过滤搜索优化
使用部分索引和复合索引策略，优化同时结合向量相似度和元数据过滤条件的 pgvector 查询。
过滤搜索优化 是 CoddyKit 上的免费 Vector Databases: Pinecone, Weaviate & pgvector 课时。 这是第 4 节课，共 4 节。 你可以在下方免费阅读本课时的完整内容 — 然后在浏览器中使用内置代码编辑器和全天候 AI 导师进行实践。 这是 Vector Databases: Pinecone, Weaviate & pgvector 学习路径的一部分，你的进度在网页和 CoddyKit 应用中同步。 Vector Databases: Pinecone, Weaviate & pgvector 课程共包含 4 节课。
本课时的部分内容尚未翻译，以英文显示。
The Filtered Search Problem
Real queries often combine a vector search with a metadata WHERE filter. Naive combinations can defeat the vector index and become slow.
SELECT id FROM docs
WHERE category = 'news'
ORDER BY embedding <=> '[...]'::vector
LIMIT 10;Why Filters Hurt Recall
ANN indexes return approximate nearest neighbors before the filter applies. If most candidates are filtered out, you may get fewer than LIMIT results — this is over-filtering.
Increasing Search Scope
Raise hnsw.ef_search (or IVFFlat probes) so the index scans more candidates before filtering, improving recall on selective filters.
SET hnsw.ef_search = 100;Partial Indexes
For a common, low-cardinality filter value, build a partial index covering only those rows. The index then perfectly matches the filtered subset.
CREATE INDEX ON docs
USING hnsw (embedding vector_cosine_ops)
WHERE category = 'news';When Partial Indexes Win
Partial indexes shine when:
- Filter values are few and known
- Each subset is large enough to matter
- Queries almost always include that filter
B-tree Support Indexes
For high-cardinality filters, add a regular B-tree index on the metadata column. The planner can combine it with the vector scan.
CREATE INDEX ON docs (category);
CREATE INDEX ON docs (published_at);Iterative Index Scans
Newer pgvector supports iterative scans, which keep fetching from the index until enough rows pass the filter. Enable it to avoid under-filling results.
SET hnsw.iterative_scan = 'relaxed_order';Pre-filtering vs Post-filtering
Pre-filter: narrow rows first, then vector search (good for very selective filters). Post-filter: vector search first, then filter (good for broad filters). Test both.
Measuring with EXPLAIN
Always confirm the plan. Look for Index Scan using ... hnsw rather than a sequential scan.
EXPLAIN ANALYZE
SELECT id FROM docs
WHERE category = 'news'
ORDER BY embedding <=> '[...]'::vector
LIMIT 10;Combining Strategies
A robust setup often uses partial indexes for hot filters, B-tree indexes for the rest, and a tuned ef_search with iterative scans as a safety net.
Checklist
Before shipping a filtered vector query:
- Verify index usage with EXPLAIN
- Confirm you reliably get LIMIT results
- Benchmark latency at p95
Quick Check
Test filtered search knowledge.
Recap
You learned why filters challenge ANN indexes and how partial indexes, B-tree support indexes, ef_search tuning, and iterative scans keep filtered vector queries both fast and accurate.
常见问题解答
「过滤搜索优化」课时是免费的吗？
是的 — 「过滤搜索优化」的完整文本可在网页上免费阅读。要进行交互式练习（内置代码编辑器和全天候 AI 导师）并解锁 Vector Databases: Pinecone, Weaviate & pgvector 课程的其余内容，请升级到 CoddyKit PRO。 Vector Databases: Pinecone, Weaviate & pgvector 课程共包含 4 节课。
「过滤搜索优化」这节课中我会学到什么？
使用部分索引和复合索引策略，优化同时结合向量相似度和元数据过滤条件的 pgvector 查询。 你通过在浏览器中直接运行的动手代码来练习 Vector Databases: Pinecone, Weaviate & pgvector，全天候 AI 导师会在你学习这节课的过程中回答你的问题。
学习 Vector Databases: Pinecone, Weaviate & pgvector 需要有经验吗？
无需任何先前经验。CoddyKit 上的 Vector Databases: Pinecone, Weaviate & pgvector 课程适合初学者到高级学习者，你可以从这里开始或从头开始，按照自己的节奏学习。 这是第 4 节课，共 4 节。
「过滤搜索优化」课时需要多长时间？
大多数 CoddyKit 课程大约需要 5–10 分钟。每节课都很精短且互动，所以你能稳步进步，并在网页和应用中从离开的地方继续。
我能在这节 Vector Databases: Pinecone, Weaviate & pgvector 课中编写并运行代码吗？
能。每节 Vector Databases: Pinecone, Weaviate & pgvector 课都包含内置代码编辑器，你可以在浏览器中直接编写并运行真实代码，并获得即时 AI 反馈 — 无需本地设置。