
|
|
8a312c3c8e | ||
|
|
5d86b177ab | ||
|
|
ea99957fae | ||
|
|
4cede1a9c9 | ||
|
|
d0dbc8b4d1 | ||
|
|
bb855e6cb4 | ||
|
|
103ac50f1a | ||
|
|
57fb706242 | ||
|
|
020d3edaa9 | ||
|
|
1e9e355175 | ||
|
|
f1d283f565 | ||
|
|
8684c2ba62 | ||
|
|
6c692ef23f | ||
|
|
bb424e96e7 | ||
|
|
30911edb7a | ||
|
|
5ae2bd9efb | ||
|
|
19215317a4 | ||
|
|
c7ca7f05de | ||
|
|
4733cf253b | ||
|
|
06d1fa1402 | ||
|
|
8c5a4bfb6c | ||
|
|
8772c8de68 | ||
|
|
d1694a93af | ||
|
|
61870a0244 |
8
.github/workflows/build.yml
vendored
8
.github/workflows/build.yml
vendored
@@ -8,17 +8,17 @@ jobs:
|
|||||||
fail-fast: false
|
fail-fast: false
|
||||||
matrix:
|
matrix:
|
||||||
include:
|
include:
|
||||||
|
# - postgres: 18
|
||||||
|
# os: ubuntu-24.04
|
||||||
- postgres: 17
|
- postgres: 17
|
||||||
os: ubuntu-22.04
|
os: ubuntu-24.04
|
||||||
- postgres: 16
|
- postgres: 16
|
||||||
os: ubuntu-22.04
|
os: ubuntu-22.04
|
||||||
- postgres: 15
|
- postgres: 15
|
||||||
os: ubuntu-22.04
|
os: ubuntu-22.04
|
||||||
- postgres: 14
|
- postgres: 14
|
||||||
os: ubuntu-22.04
|
|
||||||
- postgres: 13
|
|
||||||
os: ubuntu-20.04
|
os: ubuntu-20.04
|
||||||
- postgres: 12
|
- postgres: 13
|
||||||
os: ubuntu-20.04
|
os: ubuntu-20.04
|
||||||
steps:
|
steps:
|
||||||
- uses: actions/checkout@v4
|
- uses: actions/checkout@v4
|
||||||
|
|||||||
17
CHANGELOG.md
17
CHANGELOG.md
@@ -1,3 +1,20 @@
|
|||||||
|
## 0.8.0 (unreleased)
|
||||||
|
|
||||||
|
- Added casts for arrays to `sparsevec`
|
||||||
|
- Reduced memory usage for HNSW index scans
|
||||||
|
- Dropped support for Postgres 12
|
||||||
|
|
||||||
|
## 0.7.4 (2024-08-05)
|
||||||
|
|
||||||
|
- Fixed locking for parallel HNSW index builds
|
||||||
|
- Fixed compilation error with GCC 14 on i386 when SSE2 is not enabled
|
||||||
|
|
||||||
|
## 0.7.3 (2024-07-22)
|
||||||
|
|
||||||
|
- Fixed `failed to add index item` error with `sparsevec`
|
||||||
|
- Fixed compilation error with FreeBSD ARM
|
||||||
|
- Fixed compilation warning with MSVC and Postgres 16
|
||||||
|
|
||||||
## 0.7.2 (2024-06-11)
|
## 0.7.2 (2024-06-11)
|
||||||
|
|
||||||
- Fixed initialization fork for indexes on unlogged tables
|
- Fixed initialization fork for indexes on unlogged tables
|
||||||
|
|||||||
@@ -2,7 +2,7 @@
|
|||||||
"name": "vector",
|
"name": "vector",
|
||||||
"abstract": "Open-source vector similarity search for Postgres",
|
"abstract": "Open-source vector similarity search for Postgres",
|
||||||
"description": "Supports L2 distance, inner product, and cosine distance",
|
"description": "Supports L2 distance, inner product, and cosine distance",
|
||||||
"version": "0.7.2",
|
"version": "0.7.4",
|
||||||
"maintainer": [
|
"maintainer": [
|
||||||
"Andrew Kane <andrew@ankane.org>"
|
"Andrew Kane <andrew@ankane.org>"
|
||||||
],
|
],
|
||||||
@@ -20,7 +20,7 @@
|
|||||||
"vector": {
|
"vector": {
|
||||||
"file": "sql/vector.sql",
|
"file": "sql/vector.sql",
|
||||||
"docfile": "README.md",
|
"docfile": "README.md",
|
||||||
"version": "0.7.2",
|
"version": "0.7.4",
|
||||||
"abstract": "Open-source vector similarity search for Postgres"
|
"abstract": "Open-source vector similarity search for Postgres"
|
||||||
}
|
}
|
||||||
},
|
},
|
||||||
|
|||||||
13
Makefile
13
Makefile
@@ -1,10 +1,11 @@
|
|||||||
EXTENSION = vector
|
EXTENSION = vector
|
||||||
EXTVERSION = 0.7.2
|
EXTVERSION = 0.7.4
|
||||||
|
|
||||||
MODULE_big = vector
|
MODULE_big = vector
|
||||||
DATA = $(wildcard sql/*--*.sql)
|
DATA = $(wildcard sql/*--*--*.sql)
|
||||||
OBJS = src/bitutils.o src/bitvec.o src/halfutils.o src/halfvec.o src/hnsw.o src/hnswbuild.o src/hnswinsert.o src/hnswscan.o src/hnswutils.o src/hnswvacuum.o src/ivfbuild.o src/ivfflat.o src/ivfinsert.o src/ivfkmeans.o src/ivfscan.o src/ivfutils.o src/ivfvacuum.o src/sparsevec.o src/vector.o
|
DATA_built = sql/$(EXTENSION)--$(EXTVERSION).sql
|
||||||
HEADERS = src/halfvec.h src/sparsevec.h src/vector.h
|
OBJS = src/bitutils.o src/bitvec.o src/halfutils.o src/halfvec.o src/hnsw.o src/hnswbuild.o src/hnswinsert.o src/hnswscan.o src/hnswutils.o src/hnswvacuum.o src/ivfbuild.o src/ivfflat.o src/ivfinsert.o src/ivfkmeans.o src/ivfscan.o src/ivfutils.o src/minivec.o src/ivfvacuum.o src/sparsevec.o src/vector.o
|
||||||
|
HEADERS = src/halfvec.h src/minivec.h src/sparsevec.h src/vector.h
|
||||||
|
|
||||||
TESTS = $(wildcard test/sql/*.sql)
|
TESTS = $(wildcard test/sql/*.sql)
|
||||||
REGRESS = $(patsubst test/sql/%.sql,%,$(TESTS))
|
REGRESS = $(patsubst test/sql/%.sql,%,$(TESTS))
|
||||||
@@ -42,8 +43,6 @@ all: sql/$(EXTENSION)--$(EXTVERSION).sql
|
|||||||
sql/$(EXTENSION)--$(EXTVERSION).sql: sql/$(EXTENSION).sql
|
sql/$(EXTENSION)--$(EXTVERSION).sql: sql/$(EXTENSION).sql
|
||||||
cp $< $@
|
cp $< $@
|
||||||
|
|
||||||
EXTRA_CLEAN = sql/$(EXTENSION)--$(EXTVERSION).sql
|
|
||||||
|
|
||||||
PG_CONFIG ?= pg_config
|
PG_CONFIG ?= pg_config
|
||||||
PGXS := $(shell $(PG_CONFIG) --pgxs)
|
PGXS := $(shell $(PG_CONFIG) --pgxs)
|
||||||
include $(PGXS)
|
include $(PGXS)
|
||||||
@@ -53,7 +52,7 @@ ifeq ($(PROVE),)
|
|||||||
PROVE = prove
|
PROVE = prove
|
||||||
endif
|
endif
|
||||||
|
|
||||||
# for Postgres 15
|
# for Postgres < 15
|
||||||
PROVE_FLAGS += -I ./test/perl
|
PROVE_FLAGS += -I ./test/perl
|
||||||
|
|
||||||
prove_installcheck:
|
prove_installcheck:
|
||||||
|
|||||||
23
Makefile.win
23
Makefile.win
@@ -1,10 +1,11 @@
|
|||||||
EXTENSION = vector
|
EXTENSION = vector
|
||||||
EXTVERSION = 0.7.2
|
EXTVERSION = 0.7.4
|
||||||
|
|
||||||
OBJS = src\bitutils.obj src\bitvec.obj src\halfutils.obj src\halfvec.obj src\hnsw.obj src\hnswbuild.obj src\hnswinsert.obj src\hnswscan.obj src\hnswutils.obj src\hnswvacuum.obj src\ivfbuild.obj src\ivfflat.obj src\ivfinsert.obj src\ivfkmeans.obj src\ivfscan.obj src\ivfutils.obj src\ivfvacuum.obj src\sparsevec.obj src\vector.obj
|
DATA_built = sql\$(EXTENSION)--$(EXTVERSION).sql
|
||||||
HEADERS = src\halfvec.h src\sparsevec.h src\vector.h
|
OBJS = src\bitutils.obj src\bitvec.obj src\halfutils.obj src\halfvec.obj src\hnsw.obj src\hnswbuild.obj src\hnswinsert.obj src\hnswscan.obj src\hnswutils.obj src\hnswvacuum.obj src\ivfbuild.obj src\ivfflat.obj src\ivfinsert.obj src\ivfkmeans.obj src\ivfscan.obj src\ivfutils.obj src\ivfvacuum.obj src\minivec.obj src\sparsevec.obj src\vector.obj
|
||||||
|
HEADERS = src\halfvec.h src\minivec.h src\sparsevec.h src\vector.h
|
||||||
|
|
||||||
REGRESS = bit btree cast copy halfvec hnsw_bit hnsw_halfvec hnsw_sparsevec hnsw_vector ivfflat_bit ivfflat_halfvec ivfflat_vector sparsevec vector_type
|
REGRESS = bit btree cast copy halfvec hnsw_bit hnsw_halfvec hnsw_sparsevec hnsw_vector ivfflat_bit ivfflat_halfvec ivfflat_vector minivec sparsevec vector_type
|
||||||
REGRESS_OPTS = --inputdir=test --load-extension=$(EXTENSION)
|
REGRESS_OPTS = --inputdir=test --load-extension=$(EXTENSION)
|
||||||
|
|
||||||
# For /arch flags
|
# For /arch flags
|
||||||
@@ -19,11 +20,6 @@ PG_CFLAGS = $(PG_CFLAGS) $(OPTFLAGS) /O2 /fp:fast
|
|||||||
# https://learn.microsoft.com/en-us/cpp/error-messages/tool-errors/vectorizer-and-parallelizer-messages
|
# https://learn.microsoft.com/en-us/cpp/error-messages/tool-errors/vectorizer-and-parallelizer-messages
|
||||||
# PG_CFLAGS = $(PG_CFLAGS) /Qvec-report:2
|
# PG_CFLAGS = $(PG_CFLAGS) /Qvec-report:2
|
||||||
|
|
||||||
all: sql\$(EXTENSION)--$(EXTVERSION).sql
|
|
||||||
|
|
||||||
sql\$(EXTENSION)--$(EXTVERSION).sql: sql\$(EXTENSION).sql
|
|
||||||
copy sql\$(EXTENSION).sql $@
|
|
||||||
|
|
||||||
# TODO use pg_config
|
# TODO use pg_config
|
||||||
!ifndef PGROOT
|
!ifndef PGROOT
|
||||||
!error PGROOT is not set
|
!error PGROOT is not set
|
||||||
@@ -43,15 +39,18 @@ SHLIB = $(EXTENSION).dll
|
|||||||
|
|
||||||
LIBS = "$(LIBDIR)\postgres.lib"
|
LIBS = "$(LIBDIR)\postgres.lib"
|
||||||
|
|
||||||
|
all: $(SHLIB) $(DATA_built)
|
||||||
|
|
||||||
.c.obj:
|
.c.obj:
|
||||||
$(CC) $(CFLAGS) /c $< /Fo$@
|
$(CC) $(CFLAGS) /c $< /Fo$@
|
||||||
|
|
||||||
$(SHLIB): $(OBJS)
|
$(SHLIB): $(OBJS)
|
||||||
$(CC) $(CFLAGS) $(OBJS) $(LIBS) /link /DLL /OUT:$(SHLIB)
|
$(CC) $(CFLAGS) $(OBJS) $(LIBS) /link /DLL /OUT:$(SHLIB)
|
||||||
|
|
||||||
all: $(SHLIB)
|
sql\$(EXTENSION)--$(EXTVERSION).sql: sql\$(EXTENSION).sql
|
||||||
|
copy sql\$(EXTENSION).sql $@
|
||||||
|
|
||||||
install:
|
install: all
|
||||||
copy $(SHLIB) "$(PKGLIBDIR)"
|
copy $(SHLIB) "$(PKGLIBDIR)"
|
||||||
copy $(EXTENSION).control "$(SHAREDIR)\extension"
|
copy $(EXTENSION).control "$(SHAREDIR)\extension"
|
||||||
copy sql\$(EXTENSION)--*.sql "$(SHAREDIR)\extension"
|
copy sql\$(EXTENSION)--*.sql "$(SHAREDIR)\extension"
|
||||||
@@ -70,6 +69,6 @@ uninstall:
|
|||||||
|
|
||||||
clean:
|
clean:
|
||||||
del /f $(SHLIB) $(EXTENSION).lib $(EXTENSION).exp
|
del /f $(SHLIB) $(EXTENSION).lib $(EXTENSION).exp
|
||||||
|
del /f $(DATA_built)
|
||||||
del /f $(OBJS)
|
del /f $(OBJS)
|
||||||
del /f sql\$(EXTENSION)--$(EXTVERSION).sql
|
|
||||||
del /f /s /q results regression.diffs regression.out tmp_check tmp_check_iso log output_iso
|
del /f /s /q results regression.diffs regression.out tmp_check tmp_check_iso log output_iso
|
||||||
|
|||||||
49
README.md
49
README.md
@@ -21,7 +21,7 @@ Compile and install the extension (supports Postgres 12+)
|
|||||||
|
|
||||||
```sh
|
```sh
|
||||||
cd /tmp
|
cd /tmp
|
||||||
git clone --branch v0.7.2 https://github.com/pgvector/pgvector.git
|
git clone --branch v0.7.4 https://github.com/pgvector/pgvector.git
|
||||||
cd pgvector
|
cd pgvector
|
||||||
make
|
make
|
||||||
make install # may need sudo
|
make install # may need sudo
|
||||||
@@ -46,7 +46,7 @@ Then use `nmake` to build:
|
|||||||
```cmd
|
```cmd
|
||||||
set "PGROOT=C:\Program Files\PostgreSQL\16"
|
set "PGROOT=C:\Program Files\PostgreSQL\16"
|
||||||
cd %TEMP%
|
cd %TEMP%
|
||||||
git clone --branch v0.7.2 https://github.com/pgvector/pgvector.git
|
git clone --branch v0.7.4 https://github.com/pgvector/pgvector.git
|
||||||
cd pgvector
|
cd pgvector
|
||||||
nmake /F Makefile.win
|
nmake /F Makefile.win
|
||||||
nmake /F Makefile.win install
|
nmake /F Makefile.win install
|
||||||
@@ -106,7 +106,7 @@ Insert vectors
|
|||||||
INSERT INTO items (embedding) VALUES ('[1,2,3]'), ('[4,5,6]');
|
INSERT INTO items (embedding) VALUES ('[1,2,3]'), ('[4,5,6]');
|
||||||
```
|
```
|
||||||
|
|
||||||
Or load vectors in bulk using `COPY` ([example](https://github.com/pgvector/pgvector-python/blob/master/examples/bulk_loading.py))
|
Or load vectors in bulk using `COPY` ([example](https://github.com/pgvector/pgvector-python/blob/master/examples/loading/example.py))
|
||||||
|
|
||||||
```sql
|
```sql
|
||||||
COPY items (embedding) FROM STDIN WITH (FORMAT BINARY);
|
COPY items (embedding) FROM STDIN WITH (FORMAT BINARY);
|
||||||
@@ -202,7 +202,7 @@ You can add an index to use approximate nearest neighbor search, which trades so
|
|||||||
|
|
||||||
Supported index types are:
|
Supported index types are:
|
||||||
|
|
||||||
- [HNSW](#hnsw) - added in 0.5.0
|
- [HNSW](#hnsw)
|
||||||
- [IVFFlat](#ivfflat)
|
- [IVFFlat](#ivfflat)
|
||||||
|
|
||||||
## HNSW
|
## HNSW
|
||||||
@@ -473,7 +473,7 @@ SELECT * FROM items ORDER BY embedding::halfvec(3) <-> '[1,2,3]' LIMIT 5;
|
|||||||
|
|
||||||
## Binary Vectors
|
## Binary Vectors
|
||||||
|
|
||||||
Use the `bit` type to store binary vectors ([example](https://github.com/pgvector/pgvector-python/blob/master/examples/hash_image_search.py))
|
Use the `bit` type to store binary vectors ([example](https://github.com/pgvector/pgvector-python/blob/master/examples/imagehash/example.py))
|
||||||
|
|
||||||
```sql
|
```sql
|
||||||
CREATE TABLE items (id bigserial PRIMARY KEY, embedding bit(3));
|
CREATE TABLE items (id bigserial PRIMARY KEY, embedding bit(3));
|
||||||
@@ -551,7 +551,7 @@ SELECT id, content FROM items, plainto_tsquery('hello search') query
|
|||||||
WHERE textsearch @@ query ORDER BY ts_rank_cd(textsearch, query) DESC LIMIT 5;
|
WHERE textsearch @@ query ORDER BY ts_rank_cd(textsearch, query) DESC LIMIT 5;
|
||||||
```
|
```
|
||||||
|
|
||||||
You can use [Reciprocal Rank Fusion](https://github.com/pgvector/pgvector-python/blob/master/examples/hybrid_search_rrf.py) or a [cross-encoder](https://github.com/pgvector/pgvector-python/blob/master/examples/hybrid_search.py) to combine results.
|
You can use [Reciprocal Rank Fusion](https://github.com/pgvector/pgvector-python/blob/master/examples/hybrid_search/rrf.py) or a [cross-encoder](https://github.com/pgvector/pgvector-python/blob/master/examples/hybrid_search/cross_encoder.py) to combine results.
|
||||||
|
|
||||||
## Indexing Subvectors
|
## Indexing Subvectors
|
||||||
|
|
||||||
@@ -597,7 +597,7 @@ Be sure to restart Postgres for changes to take effect.
|
|||||||
|
|
||||||
### Loading
|
### Loading
|
||||||
|
|
||||||
Use `COPY` for bulk loading data ([example](https://github.com/pgvector/pgvector-python/blob/master/examples/bulk_loading.py)).
|
Use `COPY` for bulk loading data ([example](https://github.com/pgvector/pgvector-python/blob/master/examples/loading/example.py)).
|
||||||
|
|
||||||
```sql
|
```sql
|
||||||
COPY items (embedding) FROM STDIN WITH (FORMAT BINARY);
|
COPY items (embedding) FROM STDIN WITH (FORMAT BINARY);
|
||||||
@@ -687,7 +687,7 @@ Scale pgvector the same way you scale Postgres.
|
|||||||
|
|
||||||
Scale vertically by increasing memory, CPU, and storage on a single instance. Use existing tools to [tune parameters](#tuning) and [monitor performance](#monitoring).
|
Scale vertically by increasing memory, CPU, and storage on a single instance. Use existing tools to [tune parameters](#tuning) and [monitor performance](#monitoring).
|
||||||
|
|
||||||
Scale horizontally with [replicas](https://www.postgresql.org/docs/current/hot-standby.html), or use [Citus](https://github.com/citusdata/citus) or another approach for sharding ([example](https://github.com/pgvector/pgvector-python/blob/master/examples/citus.py)).
|
Scale horizontally with [replicas](https://www.postgresql.org/docs/current/hot-standby.html), or use [Citus](https://github.com/citusdata/citus) or another approach for sharding ([example](https://github.com/pgvector/pgvector-python/blob/master/examples/citus/example.py)).
|
||||||
|
|
||||||
## Languages
|
## Languages
|
||||||
|
|
||||||
@@ -934,6 +934,37 @@ Function | Description | Added
|
|||||||
avg(halfvec) → halfvec | average | 0.7.0
|
avg(halfvec) → halfvec | average | 0.7.0
|
||||||
sum(halfvec) → halfvec | sum | 0.7.0
|
sum(halfvec) → halfvec | sum | 0.7.0
|
||||||
|
|
||||||
|
### Minivec Type
|
||||||
|
|
||||||
|
Each mini vector takes `dimensions + 8` bytes of storage. Each element is a E4M3 8-bit floating-point number, and all elements must be finite (no `NaN`). Mini vectors can have up to 16,000 dimensions.
|
||||||
|
|
||||||
|
### Minivec Operators
|
||||||
|
|
||||||
|
Operator | Description | Added
|
||||||
|
--- | --- | ---
|
||||||
|
\+ | element-wise addition | 0.8.0
|
||||||
|
\- | element-wise subtraction | 0.8.0
|
||||||
|
\* | element-wise multiplication | 0.8.0
|
||||||
|
\|\| | concatenate | 0.8.0
|
||||||
|
<-> | Euclidean distance | 0.8.0
|
||||||
|
<#> | negative inner product | 0.8.0
|
||||||
|
<=> | cosine distance | 0.8.0
|
||||||
|
<+> | taxicab distance | 0.8.0
|
||||||
|
|
||||||
|
### Minivec Functions
|
||||||
|
|
||||||
|
Function | Description | Added
|
||||||
|
--- | --- | ---
|
||||||
|
binary_quantize(minivec) → bit | binary quantize | 0.8.0
|
||||||
|
cosine_distance(minivec, minivec) → double precision | cosine distance | 0.8.0
|
||||||
|
inner_product(minivec, minivec) → double precision | inner product | 0.8.0
|
||||||
|
l1_distance(minivec, minivec) → double precision | taxicab distance | 0.8.0
|
||||||
|
l2_distance(minivec, minivec) → double precision | Euclidean distance | 0.8.0
|
||||||
|
l2_norm(minivec) → double precision | Euclidean norm | 0.8.0
|
||||||
|
l2_normalize(minivec) → minivec | Normalize with Euclidean norm | 0.8.0
|
||||||
|
subvector(minivec, integer, integer) → minivec | subvector | 0.8.0
|
||||||
|
vector_dims(minivec) → integer | number of dimensions | 0.8.0
|
||||||
|
|
||||||
### Bit Type
|
### Bit Type
|
||||||
|
|
||||||
Each bit vector takes `dimensions / 8 + 8` bytes of storage. See the [Postgres docs](https://www.postgresql.org/docs/current/datatype-bit.html) for more info.
|
Each bit vector takes `dimensions / 8 + 8` bytes of storage. See the [Postgres docs](https://www.postgresql.org/docs/current/datatype-bit.html) for more info.
|
||||||
@@ -1051,7 +1082,7 @@ This adds pgvector to the [Postgres image](https://hub.docker.com/_/postgres) (r
|
|||||||
You can also build the image manually:
|
You can also build the image manually:
|
||||||
|
|
||||||
```sh
|
```sh
|
||||||
git clone --branch v0.7.2 https://github.com/pgvector/pgvector.git
|
git clone --branch v0.7.4 https://github.com/pgvector/pgvector.git
|
||||||
cd pgvector
|
cd pgvector
|
||||||
docker build --pull --build-arg PG_MAJOR=16 -t myuser/pgvector .
|
docker build --pull --build-arg PG_MAJOR=16 -t myuser/pgvector .
|
||||||
```
|
```
|
||||||
|
|||||||
2
sql/vector--0.7.2--0.7.3.sql
Normal file
2
sql/vector--0.7.2--0.7.3.sql
Normal file
@@ -0,0 +1,2 @@
|
|||||||
|
-- complain if script is sourced in psql, rather than via CREATE EXTENSION
|
||||||
|
\echo Use "ALTER EXTENSION vector UPDATE TO '0.7.3'" to load this file. \quit
|
||||||
2
sql/vector--0.7.3--0.7.4.sql
Normal file
2
sql/vector--0.7.3--0.7.4.sql
Normal file
@@ -0,0 +1,2 @@
|
|||||||
|
-- complain if script is sourced in psql, rather than via CREATE EXTENSION
|
||||||
|
\echo Use "ALTER EXTENSION vector UPDATE TO '0.7.4'" to load this file. \quit
|
||||||
28
sql/vector--0.7.4--0.8.0.sql
Normal file
28
sql/vector--0.7.4--0.8.0.sql
Normal file
@@ -0,0 +1,28 @@
|
|||||||
|
-- complain if script is sourced in psql, rather than via CREATE EXTENSION
|
||||||
|
\echo Use "ALTER EXTENSION vector UPDATE TO '0.8.0'" to load this file. \quit
|
||||||
|
|
||||||
|
-- TODO minivec functions
|
||||||
|
|
||||||
|
CREATE FUNCTION array_to_sparsevec(integer[], integer, boolean) RETURNS sparsevec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION array_to_sparsevec(real[], integer, boolean) RETURNS sparsevec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION array_to_sparsevec(double precision[], integer, boolean) RETURNS sparsevec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION array_to_sparsevec(numeric[], integer, boolean) RETURNS sparsevec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE CAST (integer[] AS sparsevec)
|
||||||
|
WITH FUNCTION array_to_sparsevec(integer[], integer, boolean) AS ASSIGNMENT;
|
||||||
|
|
||||||
|
CREATE CAST (real[] AS sparsevec)
|
||||||
|
WITH FUNCTION array_to_sparsevec(real[], integer, boolean) AS ASSIGNMENT;
|
||||||
|
|
||||||
|
CREATE CAST (double precision[] AS sparsevec)
|
||||||
|
WITH FUNCTION array_to_sparsevec(double precision[], integer, boolean) AS ASSIGNMENT;
|
||||||
|
|
||||||
|
CREATE CAST (numeric[] AS sparsevec)
|
||||||
|
WITH FUNCTION array_to_sparsevec(numeric[], integer, boolean) AS ASSIGNMENT;
|
||||||
319
sql/vector.sql
319
sql/vector.sql
@@ -266,12 +266,18 @@ COMMENT ON ACCESS METHOD hnsw IS 'hnsw index access method';
|
|||||||
CREATE FUNCTION ivfflat_halfvec_support(internal) RETURNS internal
|
CREATE FUNCTION ivfflat_halfvec_support(internal) RETURNS internal
|
||||||
AS 'MODULE_PATHNAME' LANGUAGE C;
|
AS 'MODULE_PATHNAME' LANGUAGE C;
|
||||||
|
|
||||||
|
CREATE FUNCTION ivfflat_minivec_support(internal) RETURNS internal
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C;
|
||||||
|
|
||||||
CREATE FUNCTION ivfflat_bit_support(internal) RETURNS internal
|
CREATE FUNCTION ivfflat_bit_support(internal) RETURNS internal
|
||||||
AS 'MODULE_PATHNAME' LANGUAGE C;
|
AS 'MODULE_PATHNAME' LANGUAGE C;
|
||||||
|
|
||||||
CREATE FUNCTION hnsw_halfvec_support(internal) RETURNS internal
|
CREATE FUNCTION hnsw_halfvec_support(internal) RETURNS internal
|
||||||
AS 'MODULE_PATHNAME' LANGUAGE C;
|
AS 'MODULE_PATHNAME' LANGUAGE C;
|
||||||
|
|
||||||
|
CREATE FUNCTION hnsw_minivec_support(internal) RETURNS internal
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C;
|
||||||
|
|
||||||
CREATE FUNCTION hnsw_bit_support(internal) RETURNS internal
|
CREATE FUNCTION hnsw_bit_support(internal) RETURNS internal
|
||||||
AS 'MODULE_PATHNAME' LANGUAGE C;
|
AS 'MODULE_PATHNAME' LANGUAGE C;
|
||||||
|
|
||||||
@@ -647,6 +653,295 @@ CREATE OPERATOR CLASS halfvec_l1_ops
|
|||||||
FUNCTION 1 l1_distance(halfvec, halfvec),
|
FUNCTION 1 l1_distance(halfvec, halfvec),
|
||||||
FUNCTION 3 hnsw_halfvec_support(internal);
|
FUNCTION 3 hnsw_halfvec_support(internal);
|
||||||
|
|
||||||
|
-- minivec type
|
||||||
|
|
||||||
|
CREATE TYPE minivec;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_in(cstring, oid, integer) RETURNS minivec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_out(minivec) RETURNS cstring
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_typmod_in(cstring[]) RETURNS integer
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_recv(internal, oid, integer) RETURNS minivec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_send(minivec) RETURNS bytea
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE TYPE minivec (
|
||||||
|
INPUT = minivec_in,
|
||||||
|
OUTPUT = minivec_out,
|
||||||
|
TYPMOD_IN = minivec_typmod_in,
|
||||||
|
RECEIVE = minivec_recv,
|
||||||
|
SEND = minivec_send,
|
||||||
|
STORAGE = external
|
||||||
|
);
|
||||||
|
|
||||||
|
-- minivec functions
|
||||||
|
|
||||||
|
CREATE FUNCTION l2_distance(minivec, minivec) RETURNS float8
|
||||||
|
AS 'MODULE_PATHNAME', 'minivec_l2_distance' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION inner_product(minivec, minivec) RETURNS float8
|
||||||
|
AS 'MODULE_PATHNAME', 'minivec_inner_product' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION cosine_distance(minivec, minivec) RETURNS float8
|
||||||
|
AS 'MODULE_PATHNAME', 'minivec_cosine_distance' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION l1_distance(minivec, minivec) RETURNS float8
|
||||||
|
AS 'MODULE_PATHNAME', 'minivec_l1_distance' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION vector_dims(minivec) RETURNS integer
|
||||||
|
AS 'MODULE_PATHNAME', 'minivec_vector_dims' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION l2_norm(minivec) RETURNS float8
|
||||||
|
AS 'MODULE_PATHNAME', 'minivec_l2_norm' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION l2_normalize(minivec) RETURNS minivec
|
||||||
|
AS 'MODULE_PATHNAME', 'minivec_l2_normalize' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION binary_quantize(minivec) RETURNS bit
|
||||||
|
AS 'MODULE_PATHNAME', 'minivec_binary_quantize' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION subvector(minivec, int, int) RETURNS minivec
|
||||||
|
AS 'MODULE_PATHNAME', 'minivec_subvector' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
-- minivec private functions
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_add(minivec, minivec) RETURNS minivec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_sub(minivec, minivec) RETURNS minivec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_mul(minivec, minivec) RETURNS minivec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_concat(minivec, minivec) RETURNS minivec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_lt(minivec, minivec) RETURNS bool
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_le(minivec, minivec) RETURNS bool
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_eq(minivec, minivec) RETURNS bool
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_ne(minivec, minivec) RETURNS bool
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_ge(minivec, minivec) RETURNS bool
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_gt(minivec, minivec) RETURNS bool
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_cmp(minivec, minivec) RETURNS int4
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_l2_squared_distance(minivec, minivec) RETURNS float8
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_negative_inner_product(minivec, minivec) RETURNS float8
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_spherical_distance(minivec, minivec) RETURNS float8
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
-- minivec cast functions
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec(minivec, integer, boolean) RETURNS minivec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_to_vector(minivec, integer, boolean) RETURNS vector
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION vector_to_minivec(vector, integer, boolean) RETURNS minivec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION array_to_minivec(integer[], integer, boolean) RETURNS minivec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION array_to_minivec(real[], integer, boolean) RETURNS minivec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION array_to_minivec(double precision[], integer, boolean) RETURNS minivec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION array_to_minivec(numeric[], integer, boolean) RETURNS minivec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION minivec_to_float4(minivec, integer, boolean) RETURNS real[]
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
-- minivec casts
|
||||||
|
|
||||||
|
CREATE CAST (minivec AS minivec)
|
||||||
|
WITH FUNCTION minivec(minivec, integer, boolean) AS IMPLICIT;
|
||||||
|
|
||||||
|
CREATE CAST (minivec AS vector)
|
||||||
|
WITH FUNCTION minivec_to_vector(minivec, integer, boolean) AS ASSIGNMENT;
|
||||||
|
|
||||||
|
CREATE CAST (vector AS minivec)
|
||||||
|
WITH FUNCTION vector_to_minivec(vector, integer, boolean) AS IMPLICIT;
|
||||||
|
|
||||||
|
CREATE CAST (minivec AS real[])
|
||||||
|
WITH FUNCTION minivec_to_float4(minivec, integer, boolean) AS ASSIGNMENT;
|
||||||
|
|
||||||
|
CREATE CAST (integer[] AS minivec)
|
||||||
|
WITH FUNCTION array_to_minivec(integer[], integer, boolean) AS ASSIGNMENT;
|
||||||
|
|
||||||
|
CREATE CAST (real[] AS minivec)
|
||||||
|
WITH FUNCTION array_to_minivec(real[], integer, boolean) AS ASSIGNMENT;
|
||||||
|
|
||||||
|
CREATE CAST (double precision[] AS minivec)
|
||||||
|
WITH FUNCTION array_to_minivec(double precision[], integer, boolean) AS ASSIGNMENT;
|
||||||
|
|
||||||
|
CREATE CAST (numeric[] AS minivec)
|
||||||
|
WITH FUNCTION array_to_minivec(numeric[], integer, boolean) AS ASSIGNMENT;
|
||||||
|
|
||||||
|
-- minivec operators
|
||||||
|
|
||||||
|
CREATE OPERATOR <-> (
|
||||||
|
LEFTARG = minivec, RIGHTARG = minivec, PROCEDURE = l2_distance,
|
||||||
|
COMMUTATOR = '<->'
|
||||||
|
);
|
||||||
|
|
||||||
|
CREATE OPERATOR <#> (
|
||||||
|
LEFTARG = minivec, RIGHTARG = minivec, PROCEDURE = minivec_negative_inner_product,
|
||||||
|
COMMUTATOR = '<#>'
|
||||||
|
);
|
||||||
|
|
||||||
|
CREATE OPERATOR <=> (
|
||||||
|
LEFTARG = minivec, RIGHTARG = minivec, PROCEDURE = cosine_distance,
|
||||||
|
COMMUTATOR = '<=>'
|
||||||
|
);
|
||||||
|
|
||||||
|
CREATE OPERATOR <+> (
|
||||||
|
LEFTARG = minivec, RIGHTARG = minivec, PROCEDURE = l1_distance,
|
||||||
|
COMMUTATOR = '<+>'
|
||||||
|
);
|
||||||
|
|
||||||
|
CREATE OPERATOR + (
|
||||||
|
LEFTARG = minivec, RIGHTARG = minivec, PROCEDURE = minivec_add,
|
||||||
|
COMMUTATOR = +
|
||||||
|
);
|
||||||
|
|
||||||
|
CREATE OPERATOR - (
|
||||||
|
LEFTARG = minivec, RIGHTARG = minivec, PROCEDURE = minivec_sub
|
||||||
|
);
|
||||||
|
|
||||||
|
CREATE OPERATOR * (
|
||||||
|
LEFTARG = minivec, RIGHTARG = minivec, PROCEDURE = minivec_mul,
|
||||||
|
COMMUTATOR = *
|
||||||
|
);
|
||||||
|
|
||||||
|
CREATE OPERATOR || (
|
||||||
|
LEFTARG = minivec, RIGHTARG = minivec, PROCEDURE = minivec_concat
|
||||||
|
);
|
||||||
|
|
||||||
|
CREATE OPERATOR < (
|
||||||
|
LEFTARG = minivec, RIGHTARG = minivec, PROCEDURE = minivec_lt,
|
||||||
|
COMMUTATOR = > , NEGATOR = >= ,
|
||||||
|
RESTRICT = scalarltsel, JOIN = scalarltjoinsel
|
||||||
|
);
|
||||||
|
|
||||||
|
CREATE OPERATOR <= (
|
||||||
|
LEFTARG = minivec, RIGHTARG = minivec, PROCEDURE = minivec_le,
|
||||||
|
COMMUTATOR = >= , NEGATOR = > ,
|
||||||
|
RESTRICT = scalarlesel, JOIN = scalarlejoinsel
|
||||||
|
);
|
||||||
|
|
||||||
|
CREATE OPERATOR = (
|
||||||
|
LEFTARG = minivec, RIGHTARG = minivec, PROCEDURE = minivec_eq,
|
||||||
|
COMMUTATOR = = , NEGATOR = <> ,
|
||||||
|
RESTRICT = eqsel, JOIN = eqjoinsel
|
||||||
|
);
|
||||||
|
|
||||||
|
CREATE OPERATOR <> (
|
||||||
|
LEFTARG = minivec, RIGHTARG = minivec, PROCEDURE = minivec_ne,
|
||||||
|
COMMUTATOR = <> , NEGATOR = = ,
|
||||||
|
RESTRICT = eqsel, JOIN = eqjoinsel
|
||||||
|
);
|
||||||
|
|
||||||
|
CREATE OPERATOR >= (
|
||||||
|
LEFTARG = minivec, RIGHTARG = minivec, PROCEDURE = minivec_ge,
|
||||||
|
COMMUTATOR = <= , NEGATOR = < ,
|
||||||
|
RESTRICT = scalargesel, JOIN = scalargejoinsel
|
||||||
|
);
|
||||||
|
|
||||||
|
CREATE OPERATOR > (
|
||||||
|
LEFTARG = minivec, RIGHTARG = minivec, PROCEDURE = minivec_gt,
|
||||||
|
COMMUTATOR = < , NEGATOR = <= ,
|
||||||
|
RESTRICT = scalargtsel, JOIN = scalargtjoinsel
|
||||||
|
);
|
||||||
|
|
||||||
|
-- minivec op classes
|
||||||
|
|
||||||
|
CREATE OPERATOR CLASS minivec_ops
|
||||||
|
DEFAULT FOR TYPE minivec USING btree AS
|
||||||
|
OPERATOR 1 < ,
|
||||||
|
OPERATOR 2 <= ,
|
||||||
|
OPERATOR 3 = ,
|
||||||
|
OPERATOR 4 >= ,
|
||||||
|
OPERATOR 5 > ,
|
||||||
|
FUNCTION 1 minivec_cmp(minivec, minivec);
|
||||||
|
|
||||||
|
CREATE OPERATOR CLASS minivec_l2_ops
|
||||||
|
FOR TYPE minivec USING ivfflat AS
|
||||||
|
OPERATOR 1 <-> (minivec, minivec) FOR ORDER BY float_ops,
|
||||||
|
FUNCTION 1 minivec_l2_squared_distance(minivec, minivec),
|
||||||
|
FUNCTION 3 l2_distance(minivec, minivec),
|
||||||
|
FUNCTION 5 ivfflat_minivec_support(internal);
|
||||||
|
|
||||||
|
CREATE OPERATOR CLASS minivec_ip_ops
|
||||||
|
FOR TYPE minivec USING ivfflat AS
|
||||||
|
OPERATOR 1 <#> (minivec, minivec) FOR ORDER BY float_ops,
|
||||||
|
FUNCTION 1 minivec_negative_inner_product(minivec, minivec),
|
||||||
|
FUNCTION 3 minivec_spherical_distance(minivec, minivec),
|
||||||
|
FUNCTION 4 l2_norm(minivec),
|
||||||
|
FUNCTION 5 ivfflat_minivec_support(internal);
|
||||||
|
|
||||||
|
CREATE OPERATOR CLASS minivec_cosine_ops
|
||||||
|
FOR TYPE minivec USING ivfflat AS
|
||||||
|
OPERATOR 1 <=> (minivec, minivec) FOR ORDER BY float_ops,
|
||||||
|
FUNCTION 1 cosine_distance(minivec, minivec),
|
||||||
|
FUNCTION 2 l2_norm(minivec),
|
||||||
|
FUNCTION 3 minivec_spherical_distance(minivec, minivec),
|
||||||
|
FUNCTION 4 l2_norm(minivec),
|
||||||
|
FUNCTION 5 ivfflat_minivec_support(internal);
|
||||||
|
|
||||||
|
CREATE OPERATOR CLASS minivec_l2_ops
|
||||||
|
FOR TYPE minivec USING hnsw AS
|
||||||
|
OPERATOR 1 <-> (minivec, minivec) FOR ORDER BY float_ops,
|
||||||
|
FUNCTION 1 minivec_l2_squared_distance(minivec, minivec),
|
||||||
|
FUNCTION 3 hnsw_minivec_support(internal);
|
||||||
|
|
||||||
|
CREATE OPERATOR CLASS minivec_ip_ops
|
||||||
|
FOR TYPE minivec USING hnsw AS
|
||||||
|
OPERATOR 1 <#> (minivec, minivec) FOR ORDER BY float_ops,
|
||||||
|
FUNCTION 1 minivec_negative_inner_product(minivec, minivec),
|
||||||
|
FUNCTION 3 hnsw_minivec_support(internal);
|
||||||
|
|
||||||
|
CREATE OPERATOR CLASS minivec_cosine_ops
|
||||||
|
FOR TYPE minivec USING hnsw AS
|
||||||
|
OPERATOR 1 <=> (minivec, minivec) FOR ORDER BY float_ops,
|
||||||
|
FUNCTION 1 cosine_distance(minivec, minivec),
|
||||||
|
FUNCTION 2 l2_norm(minivec),
|
||||||
|
FUNCTION 3 hnsw_minivec_support(internal);
|
||||||
|
|
||||||
|
CREATE OPERATOR CLASS minivec_l1_ops
|
||||||
|
FOR TYPE minivec USING hnsw AS
|
||||||
|
OPERATOR 1 <+> (minivec, minivec) FOR ORDER BY float_ops,
|
||||||
|
FUNCTION 1 l1_distance(minivec, minivec),
|
||||||
|
FUNCTION 3 hnsw_minivec_support(internal);
|
||||||
|
|
||||||
-- bit functions
|
-- bit functions
|
||||||
|
|
||||||
CREATE FUNCTION hamming_distance(bit, bit) RETURNS float8
|
CREATE FUNCTION hamming_distance(bit, bit) RETURNS float8
|
||||||
@@ -782,6 +1077,18 @@ CREATE FUNCTION halfvec_to_sparsevec(halfvec, integer, boolean) RETURNS sparseve
|
|||||||
CREATE FUNCTION sparsevec_to_halfvec(sparsevec, integer, boolean) RETURNS halfvec
|
CREATE FUNCTION sparsevec_to_halfvec(sparsevec, integer, boolean) RETURNS halfvec
|
||||||
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION array_to_sparsevec(integer[], integer, boolean) RETURNS sparsevec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION array_to_sparsevec(real[], integer, boolean) RETURNS sparsevec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION array_to_sparsevec(double precision[], integer, boolean) RETURNS sparsevec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
|
CREATE FUNCTION array_to_sparsevec(numeric[], integer, boolean) RETURNS sparsevec
|
||||||
|
AS 'MODULE_PATHNAME' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
|
||||||
|
|
||||||
-- sparsevec casts
|
-- sparsevec casts
|
||||||
|
|
||||||
CREATE CAST (sparsevec AS sparsevec)
|
CREATE CAST (sparsevec AS sparsevec)
|
||||||
@@ -799,6 +1106,18 @@ CREATE CAST (sparsevec AS halfvec)
|
|||||||
CREATE CAST (halfvec AS sparsevec)
|
CREATE CAST (halfvec AS sparsevec)
|
||||||
WITH FUNCTION halfvec_to_sparsevec(halfvec, integer, boolean) AS IMPLICIT;
|
WITH FUNCTION halfvec_to_sparsevec(halfvec, integer, boolean) AS IMPLICIT;
|
||||||
|
|
||||||
|
CREATE CAST (integer[] AS sparsevec)
|
||||||
|
WITH FUNCTION array_to_sparsevec(integer[], integer, boolean) AS ASSIGNMENT;
|
||||||
|
|
||||||
|
CREATE CAST (real[] AS sparsevec)
|
||||||
|
WITH FUNCTION array_to_sparsevec(real[], integer, boolean) AS ASSIGNMENT;
|
||||||
|
|
||||||
|
CREATE CAST (double precision[] AS sparsevec)
|
||||||
|
WITH FUNCTION array_to_sparsevec(double precision[], integer, boolean) AS ASSIGNMENT;
|
||||||
|
|
||||||
|
CREATE CAST (numeric[] AS sparsevec)
|
||||||
|
WITH FUNCTION array_to_sparsevec(numeric[], integer, boolean) AS ASSIGNMENT;
|
||||||
|
|
||||||
-- sparsevec operators
|
-- sparsevec operators
|
||||||
|
|
||||||
CREATE OPERATOR <-> (
|
CREATE OPERATOR <-> (
|
||||||
|
|||||||
@@ -4,8 +4,8 @@
|
|||||||
#include "postgres.h"
|
#include "postgres.h"
|
||||||
|
|
||||||
/* Check version in first header */
|
/* Check version in first header */
|
||||||
#if PG_VERSION_NUM < 120000
|
#if PG_VERSION_NUM < 130000
|
||||||
#error "Requires PostgreSQL 12+"
|
#error "Requires PostgreSQL 13+"
|
||||||
#endif
|
#endif
|
||||||
|
|
||||||
extern uint64 (*BitHammingDistance) (uint32 bytes, unsigned char *ax, unsigned char *bx, uint64 distance);
|
extern uint64 (*BitHammingDistance) (uint32 bytes, unsigned char *ax, unsigned char *bx, uint64 distance);
|
||||||
|
|||||||
@@ -3,6 +3,7 @@
|
|||||||
#include "bitutils.h"
|
#include "bitutils.h"
|
||||||
#include "bitvec.h"
|
#include "bitvec.h"
|
||||||
#include "utils/varbit.h"
|
#include "utils/varbit.h"
|
||||||
|
#include "vector.h"
|
||||||
|
|
||||||
#if PG_VERSION_NUM >= 160000
|
#if PG_VERSION_NUM >= 160000
|
||||||
#include "varatt.h"
|
#include "varatt.h"
|
||||||
@@ -40,7 +41,7 @@ CheckDims(VarBit *a, VarBit *b)
|
|||||||
/*
|
/*
|
||||||
* Get the Hamming distance between two bit vectors
|
* Get the Hamming distance between two bit vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(hamming_distance);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(hamming_distance);
|
||||||
Datum
|
Datum
|
||||||
hamming_distance(PG_FUNCTION_ARGS)
|
hamming_distance(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -55,7 +56,7 @@ hamming_distance(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get the Jaccard distance between two bit vectors
|
* Get the Jaccard distance between two bit vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(jaccard_distance);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(jaccard_distance);
|
||||||
Datum
|
Datum
|
||||||
jaccard_distance(PG_FUNCTION_ARGS)
|
jaccard_distance(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
|
|||||||
@@ -19,11 +19,6 @@
|
|||||||
#include "utils/numeric.h"
|
#include "utils/numeric.h"
|
||||||
#include "vector.h"
|
#include "vector.h"
|
||||||
|
|
||||||
#if PG_VERSION_NUM < 130000
|
|
||||||
#define TYPALIGN_DOUBLE 'd'
|
|
||||||
#define TYPALIGN_INT 'i'
|
|
||||||
#endif
|
|
||||||
|
|
||||||
#define STATE_DIMS(x) (ARR_DIMS(x)[0] - 1)
|
#define STATE_DIMS(x) (ARR_DIMS(x)[0] - 1)
|
||||||
#define CreateStateDatums(dim) palloc(sizeof(Datum) * (dim + 1))
|
#define CreateStateDatums(dim) palloc(sizeof(Datum) * (dim + 1))
|
||||||
|
|
||||||
@@ -164,28 +159,10 @@ CheckStateArray(ArrayType *statearray, const char *caller)
|
|||||||
return (float8 *) ARR_DATA_PTR(statearray);
|
return (float8 *) ARR_DATA_PTR(statearray);
|
||||||
}
|
}
|
||||||
|
|
||||||
#if PG_VERSION_NUM < 120003
|
|
||||||
static pg_noinline void
|
|
||||||
float_overflow_error(void)
|
|
||||||
{
|
|
||||||
ereport(ERROR,
|
|
||||||
(errcode(ERRCODE_NUMERIC_VALUE_OUT_OF_RANGE),
|
|
||||||
errmsg("value out of range: overflow")));
|
|
||||||
}
|
|
||||||
|
|
||||||
static pg_noinline void
|
|
||||||
float_underflow_error(void)
|
|
||||||
{
|
|
||||||
ereport(ERROR,
|
|
||||||
(errcode(ERRCODE_NUMERIC_VALUE_OUT_OF_RANGE),
|
|
||||||
errmsg("value out of range: underflow")));
|
|
||||||
}
|
|
||||||
#endif
|
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Convert textual representation to internal representation
|
* Convert textual representation to internal representation
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_in);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_in);
|
||||||
Datum
|
Datum
|
||||||
halfvec_in(PG_FUNCTION_ARGS)
|
halfvec_in(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -299,7 +276,7 @@ halfvec_in(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert internal representation to textual representation
|
* Convert internal representation to textual representation
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_out);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_out);
|
||||||
Datum
|
Datum
|
||||||
halfvec_out(PG_FUNCTION_ARGS)
|
halfvec_out(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -345,7 +322,7 @@ halfvec_out(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert type modifier
|
* Convert type modifier
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_typmod_in);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_typmod_in);
|
||||||
Datum
|
Datum
|
||||||
halfvec_typmod_in(PG_FUNCTION_ARGS)
|
halfvec_typmod_in(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -376,7 +353,7 @@ halfvec_typmod_in(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert external binary representation to internal representation
|
* Convert external binary representation to internal representation
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_recv);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_recv);
|
||||||
Datum
|
Datum
|
||||||
halfvec_recv(PG_FUNCTION_ARGS)
|
halfvec_recv(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -410,7 +387,7 @@ halfvec_recv(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert internal representation to the external binary representation
|
* Convert internal representation to the external binary representation
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_send);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_send);
|
||||||
Datum
|
Datum
|
||||||
halfvec_send(PG_FUNCTION_ARGS)
|
halfvec_send(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -430,7 +407,7 @@ halfvec_send(PG_FUNCTION_ARGS)
|
|||||||
* Convert half vector to half vector
|
* Convert half vector to half vector
|
||||||
* This is needed to check the type modifier
|
* This is needed to check the type modifier
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec);
|
||||||
Datum
|
Datum
|
||||||
halfvec(PG_FUNCTION_ARGS)
|
halfvec(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -445,7 +422,7 @@ halfvec(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert array to half vector
|
* Convert array to half vector
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(array_to_halfvec);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(array_to_halfvec);
|
||||||
Datum
|
Datum
|
||||||
array_to_halfvec(PG_FUNCTION_ARGS)
|
array_to_halfvec(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -519,7 +496,7 @@ array_to_halfvec(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert half vector to float4[]
|
* Convert half vector to float4[]
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_to_float4);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_to_float4);
|
||||||
Datum
|
Datum
|
||||||
halfvec_to_float4(PG_FUNCTION_ARGS)
|
halfvec_to_float4(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -543,7 +520,7 @@ halfvec_to_float4(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert vector to half vec
|
* Convert vector to half vec
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_to_halfvec);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_to_halfvec);
|
||||||
Datum
|
Datum
|
||||||
vector_to_halfvec(PG_FUNCTION_ARGS)
|
vector_to_halfvec(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -565,7 +542,7 @@ vector_to_halfvec(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get the L2 distance between half vectors
|
* Get the L2 distance between half vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_l2_distance);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_l2_distance);
|
||||||
Datum
|
Datum
|
||||||
halfvec_l2_distance(PG_FUNCTION_ARGS)
|
halfvec_l2_distance(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -580,7 +557,7 @@ halfvec_l2_distance(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get the L2 squared distance between half vectors
|
* Get the L2 squared distance between half vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_l2_squared_distance);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_l2_squared_distance);
|
||||||
Datum
|
Datum
|
||||||
halfvec_l2_squared_distance(PG_FUNCTION_ARGS)
|
halfvec_l2_squared_distance(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -595,7 +572,7 @@ halfvec_l2_squared_distance(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get the inner product of two half vectors
|
* Get the inner product of two half vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_inner_product);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_inner_product);
|
||||||
Datum
|
Datum
|
||||||
halfvec_inner_product(PG_FUNCTION_ARGS)
|
halfvec_inner_product(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -610,7 +587,7 @@ halfvec_inner_product(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get the negative inner product of two half vectors
|
* Get the negative inner product of two half vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_negative_inner_product);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_negative_inner_product);
|
||||||
Datum
|
Datum
|
||||||
halfvec_negative_inner_product(PG_FUNCTION_ARGS)
|
halfvec_negative_inner_product(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -625,7 +602,7 @@ halfvec_negative_inner_product(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get the cosine distance between two half vectors
|
* Get the cosine distance between two half vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_cosine_distance);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_cosine_distance);
|
||||||
Datum
|
Datum
|
||||||
halfvec_cosine_distance(PG_FUNCTION_ARGS)
|
halfvec_cosine_distance(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -657,7 +634,7 @@ halfvec_cosine_distance(PG_FUNCTION_ARGS)
|
|||||||
* Currently uses angular distance since needs to satisfy triangle inequality
|
* Currently uses angular distance since needs to satisfy triangle inequality
|
||||||
* Assumes inputs are unit vectors (skips norm)
|
* Assumes inputs are unit vectors (skips norm)
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_spherical_distance);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_spherical_distance);
|
||||||
Datum
|
Datum
|
||||||
halfvec_spherical_distance(PG_FUNCTION_ARGS)
|
halfvec_spherical_distance(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -681,7 +658,7 @@ halfvec_spherical_distance(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get the L1 distance between two half vectors
|
* Get the L1 distance between two half vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_l1_distance);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_l1_distance);
|
||||||
Datum
|
Datum
|
||||||
halfvec_l1_distance(PG_FUNCTION_ARGS)
|
halfvec_l1_distance(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -696,7 +673,7 @@ halfvec_l1_distance(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get the dimensions of a half vector
|
* Get the dimensions of a half vector
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_vector_dims);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_vector_dims);
|
||||||
Datum
|
Datum
|
||||||
halfvec_vector_dims(PG_FUNCTION_ARGS)
|
halfvec_vector_dims(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -708,7 +685,7 @@ halfvec_vector_dims(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get the L2 norm of a half vector
|
* Get the L2 norm of a half vector
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_l2_norm);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_l2_norm);
|
||||||
Datum
|
Datum
|
||||||
halfvec_l2_norm(PG_FUNCTION_ARGS)
|
halfvec_l2_norm(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -730,7 +707,7 @@ halfvec_l2_norm(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Normalize a half vector with the L2 norm
|
* Normalize a half vector with the L2 norm
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_l2_normalize);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_l2_normalize);
|
||||||
Datum
|
Datum
|
||||||
halfvec_l2_normalize(PG_FUNCTION_ARGS)
|
halfvec_l2_normalize(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -769,7 +746,7 @@ halfvec_l2_normalize(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Add half vectors
|
* Add half vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_add);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_add);
|
||||||
Datum
|
Datum
|
||||||
halfvec_add(PG_FUNCTION_ARGS)
|
halfvec_add(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -808,7 +785,7 @@ halfvec_add(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Subtract half vectors
|
* Subtract half vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_sub);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_sub);
|
||||||
Datum
|
Datum
|
||||||
halfvec_sub(PG_FUNCTION_ARGS)
|
halfvec_sub(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -847,7 +824,7 @@ halfvec_sub(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Multiply half vectors
|
* Multiply half vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_mul);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_mul);
|
||||||
Datum
|
Datum
|
||||||
halfvec_mul(PG_FUNCTION_ARGS)
|
halfvec_mul(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -889,7 +866,7 @@ halfvec_mul(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Concatenate half vectors
|
* Concatenate half vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_concat);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_concat);
|
||||||
Datum
|
Datum
|
||||||
halfvec_concat(PG_FUNCTION_ARGS)
|
halfvec_concat(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -913,7 +890,7 @@ halfvec_concat(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Quantize a half vector
|
* Quantize a half vector
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_binary_quantize);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_binary_quantize);
|
||||||
Datum
|
Datum
|
||||||
halfvec_binary_quantize(PG_FUNCTION_ARGS)
|
halfvec_binary_quantize(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -931,7 +908,7 @@ halfvec_binary_quantize(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get a subvector
|
* Get a subvector
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_subvector);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_subvector);
|
||||||
Datum
|
Datum
|
||||||
halfvec_subvector(PG_FUNCTION_ARGS)
|
halfvec_subvector(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1005,7 +982,7 @@ halfvec_cmp_internal(HalfVector * a, HalfVector * b)
|
|||||||
/*
|
/*
|
||||||
* Less than
|
* Less than
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_lt);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_lt);
|
||||||
Datum
|
Datum
|
||||||
halfvec_lt(PG_FUNCTION_ARGS)
|
halfvec_lt(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1018,7 +995,7 @@ halfvec_lt(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Less than or equal
|
* Less than or equal
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_le);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_le);
|
||||||
Datum
|
Datum
|
||||||
halfvec_le(PG_FUNCTION_ARGS)
|
halfvec_le(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1031,7 +1008,7 @@ halfvec_le(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Equal
|
* Equal
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_eq);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_eq);
|
||||||
Datum
|
Datum
|
||||||
halfvec_eq(PG_FUNCTION_ARGS)
|
halfvec_eq(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1044,7 +1021,7 @@ halfvec_eq(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Not equal
|
* Not equal
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_ne);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_ne);
|
||||||
Datum
|
Datum
|
||||||
halfvec_ne(PG_FUNCTION_ARGS)
|
halfvec_ne(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1057,7 +1034,7 @@ halfvec_ne(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Greater than or equal
|
* Greater than or equal
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_ge);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_ge);
|
||||||
Datum
|
Datum
|
||||||
halfvec_ge(PG_FUNCTION_ARGS)
|
halfvec_ge(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1070,7 +1047,7 @@ halfvec_ge(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Greater than
|
* Greater than
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_gt);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_gt);
|
||||||
Datum
|
Datum
|
||||||
halfvec_gt(PG_FUNCTION_ARGS)
|
halfvec_gt(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1083,7 +1060,7 @@ halfvec_gt(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Compare half vectors
|
* Compare half vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_cmp);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_cmp);
|
||||||
Datum
|
Datum
|
||||||
halfvec_cmp(PG_FUNCTION_ARGS)
|
halfvec_cmp(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1096,7 +1073,7 @@ halfvec_cmp(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Accumulate half vectors
|
* Accumulate half vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_accum);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_accum);
|
||||||
Datum
|
Datum
|
||||||
halfvec_accum(PG_FUNCTION_ARGS)
|
halfvec_accum(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1157,7 +1134,7 @@ halfvec_accum(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Average half vectors
|
* Average half vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_avg);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_avg);
|
||||||
Datum
|
Datum
|
||||||
halfvec_avg(PG_FUNCTION_ARGS)
|
halfvec_avg(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1191,7 +1168,7 @@ halfvec_avg(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert sparse vector to half vector
|
* Convert sparse vector to half vector
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_to_halfvec);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_to_halfvec);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_to_halfvec(PG_FUNCTION_ARGS)
|
sparsevec_to_halfvec(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
|
|||||||
@@ -38,7 +38,7 @@
|
|||||||
/* F16C has better performance than _Float16 (on x86-64) */
|
/* F16C has better performance than _Float16 (on x86-64) */
|
||||||
#if defined(__F16C__)
|
#if defined(__F16C__)
|
||||||
#define F16C_SUPPORT
|
#define F16C_SUPPORT
|
||||||
#elif defined(__FLT16_MAX__) && !defined(HALFVEC_DISPATCH)
|
#elif defined(__FLT16_MAX__) && !defined(HALFVEC_DISPATCH) && !defined(__FreeBSD__) && (!defined(__i386__) || defined(__SSE2__))
|
||||||
#define FLT16_SUPPORT
|
#define FLT16_SUPPORT
|
||||||
#endif
|
#endif
|
||||||
|
|
||||||
|
|||||||
49
src/hnsw.c
49
src/hnsw.c
@@ -9,6 +9,7 @@
|
|||||||
#include "commands/vacuum.h"
|
#include "commands/vacuum.h"
|
||||||
#include "hnsw.h"
|
#include "hnsw.h"
|
||||||
#include "miscadmin.h"
|
#include "miscadmin.h"
|
||||||
|
#include "utils/float.h"
|
||||||
#include "utils/guc.h"
|
#include "utils/guc.h"
|
||||||
#include "utils/selfuncs.h"
|
#include "utils/selfuncs.h"
|
||||||
|
|
||||||
@@ -59,17 +60,9 @@ HnswInit(void)
|
|||||||
|
|
||||||
hnsw_relopt_kind = add_reloption_kind();
|
hnsw_relopt_kind = add_reloption_kind();
|
||||||
add_int_reloption(hnsw_relopt_kind, "m", "Max number of connections",
|
add_int_reloption(hnsw_relopt_kind, "m", "Max number of connections",
|
||||||
HNSW_DEFAULT_M, HNSW_MIN_M, HNSW_MAX_M
|
HNSW_DEFAULT_M, HNSW_MIN_M, HNSW_MAX_M, AccessExclusiveLock);
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
,AccessExclusiveLock
|
|
||||||
#endif
|
|
||||||
);
|
|
||||||
add_int_reloption(hnsw_relopt_kind, "ef_construction", "Size of the dynamic candidate list for construction",
|
add_int_reloption(hnsw_relopt_kind, "ef_construction", "Size of the dynamic candidate list for construction",
|
||||||
HNSW_DEFAULT_EF_CONSTRUCTION, HNSW_MIN_EF_CONSTRUCTION, HNSW_MAX_EF_CONSTRUCTION
|
HNSW_DEFAULT_EF_CONSTRUCTION, HNSW_MIN_EF_CONSTRUCTION, HNSW_MAX_EF_CONSTRUCTION, AccessExclusiveLock);
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
,AccessExclusiveLock
|
|
||||||
#endif
|
|
||||||
);
|
|
||||||
|
|
||||||
DefineCustomIntVariable("hnsw.ef_search", "Sets the size of the dynamic candidate list for search",
|
DefineCustomIntVariable("hnsw.ef_search", "Sets the size of the dynamic candidate list for search",
|
||||||
"Valid range is 1..1000.", &hnsw_ef_search,
|
"Valid range is 1..1000.", &hnsw_ef_search,
|
||||||
@@ -112,8 +105,8 @@ hnswcostestimate(PlannerInfo *root, IndexPath *path, double loop_count,
|
|||||||
/* Never use index without order */
|
/* Never use index without order */
|
||||||
if (path->indexorderbys == NULL)
|
if (path->indexorderbys == NULL)
|
||||||
{
|
{
|
||||||
*indexStartupCost = DBL_MAX;
|
*indexStartupCost = get_float8_infinity();
|
||||||
*indexTotalCost = DBL_MAX;
|
*indexTotalCost = get_float8_infinity();
|
||||||
*indexSelectivity = 0;
|
*indexSelectivity = 0;
|
||||||
*indexCorrelation = 0;
|
*indexCorrelation = 0;
|
||||||
*indexPages = 0;
|
*indexPages = 0;
|
||||||
@@ -154,23 +147,10 @@ hnswoptions(Datum reloptions, bool validate)
|
|||||||
{"ef_construction", RELOPT_TYPE_INT, offsetof(HnswOptions, efConstruction)},
|
{"ef_construction", RELOPT_TYPE_INT, offsetof(HnswOptions, efConstruction)},
|
||||||
};
|
};
|
||||||
|
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
return (bytea *) build_reloptions(reloptions, validate,
|
return (bytea *) build_reloptions(reloptions, validate,
|
||||||
hnsw_relopt_kind,
|
hnsw_relopt_kind,
|
||||||
sizeof(HnswOptions),
|
sizeof(HnswOptions),
|
||||||
tab, lengthof(tab));
|
tab, lengthof(tab));
|
||||||
#else
|
|
||||||
relopt_value *options;
|
|
||||||
int numoptions;
|
|
||||||
HnswOptions *rdopts;
|
|
||||||
|
|
||||||
options = parseRelOptions(reloptions, validate, hnsw_relopt_kind, &numoptions);
|
|
||||||
rdopts = allocateReloptStruct(sizeof(HnswOptions), options, numoptions);
|
|
||||||
fillRelOptions((void *) rdopts, sizeof(HnswOptions), options, numoptions,
|
|
||||||
validate, tab, lengthof(tab));
|
|
||||||
|
|
||||||
return (bytea *) rdopts;
|
|
||||||
#endif
|
|
||||||
}
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
@@ -187,7 +167,7 @@ hnswvalidate(Oid opclassoid)
|
|||||||
*
|
*
|
||||||
* See https://www.postgresql.org/docs/current/index-api.html
|
* See https://www.postgresql.org/docs/current/index-api.html
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(hnswhandler);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(hnswhandler);
|
||||||
Datum
|
Datum
|
||||||
hnswhandler(PG_FUNCTION_ARGS)
|
hnswhandler(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -195,9 +175,7 @@ hnswhandler(PG_FUNCTION_ARGS)
|
|||||||
|
|
||||||
amroutine->amstrategies = 0;
|
amroutine->amstrategies = 0;
|
||||||
amroutine->amsupport = 3;
|
amroutine->amsupport = 3;
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
amroutine->amoptsprocnum = 0;
|
amroutine->amoptsprocnum = 0;
|
||||||
#endif
|
|
||||||
amroutine->amcanorder = false;
|
amroutine->amcanorder = false;
|
||||||
amroutine->amcanorderbyop = true;
|
amroutine->amcanorderbyop = true;
|
||||||
amroutine->amcanbackward = false; /* can change direction mid-scan */
|
amroutine->amcanbackward = false; /* can change direction mid-scan */
|
||||||
@@ -210,17 +188,24 @@ hnswhandler(PG_FUNCTION_ARGS)
|
|||||||
amroutine->amclusterable = false;
|
amroutine->amclusterable = false;
|
||||||
amroutine->ampredlocks = false;
|
amroutine->ampredlocks = false;
|
||||||
amroutine->amcanparallel = false;
|
amroutine->amcanparallel = false;
|
||||||
amroutine->amcaninclude = false;
|
#if PG_VERSION_NUM >= 170000
|
||||||
#if PG_VERSION_NUM >= 130000
|
amroutine->amcanbuildparallel = true;
|
||||||
amroutine->amusemaintenanceworkmem = false; /* not used during VACUUM */
|
|
||||||
amroutine->amparallelvacuumoptions = VACUUM_OPTION_PARALLEL_BULKDEL;
|
|
||||||
#endif
|
#endif
|
||||||
|
amroutine->amcaninclude = false;
|
||||||
|
amroutine->amusemaintenanceworkmem = false; /* not used during VACUUM */
|
||||||
|
#if PG_VERSION_NUM >= 160000
|
||||||
|
amroutine->amsummarizing = false;
|
||||||
|
#endif
|
||||||
|
amroutine->amparallelvacuumoptions = VACUUM_OPTION_PARALLEL_BULKDEL;
|
||||||
amroutine->amkeytype = InvalidOid;
|
amroutine->amkeytype = InvalidOid;
|
||||||
|
|
||||||
/* Interface functions */
|
/* Interface functions */
|
||||||
amroutine->ambuild = hnswbuild;
|
amroutine->ambuild = hnswbuild;
|
||||||
amroutine->ambuildempty = hnswbuildempty;
|
amroutine->ambuildempty = hnswbuildempty;
|
||||||
amroutine->aminsert = hnswinsert;
|
amroutine->aminsert = hnswinsert;
|
||||||
|
#if PG_VERSION_NUM >= 170000
|
||||||
|
amroutine->aminsertcleanup = NULL;
|
||||||
|
#endif
|
||||||
amroutine->ambulkdelete = hnswbulkdelete;
|
amroutine->ambulkdelete = hnswbulkdelete;
|
||||||
amroutine->amvacuumcleanup = hnswvacuumcleanup;
|
amroutine->amvacuumcleanup = hnswvacuumcleanup;
|
||||||
amroutine->amcanreturn = NULL;
|
amroutine->amcanreturn = NULL;
|
||||||
|
|||||||
17
src/hnsw.h
17
src/hnsw.h
@@ -76,11 +76,6 @@
|
|||||||
#define SeedRandom(seed) srandom(seed)
|
#define SeedRandom(seed) srandom(seed)
|
||||||
#endif
|
#endif
|
||||||
|
|
||||||
#if PG_VERSION_NUM < 130000
|
|
||||||
#define list_delete_last(list) list_truncate(list, list_length(list) - 1)
|
|
||||||
#define list_sort(list, cmp) ((list) = list_qsort(list, cmp))
|
|
||||||
#endif
|
|
||||||
|
|
||||||
#define HnswIsElementTuple(tup) ((tup)->type == HNSW_ELEMENT_TUPLE_TYPE)
|
#define HnswIsElementTuple(tup) ((tup)->type == HNSW_ELEMENT_TUPLE_TYPE)
|
||||||
#define HnswIsNeighborTuple(tup) ((tup)->type == HNSW_NEIGHBOR_TUPLE_TYPE)
|
#define HnswIsNeighborTuple(tup) ((tup)->type == HNSW_NEIGHBOR_TUPLE_TYPE)
|
||||||
|
|
||||||
@@ -160,11 +155,13 @@ struct HnswNeighborArray
|
|||||||
HnswCandidate items[FLEXIBLE_ARRAY_MEMBER];
|
HnswCandidate items[FLEXIBLE_ARRAY_MEMBER];
|
||||||
};
|
};
|
||||||
|
|
||||||
typedef struct HnswPairingHeapNode
|
typedef struct HnswSearchCandidate
|
||||||
{
|
{
|
||||||
pairingheap_node ph_node;
|
pairingheap_node c_node;
|
||||||
HnswCandidate *inner;
|
pairingheap_node w_node;
|
||||||
} HnswPairingHeapNode;
|
HnswElementPtr element;
|
||||||
|
float distance;
|
||||||
|
} HnswSearchCandidate;
|
||||||
|
|
||||||
/* HNSW index options */
|
/* HNSW index options */
|
||||||
typedef struct HnswOptions
|
typedef struct HnswOptions
|
||||||
@@ -385,7 +382,7 @@ void *HnswAlloc(HnswAllocator * allocator, Size size);
|
|||||||
HnswElement HnswInitElement(char *base, ItemPointer tid, int m, double ml, int maxLevel, HnswAllocator * alloc);
|
HnswElement HnswInitElement(char *base, ItemPointer tid, int m, double ml, int maxLevel, HnswAllocator * alloc);
|
||||||
HnswElement HnswInitElementFromBlock(BlockNumber blkno, OffsetNumber offno);
|
HnswElement HnswInitElementFromBlock(BlockNumber blkno, OffsetNumber offno);
|
||||||
void HnswFindElementNeighbors(char *base, HnswElement element, HnswElement entryPoint, Relation index, FmgrInfo *procinfo, Oid collation, int m, int efConstruction, bool existing);
|
void HnswFindElementNeighbors(char *base, HnswElement element, HnswElement entryPoint, Relation index, FmgrInfo *procinfo, Oid collation, int m, int efConstruction, bool existing);
|
||||||
HnswCandidate *HnswEntryCandidate(char *base, HnswElement em, Datum q, Relation rel, FmgrInfo *procinfo, Oid collation, bool loadVec);
|
HnswSearchCandidate *HnswEntryCandidate(char *base, HnswElement em, Datum q, Relation rel, FmgrInfo *procinfo, Oid collation, bool loadVec);
|
||||||
void HnswUpdateMetaPage(Relation index, int updateEntry, HnswElement entryPoint, BlockNumber insertPage, ForkNumber forkNum, bool building);
|
void HnswUpdateMetaPage(Relation index, int updateEntry, HnswElement entryPoint, BlockNumber insertPage, ForkNumber forkNum, bool building);
|
||||||
void HnswSetNeighborTuple(char *base, HnswNeighborTuple ntup, HnswElement e, int m);
|
void HnswSetNeighborTuple(char *base, HnswNeighborTuple ntup, HnswElement e, int m);
|
||||||
void HnswAddHeapTid(HnswElement element, ItemPointer heaptid);
|
void HnswAddHeapTid(HnswElement element, ItemPointer heaptid);
|
||||||
|
|||||||
@@ -60,12 +60,6 @@
|
|||||||
#include "pgstat.h"
|
#include "pgstat.h"
|
||||||
#endif
|
#endif
|
||||||
|
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
#define CALLBACK_ITEM_POINTER ItemPointer tid
|
|
||||||
#else
|
|
||||||
#define CALLBACK_ITEM_POINTER HeapTuple hup
|
|
||||||
#endif
|
|
||||||
|
|
||||||
#if PG_VERSION_NUM >= 140000
|
#if PG_VERSION_NUM >= 140000
|
||||||
#include "utils/backend_status.h"
|
#include "utils/backend_status.h"
|
||||||
#include "utils/wait_event.h"
|
#include "utils/wait_event.h"
|
||||||
@@ -75,10 +69,6 @@
|
|||||||
#define PARALLEL_KEY_HNSW_AREA UINT64CONST(0xA000000000000002)
|
#define PARALLEL_KEY_HNSW_AREA UINT64CONST(0xA000000000000002)
|
||||||
#define PARALLEL_KEY_QUERY_TEXT UINT64CONST(0xA000000000000003)
|
#define PARALLEL_KEY_QUERY_TEXT UINT64CONST(0xA000000000000003)
|
||||||
|
|
||||||
#if PG_VERSION_NUM < 130000
|
|
||||||
#define GENERATIONCHUNK_RAWSIZE (SIZEOF_SIZE_T + SIZEOF_VOID_P * 2)
|
|
||||||
#endif
|
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Create the metapage
|
* Create the metapage
|
||||||
*/
|
*/
|
||||||
@@ -192,7 +182,9 @@ CreateGraphPages(HnswBuildState * buildstate)
|
|||||||
|
|
||||||
/* Initial size check */
|
/* Initial size check */
|
||||||
if (etupSize > HNSW_TUPLE_ALLOC_SIZE)
|
if (etupSize > HNSW_TUPLE_ALLOC_SIZE)
|
||||||
elog(ERROR, "index tuple too large");
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
|
||||||
|
errmsg("index tuple too large")));
|
||||||
|
|
||||||
HnswSetElementTuple(base, etup, element);
|
HnswSetElementTuple(base, etup, element);
|
||||||
|
|
||||||
@@ -379,7 +371,13 @@ UpdateNeighborsInMemory(char *base, FmgrInfo *procinfo, Oid collation, HnswEleme
|
|||||||
for (int lc = e->level; lc >= 0; lc--)
|
for (int lc = e->level; lc >= 0; lc--)
|
||||||
{
|
{
|
||||||
int lm = HnswGetLayerM(m, lc);
|
int lm = HnswGetLayerM(m, lc);
|
||||||
HnswNeighborArray *neighbors = HnswGetNeighbors(base, e, lc);
|
Size neighborsSize = HNSW_NEIGHBOR_ARRAY_SIZE(lm);
|
||||||
|
HnswNeighborArray *neighbors = palloc(neighborsSize);
|
||||||
|
|
||||||
|
/* Copy neighbors to local memory */
|
||||||
|
LWLockAcquire(&e->lock, LW_SHARED);
|
||||||
|
memcpy(neighbors, HnswGetNeighbors(base, e, lc), neighborsSize);
|
||||||
|
LWLockRelease(&e->lock);
|
||||||
|
|
||||||
for (int i = 0; i < neighbors->length; i++)
|
for (int i = 0; i < neighbors->length; i++)
|
||||||
{
|
{
|
||||||
@@ -389,7 +387,6 @@ UpdateNeighborsInMemory(char *base, FmgrInfo *procinfo, Oid collation, HnswEleme
|
|||||||
/* Keep scan-build happy on Mac x86-64 */
|
/* Keep scan-build happy on Mac x86-64 */
|
||||||
Assert(neighborElement);
|
Assert(neighborElement);
|
||||||
|
|
||||||
/* Use element for lock instead of hc since hc can be replaced */
|
|
||||||
LWLockAcquire(&neighborElement->lock, LW_EXCLUSIVE);
|
LWLockAcquire(&neighborElement->lock, LW_EXCLUSIVE);
|
||||||
HnswUpdateConnection(base, e, hc, lm, lc, NULL, NULL, procinfo, collation);
|
HnswUpdateConnection(base, e, hc, lm, lc, NULL, NULL, procinfo, collation);
|
||||||
LWLockRelease(&neighborElement->lock);
|
LWLockRelease(&neighborElement->lock);
|
||||||
@@ -578,17 +575,13 @@ InsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heaptid, Hn
|
|||||||
* Callback for table_index_build_scan
|
* Callback for table_index_build_scan
|
||||||
*/
|
*/
|
||||||
static void
|
static void
|
||||||
BuildCallback(Relation index, CALLBACK_ITEM_POINTER, Datum *values,
|
BuildCallback(Relation index, ItemPointer tid, Datum *values,
|
||||||
bool *isnull, bool tupleIsAlive, void *state)
|
bool *isnull, bool tupleIsAlive, void *state)
|
||||||
{
|
{
|
||||||
HnswBuildState *buildstate = (HnswBuildState *) state;
|
HnswBuildState *buildstate = (HnswBuildState *) state;
|
||||||
HnswGraph *graph = buildstate->graph;
|
HnswGraph *graph = buildstate->graph;
|
||||||
MemoryContext oldCtx;
|
MemoryContext oldCtx;
|
||||||
|
|
||||||
#if PG_VERSION_NUM < 130000
|
|
||||||
ItemPointer tid = &hup->t_self;
|
|
||||||
#endif
|
|
||||||
|
|
||||||
/* Skip nulls */
|
/* Skip nulls */
|
||||||
if (isnull[0])
|
if (isnull[0])
|
||||||
return;
|
return;
|
||||||
@@ -651,11 +644,7 @@ HnswMemoryContextAlloc(Size size, void *state)
|
|||||||
HnswBuildState *buildstate = (HnswBuildState *) state;
|
HnswBuildState *buildstate = (HnswBuildState *) state;
|
||||||
void *chunk = MemoryContextAlloc(buildstate->graphCtx, size);
|
void *chunk = MemoryContextAlloc(buildstate->graphCtx, size);
|
||||||
|
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
buildstate->graphData.memoryUsed = MemoryContextMemAllocated(buildstate->graphCtx, false);
|
buildstate->graphData.memoryUsed = MemoryContextMemAllocated(buildstate->graphCtx, false);
|
||||||
#else
|
|
||||||
buildstate->graphData.memoryUsed += MAXALIGN(size);
|
|
||||||
#endif
|
|
||||||
|
|
||||||
return chunk;
|
return chunk;
|
||||||
}
|
}
|
||||||
@@ -691,17 +680,25 @@ InitBuildState(HnswBuildState * buildstate, Relation heap, Relation index, Index
|
|||||||
|
|
||||||
/* Disallow varbit since require fixed dimensions */
|
/* Disallow varbit since require fixed dimensions */
|
||||||
if (TupleDescAttr(index->rd_att, 0)->atttypid == VARBITOID)
|
if (TupleDescAttr(index->rd_att, 0)->atttypid == VARBITOID)
|
||||||
elog(ERROR, "type not supported for hnsw index");
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
|
||||||
|
errmsg("type not supported for hnsw index")));
|
||||||
|
|
||||||
/* Require column to have dimensions to be indexed */
|
/* Require column to have dimensions to be indexed */
|
||||||
if (buildstate->dimensions < 0)
|
if (buildstate->dimensions < 0)
|
||||||
elog(ERROR, "column does not have dimensions");
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
|
||||||
|
errmsg("column does not have dimensions")));
|
||||||
|
|
||||||
if (buildstate->dimensions > buildstate->typeInfo->maxDimensions)
|
if (buildstate->dimensions > buildstate->typeInfo->maxDimensions)
|
||||||
elog(ERROR, "column cannot have more than %d dimensions for hnsw index", buildstate->typeInfo->maxDimensions);
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
|
||||||
|
errmsg("column cannot have more than %d dimensions for hnsw index", buildstate->typeInfo->maxDimensions)));
|
||||||
|
|
||||||
if (buildstate->efConstruction < 2 * buildstate->m)
|
if (buildstate->efConstruction < 2 * buildstate->m)
|
||||||
elog(ERROR, "ef_construction must be greater than or equal to 2 * m");
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
|
||||||
|
errmsg("ef_construction must be greater than or equal to 2 * m")));
|
||||||
|
|
||||||
buildstate->reltuples = 0;
|
buildstate->reltuples = 0;
|
||||||
buildstate->indtuples = 0;
|
buildstate->indtuples = 0;
|
||||||
|
|||||||
@@ -36,14 +36,15 @@ GetInsertPage(Relation index)
|
|||||||
* Check for a free offset
|
* Check for a free offset
|
||||||
*/
|
*/
|
||||||
static bool
|
static bool
|
||||||
HnswFreeOffset(Relation index, Buffer buf, Page page, HnswElement element, Size ntupSize, Buffer *nbuf, Page *npage, OffsetNumber *freeOffno, OffsetNumber *freeNeighborOffno, BlockNumber *newInsertPage)
|
HnswFreeOffset(Relation index, Buffer buf, Page page, HnswElement element, Size etupSize, Size ntupSize, Buffer *nbuf, Page *npage, OffsetNumber *freeOffno, OffsetNumber *freeNeighborOffno, BlockNumber *newInsertPage)
|
||||||
{
|
{
|
||||||
OffsetNumber offno;
|
OffsetNumber offno;
|
||||||
OffsetNumber maxoffno = PageGetMaxOffsetNumber(page);
|
OffsetNumber maxoffno = PageGetMaxOffsetNumber(page);
|
||||||
|
|
||||||
for (offno = FirstOffsetNumber; offno <= maxoffno; offno = OffsetNumberNext(offno))
|
for (offno = FirstOffsetNumber; offno <= maxoffno; offno = OffsetNumberNext(offno))
|
||||||
{
|
{
|
||||||
HnswElementTuple etup = (HnswElementTuple) PageGetItem(page, PageGetItemId(page, offno));
|
ItemId eitemid = PageGetItemId(page, offno);
|
||||||
|
HnswElementTuple etup = (HnswElementTuple) PageGetItem(page, eitemid);
|
||||||
|
|
||||||
/* Skip neighbor tuples */
|
/* Skip neighbor tuples */
|
||||||
if (!HnswIsElementTuple(etup))
|
if (!HnswIsElementTuple(etup))
|
||||||
@@ -54,7 +55,9 @@ HnswFreeOffset(Relation index, Buffer buf, Page page, HnswElement element, Size
|
|||||||
BlockNumber elementPage = BufferGetBlockNumber(buf);
|
BlockNumber elementPage = BufferGetBlockNumber(buf);
|
||||||
BlockNumber neighborPage = ItemPointerGetBlockNumber(&etup->neighbortid);
|
BlockNumber neighborPage = ItemPointerGetBlockNumber(&etup->neighbortid);
|
||||||
OffsetNumber neighborOffno = ItemPointerGetOffsetNumber(&etup->neighbortid);
|
OffsetNumber neighborOffno = ItemPointerGetOffsetNumber(&etup->neighbortid);
|
||||||
ItemId itemid;
|
ItemId nitemid;
|
||||||
|
Size pageFree;
|
||||||
|
Size npageFree;
|
||||||
|
|
||||||
if (!BlockNumberIsValid(*newInsertPage))
|
if (!BlockNumberIsValid(*newInsertPage))
|
||||||
*newInsertPage = elementPage;
|
*newInsertPage = elementPage;
|
||||||
@@ -73,10 +76,25 @@ HnswFreeOffset(Relation index, Buffer buf, Page page, HnswElement element, Size
|
|||||||
*npage = BufferGetPage(*nbuf);
|
*npage = BufferGetPage(*nbuf);
|
||||||
}
|
}
|
||||||
|
|
||||||
itemid = PageGetItemId(*npage, neighborOffno);
|
nitemid = PageGetItemId(*npage, neighborOffno);
|
||||||
|
|
||||||
/* Check for space on neighbor tuple page */
|
/* Ensure aligned for space check */
|
||||||
if (PageGetFreeSpace(*npage) + ItemIdGetLength(itemid) - sizeof(ItemIdData) >= ntupSize)
|
Assert(etupSize == MAXALIGN(etupSize));
|
||||||
|
Assert(ntupSize == MAXALIGN(ntupSize));
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Calculate free space individually since tuples are overwritten
|
||||||
|
* individually (in separate calls to PageIndexTupleOverwrite)
|
||||||
|
*/
|
||||||
|
pageFree = ItemIdGetLength(eitemid) + PageGetExactFreeSpace(page);
|
||||||
|
npageFree = ItemIdGetLength(nitemid);
|
||||||
|
if (neighborPage != elementPage)
|
||||||
|
npageFree += PageGetExactFreeSpace(*npage);
|
||||||
|
else if (pageFree >= etupSize)
|
||||||
|
npageFree += pageFree - etupSize;
|
||||||
|
|
||||||
|
/* Check for space */
|
||||||
|
if (pageFree >= etupSize && npageFree >= ntupSize)
|
||||||
{
|
{
|
||||||
*freeOffno = offno;
|
*freeOffno = offno;
|
||||||
*freeNeighborOffno = neighborOffno;
|
*freeNeighborOffno = neighborOffno;
|
||||||
@@ -184,7 +202,7 @@ AddElementOnDisk(Relation index, HnswElement e, int m, BlockNumber insertPage, B
|
|||||||
}
|
}
|
||||||
|
|
||||||
/* Next, try space from a deleted element */
|
/* Next, try space from a deleted element */
|
||||||
if (HnswFreeOffset(index, buf, page, e, ntupSize, &nbuf, &npage, &freeOffno, &freeNeighborOffno, &newInsertPage))
|
if (HnswFreeOffset(index, buf, page, e, etupSize, ntupSize, &nbuf, &npage, &freeOffno, &freeNeighborOffno, &newInsertPage))
|
||||||
{
|
{
|
||||||
if (nbuf != buf)
|
if (nbuf != buf)
|
||||||
{
|
{
|
||||||
@@ -361,8 +379,12 @@ HnswUpdateNeighborsOnDisk(Relation index, FmgrInfo *procinfo, Oid collation, Hns
|
|||||||
HnswElement neighborElement = HnswPtrAccess(base, hc->element);
|
HnswElement neighborElement = HnswPtrAccess(base, hc->element);
|
||||||
OffsetNumber offno = neighborElement->neighborOffno;
|
OffsetNumber offno = neighborElement->neighborOffno;
|
||||||
|
|
||||||
/* Get latest neighbors since they may have changed */
|
/*
|
||||||
/* Do not lock yet since selecting neighbors can take time */
|
* Get latest neighbors since they may have changed. Do not lock
|
||||||
|
* yet since selecting neighbors can take time. Could use
|
||||||
|
* optimistic locking to retry if another update occurs before
|
||||||
|
* getting exclusive lock.
|
||||||
|
*/
|
||||||
HnswLoadNeighbors(neighborElement, index, m);
|
HnswLoadNeighbors(neighborElement, index, m);
|
||||||
|
|
||||||
/*
|
/*
|
||||||
|
|||||||
@@ -160,15 +160,15 @@ hnswgettuple(IndexScanDesc scan, ScanDirection dir)
|
|||||||
|
|
||||||
so->first = false;
|
so->first = false;
|
||||||
|
|
||||||
#if defined(HNSW_MEMORY) && PG_VERSION_NUM >= 130000
|
#if defined(HNSW_MEMORY)
|
||||||
elog(INFO, "memory: %zu MB", MemoryContextMemAllocated(so->tmpCtx, false) / (1024 * 1024));
|
elog(INFO, "memory: %zu KB", MemoryContextMemAllocated(so->tmpCtx, false) / 1024);
|
||||||
#endif
|
#endif
|
||||||
}
|
}
|
||||||
|
|
||||||
while (list_length(so->w) > 0)
|
while (list_length(so->w) > 0)
|
||||||
{
|
{
|
||||||
char *base = NULL;
|
char *base = NULL;
|
||||||
HnswCandidate *hc = llast(so->w);
|
HnswSearchCandidate *hc = llast(so->w);
|
||||||
HnswElement element = HnswPtrAccess(base, hc->element);
|
HnswElement element = HnswPtrAccess(base, hc->element);
|
||||||
ItemPointer heaptid;
|
ItemPointer heaptid;
|
||||||
|
|
||||||
|
|||||||
367
src/hnswutils.c
367
src/hnswutils.c
@@ -5,6 +5,7 @@
|
|||||||
#include "access/generic_xlog.h"
|
#include "access/generic_xlog.h"
|
||||||
#include "catalog/pg_type.h"
|
#include "catalog/pg_type.h"
|
||||||
#include "catalog/pg_type_d.h"
|
#include "catalog/pg_type_d.h"
|
||||||
|
#include "common/hashfn.h"
|
||||||
#include "fmgr.h"
|
#include "fmgr.h"
|
||||||
#include "hnsw.h"
|
#include "hnsw.h"
|
||||||
#include "lib/pairingheap.h"
|
#include "lib/pairingheap.h"
|
||||||
@@ -14,12 +15,6 @@
|
|||||||
#include "utils/memdebug.h"
|
#include "utils/memdebug.h"
|
||||||
#include "utils/rel.h"
|
#include "utils/rel.h"
|
||||||
|
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
#include "common/hashfn.h"
|
|
||||||
#else
|
|
||||||
#include "utils/hashutils.h"
|
|
||||||
#endif
|
|
||||||
|
|
||||||
#if PG_VERSION_NUM < 170000
|
#if PG_VERSION_NUM < 170000
|
||||||
static inline uint64
|
static inline uint64
|
||||||
murmurhash64(uint64 data)
|
murmurhash64(uint64 data)
|
||||||
@@ -112,6 +107,12 @@ typedef union
|
|||||||
tidhash_hash *tids;
|
tidhash_hash *tids;
|
||||||
} visited_hash;
|
} visited_hash;
|
||||||
|
|
||||||
|
typedef union
|
||||||
|
{
|
||||||
|
HnswElement element;
|
||||||
|
ItemPointerData indextid;
|
||||||
|
} HnswUnvisited;
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Get the max number of connections in an upper layer for each element in the index
|
* Get the max number of connections in an upper layer for each element in the index
|
||||||
*/
|
*/
|
||||||
@@ -158,6 +159,9 @@ HnswOptionalProcInfo(Relation index, uint16 procnum)
|
|||||||
Datum
|
Datum
|
||||||
HnswNormValue(const HnswTypeInfo * typeInfo, Oid collation, Datum value)
|
HnswNormValue(const HnswTypeInfo * typeInfo, Oid collation, Datum value)
|
||||||
{
|
{
|
||||||
|
if (!typeInfo->normalize)
|
||||||
|
return value;
|
||||||
|
|
||||||
return DirectFunctionCall1Coll(typeInfo->normalize, collation, value);
|
return DirectFunctionCall1Coll(typeInfo->normalize, collation, value);
|
||||||
}
|
}
|
||||||
|
|
||||||
@@ -547,19 +551,19 @@ HnswLoadElementFromTuple(HnswElement element, HnswElementTuple etup, bool loadHe
|
|||||||
/*
|
/*
|
||||||
* Load an element and optionally get its distance from q
|
* Load an element and optionally get its distance from q
|
||||||
*/
|
*/
|
||||||
void
|
static void
|
||||||
HnswLoadElement(HnswElement element, float *distance, Datum *q, Relation index, FmgrInfo *procinfo, Oid collation, bool loadVec, float *maxDistance)
|
HnswLoadElementImpl(BlockNumber blkno, OffsetNumber offno, float *distance, Datum *q, Relation index, FmgrInfo *procinfo, Oid collation, bool loadVec, float *maxDistance, HnswElement * element)
|
||||||
{
|
{
|
||||||
Buffer buf;
|
Buffer buf;
|
||||||
Page page;
|
Page page;
|
||||||
HnswElementTuple etup;
|
HnswElementTuple etup;
|
||||||
|
|
||||||
/* Read vector */
|
/* Read vector */
|
||||||
buf = ReadBuffer(index, element->blkno);
|
buf = ReadBuffer(index, blkno);
|
||||||
LockBuffer(buf, BUFFER_LOCK_SHARE);
|
LockBuffer(buf, BUFFER_LOCK_SHARE);
|
||||||
page = BufferGetPage(buf);
|
page = BufferGetPage(buf);
|
||||||
|
|
||||||
etup = (HnswElementTuple) PageGetItem(page, PageGetItemId(page, element->offno));
|
etup = (HnswElementTuple) PageGetItem(page, PageGetItemId(page, offno));
|
||||||
|
|
||||||
Assert(HnswIsElementTuple(etup));
|
Assert(HnswIsElementTuple(etup));
|
||||||
|
|
||||||
@@ -574,19 +578,32 @@ HnswLoadElement(HnswElement element, float *distance, Datum *q, Relation index,
|
|||||||
|
|
||||||
/* Load element */
|
/* Load element */
|
||||||
if (distance == NULL || maxDistance == NULL || *distance < *maxDistance)
|
if (distance == NULL || maxDistance == NULL || *distance < *maxDistance)
|
||||||
HnswLoadElementFromTuple(element, etup, true, loadVec);
|
{
|
||||||
|
if (*element == NULL)
|
||||||
|
*element = HnswInitElementFromBlock(blkno, offno);
|
||||||
|
|
||||||
|
HnswLoadElementFromTuple(*element, etup, true, loadVec);
|
||||||
|
}
|
||||||
|
|
||||||
UnlockReleaseBuffer(buf);
|
UnlockReleaseBuffer(buf);
|
||||||
}
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Get the distance for a candidate
|
* Load an element and optionally get its distance from q
|
||||||
|
*/
|
||||||
|
void
|
||||||
|
HnswLoadElement(HnswElement element, float *distance, Datum *q, Relation index, FmgrInfo *procinfo, Oid collation, bool loadVec, float *maxDistance)
|
||||||
|
{
|
||||||
|
HnswLoadElementImpl(element->blkno, element->offno, distance, q, index, procinfo, collation, loadVec, maxDistance, &element);
|
||||||
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Get the distance for an element
|
||||||
*/
|
*/
|
||||||
static float
|
static float
|
||||||
GetCandidateDistance(char *base, HnswCandidate * hc, Datum q, FmgrInfo *procinfo, Oid collation)
|
GetElementDistance(char *base, HnswElement element, Datum q, FmgrInfo *procinfo, Oid collation)
|
||||||
{
|
{
|
||||||
HnswElement hce = HnswPtrAccess(base, hc->element);
|
Datum value = HnswGetValue(base, element);
|
||||||
Datum value = HnswGetValue(base, hce);
|
|
||||||
|
|
||||||
return DatumGetFloat8(FunctionCall2Coll(procinfo, collation, q, value));
|
return DatumGetFloat8(FunctionCall2Coll(procinfo, collation, q, value));
|
||||||
}
|
}
|
||||||
@@ -594,29 +611,32 @@ GetCandidateDistance(char *base, HnswCandidate * hc, Datum q, FmgrInfo *procinfo
|
|||||||
/*
|
/*
|
||||||
* Create a candidate for the entry point
|
* Create a candidate for the entry point
|
||||||
*/
|
*/
|
||||||
HnswCandidate *
|
HnswSearchCandidate *
|
||||||
HnswEntryCandidate(char *base, HnswElement entryPoint, Datum q, Relation index, FmgrInfo *procinfo, Oid collation, bool loadVec)
|
HnswEntryCandidate(char *base, HnswElement entryPoint, Datum q, Relation index, FmgrInfo *procinfo, Oid collation, bool loadVec)
|
||||||
{
|
{
|
||||||
HnswCandidate *hc = palloc(sizeof(HnswCandidate));
|
HnswSearchCandidate *hc = palloc(sizeof(HnswSearchCandidate));
|
||||||
|
|
||||||
HnswPtrStore(base, hc->element, entryPoint);
|
HnswPtrStore(base, hc->element, entryPoint);
|
||||||
if (index == NULL)
|
if (index == NULL)
|
||||||
hc->distance = GetCandidateDistance(base, hc, q, procinfo, collation);
|
hc->distance = GetElementDistance(base, entryPoint, q, procinfo, collation);
|
||||||
else
|
else
|
||||||
HnswLoadElement(entryPoint, &hc->distance, &q, index, procinfo, collation, loadVec, NULL);
|
HnswLoadElement(entryPoint, &hc->distance, &q, index, procinfo, collation, loadVec, NULL);
|
||||||
return hc;
|
return hc;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
#define HnswGetSearchCandidate(membername, ptr) pairingheap_container(HnswSearchCandidate, membername, ptr)
|
||||||
|
#define HnswGetSearchCandidateConst(membername, ptr) pairingheap_const_container(HnswSearchCandidate, membername, ptr)
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Compare candidate distances
|
* Compare candidate distances
|
||||||
*/
|
*/
|
||||||
static int
|
static int
|
||||||
CompareNearestCandidates(const pairingheap_node *a, const pairingheap_node *b, void *arg)
|
CompareNearestCandidates(const pairingheap_node *a, const pairingheap_node *b, void *arg)
|
||||||
{
|
{
|
||||||
if (((const HnswPairingHeapNode *) a)->inner->distance < ((const HnswPairingHeapNode *) b)->inner->distance)
|
if (HnswGetSearchCandidateConst(c_node, a)->distance < HnswGetSearchCandidateConst(c_node, b)->distance)
|
||||||
return 1;
|
return 1;
|
||||||
|
|
||||||
if (((const HnswPairingHeapNode *) a)->inner->distance > ((const HnswPairingHeapNode *) b)->inner->distance)
|
if (HnswGetSearchCandidateConst(c_node, a)->distance > HnswGetSearchCandidateConst(c_node, b)->distance)
|
||||||
return -1;
|
return -1;
|
||||||
|
|
||||||
return 0;
|
return 0;
|
||||||
@@ -628,27 +648,15 @@ CompareNearestCandidates(const pairingheap_node *a, const pairingheap_node *b, v
|
|||||||
static int
|
static int
|
||||||
CompareFurthestCandidates(const pairingheap_node *a, const pairingheap_node *b, void *arg)
|
CompareFurthestCandidates(const pairingheap_node *a, const pairingheap_node *b, void *arg)
|
||||||
{
|
{
|
||||||
if (((const HnswPairingHeapNode *) a)->inner->distance < ((const HnswPairingHeapNode *) b)->inner->distance)
|
if (HnswGetSearchCandidateConst(w_node, a)->distance < HnswGetSearchCandidateConst(w_node, b)->distance)
|
||||||
return -1;
|
return -1;
|
||||||
|
|
||||||
if (((const HnswPairingHeapNode *) a)->inner->distance > ((const HnswPairingHeapNode *) b)->inner->distance)
|
if (HnswGetSearchCandidateConst(w_node, a)->distance > HnswGetSearchCandidateConst(w_node, b)->distance)
|
||||||
return 1;
|
return 1;
|
||||||
|
|
||||||
return 0;
|
return 0;
|
||||||
}
|
}
|
||||||
|
|
||||||
/*
|
|
||||||
* Create a pairing heap node for a candidate
|
|
||||||
*/
|
|
||||||
static HnswPairingHeapNode *
|
|
||||||
CreatePairingHeapNode(HnswCandidate * c)
|
|
||||||
{
|
|
||||||
HnswPairingHeapNode *node = palloc(sizeof(HnswPairingHeapNode));
|
|
||||||
|
|
||||||
node->inner = c;
|
|
||||||
return node;
|
|
||||||
}
|
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Init visited
|
* Init visited
|
||||||
*/
|
*/
|
||||||
@@ -667,11 +675,11 @@ InitVisited(char *base, visited_hash * v, Relation index, int ef, int m)
|
|||||||
* Add to visited
|
* Add to visited
|
||||||
*/
|
*/
|
||||||
static inline void
|
static inline void
|
||||||
AddToVisited(char *base, visited_hash * v, HnswCandidate * hc, Relation index, bool *found)
|
AddToVisited(char *base, visited_hash * v, HnswElementPtr elementPtr, Relation index, bool *found)
|
||||||
{
|
{
|
||||||
if (index != NULL)
|
if (index != NULL)
|
||||||
{
|
{
|
||||||
HnswElement element = HnswPtrAccess(base, hc->element);
|
HnswElement element = HnswPtrAccess(base, elementPtr);
|
||||||
ItemPointerData indextid;
|
ItemPointerData indextid;
|
||||||
|
|
||||||
ItemPointerSet(&indextid, element->blkno, element->offno);
|
ItemPointerSet(&indextid, element->blkno, element->offno);
|
||||||
@@ -679,23 +687,15 @@ AddToVisited(char *base, visited_hash * v, HnswCandidate * hc, Relation index, b
|
|||||||
}
|
}
|
||||||
else if (base != NULL)
|
else if (base != NULL)
|
||||||
{
|
{
|
||||||
#if PG_VERSION_NUM >= 130000
|
HnswElement element = HnswPtrAccess(base, elementPtr);
|
||||||
HnswElement element = HnswPtrAccess(base, hc->element);
|
|
||||||
|
|
||||||
offsethash_insert_hash(v->offsets, HnswPtrOffset(hc->element), element->hash, found);
|
offsethash_insert_hash(v->offsets, HnswPtrOffset(elementPtr), element->hash, found);
|
||||||
#else
|
|
||||||
offsethash_insert(v->offsets, HnswPtrOffset(hc->element), found);
|
|
||||||
#endif
|
|
||||||
}
|
}
|
||||||
else
|
else
|
||||||
{
|
{
|
||||||
#if PG_VERSION_NUM >= 130000
|
HnswElement element = HnswPtrAccess(base, elementPtr);
|
||||||
HnswElement element = HnswPtrAccess(base, hc->element);
|
|
||||||
|
|
||||||
pointerhash_insert_hash(v->pointers, (uintptr_t) HnswPtrPointer(hc->element), element->hash, found);
|
pointerhash_insert_hash(v->pointers, (uintptr_t) HnswPtrPointer(elementPtr), element->hash, found);
|
||||||
#else
|
|
||||||
pointerhash_insert(v->pointers, (uintptr_t) HnswPtrPointer(hc->element), found);
|
|
||||||
#endif
|
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
@@ -703,20 +703,89 @@ AddToVisited(char *base, visited_hash * v, HnswCandidate * hc, Relation index, b
|
|||||||
* Count element towards ef
|
* Count element towards ef
|
||||||
*/
|
*/
|
||||||
static inline bool
|
static inline bool
|
||||||
CountElement(char *base, HnswElement skipElement, HnswCandidate * hc)
|
CountElement(HnswElement skipElement, HnswElement e)
|
||||||
{
|
{
|
||||||
HnswElement e;
|
|
||||||
|
|
||||||
if (skipElement == NULL)
|
if (skipElement == NULL)
|
||||||
return true;
|
return true;
|
||||||
|
|
||||||
/* Ensure does not access heaptidsLength during in-memory build */
|
/* Ensure does not access heaptidsLength during in-memory build */
|
||||||
pg_memory_barrier();
|
pg_memory_barrier();
|
||||||
|
|
||||||
e = HnswPtrAccess(base, hc->element);
|
/* Keep scan-build happy on Mac x86-64 */
|
||||||
|
Assert(e);
|
||||||
|
|
||||||
return e->heaptidsLength != 0;
|
return e->heaptidsLength != 0;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Load unvisited neighbors from memory
|
||||||
|
*/
|
||||||
|
static void
|
||||||
|
HnswLoadUnvisitedFromMemory(char *base, HnswElement element, HnswUnvisited * unvisited, int *unvisitedLength, visited_hash * v, int lc, HnswNeighborArray * localNeighborhood, Size neighborhoodSize)
|
||||||
|
{
|
||||||
|
/* Get the neighborhood at layer lc */
|
||||||
|
HnswNeighborArray *neighborhood = HnswGetNeighbors(base, element, lc);
|
||||||
|
|
||||||
|
/* Copy neighborhood to local memory */
|
||||||
|
LWLockAcquire(&element->lock, LW_SHARED);
|
||||||
|
memcpy(localNeighborhood, neighborhood, neighborhoodSize);
|
||||||
|
LWLockRelease(&element->lock);
|
||||||
|
|
||||||
|
*unvisitedLength = 0;
|
||||||
|
|
||||||
|
for (int i = 0; i < localNeighborhood->length; i++)
|
||||||
|
{
|
||||||
|
HnswCandidate *hc = &localNeighborhood->items[i];
|
||||||
|
bool found;
|
||||||
|
|
||||||
|
AddToVisited(base, v, hc->element, NULL, &found);
|
||||||
|
|
||||||
|
if (!found)
|
||||||
|
unvisited[(*unvisitedLength)++].element = HnswPtrAccess(base, hc->element);
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Load unvisited neighbors from disk
|
||||||
|
*/
|
||||||
|
static void
|
||||||
|
HnswLoadUnvisitedFromDisk(HnswElement element, HnswUnvisited * unvisited, int *unvisitedLength, visited_hash * v, Relation index, int m, int lm, int lc)
|
||||||
|
{
|
||||||
|
Buffer buf;
|
||||||
|
Page page;
|
||||||
|
HnswNeighborTuple ntup;
|
||||||
|
int start;
|
||||||
|
ItemPointerData indextids[HNSW_MAX_M * 2];
|
||||||
|
|
||||||
|
buf = ReadBuffer(index, element->neighborPage);
|
||||||
|
LockBuffer(buf, BUFFER_LOCK_SHARE);
|
||||||
|
page = BufferGetPage(buf);
|
||||||
|
|
||||||
|
ntup = (HnswNeighborTuple) PageGetItem(page, PageGetItemId(page, element->neighborOffno));
|
||||||
|
start = (element->level - lc) * m;
|
||||||
|
|
||||||
|
/* Copy to minimize lock time */
|
||||||
|
memcpy(&indextids, ntup->indextids + start, lm * sizeof(ItemPointerData));
|
||||||
|
|
||||||
|
UnlockReleaseBuffer(buf);
|
||||||
|
|
||||||
|
*unvisitedLength = 0;
|
||||||
|
|
||||||
|
for (int i = 0; i < lm; i++)
|
||||||
|
{
|
||||||
|
ItemPointer indextid = &indextids[i];
|
||||||
|
bool found;
|
||||||
|
|
||||||
|
if (!ItemPointerIsValid(indextid))
|
||||||
|
break;
|
||||||
|
|
||||||
|
tidhash_insert(v->tids, *indextid, &found);
|
||||||
|
|
||||||
|
if (!found)
|
||||||
|
unvisited[(*unvisitedLength)++].indextid = *indextid;
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Algorithm 2 from paper
|
* Algorithm 2 from paper
|
||||||
*/
|
*/
|
||||||
@@ -729,43 +798,45 @@ HnswSearchLayer(char *base, Datum q, List *ep, int ef, int lc, Relation index, F
|
|||||||
int wlen = 0;
|
int wlen = 0;
|
||||||
visited_hash v;
|
visited_hash v;
|
||||||
ListCell *lc2;
|
ListCell *lc2;
|
||||||
HnswNeighborArray *neighborhoodData = NULL;
|
HnswNeighborArray *localNeighborhood = NULL;
|
||||||
Size neighborhoodSize;
|
Size neighborhoodSize = 0;
|
||||||
|
int lm = HnswGetLayerM(m, lc);
|
||||||
|
HnswUnvisited *unvisited = palloc(lm * sizeof(HnswUnvisited));
|
||||||
|
int unvisitedLength;
|
||||||
|
|
||||||
InitVisited(base, &v, index, ef, m);
|
InitVisited(base, &v, index, ef, m);
|
||||||
|
|
||||||
/* Create local memory for neighborhood if needed */
|
/* Create local memory for neighborhood if needed */
|
||||||
if (index == NULL)
|
if (index == NULL)
|
||||||
{
|
{
|
||||||
neighborhoodSize = HNSW_NEIGHBOR_ARRAY_SIZE(HnswGetLayerM(m, lc));
|
neighborhoodSize = HNSW_NEIGHBOR_ARRAY_SIZE(lm);
|
||||||
neighborhoodData = palloc(neighborhoodSize);
|
localNeighborhood = palloc(neighborhoodSize);
|
||||||
}
|
}
|
||||||
|
|
||||||
/* Add entry points to v, C, and W */
|
/* Add entry points to v, C, and W */
|
||||||
foreach(lc2, ep)
|
foreach(lc2, ep)
|
||||||
{
|
{
|
||||||
HnswCandidate *hc = (HnswCandidate *) lfirst(lc2);
|
HnswSearchCandidate *hc = (HnswSearchCandidate *) lfirst(lc2);
|
||||||
bool found;
|
bool found;
|
||||||
|
|
||||||
AddToVisited(base, &v, hc, index, &found);
|
AddToVisited(base, &v, hc->element, index, &found);
|
||||||
|
|
||||||
pairingheap_add(C, &(CreatePairingHeapNode(hc)->ph_node));
|
pairingheap_add(C, &hc->c_node);
|
||||||
pairingheap_add(W, &(CreatePairingHeapNode(hc)->ph_node));
|
pairingheap_add(W, &hc->w_node);
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Do not count elements being deleted towards ef when vacuuming. It
|
* Do not count elements being deleted towards ef when vacuuming. It
|
||||||
* would be ideal to do this for inserts as well, but this could
|
* would be ideal to do this for inserts as well, but this could
|
||||||
* affect insert performance.
|
* affect insert performance.
|
||||||
*/
|
*/
|
||||||
if (CountElement(base, skipElement, hc))
|
if (CountElement(skipElement, HnswPtrAccess(base, hc->element)))
|
||||||
wlen++;
|
wlen++;
|
||||||
}
|
}
|
||||||
|
|
||||||
while (!pairingheap_is_empty(C))
|
while (!pairingheap_is_empty(C))
|
||||||
{
|
{
|
||||||
HnswNeighborArray *neighborhood;
|
HnswSearchCandidate *c = HnswGetSearchCandidate(c_node, pairingheap_remove_first(C));
|
||||||
HnswCandidate *c = ((HnswPairingHeapNode *) pairingheap_remove_first(C))->inner;
|
HnswSearchCandidate *f = HnswGetSearchCandidate(w_node, pairingheap_first(W));
|
||||||
HnswCandidate *f = ((HnswPairingHeapNode *) pairingheap_first(W))->inner;
|
|
||||||
HnswElement cElement;
|
HnswElement cElement;
|
||||||
|
|
||||||
if (c->distance > f->distance)
|
if (c->distance > f->distance)
|
||||||
@@ -773,73 +844,67 @@ HnswSearchLayer(char *base, Datum q, List *ep, int ef, int lc, Relation index, F
|
|||||||
|
|
||||||
cElement = HnswPtrAccess(base, c->element);
|
cElement = HnswPtrAccess(base, c->element);
|
||||||
|
|
||||||
if (HnswPtrIsNull(base, cElement->neighbors))
|
|
||||||
HnswLoadNeighbors(cElement, index, m);
|
|
||||||
|
|
||||||
/* Get the neighborhood at layer lc */
|
|
||||||
neighborhood = HnswGetNeighbors(base, cElement, lc);
|
|
||||||
|
|
||||||
/* Copy neighborhood to local memory if needed */
|
|
||||||
if (index == NULL)
|
if (index == NULL)
|
||||||
|
HnswLoadUnvisitedFromMemory(base, cElement, unvisited, &unvisitedLength, &v, lc, localNeighborhood, neighborhoodSize);
|
||||||
|
else
|
||||||
|
HnswLoadUnvisitedFromDisk(cElement, unvisited, &unvisitedLength, &v, index, m, lm, lc);
|
||||||
|
|
||||||
|
for (int i = 0; i < unvisitedLength; i++)
|
||||||
{
|
{
|
||||||
LWLockAcquire(&cElement->lock, LW_SHARED);
|
HnswElement eElement;
|
||||||
memcpy(neighborhoodData, neighborhood, neighborhoodSize);
|
HnswSearchCandidate *e;
|
||||||
LWLockRelease(&cElement->lock);
|
float eDistance;
|
||||||
neighborhood = neighborhoodData;
|
bool alwaysAdd = wlen < ef;
|
||||||
}
|
|
||||||
|
|
||||||
for (int i = 0; i < neighborhood->length; i++)
|
f = HnswGetSearchCandidate(w_node, pairingheap_first(W));
|
||||||
{
|
|
||||||
HnswCandidate *e = &neighborhood->items[i];
|
|
||||||
bool visited;
|
|
||||||
|
|
||||||
AddToVisited(base, &v, e, index, &visited);
|
if (index == NULL)
|
||||||
|
|
||||||
if (!visited)
|
|
||||||
{
|
{
|
||||||
float eDistance;
|
eElement = unvisited[i].element;
|
||||||
HnswElement eElement = HnswPtrAccess(base, e->element);
|
eDistance = GetElementDistance(base, eElement, q, procinfo, collation);
|
||||||
bool alwaysAdd = wlen < ef;
|
}
|
||||||
|
else
|
||||||
|
{
|
||||||
|
ItemPointer indextid = &unvisited[i].indextid;
|
||||||
|
BlockNumber blkno = ItemPointerGetBlockNumber(indextid);
|
||||||
|
OffsetNumber offno = ItemPointerGetOffsetNumber(indextid);
|
||||||
|
|
||||||
f = ((HnswPairingHeapNode *) pairingheap_first(W))->inner;
|
/* Avoid any allocations if not adding */
|
||||||
|
eElement = NULL;
|
||||||
|
HnswLoadElementImpl(blkno, offno, &eDistance, &q, index, procinfo, collation, inserting, alwaysAdd ? NULL : &f->distance, &eElement);
|
||||||
|
|
||||||
if (index == NULL)
|
if (eElement == NULL)
|
||||||
eDistance = GetCandidateDistance(base, e, q, procinfo, collation);
|
continue;
|
||||||
else
|
}
|
||||||
HnswLoadElement(eElement, &eDistance, &q, index, procinfo, collation, inserting, alwaysAdd ? NULL : &f->distance);
|
|
||||||
|
|
||||||
if (eDistance < f->distance || alwaysAdd)
|
if (!(eDistance < f->distance || alwaysAdd))
|
||||||
{
|
continue;
|
||||||
HnswCandidate *ec;
|
|
||||||
|
|
||||||
Assert(!eElement->deleted);
|
Assert(!eElement->deleted);
|
||||||
|
|
||||||
/* Make robust to issues */
|
/* Make robust to issues */
|
||||||
if (eElement->level < lc)
|
if (eElement->level < lc)
|
||||||
continue;
|
continue;
|
||||||
|
|
||||||
/* Copy e */
|
/* Create a new candidate */
|
||||||
ec = palloc(sizeof(HnswCandidate));
|
e = palloc(sizeof(HnswSearchCandidate));
|
||||||
HnswPtrStore(base, ec->element, eElement);
|
HnswPtrStore(base, e->element, eElement);
|
||||||
ec->distance = eDistance;
|
e->distance = eDistance;
|
||||||
|
pairingheap_add(C, &e->c_node);
|
||||||
|
pairingheap_add(W, &e->w_node);
|
||||||
|
|
||||||
pairingheap_add(C, &(CreatePairingHeapNode(ec)->ph_node));
|
/*
|
||||||
pairingheap_add(W, &(CreatePairingHeapNode(ec)->ph_node));
|
* Do not count elements being deleted towards ef when vacuuming.
|
||||||
|
* It would be ideal to do this for inserts as well, but this
|
||||||
|
* could affect insert performance.
|
||||||
|
*/
|
||||||
|
if (CountElement(skipElement, eElement))
|
||||||
|
{
|
||||||
|
wlen++;
|
||||||
|
|
||||||
/*
|
/* No need to decrement wlen */
|
||||||
* Do not count elements being deleted towards ef when
|
if (wlen > ef)
|
||||||
* vacuuming. It would be ideal to do this for inserts as
|
pairingheap_remove_first(W);
|
||||||
* well, but this could affect insert performance.
|
|
||||||
*/
|
|
||||||
if (CountElement(base, skipElement, e))
|
|
||||||
{
|
|
||||||
wlen++;
|
|
||||||
|
|
||||||
/* No need to decrement wlen */
|
|
||||||
if (wlen > ef)
|
|
||||||
pairingheap_remove_first(W);
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
@@ -847,7 +912,7 @@ HnswSearchLayer(char *base, Datum q, List *ep, int ef, int lc, Relation index, F
|
|||||||
/* Add each element of W to w */
|
/* Add each element of W to w */
|
||||||
while (!pairingheap_is_empty(W))
|
while (!pairingheap_is_empty(W))
|
||||||
{
|
{
|
||||||
HnswCandidate *hc = ((HnswPairingHeapNode *) pairingheap_remove_first(W))->inner;
|
HnswSearchCandidate *hc = HnswGetSearchCandidate(w_node, pairingheap_remove_first(W));
|
||||||
|
|
||||||
w = lappend(w, hc);
|
w = lappend(w, hc);
|
||||||
}
|
}
|
||||||
@@ -859,17 +924,10 @@ HnswSearchLayer(char *base, Datum q, List *ep, int ef, int lc, Relation index, F
|
|||||||
* Compare candidate distances with pointer tie-breaker
|
* Compare candidate distances with pointer tie-breaker
|
||||||
*/
|
*/
|
||||||
static int
|
static int
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
CompareCandidateDistances(const ListCell *a, const ListCell *b)
|
CompareCandidateDistances(const ListCell *a, const ListCell *b)
|
||||||
{
|
{
|
||||||
HnswCandidate *hca = lfirst(a);
|
HnswCandidate *hca = lfirst(a);
|
||||||
HnswCandidate *hcb = lfirst(b);
|
HnswCandidate *hcb = lfirst(b);
|
||||||
#else
|
|
||||||
CompareCandidateDistances(const void *a, const void *b)
|
|
||||||
{
|
|
||||||
HnswCandidate *hca = lfirst(*(ListCell **) a);
|
|
||||||
HnswCandidate *hcb = lfirst(*(ListCell **) b);
|
|
||||||
#endif
|
|
||||||
|
|
||||||
if (hca->distance < hcb->distance)
|
if (hca->distance < hcb->distance)
|
||||||
return 1;
|
return 1;
|
||||||
@@ -890,17 +948,10 @@ CompareCandidateDistances(const void *a, const void *b)
|
|||||||
* Compare candidate distances with offset tie-breaker
|
* Compare candidate distances with offset tie-breaker
|
||||||
*/
|
*/
|
||||||
static int
|
static int
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
CompareCandidateDistancesOffset(const ListCell *a, const ListCell *b)
|
CompareCandidateDistancesOffset(const ListCell *a, const ListCell *b)
|
||||||
{
|
{
|
||||||
HnswCandidate *hca = lfirst(a);
|
HnswCandidate *hca = lfirst(a);
|
||||||
HnswCandidate *hcb = lfirst(b);
|
HnswCandidate *hcb = lfirst(b);
|
||||||
#else
|
|
||||||
CompareCandidateDistancesOffset(const void *a, const void *b)
|
|
||||||
{
|
|
||||||
HnswCandidate *hca = lfirst(*(ListCell **) a);
|
|
||||||
HnswCandidate *hcb = lfirst(*(ListCell **) b);
|
|
||||||
#endif
|
|
||||||
|
|
||||||
if (hca->distance < hcb->distance)
|
if (hca->distance < hcb->distance)
|
||||||
return 1;
|
return 1;
|
||||||
@@ -1110,7 +1161,7 @@ HnswUpdateConnection(char *base, HnswElement element, HnswCandidate * hc, int lm
|
|||||||
if (HnswPtrIsNull(base, hc3Element->value))
|
if (HnswPtrIsNull(base, hc3Element->value))
|
||||||
HnswLoadElement(hc3Element, &hc3->distance, &q, index, procinfo, collation, true, NULL);
|
HnswLoadElement(hc3Element, &hc3->distance, &q, index, procinfo, collation, true, NULL);
|
||||||
else
|
else
|
||||||
hc3->distance = GetCandidateDistance(base, hc3, q, procinfo, collation);
|
hc3->distance = GetElementDistance(base, hc3Element, q, procinfo, collation);
|
||||||
|
|
||||||
/* Prune element if being deleted */
|
/* Prune element if being deleted */
|
||||||
if (hc3Element->heaptidsLength == 0)
|
if (hc3Element->heaptidsLength == 0)
|
||||||
@@ -1182,7 +1233,6 @@ RemoveElements(char *base, List *w, HnswElement skipElement)
|
|||||||
return w2;
|
return w2;
|
||||||
}
|
}
|
||||||
|
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
/*
|
/*
|
||||||
* Precompute hash
|
* Precompute hash
|
||||||
*/
|
*/
|
||||||
@@ -1198,7 +1248,6 @@ PrecomputeHash(char *base, HnswElement element)
|
|||||||
else
|
else
|
||||||
element->hash = hash_offset(HnswPtrOffset(ptr));
|
element->hash = hash_offset(HnswPtrOffset(ptr));
|
||||||
}
|
}
|
||||||
#endif
|
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Algorithm 1 from paper
|
* Algorithm 1 from paper
|
||||||
@@ -1213,11 +1262,9 @@ HnswFindElementNeighbors(char *base, HnswElement element, HnswElement entryPoint
|
|||||||
Datum q = HnswGetValue(base, element);
|
Datum q = HnswGetValue(base, element);
|
||||||
HnswElement skipElement = existing ? element : NULL;
|
HnswElement skipElement = existing ? element : NULL;
|
||||||
|
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
/* Precompute hash */
|
/* Precompute hash */
|
||||||
if (index == NULL)
|
if (index == NULL)
|
||||||
PrecomputeHash(base, element);
|
PrecomputeHash(base, element);
|
||||||
#endif
|
|
||||||
|
|
||||||
/* No neighbors if no entry point */
|
/* No neighbors if no entry point */
|
||||||
if (entryPoint == NULL)
|
if (entryPoint == NULL)
|
||||||
@@ -1246,16 +1293,27 @@ HnswFindElementNeighbors(char *base, HnswElement element, HnswElement entryPoint
|
|||||||
{
|
{
|
||||||
int lm = HnswGetLayerM(m, lc);
|
int lm = HnswGetLayerM(m, lc);
|
||||||
List *neighbors;
|
List *neighbors;
|
||||||
List *lw;
|
List *lw = NIL;
|
||||||
|
ListCell *lc2;
|
||||||
|
|
||||||
w = HnswSearchLayer(base, q, ep, efConstruction, lc, index, procinfo, collation, m, true, skipElement);
|
w = HnswSearchLayer(base, q, ep, efConstruction, lc, index, procinfo, collation, m, true, skipElement);
|
||||||
|
|
||||||
|
/* Convert search candidates to candidates */
|
||||||
|
foreach(lc2, w)
|
||||||
|
{
|
||||||
|
HnswSearchCandidate *sc = lfirst(lc2);
|
||||||
|
HnswCandidate *hc = palloc(sizeof(HnswCandidate));
|
||||||
|
|
||||||
|
hc->element = sc->element;
|
||||||
|
hc->distance = sc->distance;
|
||||||
|
|
||||||
|
lw = lappend(lw, hc);
|
||||||
|
}
|
||||||
|
|
||||||
/* Elements being deleted or skipped can help with search */
|
/* Elements being deleted or skipped can help with search */
|
||||||
/* but should be removed before selecting neighbors */
|
/* but should be removed before selecting neighbors */
|
||||||
if (index != NULL)
|
if (index != NULL)
|
||||||
lw = RemoveElements(base, w, skipElement);
|
lw = RemoveElements(base, lw, skipElement);
|
||||||
else
|
|
||||||
lw = w;
|
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Candidates are sorted, but not deterministically. Could set
|
* Candidates are sorted, but not deterministically. Could set
|
||||||
@@ -1272,6 +1330,7 @@ HnswFindElementNeighbors(char *base, HnswElement element, HnswElement entryPoint
|
|||||||
|
|
||||||
PGDLLEXPORT Datum l2_normalize(PG_FUNCTION_ARGS);
|
PGDLLEXPORT Datum l2_normalize(PG_FUNCTION_ARGS);
|
||||||
PGDLLEXPORT Datum halfvec_l2_normalize(PG_FUNCTION_ARGS);
|
PGDLLEXPORT Datum halfvec_l2_normalize(PG_FUNCTION_ARGS);
|
||||||
|
PGDLLEXPORT Datum minivec_l2_normalize(PG_FUNCTION_ARGS);
|
||||||
PGDLLEXPORT Datum sparsevec_l2_normalize(PG_FUNCTION_ARGS);
|
PGDLLEXPORT Datum sparsevec_l2_normalize(PG_FUNCTION_ARGS);
|
||||||
|
|
||||||
static void
|
static void
|
||||||
@@ -1280,7 +1339,9 @@ SparsevecCheckValue(Pointer v)
|
|||||||
SparseVector *vec = (SparseVector *) v;
|
SparseVector *vec = (SparseVector *) v;
|
||||||
|
|
||||||
if (vec->nnz > HNSW_MAX_NNZ)
|
if (vec->nnz > HNSW_MAX_NNZ)
|
||||||
elog(ERROR, "sparsevec cannot have more than %d non-zero elements for hnsw index", HNSW_MAX_NNZ);
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
|
||||||
|
errmsg("sparsevec cannot have more than %d non-zero elements for hnsw index", HNSW_MAX_NNZ)));
|
||||||
}
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
@@ -1305,7 +1366,7 @@ HnswGetTypeInfo(Relation index)
|
|||||||
return (const HnswTypeInfo *) DatumGetPointer(FunctionCall0Coll(procinfo, InvalidOid));
|
return (const HnswTypeInfo *) DatumGetPointer(FunctionCall0Coll(procinfo, InvalidOid));
|
||||||
}
|
}
|
||||||
|
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(hnsw_halfvec_support);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(hnsw_halfvec_support);
|
||||||
Datum
|
Datum
|
||||||
hnsw_halfvec_support(PG_FUNCTION_ARGS)
|
hnsw_halfvec_support(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1318,7 +1379,21 @@ hnsw_halfvec_support(PG_FUNCTION_ARGS)
|
|||||||
PG_RETURN_POINTER(&typeInfo);
|
PG_RETURN_POINTER(&typeInfo);
|
||||||
};
|
};
|
||||||
|
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(hnsw_bit_support);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(hnsw_minivec_support);
|
||||||
|
Datum
|
||||||
|
hnsw_minivec_support(PG_FUNCTION_ARGS)
|
||||||
|
{
|
||||||
|
static const HnswTypeInfo typeInfo = {
|
||||||
|
.maxDimensions = HNSW_MAX_DIM * 4,
|
||||||
|
/* Do not normalize to maximize precision */
|
||||||
|
.normalize = NULL,
|
||||||
|
.checkValue = NULL
|
||||||
|
};
|
||||||
|
|
||||||
|
PG_RETURN_POINTER(&typeInfo);
|
||||||
|
};
|
||||||
|
|
||||||
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(hnsw_bit_support);
|
||||||
Datum
|
Datum
|
||||||
hnsw_bit_support(PG_FUNCTION_ARGS)
|
hnsw_bit_support(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1331,7 +1406,7 @@ hnsw_bit_support(PG_FUNCTION_ARGS)
|
|||||||
PG_RETURN_POINTER(&typeInfo);
|
PG_RETURN_POINTER(&typeInfo);
|
||||||
};
|
};
|
||||||
|
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(hnsw_sparsevec_support);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(hnsw_sparsevec_support);
|
||||||
Datum
|
Datum
|
||||||
hnsw_sparsevec_support(PG_FUNCTION_ARGS)
|
hnsw_sparsevec_support(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
|
|||||||
@@ -26,12 +26,6 @@
|
|||||||
#include "pgstat.h"
|
#include "pgstat.h"
|
||||||
#endif
|
#endif
|
||||||
|
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
#define CALLBACK_ITEM_POINTER ItemPointer tid
|
|
||||||
#else
|
|
||||||
#define CALLBACK_ITEM_POINTER HeapTuple hup
|
|
||||||
#endif
|
|
||||||
|
|
||||||
#if PG_VERSION_NUM >= 140000
|
#if PG_VERSION_NUM >= 140000
|
||||||
#include "utils/backend_status.h"
|
#include "utils/backend_status.h"
|
||||||
#include "utils/wait_event.h"
|
#include "utils/wait_event.h"
|
||||||
@@ -96,7 +90,7 @@ AddSample(Datum *values, IvfflatBuildState * buildstate)
|
|||||||
* Callback for sampling
|
* Callback for sampling
|
||||||
*/
|
*/
|
||||||
static void
|
static void
|
||||||
SampleCallback(Relation index, CALLBACK_ITEM_POINTER, Datum *values,
|
SampleCallback(Relation index, ItemPointer tid, Datum *values,
|
||||||
bool *isnull, bool tupleIsAlive, void *state)
|
bool *isnull, bool tupleIsAlive, void *state)
|
||||||
{
|
{
|
||||||
IvfflatBuildState *buildstate = (IvfflatBuildState *) state;
|
IvfflatBuildState *buildstate = (IvfflatBuildState *) state;
|
||||||
@@ -207,16 +201,12 @@ AddTupleToSort(Relation index, ItemPointer tid, Datum *values, IvfflatBuildState
|
|||||||
* Callback for table_index_build_scan
|
* Callback for table_index_build_scan
|
||||||
*/
|
*/
|
||||||
static void
|
static void
|
||||||
BuildCallback(Relation index, CALLBACK_ITEM_POINTER, Datum *values,
|
BuildCallback(Relation index, ItemPointer tid, Datum *values,
|
||||||
bool *isnull, bool tupleIsAlive, void *state)
|
bool *isnull, bool tupleIsAlive, void *state)
|
||||||
{
|
{
|
||||||
IvfflatBuildState *buildstate = (IvfflatBuildState *) state;
|
IvfflatBuildState *buildstate = (IvfflatBuildState *) state;
|
||||||
MemoryContext oldCtx;
|
MemoryContext oldCtx;
|
||||||
|
|
||||||
#if PG_VERSION_NUM < 130000
|
|
||||||
ItemPointer tid = &hup->t_self;
|
|
||||||
#endif
|
|
||||||
|
|
||||||
/* Skip nulls */
|
/* Skip nulls */
|
||||||
if (isnull[0])
|
if (isnull[0])
|
||||||
return;
|
return;
|
||||||
@@ -335,14 +325,20 @@ InitBuildState(IvfflatBuildState * buildstate, Relation heap, Relation index, In
|
|||||||
|
|
||||||
/* Disallow varbit since require fixed dimensions */
|
/* Disallow varbit since require fixed dimensions */
|
||||||
if (TupleDescAttr(index->rd_att, 0)->atttypid == VARBITOID)
|
if (TupleDescAttr(index->rd_att, 0)->atttypid == VARBITOID)
|
||||||
elog(ERROR, "type not supported for ivfflat index");
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
|
||||||
|
errmsg("type not supported for ivfflat index")));
|
||||||
|
|
||||||
/* Require column to have dimensions to be indexed */
|
/* Require column to have dimensions to be indexed */
|
||||||
if (buildstate->dimensions < 0)
|
if (buildstate->dimensions < 0)
|
||||||
elog(ERROR, "column does not have dimensions");
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
|
||||||
|
errmsg("column does not have dimensions")));
|
||||||
|
|
||||||
if (buildstate->dimensions > buildstate->typeInfo->maxDimensions)
|
if (buildstate->dimensions > buildstate->typeInfo->maxDimensions)
|
||||||
elog(ERROR, "column cannot have more than %d dimensions for ivfflat index", buildstate->typeInfo->maxDimensions);
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_PROGRAM_LIMIT_EXCEEDED),
|
||||||
|
errmsg("column cannot have more than %d dimensions for ivfflat index", buildstate->typeInfo->maxDimensions)));
|
||||||
|
|
||||||
buildstate->reltuples = 0;
|
buildstate->reltuples = 0;
|
||||||
buildstate->indtuples = 0;
|
buildstate->indtuples = 0;
|
||||||
@@ -355,7 +351,9 @@ InitBuildState(IvfflatBuildState * buildstate, Relation heap, Relation index, In
|
|||||||
|
|
||||||
/* Require more than one dimension for spherical k-means */
|
/* Require more than one dimension for spherical k-means */
|
||||||
if (buildstate->kmeansnormprocinfo != NULL && buildstate->dimensions == 1)
|
if (buildstate->kmeansnormprocinfo != NULL && buildstate->dimensions == 1)
|
||||||
elog(ERROR, "dimensions must be greater than one for this opclass");
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
|
||||||
|
errmsg("dimensions must be greater than one for this opclass")));
|
||||||
|
|
||||||
/* Create tuple description for sorting */
|
/* Create tuple description for sorting */
|
||||||
buildstate->tupdesc = CreateTemplateTupleDesc(3);
|
buildstate->tupdesc = CreateTemplateTupleDesc(3);
|
||||||
@@ -562,6 +560,20 @@ PrintKmeansMetrics(IvfflatBuildState * buildstate)
|
|||||||
}
|
}
|
||||||
#endif
|
#endif
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Initialize build sort state
|
||||||
|
*/
|
||||||
|
static Tuplesortstate *
|
||||||
|
InitBuildSortState(TupleDesc tupdesc, int memory, SortCoordinate coordinate)
|
||||||
|
{
|
||||||
|
AttrNumber attNums[] = {1};
|
||||||
|
Oid sortOperators[] = {Int4LessOperator};
|
||||||
|
Oid sortCollations[] = {InvalidOid};
|
||||||
|
bool nullsFirstFlags[] = {false};
|
||||||
|
|
||||||
|
return tuplesort_begin_heap(tupdesc, 1, attNums, sortOperators, sortCollations, nullsFirstFlags, memory, coordinate, false);
|
||||||
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Within leader, wait for end of heap scan
|
* Within leader, wait for end of heap scan
|
||||||
*/
|
*/
|
||||||
@@ -609,12 +621,6 @@ IvfflatParallelScanAndSort(IvfflatSpool * ivfspool, IvfflatShared * ivfshared, S
|
|||||||
double reltuples;
|
double reltuples;
|
||||||
IndexInfo *indexInfo;
|
IndexInfo *indexInfo;
|
||||||
|
|
||||||
/* Sort options, which must match AssignTuples */
|
|
||||||
AttrNumber attNums[] = {1};
|
|
||||||
Oid sortOperators[] = {Int4LessOperator};
|
|
||||||
Oid sortCollations[] = {InvalidOid};
|
|
||||||
bool nullsFirstFlags[] = {false};
|
|
||||||
|
|
||||||
/* Initialize local tuplesort coordination state */
|
/* Initialize local tuplesort coordination state */
|
||||||
coordinate = palloc0(sizeof(SortCoordinateData));
|
coordinate = palloc0(sizeof(SortCoordinateData));
|
||||||
coordinate->isWorker = true;
|
coordinate->isWorker = true;
|
||||||
@@ -627,7 +633,7 @@ IvfflatParallelScanAndSort(IvfflatSpool * ivfspool, IvfflatShared * ivfshared, S
|
|||||||
InitBuildState(&buildstate, ivfspool->heap, ivfspool->index, indexInfo);
|
InitBuildState(&buildstate, ivfspool->heap, ivfspool->index, indexInfo);
|
||||||
memcpy(buildstate.centers->items, ivfcenters, buildstate.centers->itemsize * buildstate.centers->maxlen);
|
memcpy(buildstate.centers->items, ivfcenters, buildstate.centers->itemsize * buildstate.centers->maxlen);
|
||||||
buildstate.centers->length = buildstate.centers->maxlen;
|
buildstate.centers->length = buildstate.centers->maxlen;
|
||||||
ivfspool->sortstate = tuplesort_begin_heap(buildstate.tupdesc, 1, attNums, sortOperators, sortCollations, nullsFirstFlags, sortmem, coordinate, false);
|
ivfspool->sortstate = InitBuildSortState(buildstate.tupdesc, sortmem, coordinate);
|
||||||
buildstate.sortstate = ivfspool->sortstate;
|
buildstate.sortstate = ivfspool->sortstate;
|
||||||
scan = table_beginscan_parallel(ivfspool->heap,
|
scan = table_beginscan_parallel(ivfspool->heap,
|
||||||
ParallelTableScanFromIvfflatShared(ivfshared));
|
ParallelTableScanFromIvfflatShared(ivfshared));
|
||||||
@@ -924,12 +930,6 @@ AssignTuples(IvfflatBuildState * buildstate)
|
|||||||
int parallel_workers = 0;
|
int parallel_workers = 0;
|
||||||
SortCoordinate coordinate = NULL;
|
SortCoordinate coordinate = NULL;
|
||||||
|
|
||||||
/* Sort options, which must match IvfflatParallelScanAndSort */
|
|
||||||
AttrNumber attNums[] = {1};
|
|
||||||
Oid sortOperators[] = {Int4LessOperator};
|
|
||||||
Oid sortCollations[] = {InvalidOid};
|
|
||||||
bool nullsFirstFlags[] = {false};
|
|
||||||
|
|
||||||
pgstat_progress_update_param(PROGRESS_CREATEIDX_SUBPHASE, PROGRESS_IVFFLAT_PHASE_ASSIGN);
|
pgstat_progress_update_param(PROGRESS_CREATEIDX_SUBPHASE, PROGRESS_IVFFLAT_PHASE_ASSIGN);
|
||||||
|
|
||||||
/* Calculate parallel workers */
|
/* Calculate parallel workers */
|
||||||
@@ -950,7 +950,7 @@ AssignTuples(IvfflatBuildState * buildstate)
|
|||||||
}
|
}
|
||||||
|
|
||||||
/* Begin serial/leader tuplesort */
|
/* Begin serial/leader tuplesort */
|
||||||
buildstate->sortstate = tuplesort_begin_heap(buildstate->tupdesc, 1, attNums, sortOperators, sortCollations, nullsFirstFlags, maintenance_work_mem, coordinate, false);
|
buildstate->sortstate = InitBuildSortState(buildstate->tupdesc, maintenance_work_mem, coordinate);
|
||||||
|
|
||||||
/* Add tuples to sort */
|
/* Add tuples to sort */
|
||||||
if (buildstate->heap != NULL)
|
if (buildstate->heap != NULL)
|
||||||
|
|||||||
@@ -7,6 +7,7 @@
|
|||||||
#include "commands/progress.h"
|
#include "commands/progress.h"
|
||||||
#include "commands/vacuum.h"
|
#include "commands/vacuum.h"
|
||||||
#include "ivfflat.h"
|
#include "ivfflat.h"
|
||||||
|
#include "utils/float.h"
|
||||||
#include "utils/guc.h"
|
#include "utils/guc.h"
|
||||||
#include "utils/selfuncs.h"
|
#include "utils/selfuncs.h"
|
||||||
#include "utils/spccache.h"
|
#include "utils/spccache.h"
|
||||||
@@ -26,11 +27,7 @@ IvfflatInit(void)
|
|||||||
{
|
{
|
||||||
ivfflat_relopt_kind = add_reloption_kind();
|
ivfflat_relopt_kind = add_reloption_kind();
|
||||||
add_int_reloption(ivfflat_relopt_kind, "lists", "Number of inverted lists",
|
add_int_reloption(ivfflat_relopt_kind, "lists", "Number of inverted lists",
|
||||||
IVFFLAT_DEFAULT_LISTS, IVFFLAT_MIN_LISTS, IVFFLAT_MAX_LISTS
|
IVFFLAT_DEFAULT_LISTS, IVFFLAT_MIN_LISTS, IVFFLAT_MAX_LISTS, AccessExclusiveLock);
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
,AccessExclusiveLock
|
|
||||||
#endif
|
|
||||||
);
|
|
||||||
|
|
||||||
DefineCustomIntVariable("ivfflat.probes", "Sets the number of probes",
|
DefineCustomIntVariable("ivfflat.probes", "Sets the number of probes",
|
||||||
"Valid range is 1..lists.", &ivfflat_probes,
|
"Valid range is 1..lists.", &ivfflat_probes,
|
||||||
@@ -78,8 +75,8 @@ ivfflatcostestimate(PlannerInfo *root, IndexPath *path, double loop_count,
|
|||||||
/* Never use index without order */
|
/* Never use index without order */
|
||||||
if (path->indexorderbys == NULL)
|
if (path->indexorderbys == NULL)
|
||||||
{
|
{
|
||||||
*indexStartupCost = DBL_MAX;
|
*indexStartupCost = get_float8_infinity();
|
||||||
*indexTotalCost = DBL_MAX;
|
*indexTotalCost = get_float8_infinity();
|
||||||
*indexSelectivity = 0;
|
*indexSelectivity = 0;
|
||||||
*indexCorrelation = 0;
|
*indexCorrelation = 0;
|
||||||
*indexPages = 0;
|
*indexPages = 0;
|
||||||
@@ -148,23 +145,10 @@ ivfflatoptions(Datum reloptions, bool validate)
|
|||||||
{"lists", RELOPT_TYPE_INT, offsetof(IvfflatOptions, lists)},
|
{"lists", RELOPT_TYPE_INT, offsetof(IvfflatOptions, lists)},
|
||||||
};
|
};
|
||||||
|
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
return (bytea *) build_reloptions(reloptions, validate,
|
return (bytea *) build_reloptions(reloptions, validate,
|
||||||
ivfflat_relopt_kind,
|
ivfflat_relopt_kind,
|
||||||
sizeof(IvfflatOptions),
|
sizeof(IvfflatOptions),
|
||||||
tab, lengthof(tab));
|
tab, lengthof(tab));
|
||||||
#else
|
|
||||||
relopt_value *options;
|
|
||||||
int numoptions;
|
|
||||||
IvfflatOptions *rdopts;
|
|
||||||
|
|
||||||
options = parseRelOptions(reloptions, validate, ivfflat_relopt_kind, &numoptions);
|
|
||||||
rdopts = allocateReloptStruct(sizeof(IvfflatOptions), options, numoptions);
|
|
||||||
fillRelOptions((void *) rdopts, sizeof(IvfflatOptions), options, numoptions,
|
|
||||||
validate, tab, lengthof(tab));
|
|
||||||
|
|
||||||
return (bytea *) rdopts;
|
|
||||||
#endif
|
|
||||||
}
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
@@ -181,7 +165,7 @@ ivfflatvalidate(Oid opclassoid)
|
|||||||
*
|
*
|
||||||
* See https://www.postgresql.org/docs/current/index-api.html
|
* See https://www.postgresql.org/docs/current/index-api.html
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(ivfflathandler);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(ivfflathandler);
|
||||||
Datum
|
Datum
|
||||||
ivfflathandler(PG_FUNCTION_ARGS)
|
ivfflathandler(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -189,9 +173,7 @@ ivfflathandler(PG_FUNCTION_ARGS)
|
|||||||
|
|
||||||
amroutine->amstrategies = 0;
|
amroutine->amstrategies = 0;
|
||||||
amroutine->amsupport = 5;
|
amroutine->amsupport = 5;
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
amroutine->amoptsprocnum = 0;
|
amroutine->amoptsprocnum = 0;
|
||||||
#endif
|
|
||||||
amroutine->amcanorder = false;
|
amroutine->amcanorder = false;
|
||||||
amroutine->amcanorderbyop = true;
|
amroutine->amcanorderbyop = true;
|
||||||
amroutine->amcanbackward = false; /* can change direction mid-scan */
|
amroutine->amcanbackward = false; /* can change direction mid-scan */
|
||||||
@@ -204,17 +186,24 @@ ivfflathandler(PG_FUNCTION_ARGS)
|
|||||||
amroutine->amclusterable = false;
|
amroutine->amclusterable = false;
|
||||||
amroutine->ampredlocks = false;
|
amroutine->ampredlocks = false;
|
||||||
amroutine->amcanparallel = false;
|
amroutine->amcanparallel = false;
|
||||||
amroutine->amcaninclude = false;
|
#if PG_VERSION_NUM >= 170000
|
||||||
#if PG_VERSION_NUM >= 130000
|
amroutine->amcanbuildparallel = true;
|
||||||
amroutine->amusemaintenanceworkmem = false; /* not used during VACUUM */
|
|
||||||
amroutine->amparallelvacuumoptions = VACUUM_OPTION_PARALLEL_BULKDEL;
|
|
||||||
#endif
|
#endif
|
||||||
|
amroutine->amcaninclude = false;
|
||||||
|
amroutine->amusemaintenanceworkmem = false; /* not used during VACUUM */
|
||||||
|
#if PG_VERSION_NUM >= 160000
|
||||||
|
amroutine->amsummarizing = false;
|
||||||
|
#endif
|
||||||
|
amroutine->amparallelvacuumoptions = VACUUM_OPTION_PARALLEL_BULKDEL;
|
||||||
amroutine->amkeytype = InvalidOid;
|
amroutine->amkeytype = InvalidOid;
|
||||||
|
|
||||||
/* Interface functions */
|
/* Interface functions */
|
||||||
amroutine->ambuild = ivfflatbuild;
|
amroutine->ambuild = ivfflatbuild;
|
||||||
amroutine->ambuildempty = ivfflatbuildempty;
|
amroutine->ambuildempty = ivfflatbuildempty;
|
||||||
amroutine->aminsert = ivfflatinsert;
|
amroutine->aminsert = ivfflatinsert;
|
||||||
|
#if PG_VERSION_NUM >= 170000
|
||||||
|
amroutine->aminsertcleanup = NULL;
|
||||||
|
#endif
|
||||||
amroutine->ambulkdelete = ivfflatbulkdelete;
|
amroutine->ambulkdelete = ivfflatbulkdelete;
|
||||||
amroutine->amvacuumcleanup = ivfflatvacuumcleanup;
|
amroutine->amvacuumcleanup = ivfflatvacuumcleanup;
|
||||||
amroutine->amcanreturn = NULL; /* tuple not included in heapsort */
|
amroutine->amcanreturn = NULL; /* tuple not included in heapsort */
|
||||||
|
|||||||
@@ -253,8 +253,9 @@ typedef struct IvfflatScanOpaqueData
|
|||||||
/* Sorting */
|
/* Sorting */
|
||||||
Tuplesortstate *sortstate;
|
Tuplesortstate *sortstate;
|
||||||
TupleDesc tupdesc;
|
TupleDesc tupdesc;
|
||||||
TupleTableSlot *slot;
|
TupleTableSlot *vslot;
|
||||||
bool isnull;
|
TupleTableSlot *mslot;
|
||||||
|
BufferAccessStrategy bas;
|
||||||
|
|
||||||
/* Support functions */
|
/* Support functions */
|
||||||
FmgrInfo *procinfo;
|
FmgrInfo *procinfo;
|
||||||
|
|||||||
@@ -151,12 +151,8 @@ RandomCenters(Relation index, VectorArray centers, const IvfflatTypeInfo * typeI
|
|||||||
static void
|
static void
|
||||||
ShowMemoryUsage(MemoryContext context, Size estimatedSize)
|
ShowMemoryUsage(MemoryContext context, Size estimatedSize)
|
||||||
{
|
{
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
elog(INFO, "total memory: %zu MB",
|
elog(INFO, "total memory: %zu MB",
|
||||||
MemoryContextMemAllocated(context, true) / (1024 * 1024));
|
MemoryContextMemAllocated(context, true) / (1024 * 1024));
|
||||||
#else
|
|
||||||
MemoryContextStats(context);
|
|
||||||
#endif
|
|
||||||
elog(INFO, "estimated memory: %zu MB", estimatedSize / (1024 * 1024));
|
elog(INFO, "estimated memory: %zu MB", estimatedSize / (1024 * 1024));
|
||||||
}
|
}
|
||||||
#endif
|
#endif
|
||||||
@@ -327,7 +323,7 @@ ElkanKmeans(Relation index, VectorArray samples, VectorArray centers, const Ivff
|
|||||||
newCenters->length = numCenters;
|
newCenters->length = numCenters;
|
||||||
|
|
||||||
#ifdef IVFFLAT_MEMORY
|
#ifdef IVFFLAT_MEMORY
|
||||||
ShowMemoryUsage(MemoryContextGetParent(CurrentMemoryContext));
|
ShowMemoryUsage(MemoryContextGetParent(CurrentMemoryContext), totalSize);
|
||||||
#endif
|
#endif
|
||||||
|
|
||||||
/* Pick initial centers */
|
/* Pick initial centers */
|
||||||
|
|||||||
@@ -11,16 +11,23 @@
|
|||||||
#include "pgstat.h"
|
#include "pgstat.h"
|
||||||
#include "storage/bufmgr.h"
|
#include "storage/bufmgr.h"
|
||||||
|
|
||||||
|
#ifdef IVFFLAT_MEMORY
|
||||||
|
#include "utils/memutils.h"
|
||||||
|
#endif
|
||||||
|
|
||||||
|
#define GetScanList(ptr) pairingheap_container(IvfflatScanList, ph_node, ptr)
|
||||||
|
#define GetScanListConst(ptr) pairingheap_const_container(IvfflatScanList, ph_node, ptr)
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Compare list distances
|
* Compare list distances
|
||||||
*/
|
*/
|
||||||
static int
|
static int
|
||||||
CompareLists(const pairingheap_node *a, const pairingheap_node *b, void *arg)
|
CompareLists(const pairingheap_node *a, const pairingheap_node *b, void *arg)
|
||||||
{
|
{
|
||||||
if (((const IvfflatScanList *) a)->distance > ((const IvfflatScanList *) b)->distance)
|
if (GetScanListConst(a)->distance > GetScanListConst(b)->distance)
|
||||||
return 1;
|
return 1;
|
||||||
|
|
||||||
if (((const IvfflatScanList *) a)->distance < ((const IvfflatScanList *) b)->distance)
|
if (GetScanListConst(a)->distance < GetScanListConst(b)->distance)
|
||||||
return -1;
|
return -1;
|
||||||
|
|
||||||
return 0;
|
return 0;
|
||||||
@@ -72,14 +79,14 @@ GetScanLists(IndexScanDesc scan, Datum value)
|
|||||||
|
|
||||||
/* Calculate max distance */
|
/* Calculate max distance */
|
||||||
if (listCount == so->probes)
|
if (listCount == so->probes)
|
||||||
maxDistance = ((IvfflatScanList *) pairingheap_first(so->listQueue))->distance;
|
maxDistance = GetScanList(pairingheap_first(so->listQueue))->distance;
|
||||||
}
|
}
|
||||||
else if (distance < maxDistance)
|
else if (distance < maxDistance)
|
||||||
{
|
{
|
||||||
IvfflatScanList *scanlist;
|
IvfflatScanList *scanlist;
|
||||||
|
|
||||||
/* Remove */
|
/* Remove */
|
||||||
scanlist = (IvfflatScanList *) pairingheap_remove_first(so->listQueue);
|
scanlist = GetScanList(pairingheap_remove_first(so->listQueue));
|
||||||
|
|
||||||
/* Reuse */
|
/* Reuse */
|
||||||
scanlist->startPage = list->startPage;
|
scanlist->startPage = list->startPage;
|
||||||
@@ -87,7 +94,7 @@ GetScanLists(IndexScanDesc scan, Datum value)
|
|||||||
pairingheap_add(so->listQueue, &scanlist->ph_node);
|
pairingheap_add(so->listQueue, &scanlist->ph_node);
|
||||||
|
|
||||||
/* Update max distance */
|
/* Update max distance */
|
||||||
maxDistance = ((IvfflatScanList *) pairingheap_first(so->listQueue))->distance;
|
maxDistance = GetScanList(pairingheap_first(so->listQueue))->distance;
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
@@ -106,19 +113,12 @@ GetScanItems(IndexScanDesc scan, Datum value)
|
|||||||
IvfflatScanOpaque so = (IvfflatScanOpaque) scan->opaque;
|
IvfflatScanOpaque so = (IvfflatScanOpaque) scan->opaque;
|
||||||
TupleDesc tupdesc = RelationGetDescr(scan->indexRelation);
|
TupleDesc tupdesc = RelationGetDescr(scan->indexRelation);
|
||||||
double tuples = 0;
|
double tuples = 0;
|
||||||
TupleTableSlot *slot = MakeSingleTupleTableSlot(so->tupdesc, &TTSOpsVirtual);
|
TupleTableSlot *slot = so->vslot;
|
||||||
|
|
||||||
/*
|
|
||||||
* Reuse same set of shared buffers for scan
|
|
||||||
*
|
|
||||||
* See postgres/src/backend/storage/buffer/README for description
|
|
||||||
*/
|
|
||||||
BufferAccessStrategy bas = GetAccessStrategy(BAS_BULKREAD);
|
|
||||||
|
|
||||||
/* Search closest probes lists */
|
/* Search closest probes lists */
|
||||||
while (!pairingheap_is_empty(so->listQueue))
|
while (!pairingheap_is_empty(so->listQueue))
|
||||||
{
|
{
|
||||||
BlockNumber searchPage = ((IvfflatScanList *) pairingheap_remove_first(so->listQueue))->startPage;
|
BlockNumber searchPage = GetScanList(pairingheap_remove_first(so->listQueue))->startPage;
|
||||||
|
|
||||||
/* Search all entry pages for list */
|
/* Search all entry pages for list */
|
||||||
while (BlockNumberIsValid(searchPage))
|
while (BlockNumberIsValid(searchPage))
|
||||||
@@ -127,7 +127,7 @@ GetScanItems(IndexScanDesc scan, Datum value)
|
|||||||
Page page;
|
Page page;
|
||||||
OffsetNumber maxoffno;
|
OffsetNumber maxoffno;
|
||||||
|
|
||||||
buf = ReadBufferExtended(scan->indexRelation, MAIN_FORKNUM, searchPage, RBM_NORMAL, bas);
|
buf = ReadBufferExtended(scan->indexRelation, MAIN_FORKNUM, searchPage, RBM_NORMAL, so->bas);
|
||||||
LockBuffer(buf, BUFFER_LOCK_SHARE);
|
LockBuffer(buf, BUFFER_LOCK_SHARE);
|
||||||
page = BufferGetPage(buf);
|
page = BufferGetPage(buf);
|
||||||
maxoffno = PageGetMaxOffsetNumber(page);
|
maxoffno = PageGetMaxOffsetNumber(page);
|
||||||
@@ -166,8 +166,6 @@ GetScanItems(IndexScanDesc scan, Datum value)
|
|||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
FreeAccessStrategy(bas);
|
|
||||||
|
|
||||||
if (tuples < 100)
|
if (tuples < 100)
|
||||||
ereport(DEBUG1,
|
ereport(DEBUG1,
|
||||||
(errmsg("index scan found few tuples"),
|
(errmsg("index scan found few tuples"),
|
||||||
@@ -217,6 +215,20 @@ GetScanValue(IndexScanDesc scan)
|
|||||||
return value;
|
return value;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Initialize scan sort state
|
||||||
|
*/
|
||||||
|
static Tuplesortstate *
|
||||||
|
InitScanSortState(TupleDesc tupdesc)
|
||||||
|
{
|
||||||
|
AttrNumber attNums[] = {1};
|
||||||
|
Oid sortOperators[] = {Float8LessOperator};
|
||||||
|
Oid sortCollations[] = {InvalidOid};
|
||||||
|
bool nullsFirstFlags[] = {false};
|
||||||
|
|
||||||
|
return tuplesort_begin_heap(tupdesc, 1, attNums, sortOperators, sortCollations, nullsFirstFlags, work_mem, NULL, false);
|
||||||
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Prepare for an index scan
|
* Prepare for an index scan
|
||||||
*/
|
*/
|
||||||
@@ -227,10 +239,6 @@ ivfflatbeginscan(Relation index, int nkeys, int norderbys)
|
|||||||
IvfflatScanOpaque so;
|
IvfflatScanOpaque so;
|
||||||
int lists;
|
int lists;
|
||||||
int dimensions;
|
int dimensions;
|
||||||
AttrNumber attNums[] = {1};
|
|
||||||
Oid sortOperators[] = {Float8LessOperator};
|
|
||||||
Oid sortCollations[] = {InvalidOid};
|
|
||||||
bool nullsFirstFlags[] = {false};
|
|
||||||
int probes = ivfflat_probes;
|
int probes = ivfflat_probes;
|
||||||
|
|
||||||
scan = RelationGetIndexScan(index, nkeys, norderbys);
|
scan = RelationGetIndexScan(index, nkeys, norderbys);
|
||||||
@@ -258,9 +266,18 @@ ivfflatbeginscan(Relation index, int nkeys, int norderbys)
|
|||||||
TupleDescInitEntry(so->tupdesc, (AttrNumber) 2, "heaptid", TIDOID, -1, 0);
|
TupleDescInitEntry(so->tupdesc, (AttrNumber) 2, "heaptid", TIDOID, -1, 0);
|
||||||
|
|
||||||
/* Prep sort */
|
/* Prep sort */
|
||||||
so->sortstate = tuplesort_begin_heap(so->tupdesc, 1, attNums, sortOperators, sortCollations, nullsFirstFlags, work_mem, NULL, false);
|
so->sortstate = InitScanSortState(so->tupdesc);
|
||||||
|
|
||||||
so->slot = MakeSingleTupleTableSlot(so->tupdesc, &TTSOpsMinimalTuple);
|
/* Need separate slots for puttuple and gettuple */
|
||||||
|
so->vslot = MakeSingleTupleTableSlot(so->tupdesc, &TTSOpsVirtual);
|
||||||
|
so->mslot = MakeSingleTupleTableSlot(so->tupdesc, &TTSOpsMinimalTuple);
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Reuse same set of shared buffers for scan
|
||||||
|
*
|
||||||
|
* See postgres/src/backend/storage/buffer/README for description
|
||||||
|
*/
|
||||||
|
so->bas = GetAccessStrategy(BAS_BULKREAD);
|
||||||
|
|
||||||
so->listQueue = pairingheap_allocate(CompareLists, scan);
|
so->listQueue = pairingheap_allocate(CompareLists, scan);
|
||||||
|
|
||||||
@@ -277,10 +294,8 @@ ivfflatrescan(IndexScanDesc scan, ScanKey keys, int nkeys, ScanKey orderbys, int
|
|||||||
{
|
{
|
||||||
IvfflatScanOpaque so = (IvfflatScanOpaque) scan->opaque;
|
IvfflatScanOpaque so = (IvfflatScanOpaque) scan->opaque;
|
||||||
|
|
||||||
#if PG_VERSION_NUM >= 130000
|
|
||||||
if (!so->first)
|
if (!so->first)
|
||||||
tuplesort_reset(so->sortstate);
|
tuplesort_reset(so->sortstate);
|
||||||
#endif
|
|
||||||
|
|
||||||
so->first = true;
|
so->first = true;
|
||||||
pairingheap_reset(so->listQueue);
|
pairingheap_reset(so->listQueue);
|
||||||
@@ -327,14 +342,19 @@ ivfflatgettuple(IndexScanDesc scan, ScanDirection dir)
|
|||||||
IvfflatBench("GetScanItems", GetScanItems(scan, value));
|
IvfflatBench("GetScanItems", GetScanItems(scan, value));
|
||||||
so->first = false;
|
so->first = false;
|
||||||
|
|
||||||
|
#if defined(IVFFLAT_MEMORY)
|
||||||
|
elog(INFO, "memory: %zu MB", MemoryContextMemAllocated(CurrentMemoryContext, true) / (1024 * 1024));
|
||||||
|
#endif
|
||||||
|
|
||||||
/* Clean up if we allocated a new value */
|
/* Clean up if we allocated a new value */
|
||||||
if (value != scan->orderByData->sk_argument)
|
if (value != scan->orderByData->sk_argument)
|
||||||
pfree(DatumGetPointer(value));
|
pfree(DatumGetPointer(value));
|
||||||
}
|
}
|
||||||
|
|
||||||
if (tuplesort_gettupleslot(so->sortstate, true, false, so->slot, NULL))
|
if (tuplesort_gettupleslot(so->sortstate, true, false, so->mslot, NULL))
|
||||||
{
|
{
|
||||||
ItemPointer heaptid = (ItemPointer) DatumGetPointer(slot_getattr(so->slot, 2, &so->isnull));
|
bool isnull;
|
||||||
|
ItemPointer heaptid = (ItemPointer) DatumGetPointer(slot_getattr(so->mslot, 2, &isnull));
|
||||||
|
|
||||||
scan->xs_heaptid = *heaptid;
|
scan->xs_heaptid = *heaptid;
|
||||||
scan->xs_recheck = false;
|
scan->xs_recheck = false;
|
||||||
@@ -355,6 +375,10 @@ ivfflatendscan(IndexScanDesc scan)
|
|||||||
|
|
||||||
pairingheap_free(so->listQueue);
|
pairingheap_free(so->listQueue);
|
||||||
tuplesort_end(so->sortstate);
|
tuplesort_end(so->sortstate);
|
||||||
|
FreeAccessStrategy(so->bas);
|
||||||
|
FreeTupleDesc(so->tupdesc);
|
||||||
|
|
||||||
|
/* TODO Free vslot and mslot without freeing TupleDesc */
|
||||||
|
|
||||||
pfree(so);
|
pfree(so);
|
||||||
scan->opaque = NULL;
|
scan->opaque = NULL;
|
||||||
|
|||||||
@@ -7,6 +7,7 @@
|
|||||||
#include "halfutils.h"
|
#include "halfutils.h"
|
||||||
#include "halfvec.h"
|
#include "halfvec.h"
|
||||||
#include "ivfflat.h"
|
#include "ivfflat.h"
|
||||||
|
#include "minivec.h"
|
||||||
#include "storage/bufmgr.h"
|
#include "storage/bufmgr.h"
|
||||||
|
|
||||||
/*
|
/*
|
||||||
@@ -70,6 +71,9 @@ IvfflatOptionalProcInfo(Relation index, uint16 procnum)
|
|||||||
Datum
|
Datum
|
||||||
IvfflatNormValue(const IvfflatTypeInfo * typeInfo, Oid collation, Datum value)
|
IvfflatNormValue(const IvfflatTypeInfo * typeInfo, Oid collation, Datum value)
|
||||||
{
|
{
|
||||||
|
if (!typeInfo->normalize)
|
||||||
|
return value;
|
||||||
|
|
||||||
return DirectFunctionCall1Coll(typeInfo->normalize, collation, value);
|
return DirectFunctionCall1Coll(typeInfo->normalize, collation, value);
|
||||||
}
|
}
|
||||||
|
|
||||||
@@ -231,6 +235,7 @@ IvfflatUpdateList(Relation index, ListInfo listInfo,
|
|||||||
|
|
||||||
PGDLLEXPORT Datum l2_normalize(PG_FUNCTION_ARGS);
|
PGDLLEXPORT Datum l2_normalize(PG_FUNCTION_ARGS);
|
||||||
PGDLLEXPORT Datum halfvec_l2_normalize(PG_FUNCTION_ARGS);
|
PGDLLEXPORT Datum halfvec_l2_normalize(PG_FUNCTION_ARGS);
|
||||||
|
PGDLLEXPORT Datum minivec_l2_normalize(PG_FUNCTION_ARGS);
|
||||||
PGDLLEXPORT Datum sparsevec_l2_normalize(PG_FUNCTION_ARGS);
|
PGDLLEXPORT Datum sparsevec_l2_normalize(PG_FUNCTION_ARGS);
|
||||||
|
|
||||||
static Size
|
static Size
|
||||||
@@ -245,6 +250,12 @@ HalfvecItemSize(int dimensions)
|
|||||||
return HALFVEC_SIZE(dimensions);
|
return HALFVEC_SIZE(dimensions);
|
||||||
}
|
}
|
||||||
|
|
||||||
|
static Size
|
||||||
|
MinivecItemSize(int dimensions)
|
||||||
|
{
|
||||||
|
return MINIVEC_SIZE(dimensions);
|
||||||
|
}
|
||||||
|
|
||||||
static Size
|
static Size
|
||||||
BitItemSize(int dimensions)
|
BitItemSize(int dimensions)
|
||||||
{
|
{
|
||||||
@@ -275,6 +286,18 @@ HalfvecUpdateCenter(Pointer v, int dimensions, float *x)
|
|||||||
vec->x[k] = Float4ToHalfUnchecked(x[k]);
|
vec->x[k] = Float4ToHalfUnchecked(x[k]);
|
||||||
}
|
}
|
||||||
|
|
||||||
|
static void
|
||||||
|
MinivecUpdateCenter(Pointer v, int dimensions, float *x)
|
||||||
|
{
|
||||||
|
MiniVector *vec = (MiniVector *) v;
|
||||||
|
|
||||||
|
SET_VARSIZE(vec, MINIVEC_SIZE(dimensions));
|
||||||
|
vec->dim = dimensions;
|
||||||
|
|
||||||
|
for (int k = 0; k < dimensions; k++)
|
||||||
|
vec->x[k] = Float4ToFp8Unchecked(x[k]);
|
||||||
|
}
|
||||||
|
|
||||||
static void
|
static void
|
||||||
BitUpdateCenter(Pointer v, int dimensions, float *x)
|
BitUpdateCenter(Pointer v, int dimensions, float *x)
|
||||||
{
|
{
|
||||||
@@ -309,6 +332,15 @@ HalfvecSumCenter(Pointer v, float *x)
|
|||||||
x[k] += HalfToFloat4(vec->x[k]);
|
x[k] += HalfToFloat4(vec->x[k]);
|
||||||
}
|
}
|
||||||
|
|
||||||
|
static void
|
||||||
|
MinivecSumCenter(Pointer v, float *x)
|
||||||
|
{
|
||||||
|
MiniVector *vec = (MiniVector *) v;
|
||||||
|
|
||||||
|
for (int k = 0; k < vec->dim; k++)
|
||||||
|
x[k] += Fp8ToFloat4(vec->x[k]);
|
||||||
|
}
|
||||||
|
|
||||||
static void
|
static void
|
||||||
BitSumCenter(Pointer v, float *x)
|
BitSumCenter(Pointer v, float *x)
|
||||||
{
|
{
|
||||||
@@ -342,7 +374,7 @@ IvfflatGetTypeInfo(Relation index)
|
|||||||
return (const IvfflatTypeInfo *) DatumGetPointer(FunctionCall0Coll(procinfo, InvalidOid));
|
return (const IvfflatTypeInfo *) DatumGetPointer(FunctionCall0Coll(procinfo, InvalidOid));
|
||||||
}
|
}
|
||||||
|
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(ivfflat_halfvec_support);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(ivfflat_halfvec_support);
|
||||||
Datum
|
Datum
|
||||||
ivfflat_halfvec_support(PG_FUNCTION_ARGS)
|
ivfflat_halfvec_support(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -357,7 +389,23 @@ ivfflat_halfvec_support(PG_FUNCTION_ARGS)
|
|||||||
PG_RETURN_POINTER(&typeInfo);
|
PG_RETURN_POINTER(&typeInfo);
|
||||||
};
|
};
|
||||||
|
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(ivfflat_bit_support);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(ivfflat_minivec_support);
|
||||||
|
Datum
|
||||||
|
ivfflat_minivec_support(PG_FUNCTION_ARGS)
|
||||||
|
{
|
||||||
|
static const IvfflatTypeInfo typeInfo = {
|
||||||
|
.maxDimensions = IVFFLAT_MAX_DIM * 4,
|
||||||
|
/* Do not normalize to maximize precision */
|
||||||
|
.normalize = NULL,
|
||||||
|
.itemSize = MinivecItemSize,
|
||||||
|
.updateCenter = MinivecUpdateCenter,
|
||||||
|
.sumCenter = MinivecSumCenter
|
||||||
|
};
|
||||||
|
|
||||||
|
PG_RETURN_POINTER(&typeInfo);
|
||||||
|
};
|
||||||
|
|
||||||
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(ivfflat_bit_support);
|
||||||
Datum
|
Datum
|
||||||
ivfflat_bit_support(PG_FUNCTION_ARGS)
|
ivfflat_bit_support(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
|
|||||||
1051
src/minivec.c
Normal file
1051
src/minivec.c
Normal file
File diff suppressed because it is too large
Load Diff
163
src/minivec.h
Normal file
163
src/minivec.h
Normal file
@@ -0,0 +1,163 @@
|
|||||||
|
#ifndef MINIVEC_H
|
||||||
|
#define MINIVEC_H
|
||||||
|
|
||||||
|
#define MINIVEC_MAX_DIM 16000
|
||||||
|
|
||||||
|
#define fp8 uint8
|
||||||
|
|
||||||
|
#define MINIVEC_SIZE(_dim) (offsetof(MiniVector, x) + sizeof(fp8)*(_dim))
|
||||||
|
#define DatumGetMiniVector(x) ((MiniVector *) PG_DETOAST_DATUM(x))
|
||||||
|
#define PG_GETARG_MINIVEC_P(x) DatumGetMiniVector(PG_GETARG_DATUM(x))
|
||||||
|
#define PG_RETURN_MINIVEC_P(x) PG_RETURN_POINTER(x)
|
||||||
|
|
||||||
|
typedef struct MiniVector
|
||||||
|
{
|
||||||
|
int32 vl_len_; /* varlena header (do not touch directly!) */
|
||||||
|
int16 dim; /* number of dimensions */
|
||||||
|
int16 unused; /* reserved for future use, always zero */
|
||||||
|
fp8 x[FLEXIBLE_ARRAY_MEMBER];
|
||||||
|
} MiniVector;
|
||||||
|
|
||||||
|
MiniVector *InitMiniVector(int dim);
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Check if fp8 is NaN
|
||||||
|
*/
|
||||||
|
static inline bool
|
||||||
|
Fp8IsNan(fp8 num)
|
||||||
|
{
|
||||||
|
return (num & 0x7F) == 0x7F;
|
||||||
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Check if fp8 is zero
|
||||||
|
*/
|
||||||
|
static inline bool
|
||||||
|
Fp8IsZero(fp8 num)
|
||||||
|
{
|
||||||
|
return num == 0;
|
||||||
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Convert a fp8 to a float4
|
||||||
|
*/
|
||||||
|
static inline float
|
||||||
|
Fp8ToFloat4(fp8 num)
|
||||||
|
{
|
||||||
|
/* Lookup table for non-sign bits */
|
||||||
|
/* Uses uint32 for correctness */
|
||||||
|
uint32 lookup[128] = {0, 989855744, 998244352, 1002438656, 1006632960, 1008730112, 1010827264, 1012924416, 1015021568, 1016070144, 1017118720, 1018167296, 1019215872, 1020264448, 1021313024, 1022361600, 1023410176, 1024458752, 1025507328, 1026555904, 1027604480, 1028653056, 1029701632, 1030750208, 1031798784, 1032847360, 1033895936, 1034944512, 1035993088, 1037041664, 1038090240, 1039138816, 1040187392, 1041235968, 1042284544, 1043333120, 1044381696, 1045430272, 1046478848, 1047527424, 1048576000, 1049624576, 1050673152, 1051721728, 1052770304, 1053818880, 1054867456, 1055916032, 1056964608, 1058013184, 1059061760, 1060110336, 1061158912, 1062207488, 1063256064, 1064304640, 1065353216, 1066401792, 1067450368, 1068498944, 1069547520, 1070596096, 1071644672, 1072693248, 1073741824, 1074790400, 1075838976, 1076887552, 1077936128, 1078984704, 1080033280, 1081081856, 1082130432, 1083179008, 1084227584, 1085276160, 1086324736, 1087373312, 1088421888, 1089470464, 1090519040, 1091567616, 1092616192, 1093664768, 1094713344, 1095761920, 1096810496, 1097859072, 1098907648, 1099956224, 1101004800, 1102053376, 1103101952, 1104150528, 1105199104, 1106247680, 1107296256, 1108344832, 1109393408, 1110441984, 1111490560, 1112539136, 1113587712, 1114636288, 1115684864, 1116733440, 1117782016, 1118830592, 1119879168, 1120927744, 1121976320, 1123024896, 1124073472, 1125122048, 1126170624, 1127219200, 1128267776, 1129316352, 1130364928, 1131413504, 1132462080, 1133510656, 1134559232, 1135607808, 1136656384, 1137704960, 1138753536, 2146435072};
|
||||||
|
|
||||||
|
union
|
||||||
|
{
|
||||||
|
float f;
|
||||||
|
uint32 i;
|
||||||
|
} swap;
|
||||||
|
|
||||||
|
swap.i = lookup[num & 0x7F];
|
||||||
|
|
||||||
|
return (num & 0x80) == 0x80 ? -swap.f : swap.f;
|
||||||
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Convert a float4 to a fp8
|
||||||
|
*/
|
||||||
|
static inline fp8
|
||||||
|
Float4ToFp8Unchecked(float num)
|
||||||
|
{
|
||||||
|
union
|
||||||
|
{
|
||||||
|
float f;
|
||||||
|
uint32 i;
|
||||||
|
} swap;
|
||||||
|
|
||||||
|
uint32 bin;
|
||||||
|
int exponent;
|
||||||
|
int mantissa;
|
||||||
|
uint8 result;
|
||||||
|
|
||||||
|
swap.f = num;
|
||||||
|
bin = swap.i;
|
||||||
|
exponent = (bin & 0x7F800000) >> 23;
|
||||||
|
mantissa = bin & 0x007FFFFF;
|
||||||
|
|
||||||
|
/* Sign */
|
||||||
|
result = (bin & 0x80000000) >> 24;
|
||||||
|
|
||||||
|
if (isinf(num) || isnan(num))
|
||||||
|
{
|
||||||
|
/* NaN */
|
||||||
|
result |= 0x7F;
|
||||||
|
}
|
||||||
|
else if (exponent > 114)
|
||||||
|
{
|
||||||
|
int m;
|
||||||
|
int gr;
|
||||||
|
int s;
|
||||||
|
|
||||||
|
exponent -= 127;
|
||||||
|
s = mantissa & 0x0007FFFF;
|
||||||
|
|
||||||
|
/* Subnormal */
|
||||||
|
if (exponent < -6)
|
||||||
|
{
|
||||||
|
int diff = -exponent - 6;
|
||||||
|
|
||||||
|
mantissa >>= diff;
|
||||||
|
mantissa += 1 << (23 - diff);
|
||||||
|
s |= mantissa & 0x0007FFFF;
|
||||||
|
}
|
||||||
|
|
||||||
|
m = mantissa >> 20;
|
||||||
|
|
||||||
|
/* Round */
|
||||||
|
gr = (mantissa >> 19) % 4;
|
||||||
|
if (gr == 3 || (gr == 1 && s != 0))
|
||||||
|
m += 1;
|
||||||
|
|
||||||
|
if (m == 8)
|
||||||
|
{
|
||||||
|
m = 0;
|
||||||
|
exponent += 1;
|
||||||
|
}
|
||||||
|
|
||||||
|
if (exponent > 8)
|
||||||
|
{
|
||||||
|
/* Infinite, which is NaN */
|
||||||
|
result |= 0x7F;
|
||||||
|
}
|
||||||
|
else
|
||||||
|
{
|
||||||
|
if (exponent >= -6)
|
||||||
|
result |= (exponent + 7) << 3;
|
||||||
|
|
||||||
|
result |= m;
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
return result;
|
||||||
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Convert a float4 to a fp8
|
||||||
|
*/
|
||||||
|
static inline fp8
|
||||||
|
Float4ToFp8(float num)
|
||||||
|
{
|
||||||
|
fp8 result = Float4ToFp8Unchecked(num);
|
||||||
|
|
||||||
|
if (unlikely(Fp8IsNan(result)) && !isinf(num))
|
||||||
|
{
|
||||||
|
char *buf = palloc(FLOAT_SHORTEST_DECIMAL_LEN);
|
||||||
|
|
||||||
|
float_to_shortest_decimal_buf(num, buf);
|
||||||
|
|
||||||
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_NUMERIC_VALUE_OUT_OF_RANGE),
|
||||||
|
errmsg("\"%s\" is out of range for type minivec", buf)));
|
||||||
|
}
|
||||||
|
|
||||||
|
return result;
|
||||||
|
}
|
||||||
|
|
||||||
|
#endif
|
||||||
179
src/sparsevec.c
179
src/sparsevec.c
@@ -3,6 +3,7 @@
|
|||||||
#include <limits.h>
|
#include <limits.h>
|
||||||
#include <math.h>
|
#include <math.h>
|
||||||
|
|
||||||
|
#include "catalog/pg_type.h"
|
||||||
#include "common/string.h"
|
#include "common/string.h"
|
||||||
#include "fmgr.h"
|
#include "fmgr.h"
|
||||||
#include "halfutils.h"
|
#include "halfutils.h"
|
||||||
@@ -11,6 +12,7 @@
|
|||||||
#include "sparsevec.h"
|
#include "sparsevec.h"
|
||||||
#include "utils/array.h"
|
#include "utils/array.h"
|
||||||
#include "utils/builtins.h"
|
#include "utils/builtins.h"
|
||||||
|
#include "utils/lsyscache.h"
|
||||||
#include "vector.h"
|
#include "vector.h"
|
||||||
|
|
||||||
#if PG_VERSION_NUM >= 120000
|
#if PG_VERSION_NUM >= 120000
|
||||||
@@ -188,7 +190,7 @@ CompareIndices(const void *a, const void *b)
|
|||||||
/*
|
/*
|
||||||
* Convert textual representation to internal representation
|
* Convert textual representation to internal representation
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_in);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_in);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_in(PG_FUNCTION_ARGS)
|
sparsevec_in(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -409,7 +411,7 @@ sparsevec_in(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert internal representation to textual representation
|
* Convert internal representation to textual representation
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_out);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_out);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_out(PG_FUNCTION_ARGS)
|
sparsevec_out(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -462,7 +464,7 @@ sparsevec_out(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert type modifier
|
* Convert type modifier
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_typmod_in);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_typmod_in);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_typmod_in(PG_FUNCTION_ARGS)
|
sparsevec_typmod_in(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -493,7 +495,7 @@ sparsevec_typmod_in(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert external binary representation to internal representation
|
* Convert external binary representation to internal representation
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_recv);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_recv);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_recv(PG_FUNCTION_ARGS)
|
sparsevec_recv(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -545,7 +547,7 @@ sparsevec_recv(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert internal representation to the external binary representation
|
* Convert internal representation to the external binary representation
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_send);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_send);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_send(PG_FUNCTION_ARGS)
|
sparsevec_send(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -572,7 +574,7 @@ sparsevec_send(PG_FUNCTION_ARGS)
|
|||||||
* Convert sparse vector to sparse vector
|
* Convert sparse vector to sparse vector
|
||||||
* This is needed to check the type modifier
|
* This is needed to check the type modifier
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec);
|
||||||
Datum
|
Datum
|
||||||
sparsevec(PG_FUNCTION_ARGS)
|
sparsevec(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -587,7 +589,7 @@ sparsevec(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert dense vector to sparse vector
|
* Convert dense vector to sparse vector
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_to_sparsevec);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_to_sparsevec);
|
||||||
Datum
|
Datum
|
||||||
vector_to_sparsevec(PG_FUNCTION_ARGS)
|
vector_to_sparsevec(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -630,7 +632,7 @@ vector_to_sparsevec(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert half vector to sparse vector
|
* Convert half vector to sparse vector
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_to_sparsevec);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_to_sparsevec);
|
||||||
Datum
|
Datum
|
||||||
halfvec_to_sparsevec(PG_FUNCTION_ARGS)
|
halfvec_to_sparsevec(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -670,6 +672,137 @@ halfvec_to_sparsevec(PG_FUNCTION_ARGS)
|
|||||||
PG_RETURN_POINTER(result);
|
PG_RETURN_POINTER(result);
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Convert array to sparse vector
|
||||||
|
*/
|
||||||
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(array_to_sparsevec);
|
||||||
|
Datum
|
||||||
|
array_to_sparsevec(PG_FUNCTION_ARGS)
|
||||||
|
{
|
||||||
|
ArrayType *array = PG_GETARG_ARRAYTYPE_P(0);
|
||||||
|
int32 typmod = PG_GETARG_INT32(1);
|
||||||
|
SparseVector *result;
|
||||||
|
int16 typlen;
|
||||||
|
bool typbyval;
|
||||||
|
char typalign;
|
||||||
|
Datum *elemsp;
|
||||||
|
int nelemsp;
|
||||||
|
int nnz = 0;
|
||||||
|
float *values;
|
||||||
|
int j = 0;
|
||||||
|
|
||||||
|
if (ARR_NDIM(array) > 1)
|
||||||
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_DATA_EXCEPTION),
|
||||||
|
errmsg("array must be 1-D")));
|
||||||
|
|
||||||
|
if (ARR_HASNULL(array) && array_contains_nulls(array))
|
||||||
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_NULL_VALUE_NOT_ALLOWED),
|
||||||
|
errmsg("array must not contain nulls")));
|
||||||
|
|
||||||
|
get_typlenbyvalalign(ARR_ELEMTYPE(array), &typlen, &typbyval, &typalign);
|
||||||
|
deconstruct_array(array, ARR_ELEMTYPE(array), typlen, typbyval, typalign, &elemsp, NULL, &nelemsp);
|
||||||
|
|
||||||
|
CheckDim(nelemsp);
|
||||||
|
CheckExpectedDim(typmod, nelemsp);
|
||||||
|
|
||||||
|
#ifdef _MSC_VER
|
||||||
|
/* /fp:fast may not propagate +/-Infinity or NaN */
|
||||||
|
#define IS_NOT_ZERO(v) (isnan((float) (v)) || isinf((float) (v)) || ((float) (v)) != 0)
|
||||||
|
#else
|
||||||
|
#define IS_NOT_ZERO(v) (((float) (v)) != 0)
|
||||||
|
#endif
|
||||||
|
|
||||||
|
if (ARR_ELEMTYPE(array) == INT4OID)
|
||||||
|
{
|
||||||
|
for (int i = 0; i < nelemsp; i++)
|
||||||
|
nnz += IS_NOT_ZERO(DatumGetInt32(elemsp[i]));
|
||||||
|
}
|
||||||
|
else if (ARR_ELEMTYPE(array) == FLOAT8OID)
|
||||||
|
{
|
||||||
|
for (int i = 0; i < nelemsp; i++)
|
||||||
|
nnz += IS_NOT_ZERO(DatumGetFloat8(elemsp[i]));
|
||||||
|
}
|
||||||
|
else if (ARR_ELEMTYPE(array) == FLOAT4OID)
|
||||||
|
{
|
||||||
|
for (int i = 0; i < nelemsp; i++)
|
||||||
|
nnz += IS_NOT_ZERO(DatumGetFloat4(elemsp[i]));
|
||||||
|
}
|
||||||
|
else if (ARR_ELEMTYPE(array) == NUMERICOID)
|
||||||
|
{
|
||||||
|
for (int i = 0; i < nelemsp; i++)
|
||||||
|
nnz += IS_NOT_ZERO(DirectFunctionCall1(numeric_float4, elemsp[i]));
|
||||||
|
}
|
||||||
|
else
|
||||||
|
{
|
||||||
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_DATA_EXCEPTION),
|
||||||
|
errmsg("unsupported array type")));
|
||||||
|
}
|
||||||
|
|
||||||
|
result = InitSparseVector(nelemsp, nnz);
|
||||||
|
values = SPARSEVEC_VALUES(result);
|
||||||
|
|
||||||
|
#define PROCESS_ARRAY_ELEM(elem) \
|
||||||
|
do { \
|
||||||
|
float v = (float) (elem); \
|
||||||
|
if (IS_NOT_ZERO(v)) { \
|
||||||
|
/* Safety check */ \
|
||||||
|
if (j >= result->nnz) \
|
||||||
|
elog(ERROR, "safety check failed"); \
|
||||||
|
result->indices[j] = i; \
|
||||||
|
values[j] = v; \
|
||||||
|
j++; \
|
||||||
|
} \
|
||||||
|
} while (0)
|
||||||
|
|
||||||
|
if (ARR_ELEMTYPE(array) == INT4OID)
|
||||||
|
{
|
||||||
|
for (int i = 0; i < nelemsp; i++)
|
||||||
|
PROCESS_ARRAY_ELEM(DatumGetInt32(elemsp[i]));
|
||||||
|
}
|
||||||
|
else if (ARR_ELEMTYPE(array) == FLOAT8OID)
|
||||||
|
{
|
||||||
|
for (int i = 0; i < nelemsp; i++)
|
||||||
|
PROCESS_ARRAY_ELEM(DatumGetFloat8(elemsp[i]));
|
||||||
|
}
|
||||||
|
else if (ARR_ELEMTYPE(array) == FLOAT4OID)
|
||||||
|
{
|
||||||
|
for (int i = 0; i < nelemsp; i++)
|
||||||
|
PROCESS_ARRAY_ELEM(DatumGetFloat4(elemsp[i]));
|
||||||
|
}
|
||||||
|
else if (ARR_ELEMTYPE(array) == NUMERICOID)
|
||||||
|
{
|
||||||
|
for (int i = 0; i < nelemsp; i++)
|
||||||
|
PROCESS_ARRAY_ELEM(DatumGetFloat4(DirectFunctionCall1(numeric_float4, elemsp[i])));
|
||||||
|
}
|
||||||
|
else
|
||||||
|
{
|
||||||
|
ereport(ERROR,
|
||||||
|
(errcode(ERRCODE_DATA_EXCEPTION),
|
||||||
|
errmsg("unsupported array type")));
|
||||||
|
}
|
||||||
|
|
||||||
|
#undef PROCESS_ARRAY_ELEM
|
||||||
|
#undef IS_NOT_ZERO
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Free allocation from deconstruct_array. Do not free individual elements
|
||||||
|
* when pass-by-reference since they point to original array.
|
||||||
|
*/
|
||||||
|
pfree(elemsp);
|
||||||
|
|
||||||
|
if (j != result->nnz)
|
||||||
|
elog(ERROR, "correctness check failed");
|
||||||
|
|
||||||
|
/* Check elements */
|
||||||
|
for (int i = 0; i < result->nnz; i++)
|
||||||
|
CheckElement(values[i]);
|
||||||
|
|
||||||
|
PG_RETURN_POINTER(result);
|
||||||
|
}
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Get the L2 squared distance between sparse vectors
|
* Get the L2 squared distance between sparse vectors
|
||||||
*/
|
*/
|
||||||
@@ -721,7 +854,7 @@ SparsevecL2SquaredDistance(SparseVector * a, SparseVector * b)
|
|||||||
/*
|
/*
|
||||||
* Get the L2 distance between sparse vectors
|
* Get the L2 distance between sparse vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_l2_distance);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_l2_distance);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_l2_distance(PG_FUNCTION_ARGS)
|
sparsevec_l2_distance(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -737,7 +870,7 @@ sparsevec_l2_distance(PG_FUNCTION_ARGS)
|
|||||||
* Get the L2 squared distance between sparse vectors
|
* Get the L2 squared distance between sparse vectors
|
||||||
* This saves a sqrt calculation
|
* This saves a sqrt calculation
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_l2_squared_distance);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_l2_squared_distance);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_l2_squared_distance(PG_FUNCTION_ARGS)
|
sparsevec_l2_squared_distance(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -788,7 +921,7 @@ SparsevecInnerProduct(SparseVector * a, SparseVector * b)
|
|||||||
/*
|
/*
|
||||||
* Get the inner product of two sparse vectors
|
* Get the inner product of two sparse vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_inner_product);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_inner_product);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_inner_product(PG_FUNCTION_ARGS)
|
sparsevec_inner_product(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -803,7 +936,7 @@ sparsevec_inner_product(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get the negative inner product of two sparse vectors
|
* Get the negative inner product of two sparse vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_negative_inner_product);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_negative_inner_product);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_negative_inner_product(PG_FUNCTION_ARGS)
|
sparsevec_negative_inner_product(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -818,7 +951,7 @@ sparsevec_negative_inner_product(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get the cosine distance between two sparse vectors
|
* Get the cosine distance between two sparse vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_cosine_distance);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_cosine_distance);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_cosine_distance(PG_FUNCTION_ARGS)
|
sparsevec_cosine_distance(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -863,7 +996,7 @@ sparsevec_cosine_distance(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get the L1 distance between two sparse vectors
|
* Get the L1 distance between two sparse vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_l1_distance);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_l1_distance);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_l1_distance(PG_FUNCTION_ARGS)
|
sparsevec_l1_distance(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -912,7 +1045,7 @@ sparsevec_l1_distance(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get the L2 norm of a sparse vector
|
* Get the L2 norm of a sparse vector
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_l2_norm);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_l2_norm);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_l2_norm(PG_FUNCTION_ARGS)
|
sparsevec_l2_norm(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -930,7 +1063,7 @@ sparsevec_l2_norm(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Normalize a sparse vector with the L2 norm
|
* Normalize a sparse vector with the L2 norm
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_l2_normalize);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_l2_normalize);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_l2_normalize(PG_FUNCTION_ARGS)
|
sparsevec_l2_normalize(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1040,7 +1173,7 @@ sparsevec_cmp_internal(SparseVector * a, SparseVector * b)
|
|||||||
/*
|
/*
|
||||||
* Less than
|
* Less than
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_lt);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_lt);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_lt(PG_FUNCTION_ARGS)
|
sparsevec_lt(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1053,7 +1186,7 @@ sparsevec_lt(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Less than or equal
|
* Less than or equal
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_le);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_le);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_le(PG_FUNCTION_ARGS)
|
sparsevec_le(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1066,7 +1199,7 @@ sparsevec_le(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Equal
|
* Equal
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_eq);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_eq);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_eq(PG_FUNCTION_ARGS)
|
sparsevec_eq(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1079,7 +1212,7 @@ sparsevec_eq(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Not equal
|
* Not equal
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_ne);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_ne);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_ne(PG_FUNCTION_ARGS)
|
sparsevec_ne(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1092,7 +1225,7 @@ sparsevec_ne(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Greater than or equal
|
* Greater than or equal
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_ge);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_ge);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_ge(PG_FUNCTION_ARGS)
|
sparsevec_ge(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1105,7 +1238,7 @@ sparsevec_ge(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Greater than
|
* Greater than
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_gt);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_gt);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_gt(PG_FUNCTION_ARGS)
|
sparsevec_gt(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1118,7 +1251,7 @@ sparsevec_gt(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Compare sparse vectors
|
* Compare sparse vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_cmp);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_cmp);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_cmp(PG_FUNCTION_ARGS)
|
sparsevec_cmp(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
|
|||||||
118
src/vector.c
118
src/vector.c
@@ -13,6 +13,7 @@
|
|||||||
#include "ivfflat.h"
|
#include "ivfflat.h"
|
||||||
#include "lib/stringinfo.h"
|
#include "lib/stringinfo.h"
|
||||||
#include "libpq/pqformat.h"
|
#include "libpq/pqformat.h"
|
||||||
|
#include "minivec.h"
|
||||||
#include "port.h" /* for strtof() */
|
#include "port.h" /* for strtof() */
|
||||||
#include "sparsevec.h"
|
#include "sparsevec.h"
|
||||||
#include "utils/array.h"
|
#include "utils/array.h"
|
||||||
@@ -26,11 +27,6 @@
|
|||||||
#include "varatt.h"
|
#include "varatt.h"
|
||||||
#endif
|
#endif
|
||||||
|
|
||||||
#if PG_VERSION_NUM < 130000
|
|
||||||
#define TYPALIGN_DOUBLE 'd'
|
|
||||||
#define TYPALIGN_INT 'i'
|
|
||||||
#endif
|
|
||||||
|
|
||||||
#define STATE_DIMS(x) (ARR_DIMS(x)[0] - 1)
|
#define STATE_DIMS(x) (ARR_DIMS(x)[0] - 1)
|
||||||
#define CreateStateDatums(dim) palloc(sizeof(Datum) * (dim + 1))
|
#define CreateStateDatums(dim) palloc(sizeof(Datum) * (dim + 1))
|
||||||
|
|
||||||
@@ -160,28 +156,10 @@ CheckStateArray(ArrayType *statearray, const char *caller)
|
|||||||
return (float8 *) ARR_DATA_PTR(statearray);
|
return (float8 *) ARR_DATA_PTR(statearray);
|
||||||
}
|
}
|
||||||
|
|
||||||
#if PG_VERSION_NUM < 120003
|
|
||||||
static pg_noinline void
|
|
||||||
float_overflow_error(void)
|
|
||||||
{
|
|
||||||
ereport(ERROR,
|
|
||||||
(errcode(ERRCODE_NUMERIC_VALUE_OUT_OF_RANGE),
|
|
||||||
errmsg("value out of range: overflow")));
|
|
||||||
}
|
|
||||||
|
|
||||||
static pg_noinline void
|
|
||||||
float_underflow_error(void)
|
|
||||||
{
|
|
||||||
ereport(ERROR,
|
|
||||||
(errcode(ERRCODE_NUMERIC_VALUE_OUT_OF_RANGE),
|
|
||||||
errmsg("value out of range: underflow")));
|
|
||||||
}
|
|
||||||
#endif
|
|
||||||
|
|
||||||
/*
|
/*
|
||||||
* Convert textual representation to internal representation
|
* Convert textual representation to internal representation
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_in);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_in);
|
||||||
Datum
|
Datum
|
||||||
vector_in(PG_FUNCTION_ARGS)
|
vector_in(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -294,7 +272,7 @@ vector_in(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert internal representation to textual representation
|
* Convert internal representation to textual representation
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_out);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_out);
|
||||||
Datum
|
Datum
|
||||||
vector_out(PG_FUNCTION_ARGS)
|
vector_out(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -348,7 +326,7 @@ PrintVector(char *msg, Vector * vector)
|
|||||||
/*
|
/*
|
||||||
* Convert type modifier
|
* Convert type modifier
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_typmod_in);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_typmod_in);
|
||||||
Datum
|
Datum
|
||||||
vector_typmod_in(PG_FUNCTION_ARGS)
|
vector_typmod_in(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -379,7 +357,7 @@ vector_typmod_in(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert external binary representation to internal representation
|
* Convert external binary representation to internal representation
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_recv);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_recv);
|
||||||
Datum
|
Datum
|
||||||
vector_recv(PG_FUNCTION_ARGS)
|
vector_recv(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -413,7 +391,7 @@ vector_recv(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert internal representation to the external binary representation
|
* Convert internal representation to the external binary representation
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_send);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_send);
|
||||||
Datum
|
Datum
|
||||||
vector_send(PG_FUNCTION_ARGS)
|
vector_send(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -433,7 +411,7 @@ vector_send(PG_FUNCTION_ARGS)
|
|||||||
* Convert vector to vector
|
* Convert vector to vector
|
||||||
* This is needed to check the type modifier
|
* This is needed to check the type modifier
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector);
|
||||||
Datum
|
Datum
|
||||||
vector(PG_FUNCTION_ARGS)
|
vector(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -448,7 +426,7 @@ vector(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert array to vector
|
* Convert array to vector
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(array_to_vector);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(array_to_vector);
|
||||||
Datum
|
Datum
|
||||||
array_to_vector(PG_FUNCTION_ARGS)
|
array_to_vector(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -522,7 +500,7 @@ array_to_vector(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert vector to float4[]
|
* Convert vector to float4[]
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_to_float4);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_to_float4);
|
||||||
Datum
|
Datum
|
||||||
vector_to_float4(PG_FUNCTION_ARGS)
|
vector_to_float4(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -546,7 +524,7 @@ vector_to_float4(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert half vector to vector
|
* Convert half vector to vector
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(halfvec_to_vector);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(halfvec_to_vector);
|
||||||
Datum
|
Datum
|
||||||
halfvec_to_vector(PG_FUNCTION_ARGS)
|
halfvec_to_vector(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -565,6 +543,28 @@ halfvec_to_vector(PG_FUNCTION_ARGS)
|
|||||||
PG_RETURN_POINTER(result);
|
PG_RETURN_POINTER(result);
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/*
|
||||||
|
* Convert fp8 vector to vector
|
||||||
|
*/
|
||||||
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(minivec_to_vector);
|
||||||
|
Datum
|
||||||
|
minivec_to_vector(PG_FUNCTION_ARGS)
|
||||||
|
{
|
||||||
|
MiniVector *vec = PG_GETARG_MINIVEC_P(0);
|
||||||
|
int32 typmod = PG_GETARG_INT32(1);
|
||||||
|
Vector *result;
|
||||||
|
|
||||||
|
CheckDim(vec->dim);
|
||||||
|
CheckExpectedDim(typmod, vec->dim);
|
||||||
|
|
||||||
|
result = InitVector(vec->dim);
|
||||||
|
|
||||||
|
for (int i = 0; i < vec->dim; i++)
|
||||||
|
result->x[i] = Fp8ToFloat4(vec->x[i]);
|
||||||
|
|
||||||
|
PG_RETURN_POINTER(result);
|
||||||
|
}
|
||||||
|
|
||||||
VECTOR_TARGET_CLONES static float
|
VECTOR_TARGET_CLONES static float
|
||||||
VectorL2SquaredDistance(int dim, float *ax, float *bx)
|
VectorL2SquaredDistance(int dim, float *ax, float *bx)
|
||||||
{
|
{
|
||||||
@@ -584,7 +584,7 @@ VectorL2SquaredDistance(int dim, float *ax, float *bx)
|
|||||||
/*
|
/*
|
||||||
* Get the L2 distance between vectors
|
* Get the L2 distance between vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(l2_distance);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(l2_distance);
|
||||||
Datum
|
Datum
|
||||||
l2_distance(PG_FUNCTION_ARGS)
|
l2_distance(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -600,7 +600,7 @@ l2_distance(PG_FUNCTION_ARGS)
|
|||||||
* Get the L2 squared distance between vectors
|
* Get the L2 squared distance between vectors
|
||||||
* This saves a sqrt calculation
|
* This saves a sqrt calculation
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_l2_squared_distance);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_l2_squared_distance);
|
||||||
Datum
|
Datum
|
||||||
vector_l2_squared_distance(PG_FUNCTION_ARGS)
|
vector_l2_squared_distance(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -627,7 +627,7 @@ VectorInnerProduct(int dim, float *ax, float *bx)
|
|||||||
/*
|
/*
|
||||||
* Get the inner product of two vectors
|
* Get the inner product of two vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(inner_product);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(inner_product);
|
||||||
Datum
|
Datum
|
||||||
inner_product(PG_FUNCTION_ARGS)
|
inner_product(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -642,7 +642,7 @@ inner_product(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get the negative inner product of two vectors
|
* Get the negative inner product of two vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_negative_inner_product);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_negative_inner_product);
|
||||||
Datum
|
Datum
|
||||||
vector_negative_inner_product(PG_FUNCTION_ARGS)
|
vector_negative_inner_product(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -676,7 +676,7 @@ VectorCosineSimilarity(int dim, float *ax, float *bx)
|
|||||||
/*
|
/*
|
||||||
* Get the cosine distance between two vectors
|
* Get the cosine distance between two vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(cosine_distance);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(cosine_distance);
|
||||||
Datum
|
Datum
|
||||||
cosine_distance(PG_FUNCTION_ARGS)
|
cosine_distance(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -708,7 +708,7 @@ cosine_distance(PG_FUNCTION_ARGS)
|
|||||||
* Currently uses angular distance since needs to satisfy triangle inequality
|
* Currently uses angular distance since needs to satisfy triangle inequality
|
||||||
* Assumes inputs are unit vectors (skips norm)
|
* Assumes inputs are unit vectors (skips norm)
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_spherical_distance);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_spherical_distance);
|
||||||
Datum
|
Datum
|
||||||
vector_spherical_distance(PG_FUNCTION_ARGS)
|
vector_spherical_distance(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -745,7 +745,7 @@ VectorL1Distance(int dim, float *ax, float *bx)
|
|||||||
/*
|
/*
|
||||||
* Get the L1 distance between two vectors
|
* Get the L1 distance between two vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(l1_distance);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(l1_distance);
|
||||||
Datum
|
Datum
|
||||||
l1_distance(PG_FUNCTION_ARGS)
|
l1_distance(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -760,7 +760,7 @@ l1_distance(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get the dimensions of a vector
|
* Get the dimensions of a vector
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_dims);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_dims);
|
||||||
Datum
|
Datum
|
||||||
vector_dims(PG_FUNCTION_ARGS)
|
vector_dims(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -772,7 +772,7 @@ vector_dims(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get the L2 norm of a vector
|
* Get the L2 norm of a vector
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_norm);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_norm);
|
||||||
Datum
|
Datum
|
||||||
vector_norm(PG_FUNCTION_ARGS)
|
vector_norm(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -790,7 +790,7 @@ vector_norm(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Normalize a vector with the L2 norm
|
* Normalize a vector with the L2 norm
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(l2_normalize);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(l2_normalize);
|
||||||
Datum
|
Datum
|
||||||
l2_normalize(PG_FUNCTION_ARGS)
|
l2_normalize(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -829,7 +829,7 @@ l2_normalize(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Add vectors
|
* Add vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_add);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_add);
|
||||||
Datum
|
Datum
|
||||||
vector_add(PG_FUNCTION_ARGS)
|
vector_add(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -862,7 +862,7 @@ vector_add(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Subtract vectors
|
* Subtract vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_sub);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_sub);
|
||||||
Datum
|
Datum
|
||||||
vector_sub(PG_FUNCTION_ARGS)
|
vector_sub(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -895,7 +895,7 @@ vector_sub(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Multiply vectors
|
* Multiply vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_mul);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_mul);
|
||||||
Datum
|
Datum
|
||||||
vector_mul(PG_FUNCTION_ARGS)
|
vector_mul(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -931,7 +931,7 @@ vector_mul(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Concatenate vectors
|
* Concatenate vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_concat);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_concat);
|
||||||
Datum
|
Datum
|
||||||
vector_concat(PG_FUNCTION_ARGS)
|
vector_concat(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -955,7 +955,7 @@ vector_concat(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Quantize a vector
|
* Quantize a vector
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(binary_quantize);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(binary_quantize);
|
||||||
Datum
|
Datum
|
||||||
binary_quantize(PG_FUNCTION_ARGS)
|
binary_quantize(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -973,7 +973,7 @@ binary_quantize(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Get a subvector
|
* Get a subvector
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(subvector);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(subvector);
|
||||||
Datum
|
Datum
|
||||||
subvector(PG_FUNCTION_ARGS)
|
subvector(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1047,7 +1047,7 @@ vector_cmp_internal(Vector * a, Vector * b)
|
|||||||
/*
|
/*
|
||||||
* Less than
|
* Less than
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_lt);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_lt);
|
||||||
Datum
|
Datum
|
||||||
vector_lt(PG_FUNCTION_ARGS)
|
vector_lt(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1060,7 +1060,7 @@ vector_lt(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Less than or equal
|
* Less than or equal
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_le);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_le);
|
||||||
Datum
|
Datum
|
||||||
vector_le(PG_FUNCTION_ARGS)
|
vector_le(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1073,7 +1073,7 @@ vector_le(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Equal
|
* Equal
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_eq);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_eq);
|
||||||
Datum
|
Datum
|
||||||
vector_eq(PG_FUNCTION_ARGS)
|
vector_eq(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1086,7 +1086,7 @@ vector_eq(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Not equal
|
* Not equal
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_ne);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_ne);
|
||||||
Datum
|
Datum
|
||||||
vector_ne(PG_FUNCTION_ARGS)
|
vector_ne(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1099,7 +1099,7 @@ vector_ne(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Greater than or equal
|
* Greater than or equal
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_ge);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_ge);
|
||||||
Datum
|
Datum
|
||||||
vector_ge(PG_FUNCTION_ARGS)
|
vector_ge(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1112,7 +1112,7 @@ vector_ge(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Greater than
|
* Greater than
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_gt);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_gt);
|
||||||
Datum
|
Datum
|
||||||
vector_gt(PG_FUNCTION_ARGS)
|
vector_gt(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1125,7 +1125,7 @@ vector_gt(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Compare vectors
|
* Compare vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_cmp);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_cmp);
|
||||||
Datum
|
Datum
|
||||||
vector_cmp(PG_FUNCTION_ARGS)
|
vector_cmp(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1138,7 +1138,7 @@ vector_cmp(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Accumulate vectors
|
* Accumulate vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_accum);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_accum);
|
||||||
Datum
|
Datum
|
||||||
vector_accum(PG_FUNCTION_ARGS)
|
vector_accum(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1199,7 +1199,7 @@ vector_accum(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Combine vectors or half vectors (also used for halfvec_combine)
|
* Combine vectors or half vectors (also used for halfvec_combine)
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_combine);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_combine);
|
||||||
Datum
|
Datum
|
||||||
vector_combine(PG_FUNCTION_ARGS)
|
vector_combine(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1270,7 +1270,7 @@ vector_combine(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Average vectors
|
* Average vectors
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_avg);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(vector_avg);
|
||||||
Datum
|
Datum
|
||||||
vector_avg(PG_FUNCTION_ARGS)
|
vector_avg(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
@@ -1304,7 +1304,7 @@ vector_avg(PG_FUNCTION_ARGS)
|
|||||||
/*
|
/*
|
||||||
* Convert sparse vector to dense vector
|
* Convert sparse vector to dense vector
|
||||||
*/
|
*/
|
||||||
PGDLLEXPORT PG_FUNCTION_INFO_V1(sparsevec_to_vector);
|
FUNCTION_PREFIX PG_FUNCTION_INFO_V1(sparsevec_to_vector);
|
||||||
Datum
|
Datum
|
||||||
sparsevec_to_vector(PG_FUNCTION_ARGS)
|
sparsevec_to_vector(PG_FUNCTION_ARGS)
|
||||||
{
|
{
|
||||||
|
|||||||
@@ -20,4 +20,11 @@ Vector *InitVector(int dim);
|
|||||||
void PrintVector(char *msg, Vector * vector);
|
void PrintVector(char *msg, Vector * vector);
|
||||||
int vector_cmp_internal(Vector * a, Vector * b);
|
int vector_cmp_internal(Vector * a, Vector * b);
|
||||||
|
|
||||||
|
/* TODO Move to better place */
|
||||||
|
#if PG_VERSION_NUM >= 160000
|
||||||
|
#define FUNCTION_PREFIX
|
||||||
|
#else
|
||||||
|
#define FUNCTION_PREFIX PGDLLEXPORT
|
||||||
|
#endif
|
||||||
|
|
||||||
#endif
|
#endif
|
||||||
|
|||||||
@@ -38,6 +38,26 @@ SELECT * FROM t ORDER BY val;
|
|||||||
|
|
||||||
(4 rows)
|
(4 rows)
|
||||||
|
|
||||||
|
DROP TABLE t;
|
||||||
|
-- minivec
|
||||||
|
CREATE TABLE t (val minivec(3));
|
||||||
|
INSERT INTO t (val) VALUES ('[0,0,0]'), ('[1,2,3]'), ('[1,1,1]'), (NULL);
|
||||||
|
CREATE INDEX ON t (val);
|
||||||
|
SELECT * FROM t WHERE val = '[1,2,3]';
|
||||||
|
val
|
||||||
|
---------
|
||||||
|
[1,2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT * FROM t ORDER BY val;
|
||||||
|
val
|
||||||
|
---------
|
||||||
|
[0,0,0]
|
||||||
|
[1,1,1]
|
||||||
|
[1,2,3]
|
||||||
|
|
||||||
|
(4 rows)
|
||||||
|
|
||||||
DROP TABLE t;
|
DROP TABLE t;
|
||||||
-- sparsevec
|
-- sparsevec
|
||||||
CREATE TABLE t (val sparsevec(3));
|
CREATE TABLE t (val sparsevec(3));
|
||||||
|
|||||||
@@ -140,6 +140,64 @@ SELECT '{1e-8,-1e-8}'::real[]::halfvec;
|
|||||||
[0,-0]
|
[0,-0]
|
||||||
(1 row)
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::vector::minivec;
|
||||||
|
minivec
|
||||||
|
---------
|
||||||
|
[1,2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::vector::minivec(3);
|
||||||
|
minivec
|
||||||
|
---------
|
||||||
|
[1,2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::vector::minivec(2);
|
||||||
|
ERROR: expected 2 dimensions, not 3
|
||||||
|
SELECT '[465]'::vector::minivec;
|
||||||
|
ERROR: "465" is out of range for type minivec
|
||||||
|
SELECT '[1e-8]'::vector::minivec;
|
||||||
|
minivec
|
||||||
|
---------
|
||||||
|
[0]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec::vector;
|
||||||
|
vector
|
||||||
|
---------
|
||||||
|
[1,2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec::vector(3);
|
||||||
|
vector
|
||||||
|
---------
|
||||||
|
[1,2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec::vector(2);
|
||||||
|
ERROR: expected 2 dimensions, not 3
|
||||||
|
SELECT '{1,2,3}'::real[]::minivec;
|
||||||
|
minivec
|
||||||
|
---------
|
||||||
|
[1,2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '{1,2,3}'::real[]::minivec(3);
|
||||||
|
minivec
|
||||||
|
---------
|
||||||
|
[1,2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '{1,2,3}'::real[]::minivec(2);
|
||||||
|
ERROR: expected 2 dimensions, not 3
|
||||||
|
SELECT '{465,-465}'::real[]::minivec;
|
||||||
|
ERROR: "465" is out of range for type minivec
|
||||||
|
SELECT '{1e-8,-1e-8}'::real[]::minivec;
|
||||||
|
minivec
|
||||||
|
---------
|
||||||
|
[0,-0]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
SELECT '[0,1.5,0,3.5,0]'::vector::sparsevec;
|
SELECT '[0,1.5,0,3.5,0]'::vector::sparsevec;
|
||||||
sparsevec
|
sparsevec
|
||||||
-----------------
|
-----------------
|
||||||
@@ -208,6 +266,62 @@ SELECT '{1:1e-8}/1'::sparsevec::halfvec;
|
|||||||
[0]
|
[0]
|
||||||
(1 row)
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ARRAY[1,0,2,0,3,0]::sparsevec;
|
||||||
|
array
|
||||||
|
-----------------
|
||||||
|
{1:1,3:2,5:3}/6
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ARRAY[1.0,0.0,2.0,0.0,3.0,0.0]::sparsevec;
|
||||||
|
array
|
||||||
|
-----------------
|
||||||
|
{1:1,3:2,5:3}/6
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ARRAY[1,0,2,0,3,0]::float4[]::sparsevec;
|
||||||
|
array
|
||||||
|
-----------------
|
||||||
|
{1:1,3:2,5:3}/6
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ARRAY[1,0,2,0,3,0]::float8[]::sparsevec;
|
||||||
|
array
|
||||||
|
-----------------
|
||||||
|
{1:1,3:2,5:3}/6
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ARRAY[1,0,2,0,3,0]::numeric[]::sparsevec;
|
||||||
|
array
|
||||||
|
-----------------
|
||||||
|
{1:1,3:2,5:3}/6
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '{1,0,2,0,3,0}'::real[]::sparsevec;
|
||||||
|
sparsevec
|
||||||
|
-----------------
|
||||||
|
{1:1,3:2,5:3}/6
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '{1,0,2,0,3,0}'::real[]::sparsevec(6);
|
||||||
|
sparsevec
|
||||||
|
-----------------
|
||||||
|
{1:1,3:2,5:3}/6
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '{1,0,2,0,3,0}'::real[]::sparsevec(5);
|
||||||
|
ERROR: expected 5 dimensions, not 6
|
||||||
|
SELECT '{NULL}'::real[]::sparsevec;
|
||||||
|
ERROR: array must not contain nulls
|
||||||
|
SELECT '{NaN}'::real[]::sparsevec;
|
||||||
|
ERROR: NaN not allowed in sparsevec
|
||||||
|
SELECT '{Infinity}'::real[]::sparsevec;
|
||||||
|
ERROR: infinite value not allowed in sparsevec
|
||||||
|
SELECT '{-Infinity}'::real[]::sparsevec;
|
||||||
|
ERROR: infinite value not allowed in sparsevec
|
||||||
|
SELECT '{}'::real[]::sparsevec;
|
||||||
|
ERROR: sparsevec must have at least 1 dimension
|
||||||
|
SELECT '{{1}}'::real[]::sparsevec;
|
||||||
|
ERROR: array must be 1-D
|
||||||
SELECT array_agg(n)::vector FROM generate_series(1, 16001) n;
|
SELECT array_agg(n)::vector FROM generate_series(1, 16001) n;
|
||||||
ERROR: vector cannot have more than 16000 dimensions
|
ERROR: vector cannot have more than 16000 dimensions
|
||||||
SELECT array_to_vector(array_agg(n), 16001, false) FROM generate_series(1, 16001) n;
|
SELECT array_to_vector(array_agg(n), 16001, false) FROM generate_series(1, 16001) n;
|
||||||
|
|||||||
@@ -30,6 +30,23 @@ SELECT * FROM t2 ORDER BY val;
|
|||||||
|
|
||||||
(4 rows)
|
(4 rows)
|
||||||
|
|
||||||
|
DROP TABLE t;
|
||||||
|
DROP TABLE t2;
|
||||||
|
-- minivec
|
||||||
|
CREATE TABLE t (val minivec(3));
|
||||||
|
INSERT INTO t (val) VALUES ('[0,0,0]'), ('[1,2,3]'), ('[1,1,1]'), (NULL);
|
||||||
|
CREATE TABLE t2 (val minivec(3));
|
||||||
|
\copy t TO 'results/minivec.bin' WITH (FORMAT binary)
|
||||||
|
\copy t2 FROM 'results/minivec.bin' WITH (FORMAT binary)
|
||||||
|
SELECT * FROM t2 ORDER BY val;
|
||||||
|
val
|
||||||
|
---------
|
||||||
|
[0,0,0]
|
||||||
|
[1,1,1]
|
||||||
|
[1,2,3]
|
||||||
|
|
||||||
|
(4 rows)
|
||||||
|
|
||||||
DROP TABLE t;
|
DROP TABLE t;
|
||||||
DROP TABLE t2;
|
DROP TABLE t2;
|
||||||
-- sparsevec
|
-- sparsevec
|
||||||
|
|||||||
84
test/expected/ivfflat_minivec.out
Normal file
84
test/expected/ivfflat_minivec.out
Normal file
@@ -0,0 +1,84 @@
|
|||||||
|
SET enable_seqscan = off;
|
||||||
|
-- L2
|
||||||
|
CREATE TABLE t (val minivec(3));
|
||||||
|
INSERT INTO t (val) VALUES ('[0,0,0]'), ('[1,2,3]'), ('[1,1,1]'), (NULL);
|
||||||
|
CREATE INDEX ON t USING ivfflat (val minivec_l2_ops) WITH (lists = 1);
|
||||||
|
INSERT INTO t (val) VALUES ('[1,2,4]');
|
||||||
|
SELECT * FROM t ORDER BY val <-> '[3,3,3]';
|
||||||
|
val
|
||||||
|
---------
|
||||||
|
[1,2,3]
|
||||||
|
[1,2,4]
|
||||||
|
[1,1,1]
|
||||||
|
[0,0,0]
|
||||||
|
(4 rows)
|
||||||
|
|
||||||
|
SELECT COUNT(*) FROM (SELECT * FROM t ORDER BY val <-> (SELECT NULL::minivec)) t2;
|
||||||
|
count
|
||||||
|
-------
|
||||||
|
4
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT COUNT(*) FROM t;
|
||||||
|
count
|
||||||
|
-------
|
||||||
|
5
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
TRUNCATE t;
|
||||||
|
NOTICE: ivfflat index created with little data
|
||||||
|
DETAIL: This will cause low recall.
|
||||||
|
HINT: Drop the index until the table has more data.
|
||||||
|
SELECT * FROM t ORDER BY val <-> '[3,3,3]';
|
||||||
|
val
|
||||||
|
-----
|
||||||
|
(0 rows)
|
||||||
|
|
||||||
|
DROP TABLE t;
|
||||||
|
-- inner product
|
||||||
|
CREATE TABLE t (val minivec(3));
|
||||||
|
INSERT INTO t (val) VALUES ('[0,0,0]'), ('[1,2,3]'), ('[1,1,1]'), (NULL);
|
||||||
|
CREATE INDEX ON t USING ivfflat (val minivec_ip_ops) WITH (lists = 1);
|
||||||
|
INSERT INTO t (val) VALUES ('[1,2,4]');
|
||||||
|
SELECT * FROM t ORDER BY val <#> '[3,3,3]';
|
||||||
|
val
|
||||||
|
---------
|
||||||
|
[1,2,4]
|
||||||
|
[1,2,3]
|
||||||
|
[1,1,1]
|
||||||
|
[0,0,0]
|
||||||
|
(4 rows)
|
||||||
|
|
||||||
|
SELECT COUNT(*) FROM (SELECT * FROM t ORDER BY val <#> (SELECT NULL::minivec)) t2;
|
||||||
|
count
|
||||||
|
-------
|
||||||
|
4
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
DROP TABLE t;
|
||||||
|
-- cosine
|
||||||
|
CREATE TABLE t (val minivec(3));
|
||||||
|
INSERT INTO t (val) VALUES ('[0,0,0]'), ('[1,2,3]'), ('[1,1,1]'), (NULL);
|
||||||
|
CREATE INDEX ON t USING ivfflat (val minivec_cosine_ops) WITH (lists = 1);
|
||||||
|
INSERT INTO t (val) VALUES ('[1,2,4]');
|
||||||
|
SELECT * FROM t ORDER BY val <=> '[3,3,3]';
|
||||||
|
val
|
||||||
|
---------
|
||||||
|
[1,1,1]
|
||||||
|
[1,2,3]
|
||||||
|
[1,2,4]
|
||||||
|
(3 rows)
|
||||||
|
|
||||||
|
SELECT COUNT(*) FROM (SELECT * FROM t ORDER BY val <=> '[0,0,0]') t2;
|
||||||
|
count
|
||||||
|
-------
|
||||||
|
3
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT COUNT(*) FROM (SELECT * FROM t ORDER BY val <=> (SELECT NULL::minivec)) t2;
|
||||||
|
count
|
||||||
|
-------
|
||||||
|
3
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
DROP TABLE t;
|
||||||
588
test/expected/minivec.out
Normal file
588
test/expected/minivec.out
Normal file
@@ -0,0 +1,588 @@
|
|||||||
|
SELECT '[1,2,3]'::minivec;
|
||||||
|
minivec
|
||||||
|
---------
|
||||||
|
[1,2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[-1,-2,-3]'::minivec;
|
||||||
|
minivec
|
||||||
|
------------
|
||||||
|
[-1,-2,-3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1.,2.,3.]'::minivec;
|
||||||
|
minivec
|
||||||
|
---------
|
||||||
|
[1,2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT ' [ 1, 2 , 3 ] '::minivec;
|
||||||
|
minivec
|
||||||
|
---------
|
||||||
|
[1,2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1.23456]'::minivec;
|
||||||
|
minivec
|
||||||
|
---------
|
||||||
|
[1.25]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[hello,1]'::minivec;
|
||||||
|
ERROR: invalid input syntax for type minivec: "[hello,1]"
|
||||||
|
LINE 1: SELECT '[hello,1]'::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[NaN,1]'::minivec;
|
||||||
|
ERROR: NaN not allowed in minivec
|
||||||
|
LINE 1: SELECT '[NaN,1]'::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[Infinity,1]'::minivec;
|
||||||
|
ERROR: "Infinity" is out of range for type minivec
|
||||||
|
LINE 1: SELECT '[Infinity,1]'::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[-Infinity,1]'::minivec;
|
||||||
|
ERROR: "-Infinity" is out of range for type minivec
|
||||||
|
LINE 1: SELECT '[-Infinity,1]'::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[65519,-65519]'::minivec;
|
||||||
|
ERROR: "65519" is out of range for type minivec
|
||||||
|
LINE 1: SELECT '[65519,-65519]'::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[65520,-65520]'::minivec;
|
||||||
|
ERROR: "65520" is out of range for type minivec
|
||||||
|
LINE 1: SELECT '[65520,-65520]'::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[1e-8,-1e-8]'::minivec;
|
||||||
|
minivec
|
||||||
|
---------
|
||||||
|
[0,-0]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[4e38,1]'::minivec;
|
||||||
|
ERROR: "4e38" is out of range for type minivec
|
||||||
|
LINE 1: SELECT '[4e38,1]'::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[1e-46,1]'::minivec;
|
||||||
|
minivec
|
||||||
|
---------
|
||||||
|
[0,1]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3'::minivec;
|
||||||
|
ERROR: invalid input syntax for type minivec: "[1,2,3"
|
||||||
|
LINE 1: SELECT '[1,2,3'::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[1,2,3]9'::minivec;
|
||||||
|
ERROR: invalid input syntax for type minivec: "[1,2,3]9"
|
||||||
|
LINE 1: SELECT '[1,2,3]9'::minivec;
|
||||||
|
^
|
||||||
|
DETAIL: Junk after closing right brace.
|
||||||
|
SELECT '1,2,3'::minivec;
|
||||||
|
ERROR: invalid input syntax for type minivec: "1,2,3"
|
||||||
|
LINE 1: SELECT '1,2,3'::minivec;
|
||||||
|
^
|
||||||
|
DETAIL: Vector contents must start with "[".
|
||||||
|
SELECT ''::minivec;
|
||||||
|
ERROR: invalid input syntax for type minivec: ""
|
||||||
|
LINE 1: SELECT ''::minivec;
|
||||||
|
^
|
||||||
|
DETAIL: Vector contents must start with "[".
|
||||||
|
SELECT '['::minivec;
|
||||||
|
ERROR: invalid input syntax for type minivec: "["
|
||||||
|
LINE 1: SELECT '['::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[ '::minivec;
|
||||||
|
ERROR: invalid input syntax for type minivec: "[ "
|
||||||
|
LINE 1: SELECT '[ '::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[,'::minivec;
|
||||||
|
ERROR: invalid input syntax for type minivec: "[,"
|
||||||
|
LINE 1: SELECT '[,'::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[]'::minivec;
|
||||||
|
ERROR: minivec must have at least 1 dimension
|
||||||
|
LINE 1: SELECT '[]'::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[ ]'::minivec;
|
||||||
|
ERROR: minivec must have at least 1 dimension
|
||||||
|
LINE 1: SELECT '[ ]'::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[,]'::minivec;
|
||||||
|
ERROR: invalid input syntax for type minivec: "[,]"
|
||||||
|
LINE 1: SELECT '[,]'::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[1,]'::minivec;
|
||||||
|
ERROR: invalid input syntax for type minivec: "[1,]"
|
||||||
|
LINE 1: SELECT '[1,]'::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[1a]'::minivec;
|
||||||
|
ERROR: invalid input syntax for type minivec: "[1a]"
|
||||||
|
LINE 1: SELECT '[1a]'::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[1,,3]'::minivec;
|
||||||
|
ERROR: invalid input syntax for type minivec: "[1,,3]"
|
||||||
|
LINE 1: SELECT '[1,,3]'::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[1, ,3]'::minivec;
|
||||||
|
ERROR: invalid input syntax for type minivec: "[1, ,3]"
|
||||||
|
LINE 1: SELECT '[1, ,3]'::minivec;
|
||||||
|
^
|
||||||
|
SELECT '[1,2,3]'::minivec(3);
|
||||||
|
minivec
|
||||||
|
---------
|
||||||
|
[1,2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec(2);
|
||||||
|
ERROR: expected 2 dimensions, not 3
|
||||||
|
SELECT '[1,2,3]'::minivec(3, 2);
|
||||||
|
ERROR: invalid type modifier
|
||||||
|
LINE 1: SELECT '[1,2,3]'::minivec(3, 2);
|
||||||
|
^
|
||||||
|
SELECT '[1,2,3]'::minivec('a');
|
||||||
|
ERROR: invalid input syntax for type integer: "a"
|
||||||
|
LINE 1: SELECT '[1,2,3]'::minivec('a');
|
||||||
|
^
|
||||||
|
SELECT '[1,2,3]'::minivec(0);
|
||||||
|
ERROR: dimensions for type minivec must be at least 1
|
||||||
|
LINE 1: SELECT '[1,2,3]'::minivec(0);
|
||||||
|
^
|
||||||
|
SELECT '[1,2,3]'::minivec(16001);
|
||||||
|
ERROR: dimensions for type minivec cannot exceed 16000
|
||||||
|
LINE 1: SELECT '[1,2,3]'::minivec(16001);
|
||||||
|
^
|
||||||
|
SELECT unnest('{"[1,2,3]", "[4,5,6]"}'::minivec[]);
|
||||||
|
unnest
|
||||||
|
---------
|
||||||
|
[1,2,3]
|
||||||
|
[4,5,6]
|
||||||
|
(2 rows)
|
||||||
|
|
||||||
|
SELECT '{"[1,2,3]"}'::minivec(2)[];
|
||||||
|
ERROR: expected 2 dimensions, not 3
|
||||||
|
SELECT '[1,2,3]'::minivec + '[4,5,6]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
[5,7,9]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[448]'::minivec + '[448]';
|
||||||
|
ERROR: value out of range: overflow
|
||||||
|
SELECT '[1,2]'::minivec + '[3]';
|
||||||
|
ERROR: different minivec dimensions 2 and 1
|
||||||
|
SELECT '[1,2,3]'::minivec - '[4,5,6]';
|
||||||
|
?column?
|
||||||
|
------------
|
||||||
|
[-3,-3,-3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[-448]'::minivec - '[448]';
|
||||||
|
ERROR: value out of range: overflow
|
||||||
|
SELECT '[1,2]'::minivec - '[3]';
|
||||||
|
ERROR: different minivec dimensions 2 and 1
|
||||||
|
SELECT '[1,2,3]'::minivec * '[4,5,6]';
|
||||||
|
?column?
|
||||||
|
-----------
|
||||||
|
[4,10,18]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[448]'::minivec * '[448]';
|
||||||
|
ERROR: value out of range: overflow
|
||||||
|
SELECT '[1e-7]'::minivec * '[1e-7]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
[0]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2]'::minivec * '[3]';
|
||||||
|
ERROR: different minivec dimensions 2 and 1
|
||||||
|
SELECT '[1,2,3]'::minivec || '[4,5]';
|
||||||
|
?column?
|
||||||
|
-------------
|
||||||
|
[1,2,3,4,5]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT array_fill(0, ARRAY[16000])::minivec || '[1]';
|
||||||
|
ERROR: minivec cannot have more than 16000 dimensions
|
||||||
|
SELECT '[1,2,3]'::minivec < '[1,2,3]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
f
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec < '[1,2]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
f
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec <= '[1,2,3]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
t
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec <= '[1,2]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
f
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec = '[1,2,3]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
t
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec = '[1,2]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
f
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec != '[1,2,3]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
f
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec != '[1,2]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
t
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec >= '[1,2,3]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
t
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec >= '[1,2]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
t
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec > '[1,2,3]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
f
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec > '[1,2]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
t
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT minivec_cmp('[1,2,3]', '[1,2,3]');
|
||||||
|
minivec_cmp
|
||||||
|
-------------
|
||||||
|
0
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT minivec_cmp('[1,2,3]', '[0,0,0]');
|
||||||
|
minivec_cmp
|
||||||
|
-------------
|
||||||
|
1
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT minivec_cmp('[0,0,0]', '[1,2,3]');
|
||||||
|
minivec_cmp
|
||||||
|
-------------
|
||||||
|
-1
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT minivec_cmp('[1,2]', '[1,2,3]');
|
||||||
|
minivec_cmp
|
||||||
|
-------------
|
||||||
|
-1
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT minivec_cmp('[1,2,3]', '[1,2]');
|
||||||
|
minivec_cmp
|
||||||
|
-------------
|
||||||
|
1
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT minivec_cmp('[1,2]', '[2,3,4]');
|
||||||
|
minivec_cmp
|
||||||
|
-------------
|
||||||
|
-1
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT minivec_cmp('[2,3]', '[1,2,3]');
|
||||||
|
minivec_cmp
|
||||||
|
-------------
|
||||||
|
1
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT vector_dims('[1,2,3]'::minivec);
|
||||||
|
vector_dims
|
||||||
|
-------------
|
||||||
|
3
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT round(l2_norm('[1,1]'::minivec)::numeric, 5);
|
||||||
|
round
|
||||||
|
---------
|
||||||
|
1.41421
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT l2_norm('[3,4]'::minivec);
|
||||||
|
l2_norm
|
||||||
|
---------
|
||||||
|
5
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT l2_norm('[0,1]'::minivec);
|
||||||
|
l2_norm
|
||||||
|
---------
|
||||||
|
1
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT l2_norm('[0,0]'::minivec);
|
||||||
|
l2_norm
|
||||||
|
---------
|
||||||
|
0
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT l2_norm('[2]'::minivec);
|
||||||
|
l2_norm
|
||||||
|
---------
|
||||||
|
2
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT l2_distance('[0,0]'::minivec, '[3,4]');
|
||||||
|
l2_distance
|
||||||
|
-------------
|
||||||
|
5
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT l2_distance('[0,0]'::minivec, '[0,1]');
|
||||||
|
l2_distance
|
||||||
|
-------------
|
||||||
|
1
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT l2_distance('[1,2]'::minivec, '[3]');
|
||||||
|
ERROR: different minivec dimensions 2 and 1
|
||||||
|
SELECT l2_distance('[1,1,1,1,1,1,1,1,1]'::minivec, '[1,1,1,1,1,1,1,4,5]');
|
||||||
|
l2_distance
|
||||||
|
-------------
|
||||||
|
5
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[0,0]'::minivec <-> '[3,4]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
5
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT inner_product('[1,2]'::minivec, '[3,4]');
|
||||||
|
inner_product
|
||||||
|
---------------
|
||||||
|
11
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT inner_product('[1,2]'::minivec, '[3]');
|
||||||
|
ERROR: different minivec dimensions 2 and 1
|
||||||
|
SELECT inner_product('[448]'::minivec, '[448]');
|
||||||
|
inner_product
|
||||||
|
---------------
|
||||||
|
200704
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT inner_product('[1,1,1,1,1,1,1,1,1]'::minivec, '[1,2,3,4,5,6,7,8,9]');
|
||||||
|
inner_product
|
||||||
|
---------------
|
||||||
|
45
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2]'::minivec <#> '[3,4]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
-11
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT cosine_distance('[1,2]'::minivec, '[2,4]');
|
||||||
|
cosine_distance
|
||||||
|
-----------------
|
||||||
|
0
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT cosine_distance('[1,2]'::minivec, '[0,0]');
|
||||||
|
cosine_distance
|
||||||
|
-----------------
|
||||||
|
NaN
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT cosine_distance('[1,1]'::minivec, '[1,1]');
|
||||||
|
cosine_distance
|
||||||
|
-----------------
|
||||||
|
0
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT cosine_distance('[1,0]'::minivec, '[0,2]');
|
||||||
|
cosine_distance
|
||||||
|
-----------------
|
||||||
|
1
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT cosine_distance('[1,1]'::minivec, '[-1,-1]');
|
||||||
|
cosine_distance
|
||||||
|
-----------------
|
||||||
|
2
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT cosine_distance('[1,2]'::minivec, '[3]');
|
||||||
|
ERROR: different minivec dimensions 2 and 1
|
||||||
|
SELECT cosine_distance('[1,1]'::minivec, '[1.1,1.1]');
|
||||||
|
cosine_distance
|
||||||
|
-----------------
|
||||||
|
0
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT cosine_distance('[1,1]'::minivec, '[-1.1,-1.1]');
|
||||||
|
cosine_distance
|
||||||
|
-----------------
|
||||||
|
2
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT cosine_distance('[1,2,3,4,5,6,7,8,9]'::minivec, '[1,2,3,4,5,6,7,8,9]');
|
||||||
|
cosine_distance
|
||||||
|
-----------------
|
||||||
|
0
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT cosine_distance('[1,2,3,4,5,6,7,8,9]'::minivec, '[-1,-2,-3,-4,-5,-6,-7,-8,-9]');
|
||||||
|
cosine_distance
|
||||||
|
-----------------
|
||||||
|
2
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[1,2]'::minivec <=> '[2,4]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
0
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT l1_distance('[0,0]'::minivec, '[3,4]');
|
||||||
|
l1_distance
|
||||||
|
-------------
|
||||||
|
7
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT l1_distance('[0,0]'::minivec, '[0,1]');
|
||||||
|
l1_distance
|
||||||
|
-------------
|
||||||
|
1
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT l1_distance('[1,2]'::minivec, '[3]');
|
||||||
|
ERROR: different minivec dimensions 2 and 1
|
||||||
|
SELECT l1_distance('[1,2,3,4,5,6,7,8,9]'::minivec, '[1,2,3,4,5,6,7,8,9]');
|
||||||
|
l1_distance
|
||||||
|
-------------
|
||||||
|
0
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT l1_distance('[1,2,3,4,5,6,7,8,9]'::minivec, '[0,3,2,5,4,7,6,9,8]');
|
||||||
|
l1_distance
|
||||||
|
-------------
|
||||||
|
9
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT '[0,0]'::minivec <+> '[3,4]';
|
||||||
|
?column?
|
||||||
|
----------
|
||||||
|
7
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT l2_normalize('[3,4]'::minivec);
|
||||||
|
l2_normalize
|
||||||
|
----------------
|
||||||
|
[0.625,0.8125]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT l2_normalize('[3,0]'::minivec);
|
||||||
|
l2_normalize
|
||||||
|
--------------
|
||||||
|
[1,0]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT l2_normalize('[0,0.1]'::minivec);
|
||||||
|
l2_normalize
|
||||||
|
--------------
|
||||||
|
[0,1]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT l2_normalize('[0,0]'::minivec);
|
||||||
|
l2_normalize
|
||||||
|
--------------
|
||||||
|
[0,0]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT l2_normalize('[448]'::minivec);
|
||||||
|
l2_normalize
|
||||||
|
--------------
|
||||||
|
[1]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT binary_quantize('[1,0,-1]'::minivec);
|
||||||
|
binary_quantize
|
||||||
|
-----------------
|
||||||
|
100
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT binary_quantize('[0,0.1,-0.2,-0.3,0.4,0.5,0.6,-0.7,0.8,-0.9,1]'::minivec);
|
||||||
|
binary_quantize
|
||||||
|
-----------------
|
||||||
|
01001110101
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, 1, 3);
|
||||||
|
subvector
|
||||||
|
-----------
|
||||||
|
[1,2,3]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, 3, 2);
|
||||||
|
subvector
|
||||||
|
-----------
|
||||||
|
[3,4]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, -1, 3);
|
||||||
|
subvector
|
||||||
|
-----------
|
||||||
|
[1]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, 3, 9);
|
||||||
|
subvector
|
||||||
|
-----------
|
||||||
|
[3,4,5]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, 1, 0);
|
||||||
|
ERROR: minivec must have at least 1 dimension
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, 3, -1);
|
||||||
|
ERROR: minivec must have at least 1 dimension
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, -1, 2);
|
||||||
|
ERROR: minivec must have at least 1 dimension
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, 2147483647, 10);
|
||||||
|
ERROR: minivec must have at least 1 dimension
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, 3, 2147483647);
|
||||||
|
subvector
|
||||||
|
-----------
|
||||||
|
[3,4,5]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, -2147483644, 2147483647);
|
||||||
|
subvector
|
||||||
|
-----------
|
||||||
|
[1,2]
|
||||||
|
(1 row)
|
||||||
|
|
||||||
11
test/perl/PostgreSQL/Test/Cluster.pm
Normal file
11
test/perl/PostgreSQL/Test/Cluster.pm
Normal file
@@ -0,0 +1,11 @@
|
|||||||
|
package PostgreSQL::Test::Cluster;
|
||||||
|
|
||||||
|
use PostgresNode;
|
||||||
|
|
||||||
|
sub new
|
||||||
|
{
|
||||||
|
my ($class, $name) = @_;
|
||||||
|
return get_new_node($name);
|
||||||
|
}
|
||||||
|
|
||||||
|
1;
|
||||||
5
test/perl/PostgreSQL/Test/Utils.pm
Normal file
5
test/perl/PostgreSQL/Test/Utils.pm
Normal file
@@ -0,0 +1,5 @@
|
|||||||
|
package PostgreSQL::Test::Utils;
|
||||||
|
|
||||||
|
use TestLib;
|
||||||
|
|
||||||
|
1;
|
||||||
@@ -1,8 +0,0 @@
|
|||||||
use PostgreSQL::Test::Cluster;
|
|
||||||
|
|
||||||
sub get_new_node
|
|
||||||
{
|
|
||||||
return PostgreSQL::Test::Cluster->new(@_);
|
|
||||||
}
|
|
||||||
|
|
||||||
1;
|
|
||||||
@@ -1,3 +0,0 @@
|
|||||||
use PostgreSQL::Test::Utils;
|
|
||||||
|
|
||||||
1;
|
|
||||||
@@ -22,6 +22,17 @@ SELECT * FROM t ORDER BY val;
|
|||||||
|
|
||||||
DROP TABLE t;
|
DROP TABLE t;
|
||||||
|
|
||||||
|
-- minivec
|
||||||
|
|
||||||
|
CREATE TABLE t (val minivec(3));
|
||||||
|
INSERT INTO t (val) VALUES ('[0,0,0]'), ('[1,2,3]'), ('[1,1,1]'), (NULL);
|
||||||
|
CREATE INDEX ON t (val);
|
||||||
|
|
||||||
|
SELECT * FROM t WHERE val = '[1,2,3]';
|
||||||
|
SELECT * FROM t ORDER BY val;
|
||||||
|
|
||||||
|
DROP TABLE t;
|
||||||
|
|
||||||
-- sparsevec
|
-- sparsevec
|
||||||
|
|
||||||
CREATE TABLE t (val sparsevec(3));
|
CREATE TABLE t (val sparsevec(3));
|
||||||
|
|||||||
@@ -38,6 +38,22 @@ SELECT '{1,2,3}'::real[]::halfvec(2);
|
|||||||
SELECT '{65520,-65520}'::real[]::halfvec;
|
SELECT '{65520,-65520}'::real[]::halfvec;
|
||||||
SELECT '{1e-8,-1e-8}'::real[]::halfvec;
|
SELECT '{1e-8,-1e-8}'::real[]::halfvec;
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::vector::minivec;
|
||||||
|
SELECT '[1,2,3]'::vector::minivec(3);
|
||||||
|
SELECT '[1,2,3]'::vector::minivec(2);
|
||||||
|
SELECT '[465]'::vector::minivec;
|
||||||
|
SELECT '[1e-8]'::vector::minivec;
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec::vector;
|
||||||
|
SELECT '[1,2,3]'::minivec::vector(3);
|
||||||
|
SELECT '[1,2,3]'::minivec::vector(2);
|
||||||
|
|
||||||
|
SELECT '{1,2,3}'::real[]::minivec;
|
||||||
|
SELECT '{1,2,3}'::real[]::minivec(3);
|
||||||
|
SELECT '{1,2,3}'::real[]::minivec(2);
|
||||||
|
SELECT '{465,-465}'::real[]::minivec;
|
||||||
|
SELECT '{1e-8,-1e-8}'::real[]::minivec;
|
||||||
|
|
||||||
SELECT '[0,1.5,0,3.5,0]'::vector::sparsevec;
|
SELECT '[0,1.5,0,3.5,0]'::vector::sparsevec;
|
||||||
SELECT '[0,1.5,0,3.5,0]'::vector::sparsevec(5);
|
SELECT '[0,1.5,0,3.5,0]'::vector::sparsevec(5);
|
||||||
SELECT '[0,1.5,0,3.5,0]'::vector::sparsevec(4);
|
SELECT '[0,1.5,0,3.5,0]'::vector::sparsevec(4);
|
||||||
@@ -58,6 +74,22 @@ SELECT '{}/16001'::sparsevec::halfvec;
|
|||||||
SELECT '{1:65520}/1'::sparsevec::halfvec;
|
SELECT '{1:65520}/1'::sparsevec::halfvec;
|
||||||
SELECT '{1:1e-8}/1'::sparsevec::halfvec;
|
SELECT '{1:1e-8}/1'::sparsevec::halfvec;
|
||||||
|
|
||||||
|
SELECT ARRAY[1,0,2,0,3,0]::sparsevec;
|
||||||
|
SELECT ARRAY[1.0,0.0,2.0,0.0,3.0,0.0]::sparsevec;
|
||||||
|
SELECT ARRAY[1,0,2,0,3,0]::float4[]::sparsevec;
|
||||||
|
SELECT ARRAY[1,0,2,0,3,0]::float8[]::sparsevec;
|
||||||
|
SELECT ARRAY[1,0,2,0,3,0]::numeric[]::sparsevec;
|
||||||
|
|
||||||
|
SELECT '{1,0,2,0,3,0}'::real[]::sparsevec;
|
||||||
|
SELECT '{1,0,2,0,3,0}'::real[]::sparsevec(6);
|
||||||
|
SELECT '{1,0,2,0,3,0}'::real[]::sparsevec(5);
|
||||||
|
SELECT '{NULL}'::real[]::sparsevec;
|
||||||
|
SELECT '{NaN}'::real[]::sparsevec;
|
||||||
|
SELECT '{Infinity}'::real[]::sparsevec;
|
||||||
|
SELECT '{-Infinity}'::real[]::sparsevec;
|
||||||
|
SELECT '{}'::real[]::sparsevec;
|
||||||
|
SELECT '{{1}}'::real[]::sparsevec;
|
||||||
|
|
||||||
SELECT array_agg(n)::vector FROM generate_series(1, 16001) n;
|
SELECT array_agg(n)::vector FROM generate_series(1, 16001) n;
|
||||||
SELECT array_to_vector(array_agg(n), 16001, false) FROM generate_series(1, 16001) n;
|
SELECT array_to_vector(array_agg(n), 16001, false) FROM generate_series(1, 16001) n;
|
||||||
|
|
||||||
|
|||||||
@@ -28,6 +28,21 @@ SELECT * FROM t2 ORDER BY val;
|
|||||||
DROP TABLE t;
|
DROP TABLE t;
|
||||||
DROP TABLE t2;
|
DROP TABLE t2;
|
||||||
|
|
||||||
|
-- minivec
|
||||||
|
|
||||||
|
CREATE TABLE t (val minivec(3));
|
||||||
|
INSERT INTO t (val) VALUES ('[0,0,0]'), ('[1,2,3]'), ('[1,1,1]'), (NULL);
|
||||||
|
|
||||||
|
CREATE TABLE t2 (val minivec(3));
|
||||||
|
|
||||||
|
\copy t TO 'results/minivec.bin' WITH (FORMAT binary)
|
||||||
|
\copy t2 FROM 'results/minivec.bin' WITH (FORMAT binary)
|
||||||
|
|
||||||
|
SELECT * FROM t2 ORDER BY val;
|
||||||
|
|
||||||
|
DROP TABLE t;
|
||||||
|
DROP TABLE t2;
|
||||||
|
|
||||||
-- sparsevec
|
-- sparsevec
|
||||||
|
|
||||||
CREATE TABLE t (val sparsevec(3));
|
CREATE TABLE t (val sparsevec(3));
|
||||||
|
|||||||
45
test/sql/ivfflat_minivec.sql
Normal file
45
test/sql/ivfflat_minivec.sql
Normal file
@@ -0,0 +1,45 @@
|
|||||||
|
SET enable_seqscan = off;
|
||||||
|
|
||||||
|
-- L2
|
||||||
|
|
||||||
|
CREATE TABLE t (val minivec(3));
|
||||||
|
INSERT INTO t (val) VALUES ('[0,0,0]'), ('[1,2,3]'), ('[1,1,1]'), (NULL);
|
||||||
|
CREATE INDEX ON t USING ivfflat (val minivec_l2_ops) WITH (lists = 1);
|
||||||
|
|
||||||
|
INSERT INTO t (val) VALUES ('[1,2,4]');
|
||||||
|
|
||||||
|
SELECT * FROM t ORDER BY val <-> '[3,3,3]';
|
||||||
|
SELECT COUNT(*) FROM (SELECT * FROM t ORDER BY val <-> (SELECT NULL::minivec)) t2;
|
||||||
|
SELECT COUNT(*) FROM t;
|
||||||
|
|
||||||
|
TRUNCATE t;
|
||||||
|
SELECT * FROM t ORDER BY val <-> '[3,3,3]';
|
||||||
|
|
||||||
|
DROP TABLE t;
|
||||||
|
|
||||||
|
-- inner product
|
||||||
|
|
||||||
|
CREATE TABLE t (val minivec(3));
|
||||||
|
INSERT INTO t (val) VALUES ('[0,0,0]'), ('[1,2,3]'), ('[1,1,1]'), (NULL);
|
||||||
|
CREATE INDEX ON t USING ivfflat (val minivec_ip_ops) WITH (lists = 1);
|
||||||
|
|
||||||
|
INSERT INTO t (val) VALUES ('[1,2,4]');
|
||||||
|
|
||||||
|
SELECT * FROM t ORDER BY val <#> '[3,3,3]';
|
||||||
|
SELECT COUNT(*) FROM (SELECT * FROM t ORDER BY val <#> (SELECT NULL::minivec)) t2;
|
||||||
|
|
||||||
|
DROP TABLE t;
|
||||||
|
|
||||||
|
-- cosine
|
||||||
|
|
||||||
|
CREATE TABLE t (val minivec(3));
|
||||||
|
INSERT INTO t (val) VALUES ('[0,0,0]'), ('[1,2,3]'), ('[1,1,1]'), (NULL);
|
||||||
|
CREATE INDEX ON t USING ivfflat (val minivec_cosine_ops) WITH (lists = 1);
|
||||||
|
|
||||||
|
INSERT INTO t (val) VALUES ('[1,2,4]');
|
||||||
|
|
||||||
|
SELECT * FROM t ORDER BY val <=> '[3,3,3]';
|
||||||
|
SELECT COUNT(*) FROM (SELECT * FROM t ORDER BY val <=> '[0,0,0]') t2;
|
||||||
|
SELECT COUNT(*) FROM (SELECT * FROM t ORDER BY val <=> (SELECT NULL::minivec)) t2;
|
||||||
|
|
||||||
|
DROP TABLE t;
|
||||||
134
test/sql/minivec.sql
Normal file
134
test/sql/minivec.sql
Normal file
@@ -0,0 +1,134 @@
|
|||||||
|
SELECT '[1,2,3]'::minivec;
|
||||||
|
SELECT '[-1,-2,-3]'::minivec;
|
||||||
|
SELECT '[1.,2.,3.]'::minivec;
|
||||||
|
SELECT ' [ 1, 2 , 3 ] '::minivec;
|
||||||
|
SELECT '[1.23456]'::minivec;
|
||||||
|
SELECT '[hello,1]'::minivec;
|
||||||
|
SELECT '[NaN,1]'::minivec;
|
||||||
|
SELECT '[Infinity,1]'::minivec;
|
||||||
|
SELECT '[-Infinity,1]'::minivec;
|
||||||
|
SELECT '[65519,-65519]'::minivec;
|
||||||
|
SELECT '[65520,-65520]'::minivec;
|
||||||
|
SELECT '[1e-8,-1e-8]'::minivec;
|
||||||
|
SELECT '[4e38,1]'::minivec;
|
||||||
|
SELECT '[1e-46,1]'::minivec;
|
||||||
|
SELECT '[1,2,3'::minivec;
|
||||||
|
SELECT '[1,2,3]9'::minivec;
|
||||||
|
SELECT '1,2,3'::minivec;
|
||||||
|
SELECT ''::minivec;
|
||||||
|
SELECT '['::minivec;
|
||||||
|
SELECT '[ '::minivec;
|
||||||
|
SELECT '[,'::minivec;
|
||||||
|
SELECT '[]'::minivec;
|
||||||
|
SELECT '[ ]'::minivec;
|
||||||
|
SELECT '[,]'::minivec;
|
||||||
|
SELECT '[1,]'::minivec;
|
||||||
|
SELECT '[1a]'::minivec;
|
||||||
|
SELECT '[1,,3]'::minivec;
|
||||||
|
SELECT '[1, ,3]'::minivec;
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec(3);
|
||||||
|
SELECT '[1,2,3]'::minivec(2);
|
||||||
|
SELECT '[1,2,3]'::minivec(3, 2);
|
||||||
|
SELECT '[1,2,3]'::minivec('a');
|
||||||
|
SELECT '[1,2,3]'::minivec(0);
|
||||||
|
SELECT '[1,2,3]'::minivec(16001);
|
||||||
|
|
||||||
|
SELECT unnest('{"[1,2,3]", "[4,5,6]"}'::minivec[]);
|
||||||
|
SELECT '{"[1,2,3]"}'::minivec(2)[];
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec + '[4,5,6]';
|
||||||
|
SELECT '[448]'::minivec + '[448]';
|
||||||
|
SELECT '[1,2]'::minivec + '[3]';
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec - '[4,5,6]';
|
||||||
|
SELECT '[-448]'::minivec - '[448]';
|
||||||
|
SELECT '[1,2]'::minivec - '[3]';
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec * '[4,5,6]';
|
||||||
|
SELECT '[448]'::minivec * '[448]';
|
||||||
|
SELECT '[1e-7]'::minivec * '[1e-7]';
|
||||||
|
SELECT '[1,2]'::minivec * '[3]';
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec || '[4,5]';
|
||||||
|
SELECT array_fill(0, ARRAY[16000])::minivec || '[1]';
|
||||||
|
|
||||||
|
SELECT '[1,2,3]'::minivec < '[1,2,3]';
|
||||||
|
SELECT '[1,2,3]'::minivec < '[1,2]';
|
||||||
|
SELECT '[1,2,3]'::minivec <= '[1,2,3]';
|
||||||
|
SELECT '[1,2,3]'::minivec <= '[1,2]';
|
||||||
|
SELECT '[1,2,3]'::minivec = '[1,2,3]';
|
||||||
|
SELECT '[1,2,3]'::minivec = '[1,2]';
|
||||||
|
SELECT '[1,2,3]'::minivec != '[1,2,3]';
|
||||||
|
SELECT '[1,2,3]'::minivec != '[1,2]';
|
||||||
|
SELECT '[1,2,3]'::minivec >= '[1,2,3]';
|
||||||
|
SELECT '[1,2,3]'::minivec >= '[1,2]';
|
||||||
|
SELECT '[1,2,3]'::minivec > '[1,2,3]';
|
||||||
|
SELECT '[1,2,3]'::minivec > '[1,2]';
|
||||||
|
|
||||||
|
SELECT minivec_cmp('[1,2,3]', '[1,2,3]');
|
||||||
|
SELECT minivec_cmp('[1,2,3]', '[0,0,0]');
|
||||||
|
SELECT minivec_cmp('[0,0,0]', '[1,2,3]');
|
||||||
|
SELECT minivec_cmp('[1,2]', '[1,2,3]');
|
||||||
|
SELECT minivec_cmp('[1,2,3]', '[1,2]');
|
||||||
|
SELECT minivec_cmp('[1,2]', '[2,3,4]');
|
||||||
|
SELECT minivec_cmp('[2,3]', '[1,2,3]');
|
||||||
|
|
||||||
|
SELECT vector_dims('[1,2,3]'::minivec);
|
||||||
|
|
||||||
|
SELECT round(l2_norm('[1,1]'::minivec)::numeric, 5);
|
||||||
|
SELECT l2_norm('[3,4]'::minivec);
|
||||||
|
SELECT l2_norm('[0,1]'::minivec);
|
||||||
|
SELECT l2_norm('[0,0]'::minivec);
|
||||||
|
SELECT l2_norm('[2]'::minivec);
|
||||||
|
|
||||||
|
SELECT l2_distance('[0,0]'::minivec, '[3,4]');
|
||||||
|
SELECT l2_distance('[0,0]'::minivec, '[0,1]');
|
||||||
|
SELECT l2_distance('[1,2]'::minivec, '[3]');
|
||||||
|
SELECT l2_distance('[1,1,1,1,1,1,1,1,1]'::minivec, '[1,1,1,1,1,1,1,4,5]');
|
||||||
|
SELECT '[0,0]'::minivec <-> '[3,4]';
|
||||||
|
|
||||||
|
SELECT inner_product('[1,2]'::minivec, '[3,4]');
|
||||||
|
SELECT inner_product('[1,2]'::minivec, '[3]');
|
||||||
|
SELECT inner_product('[448]'::minivec, '[448]');
|
||||||
|
SELECT inner_product('[1,1,1,1,1,1,1,1,1]'::minivec, '[1,2,3,4,5,6,7,8,9]');
|
||||||
|
SELECT '[1,2]'::minivec <#> '[3,4]';
|
||||||
|
|
||||||
|
SELECT cosine_distance('[1,2]'::minivec, '[2,4]');
|
||||||
|
SELECT cosine_distance('[1,2]'::minivec, '[0,0]');
|
||||||
|
SELECT cosine_distance('[1,1]'::minivec, '[1,1]');
|
||||||
|
SELECT cosine_distance('[1,0]'::minivec, '[0,2]');
|
||||||
|
SELECT cosine_distance('[1,1]'::minivec, '[-1,-1]');
|
||||||
|
SELECT cosine_distance('[1,2]'::minivec, '[3]');
|
||||||
|
SELECT cosine_distance('[1,1]'::minivec, '[1.1,1.1]');
|
||||||
|
SELECT cosine_distance('[1,1]'::minivec, '[-1.1,-1.1]');
|
||||||
|
SELECT cosine_distance('[1,2,3,4,5,6,7,8,9]'::minivec, '[1,2,3,4,5,6,7,8,9]');
|
||||||
|
SELECT cosine_distance('[1,2,3,4,5,6,7,8,9]'::minivec, '[-1,-2,-3,-4,-5,-6,-7,-8,-9]');
|
||||||
|
SELECT '[1,2]'::minivec <=> '[2,4]';
|
||||||
|
|
||||||
|
SELECT l1_distance('[0,0]'::minivec, '[3,4]');
|
||||||
|
SELECT l1_distance('[0,0]'::minivec, '[0,1]');
|
||||||
|
SELECT l1_distance('[1,2]'::minivec, '[3]');
|
||||||
|
SELECT l1_distance('[1,2,3,4,5,6,7,8,9]'::minivec, '[1,2,3,4,5,6,7,8,9]');
|
||||||
|
SELECT l1_distance('[1,2,3,4,5,6,7,8,9]'::minivec, '[0,3,2,5,4,7,6,9,8]');
|
||||||
|
SELECT '[0,0]'::minivec <+> '[3,4]';
|
||||||
|
|
||||||
|
SELECT l2_normalize('[3,4]'::minivec);
|
||||||
|
SELECT l2_normalize('[3,0]'::minivec);
|
||||||
|
SELECT l2_normalize('[0,0.1]'::minivec);
|
||||||
|
SELECT l2_normalize('[0,0]'::minivec);
|
||||||
|
SELECT l2_normalize('[448]'::minivec);
|
||||||
|
|
||||||
|
SELECT binary_quantize('[1,0,-1]'::minivec);
|
||||||
|
SELECT binary_quantize('[0,0.1,-0.2,-0.3,0.4,0.5,0.6,-0.7,0.8,-0.9,1]'::minivec);
|
||||||
|
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, 1, 3);
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, 3, 2);
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, -1, 3);
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, 3, 9);
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, 1, 0);
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, 3, -1);
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, -1, 2);
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, 2147483647, 10);
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, 3, 2147483647);
|
||||||
|
SELECT subvector('[1,2,3,4,5]'::minivec, -2147483644, 2147483647);
|
||||||
@@ -2,9 +2,9 @@
|
|||||||
|
|
||||||
# Test generic xlog record work for ivfflat index replication.
|
# Test generic xlog record work for ivfflat index replication.
|
||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $dim = 32;
|
my $dim = 32;
|
||||||
@@ -49,7 +49,7 @@ sub test_index_replay
|
|||||||
my $array_sql = join(",", ('random()') x $dim);
|
my $array_sql = join(",", ('random()') x $dim);
|
||||||
|
|
||||||
# Initialize primary node
|
# Initialize primary node
|
||||||
$node_primary = get_new_node('primary');
|
$node_primary = PostgreSQL::Test::Cluster->new('primary');
|
||||||
$node_primary->init(allows_streaming => 1);
|
$node_primary->init(allows_streaming => 1);
|
||||||
if ($dim > 32)
|
if ($dim > 32)
|
||||||
{
|
{
|
||||||
@@ -67,7 +67,7 @@ my $backup_name = 'my_backup';
|
|||||||
$node_primary->backup($backup_name);
|
$node_primary->backup($backup_name);
|
||||||
|
|
||||||
# Create streaming replica linking to primary
|
# Create streaming replica linking to primary
|
||||||
$node_replica = get_new_node('replica');
|
$node_replica = PostgreSQL::Test::Cluster->new('replica');
|
||||||
$node_replica->init_from_backup($node_primary, $backup_name, has_streaming => 1);
|
$node_replica->init_from_backup($node_primary, $backup_name, has_streaming => 1);
|
||||||
$node_replica->start;
|
$node_replica->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $dim = 3;
|
my $dim = 3;
|
||||||
@@ -15,7 +15,7 @@ for (1 .. $dim)
|
|||||||
my $array_sql = join(", ", @r);
|
my $array_sql = join(", ", @r);
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -49,7 +49,7 @@ sub test_recall
|
|||||||
}
|
}
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -48,7 +48,7 @@ sub test_recall
|
|||||||
}
|
}
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,11 +1,11 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,11 +1,11 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $dim = 768;
|
my $dim = 768;
|
||||||
@@ -9,7 +9,7 @@ my $dim = 768;
|
|||||||
my $array_sql = join(",", ('random()') x $dim);
|
my $array_sql = join(",", ('random()') x $dim);
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,11 +1,11 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $dim = 3;
|
my $dim = 3;
|
||||||
@@ -11,7 +11,7 @@ my $limit = 20;
|
|||||||
my $array_sql = join(",", ('random()') x $dim);
|
my $array_sql = join(",", ('random()') x $dim);
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -2,9 +2,9 @@
|
|||||||
|
|
||||||
# Test generic xlog record work for hnsw index replication.
|
# Test generic xlog record work for hnsw index replication.
|
||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $dim = 32;
|
my $dim = 32;
|
||||||
@@ -49,7 +49,7 @@ sub test_index_replay
|
|||||||
my $array_sql = join(",", ('random()') x $dim);
|
my $array_sql = join(",", ('random()') x $dim);
|
||||||
|
|
||||||
# Initialize primary node
|
# Initialize primary node
|
||||||
$node_primary = get_new_node('primary');
|
$node_primary = PostgreSQL::Test::Cluster->new('primary');
|
||||||
$node_primary->init(allows_streaming => 1);
|
$node_primary->init(allows_streaming => 1);
|
||||||
if ($dim > 32)
|
if ($dim > 32)
|
||||||
{
|
{
|
||||||
@@ -67,7 +67,7 @@ my $backup_name = 'my_backup';
|
|||||||
$node_primary->backup($backup_name);
|
$node_primary->backup($backup_name);
|
||||||
|
|
||||||
# Create streaming replica linking to primary
|
# Create streaming replica linking to primary
|
||||||
$node_replica = get_new_node('replica');
|
$node_replica = PostgreSQL::Test::Cluster->new('replica');
|
||||||
$node_replica->init_from_backup($node_primary, $backup_name, has_streaming => 1);
|
$node_replica->init_from_backup($node_primary, $backup_name, has_streaming => 1);
|
||||||
$node_replica->start;
|
$node_replica->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $dim = 3;
|
my $dim = 3;
|
||||||
@@ -15,7 +15,7 @@ for (1 .. $dim)
|
|||||||
my $array_sql = join(", ", @r);
|
my $array_sql = join(", ", @r);
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -47,7 +47,7 @@ sub test_recall
|
|||||||
}
|
}
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -47,7 +47,7 @@ sub test_recall
|
|||||||
}
|
}
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -48,7 +48,7 @@ sub test_recall
|
|||||||
}
|
}
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,11 +1,11 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
# Ensures elements and neighbors on both same and different pages
|
# Ensures elements and neighbors on both same and different pages
|
||||||
@@ -10,7 +10,7 @@ my $dim = 1900;
|
|||||||
my $array_sql = join(",", ('random()') x $dim);
|
my $array_sql = join(",", ('random()') x $dim);
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $dim = 3;
|
my $dim = 3;
|
||||||
@@ -11,7 +11,7 @@ my $limit = 20;
|
|||||||
my $array_sql = join(",", ('random()') x $dim);
|
my $array_sql = join(",", ('random()') x $dim);
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,11 +1,11 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
@@ -53,7 +53,7 @@ sub test_aggregate
|
|||||||
else
|
else
|
||||||
{
|
{
|
||||||
# Does not raise overflow error in this instance due to loss of precision
|
# Does not raise overflow error in this instance due to loss of precision
|
||||||
is($res, "[24576,24576,49152]")
|
is($res, "[24576,24576,49152]");
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
|||||||
@@ -1,13 +1,13 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $dim = 1024;
|
my $dim = 1024;
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -51,7 +51,7 @@ sub test_recall
|
|||||||
}
|
}
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -51,7 +51,7 @@ sub test_recall
|
|||||||
}
|
}
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -51,7 +51,7 @@ sub test_recall
|
|||||||
}
|
}
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,11 +1,11 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -48,7 +48,7 @@ sub test_recall
|
|||||||
}
|
}
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -48,7 +48,7 @@ sub test_recall
|
|||||||
}
|
}
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -48,7 +48,7 @@ sub test_recall
|
|||||||
}
|
}
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,11 +1,11 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -47,7 +47,7 @@ sub test_recall
|
|||||||
}
|
}
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -47,7 +47,7 @@ sub test_recall
|
|||||||
}
|
}
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -48,7 +48,7 @@ sub test_recall
|
|||||||
}
|
}
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,11 +1,11 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -51,7 +51,7 @@ sub test_recall
|
|||||||
}
|
}
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,14 +1,14 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
my $array_sql = join(",", ('floor(random() * 2)::int - 1') x 3);
|
my $array_sql = join(",", ('floor(random() * 2)::int - 1') x 3);
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
@@ -40,6 +40,10 @@ for (1 .. 50)
|
|||||||
$actual = $node->safe_psql("postgres", "SELECT halfvec_cmp(v::halfvec, '$query'::real[]::halfvec) FROM tst");
|
$actual = $node->safe_psql("postgres", "SELECT halfvec_cmp(v::halfvec, '$query'::real[]::halfvec) FROM tst");
|
||||||
is($expected, $actual);
|
is($expected, $actual);
|
||||||
|
|
||||||
|
# Test minivec
|
||||||
|
$actual = $node->safe_psql("postgres", "SELECT minivec_cmp(v::minivec, '$query'::real[]::minivec) FROM tst");
|
||||||
|
is($expected, $actual);
|
||||||
|
|
||||||
# Test sparsevec
|
# Test sparsevec
|
||||||
$actual = $node->safe_psql("postgres", "SELECT sparsevec_cmp(v::vector::sparsevec, '$query'::real[]::vector::sparsevec) FROM tst");
|
$actual = $node->safe_psql("postgres", "SELECT sparsevec_cmp(v::vector::sparsevec, '$query'::real[]::vector::sparsevec) FROM tst");
|
||||||
is($expected, $actual);
|
is($expected, $actual);
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -10,7 +10,7 @@ my $dim = 5;
|
|||||||
my $array_sql = join(",", ('floor(random() * 4)::int - 2') x $dim);
|
my $array_sql = join(",", ('floor(random() * 4)::int - 2') x $dim);
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
@@ -45,6 +45,10 @@ for my $function (@functions)
|
|||||||
my $actual = $node->safe_psql("postgres", "SELECT $function(v::halfvec, '$query'::vector::halfvec) FROM tst");
|
my $actual = $node->safe_psql("postgres", "SELECT $function(v::halfvec, '$query'::vector::halfvec) FROM tst");
|
||||||
is($expected, $actual, "halfvec $function");
|
is($expected, $actual, "halfvec $function");
|
||||||
|
|
||||||
|
# Test minivec
|
||||||
|
$actual = $node->safe_psql("postgres", "SELECT $function(v::minivec, '$query'::vector::minivec) FROM tst");
|
||||||
|
is($expected, $actual, "minivec $function");
|
||||||
|
|
||||||
# Test sparsevec
|
# Test sparsevec
|
||||||
$actual = $node->safe_psql("postgres", "SELECT $function(v::sparsevec, '$query'::vector::sparsevec) FROM tst");
|
$actual = $node->safe_psql("postgres", "SELECT $function(v::sparsevec, '$query'::vector::sparsevec) FROM tst");
|
||||||
is($expected, $actual, "sparsevec $function");
|
is($expected, $actual, "sparsevec $function");
|
||||||
|
|||||||
@@ -1,7 +1,7 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
my $node;
|
my $node;
|
||||||
@@ -51,7 +51,7 @@ sub test_recall
|
|||||||
}
|
}
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
$node = get_new_node('node');
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,11 +1,11 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
|
|||||||
@@ -1,19 +1,19 @@
|
|||||||
use strict;
|
use strict;
|
||||||
use warnings;
|
use warnings FATAL => 'all';
|
||||||
use PostgresNode;
|
use PostgreSQL::Test::Cluster;
|
||||||
use TestLib;
|
use PostgreSQL::Test::Utils;
|
||||||
use Test::More;
|
use Test::More;
|
||||||
|
|
||||||
# Initialize node
|
# Initialize node
|
||||||
my $node = get_new_node('node');
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
$node->init;
|
$node->init;
|
||||||
$node->start;
|
$node->start;
|
||||||
|
|
||||||
# Create extension
|
# Create extension
|
||||||
$node->safe_psql("postgres", "CREATE EXTENSION vector;");
|
$node->safe_psql("postgres", "CREATE EXTENSION vector;");
|
||||||
|
|
||||||
my @types = ("vector", "halfvec", "sparsevec");
|
my @types = ("vector", "halfvec", "minivec", "sparsevec");
|
||||||
my @inputs = ("[1.23,4.56,7.89]", "[1.23,4.56,7.89]", "{1:1.23,2:4.56,3:7.89}/3");
|
my @inputs = ("[1.23,4.56,7.89]", "[1.23,4.56,7.89]", "[1.23,4.56,7.89]", "{1:1.23,2:4.56,3:7.89}/3");
|
||||||
my @subs = (" ", " ", ",", ":", "-", "1", "9", "\0", "2147483648", "-2147483649");
|
my @subs = (" ", " ", ",", ":", "-", "1", "9", "\0", "2147483648", "-2147483649");
|
||||||
|
|
||||||
for my $i (0 .. $#types)
|
for my $i (0 .. $#types)
|
||||||
|
|||||||
42
test/t/038_hnsw_sparsevec_vacuum_insert.pl
Normal file
42
test/t/038_hnsw_sparsevec_vacuum_insert.pl
Normal file
@@ -0,0 +1,42 @@
|
|||||||
|
use strict;
|
||||||
|
use warnings FATAL => 'all';
|
||||||
|
use PostgreSQL::Test::Cluster;
|
||||||
|
use PostgreSQL::Test::Utils;
|
||||||
|
use Test::More;
|
||||||
|
|
||||||
|
# Initialize node
|
||||||
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
|
$node->init;
|
||||||
|
$node->start;
|
||||||
|
|
||||||
|
# Create table and index
|
||||||
|
$node->safe_psql("postgres", "CREATE EXTENSION vector;");
|
||||||
|
$node->safe_psql("postgres", "CREATE TABLE tst (i serial, v sparsevec(100000));");
|
||||||
|
$node->safe_psql("postgres", "CREATE INDEX ON tst USING hnsw (v sparsevec_l2_ops);");
|
||||||
|
|
||||||
|
for (1 .. 3)
|
||||||
|
{
|
||||||
|
for (1 .. 100)
|
||||||
|
{
|
||||||
|
my @elements;
|
||||||
|
my %indices;
|
||||||
|
for (1 .. int(rand() * 100))
|
||||||
|
{
|
||||||
|
my $index = int(rand() * (100000 - 1)) + 1;
|
||||||
|
if (!exists($indices{$index}))
|
||||||
|
{
|
||||||
|
my $value = rand();
|
||||||
|
push(@elements, "$index:$value");
|
||||||
|
$indices{$index} = 1;
|
||||||
|
}
|
||||||
|
}
|
||||||
|
my $embedding = "{" . join(",", @elements) . "}/100000";
|
||||||
|
$node->safe_psql("postgres", "INSERT INTO tst (v) VALUES ('$embedding');");
|
||||||
|
}
|
||||||
|
|
||||||
|
$node->safe_psql("postgres", "DELETE FROM tst WHERE i % 2 = 0;");
|
||||||
|
$node->safe_psql("postgres", "VACUUM tst;");
|
||||||
|
is(1, 1);
|
||||||
|
}
|
||||||
|
|
||||||
|
done_testing();
|
||||||
132
test/t/039_hnsw_minivec_build_recall.pl
Normal file
132
test/t/039_hnsw_minivec_build_recall.pl
Normal file
@@ -0,0 +1,132 @@
|
|||||||
|
use strict;
|
||||||
|
use warnings FATAL => 'all';
|
||||||
|
use PostgreSQL::Test::Cluster;
|
||||||
|
use PostgreSQL::Test::Utils;
|
||||||
|
use Test::More;
|
||||||
|
|
||||||
|
my $node;
|
||||||
|
my @queries = ();
|
||||||
|
my @expected;
|
||||||
|
my $limit = 20;
|
||||||
|
my $dim = 10;
|
||||||
|
my $array_sql = join(",", ('2 * random() * random()') x $dim);
|
||||||
|
|
||||||
|
sub test_recall
|
||||||
|
{
|
||||||
|
my ($min, $operator) = @_;
|
||||||
|
my $correct = 0;
|
||||||
|
my $total = 0;
|
||||||
|
|
||||||
|
my $explain = $node->safe_psql("postgres", qq(
|
||||||
|
SET enable_seqscan = off;
|
||||||
|
EXPLAIN ANALYZE SELECT i FROM tst ORDER BY v $operator '$queries[0]' LIMIT $limit;
|
||||||
|
));
|
||||||
|
like($explain, qr/Index Scan/);
|
||||||
|
|
||||||
|
for my $i (0 .. $#queries)
|
||||||
|
{
|
||||||
|
my $actual = $node->safe_psql("postgres", qq(
|
||||||
|
SET enable_seqscan = off;
|
||||||
|
SELECT i FROM tst ORDER BY v $operator '$queries[$i]' LIMIT $limit;
|
||||||
|
));
|
||||||
|
my @actual_ids = split("\n", $actual);
|
||||||
|
my %actual_set = map { $_ => 1 } @actual_ids;
|
||||||
|
|
||||||
|
my @expected_ids = split("\n", $expected[$i]);
|
||||||
|
|
||||||
|
foreach (@expected_ids)
|
||||||
|
{
|
||||||
|
if (exists($actual_set{$_}))
|
||||||
|
{
|
||||||
|
$correct++;
|
||||||
|
}
|
||||||
|
$total++;
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
cmp_ok($correct / $total, ">=", $min, $operator);
|
||||||
|
}
|
||||||
|
|
||||||
|
# Initialize node
|
||||||
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
|
$node->init;
|
||||||
|
$node->start;
|
||||||
|
|
||||||
|
# Create table
|
||||||
|
$node->safe_psql("postgres", "CREATE EXTENSION vector;");
|
||||||
|
$node->safe_psql("postgres", "CREATE TABLE tst (i int4, v minivec($dim));");
|
||||||
|
$node->safe_psql("postgres",
|
||||||
|
"INSERT INTO tst SELECT i, ARRAY[$array_sql] FROM generate_series(1, 10000) i;"
|
||||||
|
);
|
||||||
|
|
||||||
|
# Generate queries
|
||||||
|
for (1 .. 20)
|
||||||
|
{
|
||||||
|
my @r = ();
|
||||||
|
for (1 .. $dim)
|
||||||
|
{
|
||||||
|
push(@r, rand());
|
||||||
|
}
|
||||||
|
push(@queries, "[" . join(",", @r) . "]");
|
||||||
|
}
|
||||||
|
|
||||||
|
# Check each index type
|
||||||
|
my @operators = ("<->", "<#>", "<=>", "<+>");
|
||||||
|
my @opclasses = ("minivec_l2_ops", "minivec_ip_ops", "minivec_cosine_ops", "minivec_l1_ops");
|
||||||
|
|
||||||
|
for my $i (0 .. $#operators)
|
||||||
|
{
|
||||||
|
my $operator = $operators[$i];
|
||||||
|
my $opclass = $opclasses[$i];
|
||||||
|
|
||||||
|
# Get exact results
|
||||||
|
@expected = ();
|
||||||
|
foreach (@queries)
|
||||||
|
{
|
||||||
|
my $res = $node->safe_psql("postgres", "SELECT i FROM tst ORDER BY v $operator '$_' LIMIT $limit;");
|
||||||
|
push(@expected, $res);
|
||||||
|
}
|
||||||
|
|
||||||
|
# Build index serially
|
||||||
|
$node->safe_psql("postgres", qq(
|
||||||
|
SET max_parallel_maintenance_workers = 0;
|
||||||
|
CREATE INDEX idx ON tst USING hnsw (v $opclass);
|
||||||
|
));
|
||||||
|
|
||||||
|
# Test approximate results
|
||||||
|
my $min = 0.98;
|
||||||
|
test_recall($min, $operator);
|
||||||
|
|
||||||
|
$node->safe_psql("postgres", "DROP INDEX idx;");
|
||||||
|
|
||||||
|
# Build index in parallel in memory
|
||||||
|
my ($ret, $stdout, $stderr) = $node->psql("postgres", qq(
|
||||||
|
SET client_min_messages = DEBUG;
|
||||||
|
SET min_parallel_table_scan_size = 1;
|
||||||
|
CREATE INDEX idx ON tst USING hnsw (v $opclass);
|
||||||
|
));
|
||||||
|
is($ret, 0, $stderr);
|
||||||
|
like($stderr, qr/using \d+ parallel workers/);
|
||||||
|
|
||||||
|
# Test approximate results
|
||||||
|
test_recall($min, $operator);
|
||||||
|
|
||||||
|
$node->safe_psql("postgres", "DROP INDEX idx;");
|
||||||
|
|
||||||
|
# Build index in parallel on disk
|
||||||
|
# Set parallel_workers on table to use workers with low maintenance_work_mem
|
||||||
|
($ret, $stdout, $stderr) = $node->psql("postgres", qq(
|
||||||
|
ALTER TABLE tst SET (parallel_workers = 2);
|
||||||
|
SET client_min_messages = DEBUG;
|
||||||
|
SET maintenance_work_mem = '4MB';
|
||||||
|
CREATE INDEX idx ON tst USING hnsw (v $opclass);
|
||||||
|
ALTER TABLE tst RESET (parallel_workers);
|
||||||
|
));
|
||||||
|
is($ret, 0, $stderr);
|
||||||
|
like($stderr, qr/using \d+ parallel workers/);
|
||||||
|
like($stderr, qr/hnsw graph no longer fits into maintenance_work_mem/);
|
||||||
|
|
||||||
|
$node->safe_psql("postgres", "DROP INDEX idx;");
|
||||||
|
}
|
||||||
|
|
||||||
|
done_testing();
|
||||||
113
test/t/040_hnsw_minivec_insert_recall.pl
Normal file
113
test/t/040_hnsw_minivec_insert_recall.pl
Normal file
@@ -0,0 +1,113 @@
|
|||||||
|
use strict;
|
||||||
|
use warnings FATAL => 'all';
|
||||||
|
use PostgreSQL::Test::Cluster;
|
||||||
|
use PostgreSQL::Test::Utils;
|
||||||
|
use Test::More;
|
||||||
|
|
||||||
|
my $node;
|
||||||
|
my @queries = ();
|
||||||
|
my @expected;
|
||||||
|
my $limit = 20;
|
||||||
|
my $dim = 10;
|
||||||
|
my $array_sql = join(",", ('2 * random() * random()') x $dim);
|
||||||
|
|
||||||
|
sub test_recall
|
||||||
|
{
|
||||||
|
my ($min, $operator) = @_;
|
||||||
|
my $correct = 0;
|
||||||
|
my $total = 0;
|
||||||
|
|
||||||
|
my $explain = $node->safe_psql("postgres", qq(
|
||||||
|
SET enable_seqscan = off;
|
||||||
|
EXPLAIN ANALYZE SELECT i FROM tst ORDER BY v $operator '$queries[0]' LIMIT $limit;
|
||||||
|
));
|
||||||
|
like($explain, qr/Index Scan/);
|
||||||
|
|
||||||
|
for my $i (0 .. $#queries)
|
||||||
|
{
|
||||||
|
my $actual = $node->safe_psql("postgres", qq(
|
||||||
|
SET enable_seqscan = off;
|
||||||
|
SELECT i FROM tst ORDER BY v $operator '$queries[$i]' LIMIT $limit;
|
||||||
|
));
|
||||||
|
my @actual_ids = split("\n", $actual);
|
||||||
|
my %actual_set = map { $_ => 1 } @actual_ids;
|
||||||
|
|
||||||
|
my @expected_ids = split("\n", $expected[$i]);
|
||||||
|
|
||||||
|
foreach (@expected_ids)
|
||||||
|
{
|
||||||
|
if (exists($actual_set{$_}))
|
||||||
|
{
|
||||||
|
$correct++;
|
||||||
|
}
|
||||||
|
$total++;
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
cmp_ok($correct / $total, ">=", $min, $operator);
|
||||||
|
}
|
||||||
|
|
||||||
|
# Initialize node
|
||||||
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
|
$node->init;
|
||||||
|
$node->start;
|
||||||
|
|
||||||
|
# Create table
|
||||||
|
$node->safe_psql("postgres", "CREATE EXTENSION vector;");
|
||||||
|
$node->safe_psql("postgres", "CREATE TABLE tst (i serial, v minivec($dim));");
|
||||||
|
|
||||||
|
# Generate queries
|
||||||
|
for (1 .. 20)
|
||||||
|
{
|
||||||
|
my @r = ();
|
||||||
|
for (1 .. $dim)
|
||||||
|
{
|
||||||
|
push(@r, rand());
|
||||||
|
}
|
||||||
|
push(@queries, "[" . join(",", @r) . "]");
|
||||||
|
}
|
||||||
|
|
||||||
|
# Check each index type
|
||||||
|
my @operators = ("<->", "<#>", "<=>", "<+>");
|
||||||
|
my @opclasses = ("minivec_l2_ops", "minivec_ip_ops", "minivec_cosine_ops", "minivec_l1_ops");
|
||||||
|
|
||||||
|
for my $i (0 .. $#operators)
|
||||||
|
{
|
||||||
|
my $operator = $operators[$i];
|
||||||
|
my $opclass = $opclasses[$i];
|
||||||
|
|
||||||
|
# Add index
|
||||||
|
$node->safe_psql("postgres", "CREATE INDEX idx ON tst USING hnsw (v $opclass);");
|
||||||
|
|
||||||
|
# Use concurrent inserts
|
||||||
|
$node->pgbench(
|
||||||
|
"--no-vacuum --client=10 --transactions=1000",
|
||||||
|
0,
|
||||||
|
[qr{actually processed}],
|
||||||
|
[qr{^$}],
|
||||||
|
"concurrent INSERTs",
|
||||||
|
{
|
||||||
|
"040_hnsw_minivec_insert_recall_$opclass" => "INSERT INTO tst (v) VALUES (ARRAY[$array_sql]);"
|
||||||
|
}
|
||||||
|
);
|
||||||
|
|
||||||
|
# Get exact results
|
||||||
|
@expected = ();
|
||||||
|
foreach (@queries)
|
||||||
|
{
|
||||||
|
my $res = $node->safe_psql("postgres", qq(
|
||||||
|
SET enable_indexscan = off;
|
||||||
|
SELECT i FROM tst ORDER BY v $operator '$_' LIMIT $limit;
|
||||||
|
));
|
||||||
|
push(@expected, $res);
|
||||||
|
}
|
||||||
|
|
||||||
|
# Test approximate results
|
||||||
|
my $min = 0.98;
|
||||||
|
test_recall($min, $operator);
|
||||||
|
|
||||||
|
$node->safe_psql("postgres", "DROP INDEX idx;");
|
||||||
|
$node->safe_psql("postgres", "TRUNCATE tst;");
|
||||||
|
}
|
||||||
|
|
||||||
|
done_testing();
|
||||||
97
test/t/041_hnsw_minivec_vacuum_recall.pl
Normal file
97
test/t/041_hnsw_minivec_vacuum_recall.pl
Normal file
@@ -0,0 +1,97 @@
|
|||||||
|
use strict;
|
||||||
|
use warnings FATAL => 'all';
|
||||||
|
use PostgreSQL::Test::Cluster;
|
||||||
|
use PostgreSQL::Test::Utils;
|
||||||
|
use Test::More;
|
||||||
|
|
||||||
|
my $node;
|
||||||
|
my @queries = ();
|
||||||
|
my @expected;
|
||||||
|
my $limit = 20;
|
||||||
|
|
||||||
|
sub test_recall
|
||||||
|
{
|
||||||
|
my ($min, $ef_search, $test_name) = @_;
|
||||||
|
my $correct = 0;
|
||||||
|
my $total = 0;
|
||||||
|
|
||||||
|
my $explain = $node->safe_psql("postgres", qq(
|
||||||
|
SET enable_seqscan = off;
|
||||||
|
SET hnsw.ef_search = $ef_search;
|
||||||
|
EXPLAIN ANALYZE SELECT i FROM tst ORDER BY v <-> '$queries[0]' LIMIT $limit;
|
||||||
|
));
|
||||||
|
like($explain, qr/Index Scan/);
|
||||||
|
|
||||||
|
for my $i (0 .. $#queries)
|
||||||
|
{
|
||||||
|
my $actual = $node->safe_psql("postgres", qq(
|
||||||
|
SET enable_seqscan = off;
|
||||||
|
SET hnsw.ef_search = $ef_search;
|
||||||
|
SELECT i FROM tst ORDER BY v <-> '$queries[$i]' LIMIT $limit;
|
||||||
|
));
|
||||||
|
my @actual_ids = split("\n", $actual);
|
||||||
|
my %actual_set = map { $_ => 1 } @actual_ids;
|
||||||
|
|
||||||
|
my @expected_ids = split("\n", $expected[$i]);
|
||||||
|
|
||||||
|
foreach (@expected_ids)
|
||||||
|
{
|
||||||
|
if (exists($actual_set{$_}))
|
||||||
|
{
|
||||||
|
$correct++;
|
||||||
|
}
|
||||||
|
$total++;
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
cmp_ok($correct / $total, ">=", $min, $test_name);
|
||||||
|
}
|
||||||
|
|
||||||
|
# Initialize node
|
||||||
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
|
$node->init;
|
||||||
|
$node->start;
|
||||||
|
|
||||||
|
# Create table
|
||||||
|
$node->safe_psql("postgres", "CREATE EXTENSION vector;");
|
||||||
|
$node->safe_psql("postgres", "CREATE TABLE tst (i int4, v minivec(3));");
|
||||||
|
$node->safe_psql("postgres", "ALTER TABLE tst SET (autovacuum_enabled = false);");
|
||||||
|
$node->safe_psql("postgres",
|
||||||
|
"INSERT INTO tst SELECT i, ARRAY[random(), random(), random()] FROM generate_series(1, 10000) i;"
|
||||||
|
);
|
||||||
|
|
||||||
|
# Add index
|
||||||
|
$node->safe_psql("postgres", "CREATE INDEX ON tst USING hnsw (v minivec_l2_ops) WITH (m = 4, ef_construction = 8);");
|
||||||
|
|
||||||
|
# Delete data
|
||||||
|
$node->safe_psql("postgres", "DELETE FROM tst WHERE i > 2500;");
|
||||||
|
|
||||||
|
# Generate queries
|
||||||
|
for (1 .. 20)
|
||||||
|
{
|
||||||
|
my $r1 = rand();
|
||||||
|
my $r2 = rand();
|
||||||
|
my $r3 = rand();
|
||||||
|
push(@queries, "[$r1,$r2,$r3]");
|
||||||
|
}
|
||||||
|
|
||||||
|
# Get exact results
|
||||||
|
@expected = ();
|
||||||
|
foreach (@queries)
|
||||||
|
{
|
||||||
|
my $res = $node->safe_psql("postgres", qq(
|
||||||
|
SET enable_indexscan = off;
|
||||||
|
SELECT i FROM tst ORDER BY v <-> '$_' LIMIT $limit;
|
||||||
|
));
|
||||||
|
push(@expected, $res);
|
||||||
|
}
|
||||||
|
|
||||||
|
test_recall(0.18, $limit, "before vacuum");
|
||||||
|
test_recall(0.93, 100, "before vacuum");
|
||||||
|
|
||||||
|
# TODO Test concurrent inserts with vacuum
|
||||||
|
$node->safe_psql("postgres", "VACUUM tst;");
|
||||||
|
|
||||||
|
test_recall(0.95, $limit, "after vacuum");
|
||||||
|
|
||||||
|
done_testing();
|
||||||
58
test/t/042_hnsw_minivec_duplicates.pl
Normal file
58
test/t/042_hnsw_minivec_duplicates.pl
Normal file
@@ -0,0 +1,58 @@
|
|||||||
|
use strict;
|
||||||
|
use warnings FATAL => 'all';
|
||||||
|
use PostgreSQL::Test::Cluster;
|
||||||
|
use PostgreSQL::Test::Utils;
|
||||||
|
use Test::More;
|
||||||
|
|
||||||
|
# Initialize node
|
||||||
|
my $node = PostgreSQL::Test::Cluster->new('node');
|
||||||
|
$node->init;
|
||||||
|
$node->start;
|
||||||
|
|
||||||
|
# Create table
|
||||||
|
$node->safe_psql("postgres", "CREATE EXTENSION vector;");
|
||||||
|
$node->safe_psql("postgres", "CREATE TABLE tst (v minivec(3));");
|
||||||
|
|
||||||
|
sub insert_vectors
|
||||||
|
{
|
||||||
|
for my $i (1 .. 20)
|
||||||
|
{
|
||||||
|
$node->safe_psql("postgres", "INSERT INTO tst VALUES ('[1,1,1]');");
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
sub test_duplicates
|
||||||
|
{
|
||||||
|
my $res = $node->safe_psql("postgres", qq(
|
||||||
|
SET enable_seqscan = off;
|
||||||
|
SET hnsw.ef_search = 1;
|
||||||
|
SELECT COUNT(*) FROM (SELECT * FROM tst ORDER BY v <-> '[1,1,1]') t;
|
||||||
|
));
|
||||||
|
is($res, 10);
|
||||||
|
}
|
||||||
|
|
||||||
|
# Test duplicates with build
|
||||||
|
insert_vectors();
|
||||||
|
$node->safe_psql("postgres", "CREATE INDEX idx ON tst USING hnsw (v minivec_l2_ops);");
|
||||||
|
test_duplicates();
|
||||||
|
|
||||||
|
# Reset
|
||||||
|
$node->safe_psql("postgres", "TRUNCATE tst;");
|
||||||
|
|
||||||
|
# Test duplicates with inserts
|
||||||
|
insert_vectors();
|
||||||
|
test_duplicates();
|
||||||
|
|
||||||
|
# Test fallback path for inserts
|
||||||
|
$node->pgbench(
|
||||||
|
"--no-vacuum --client=5 --transactions=100",
|
||||||
|
0,
|
||||||
|
[qr{actually processed}],
|
||||||
|
[qr{^$}],
|
||||||
|
"concurrent INSERTs",
|
||||||
|
{
|
||||||
|
"042_hnsw_minivec_duplicates" => "INSERT INTO tst VALUES ('[1,1,1]');"
|
||||||
|
}
|
||||||
|
);
|
||||||
|
|
||||||
|
done_testing();
|
||||||
154
test/t/043_ivfflat_minivec_build_recall.pl
Normal file
154
test/t/043_ivfflat_minivec_build_recall.pl
Normal file
@@ -0,0 +1,154 @@
|
|||||||
|
use strict;
|
||||||
|
use warnings FATAL => 'all';
|
||||||
|
use PostgreSQL::Test::Cluster;
|
||||||
|
use PostgreSQL::Test::Utils;
|
||||||
|
use Test::More;
|
||||||
|
|
||||||
|
my $node;
|
||||||
|
my @queries = ();
|
||||||
|
my @expected;
|
||||||
|
my $limit = 20;
|
||||||
|
my $dim = 10;
|
||||||
|
my $array_sql = join(",", ('random()') x $dim);
|
||||||
|
|
||||||
|
sub test_recall
|
||||||
|
{
|
||||||
|
my ($probes, $min, $operator) = @_;
|
||||||
|
my $correct = 0;
|
||||||
|
my $total = 0;
|
||||||
|
|
||||||
|
my $explain = $node->safe_psql("postgres", qq(
|
||||||
|
SET enable_seqscan = off;
|
||||||
|
SET ivfflat.probes = $probes;
|
||||||
|
EXPLAIN ANALYZE SELECT i FROM tst ORDER BY v $operator '$queries[0]' LIMIT $limit;
|
||||||
|
));
|
||||||
|
like($explain, qr/Index Scan using idx on tst/);
|
||||||
|
|
||||||
|
for my $i (0 .. $#queries)
|
||||||
|
{
|
||||||
|
my $actual = $node->safe_psql("postgres", qq(
|
||||||
|
SET enable_seqscan = off;
|
||||||
|
SET ivfflat.probes = $probes;
|
||||||
|
SELECT i FROM tst ORDER BY v $operator '$queries[$i]' LIMIT $limit;
|
||||||
|
));
|
||||||
|
my @actual_ids = split("\n", $actual);
|
||||||
|
|
||||||
|
my @expected_ids = split("\n", $expected[$i]);
|
||||||
|
my %expected_set = map { $_ => 1 } @expected_ids;
|
||||||
|
|
||||||
|
foreach (@actual_ids)
|
||||||
|
{
|
||||||
|
if (exists($expected_set{$_}))
|
||||||
|
{
|
||||||
|
$correct++;
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
$total += $limit;
|
||||||
|
}
|
||||||
|
|
||||||
|
cmp_ok($correct / $total, ">=", $min, $operator);
|
||||||
|
}
|
||||||
|
|
||||||
|
# Initialize node
|
||||||
|
$node = PostgreSQL::Test::Cluster->new('node');
|
||||||
|
$node->init;
|
||||||
|
$node->start;
|
||||||
|
|
||||||
|
# Create table
|
||||||
|
$node->safe_psql("postgres", "CREATE EXTENSION vector;");
|
||||||
|
$node->safe_psql("postgres", "CREATE TABLE tst (i int4, v minivec($dim));");
|
||||||
|
$node->safe_psql("postgres",
|
||||||
|
"INSERT INTO tst SELECT i, ARRAY[$array_sql] FROM generate_series(1, 100000) i;"
|
||||||
|
);
|
||||||
|
|
||||||
|
# Generate queries
|
||||||
|
for (1 .. 20)
|
||||||
|
{
|
||||||
|
my @r = ();
|
||||||
|
for (1 .. $dim)
|
||||||
|
{
|
||||||
|
push(@r, rand());
|
||||||
|
}
|
||||||
|
push(@queries, "[" . join(",", @r) . "]");
|
||||||
|
}
|
||||||
|
|
||||||
|
# Check each index type
|
||||||
|
my @operators = ("<->", "<#>", "<=>");
|
||||||
|
my @opclasses = ("minivec_l2_ops", "minivec_ip_ops", "minivec_cosine_ops");
|
||||||
|
|
||||||
|
for my $i (0 .. $#operators)
|
||||||
|
{
|
||||||
|
my $operator = $operators[$i];
|
||||||
|
my $opclass = $opclasses[$i];
|
||||||
|
|
||||||
|
# Get exact results
|
||||||
|
@expected = ();
|
||||||
|
foreach (@queries)
|
||||||
|
{
|
||||||
|
my $res = $node->safe_psql("postgres", qq(
|
||||||
|
WITH top AS (
|
||||||
|
SELECT v $operator '$_' AS distance FROM tst ORDER BY distance LIMIT $limit
|
||||||
|
)
|
||||||
|
SELECT i FROM tst WHERE (v $operator '$_') <= (SELECT MAX(distance) FROM top)
|
||||||
|
));
|
||||||
|
push(@expected, $res);
|
||||||
|
}
|
||||||
|
|
||||||
|
# Build index serially
|
||||||
|
$node->safe_psql("postgres", qq(
|
||||||
|
SET max_parallel_maintenance_workers = 0;
|
||||||
|
CREATE INDEX idx ON tst USING ivfflat (v $opclass);
|
||||||
|
));
|
||||||
|
|
||||||
|
# Test approximate results
|
||||||
|
if ($operator ne "<#>")
|
||||||
|
{
|
||||||
|
# TODO Fix test (uniform random vectors all have similar inner product)
|
||||||
|
test_recall(1, 0.33, $operator);
|
||||||
|
test_recall(10, 0.93, $operator);
|
||||||
|
}
|
||||||
|
|
||||||
|
# Test probes equals lists
|
||||||
|
if ($operator eq "<=>")
|
||||||
|
{
|
||||||
|
test_recall(100, 0.98, $operator);
|
||||||
|
}
|
||||||
|
else
|
||||||
|
{
|
||||||
|
test_recall(100, 1.00, $operator);
|
||||||
|
}
|
||||||
|
|
||||||
|
$node->safe_psql("postgres", "DROP INDEX idx;");
|
||||||
|
|
||||||
|
# Build index in parallel
|
||||||
|
my ($ret, $stdout, $stderr) = $node->psql("postgres", qq(
|
||||||
|
SET client_min_messages = DEBUG;
|
||||||
|
SET min_parallel_table_scan_size = 1;
|
||||||
|
CREATE INDEX idx ON tst USING ivfflat (v $opclass);
|
||||||
|
));
|
||||||
|
is($ret, 0, $stderr);
|
||||||
|
like($stderr, qr/using \d+ parallel workers/);
|
||||||
|
|
||||||
|
# Test approximate results
|
||||||
|
if ($operator ne "<#>")
|
||||||
|
{
|
||||||
|
# TODO Fix test (uniform random vectors all have similar inner product)
|
||||||
|
test_recall(1, 0.33, $operator);
|
||||||
|
test_recall(10, 0.93, $operator);
|
||||||
|
}
|
||||||
|
|
||||||
|
# Test probes equals lists
|
||||||
|
if ($operator eq "<=>")
|
||||||
|
{
|
||||||
|
test_recall(100, 0.98, $operator);
|
||||||
|
}
|
||||||
|
else
|
||||||
|
{
|
||||||
|
test_recall(100, 1.00, $operator);
|
||||||
|
}
|
||||||
|
|
||||||
|
$node->safe_psql("postgres", "DROP INDEX idx;");
|
||||||
|
}
|
||||||
|
|
||||||
|
done_testing();
|
||||||
@@ -1,4 +1,4 @@
|
|||||||
comment = 'vector data type and ivfflat and hnsw access methods'
|
comment = 'vector data type and ivfflat and hnsw access methods'
|
||||||
default_version = '0.7.2'
|
default_version = '0.7.4'
|
||||||
module_pathname = '$libdir/vector'
|
module_pathname = '$libdir/vector'
|
||||||
relocatable = true
|
relocatable = true
|
||||||
|
|||||||
Reference in New Issue
Block a user