DuckDB向量存储

🌐 DuckDBVector Store

DuckDB 的存储实现提供了一种嵌入式高性能向量搜索解决方案，使用 DuckDB —— 一种进程内分析型数据库。它使用 VSS 扩展进行向量相似性搜索，并采用 HNSW 索引，提供轻量且高效的向量数据库，无需外部服务器。

🌐 The DuckDB storage implementation provides an embedded high-performance vector search solution using DuckDB, an in-process analytical database. It uses the VSS extension for vector similarity search with HNSW indexing, offering a lightweight and efficient vector database that requires no external server.

它是 @mastra/duckdb 包的一部分，提供带有元数据过滤的高效向量相似度搜索。

🌐 It's part of the @mastra/duckdb package and offers efficient vector similarity search with metadata filtering.

安装

🌐 Installation

npm

npm install @mastra/duckdb@latest

pnpm

pnpm add @mastra/duckdb@latest

Yarn

yarn add @mastra/duckdb@latest

Bun

bun add @mastra/duckdb@latest

用法

🌐 Usage

import { DuckDBVector } from "@mastra/duckdb";

// Create a new vector store instance
const store = new DuckDBVector({
  id: "duckdb-vector",
  path: ":memory:", // or './vectors.duckdb' for file persistence
});

// Create an index
await store.createIndex({
  indexName: "myCollection",
  dimension: 1536,
  metric: "cosine",
});

// Add vectors with metadata
const vectors = [[0.1, 0.2, ...], [0.3, 0.4, ...]];
const metadata = [
  { text: "first document", category: "A" },
  { text: "second document", category: "B" },
];
await store.upsert({
  indexName: "myCollection",
  vectors,
  metadata,
});

// Query similar vectors
const queryVector = [0.1, 0.2, ...];
const results = await store.query({
  indexName: "myCollection",
  queryVector,
  topK: 10,
  filter: { category: "A" },
});

// Clean up
await store.close();

构造函数选项

🌐 Constructor Options

id:

string

Unique identifier for the vector store instance

path?:

string

= ':memory:'

Database file path. Use ':memory:' for in-memory database, or a file path like './vectors.duckdb' for persistence.

dimensions?:

number

= 1536

Default dimension for vector embeddings

metric?:

'cosine' | 'euclidean' | 'dotproduct'

= cosine

Default distance metric for similarity search

方法

🌐 Methods

createIndex()

创建一个新的向量集合，并可选择使用 HNSW 索引以实现快速近似最近邻搜索。

🌐 Creates a new vector collection with optional HNSW index for fast approximate nearest neighbor search.

indexName:

string

Name of the index to create

dimension:

number

Vector dimension size (must match your embedding model)

metric?:

'cosine' | 'euclidean' | 'dotproduct'

= cosine

Distance metric for similarity search

upsert()

在索引中添加或更新向量及其元数据。

🌐 Adds or updates vectors and their metadata in the index.

indexName:

string

Name of the index to insert into

vectors:

number[][]

Array of embedding vectors

metadata?:

Record<string, any>[]

Metadata for each vector

ids?:

string[]

Optional vector IDs (auto-generated UUIDs if not provided)

query()

搜索具有可选元数据过滤的相似向量。

🌐 Searches for similar vectors with optional metadata filtering.

indexName:

string

Name of the index to search in

queryVector:

number[]

Query vector to find similar vectors for

topK?:

number

= 10

Number of results to return

filter?:

Filter

Metadata filters using MongoDB-like query syntax

includeVector?:

boolean

= false

Whether to include vector data in results

describeIndex()

获取有关索引的信息。

🌐 Gets information about an index.

indexName:

string

Name of the index to describe

返回：

🌐 Returns:

interface IndexStats {
  dimension: number;
  count: number;
  metric: "cosine" | "euclidean" | "dotproduct";
}

deleteIndex()

删除索引及其所有数据。

🌐 Deletes an index and all its data.

indexName:

string

Name of the index to delete

listIndexes()

列出数据库中所有的向量索引。

🌐 Lists all vector indexes in the database.

返回：Promise<string[]>

🌐 Returns: Promise<string[]>

updateVector()

通过 ID 或元数据过滤器更新单个向量。必须提供 id 或 filter 中的一个，但不能同时提供两者。

🌐 Update a single vector by ID or by metadata filter. Either id or filter must be provided, but not both.

indexName:

string

Name of the index containing the vector

id?:

string

ID of the vector entry to update (mutually exclusive with filter)

filter?:

Record<string, any>

Metadata filter to identify vector(s) to update (mutually exclusive with id)

update:

object

Update data containing vector and/or metadata

update.vector?:

number[]

New vector data to update

update.metadata?:

Record<string, any>

New metadata to update

deleteVector()

通过其 ID 从索引中删除特定的向量条目。

🌐 Deletes a specific vector entry from an index by its ID.

indexName:

string

Name of the index containing the vector

id:

string

ID of the vector entry to delete

deleteVectors()

通过 ID 或元数据过滤器删除多个向量。必须提供 ids 或 filter 中的一个，但不能同时提供两者。

🌐 Delete multiple vectors by IDs or by metadata filter. Either ids or filter must be provided, but not both.

indexName:

string

Name of the index containing the vectors to delete

ids?:

string[]

Array of vector IDs to delete (mutually exclusive with filter)

filter?:

Record<string, any>

Metadata filter to identify vectors to delete (mutually exclusive with ids)

close()

关闭数据库连接并释放资源。

🌐 Closes the database connection and releases resources.

await store.close();

响应类型

🌐 Response Types

查询结果以此格式返回：

🌐 Query results are returned in this format:

interface QueryResult {
  id: string;
  score: number;
  metadata: Record<string, any>;
  vector?: number[]; // Only included if includeVector is true
}

筛选运算符

🌐 Filter Operators

DuckDB 向量存储支持类似 MongoDB 的过滤操作符：

🌐 DuckDB vector store supports MongoDB-like filter operators:

类别	运算符
比较	$eq, $ne, $gt, $gte, $lt, $lte
逻辑	$and, $or, $not, $nor
数组	$in, $nin
元素	$exists
文本	$contains

筛选示例

🌐 Filter Examples

// Allegato operators
const results = await store.query({
  indexName: "docs",
  queryVector: [...],
  filter: {
    $and: [
      { category: "electronics" },
      { price: { $gte: 100, $lte: 500 } },
    ],
  },
});

// Nested field access
const results = await store.query({
  indexName: "docs",
  queryVector: [...],
  filter: { "user.profile.tier": "premium" },
});

距离度量

🌐 Distance Metrics

指标	描述	分数解释	适用场景
cosine	余弦相似度	0-1（1 = 最相似）	文本向量、归一化向量
euclidean	L2 距离	0-∞（0 = 最相似）	图片向量、空间数据
dotproduct	内积	越大 = 越相似	当向量大小很重要时

错误处理

🌐 Error Handling

针对不同的失败情况，存储会抛出特定的错误：

🌐 The store throws specific errors for different failure cases:

try {
  await store.query({
    indexName: "my-collection",
    queryVector: queryVector,
  });
} catch (error) {
  if (error.message.includes("not found")) {
    console.error("The specified index does not exist");
  } else if (error.message.includes("Invalid identifier")) {
    console.error("Index name contains invalid characters");
  } else {
    console.error("Vector store error:", error.message);
  }
}

常见错误情况包括：

🌐 Common error cases include:

• 索引名称格式无效
• 未找到索引/表
• 查询向量与索引的维度不匹配
• 删除/更新操作中的过滤器或 ID 数组为空
• 互斥冲突（同时提供 id 和 filter）

用例

🌐 Use Cases

嵌入式语义搜索

🌐 Embedded Semantic Search

构建具有语义搜索功能的离线 AI 应用，该搜索完全在进程内运行：

🌐 Build offline-capable AI applications with semantic search that runs entirely in-process:

const store = new DuckDBVector({
  id: "offline-search",
  path: "./search.duckdb",
});

本地 RAG 流水线

🌐 Local RAG Pipelines

在本地处理敏感文档，无需将数据发送到云端向量数据库：

🌐 Process sensitive documents locally without sending data to cloud vector databases:

const store = new DuckDBVector({
  id: "private-rag",
  path: "./confidential.duckdb",
  dimensions: 1536,
});

开发与测试

🌐 Development and Testing

无需基础设施即可快速原型化向量搜索功能：

🌐 Rapidly prototype vector search features with zero infrastructure:

const store = new DuckDBVector({
  id: "dev-store",
  path: ":memory:", // Fast in-memory for tests
});

相关

🌐 Related

• 元数据过滤器
• DuckDB 文档