> ## Documentation Index
> Fetch the complete documentation index at: https://langchain-zh.cn/llms.txt
> Use this file to discover all available pages before exploring further.

# Amazon MemoryDB 集成

> 使用 LangChain Python 与 Amazon MemoryDB 向量存储进行集成。

> [向量搜索](https://docs.aws.amazon.com/memorydb/latest/devguide/vector-search.html/) 介绍和 LangChain 集成指南。

## 什么是 Amazon MemoryDB？

MemoryDB 兼容 Redis OSS，这是一个流行的开源数据存储，使您能够使用他们今天已经使用的相同灵活且友好的 Redis OSS 数据结构、API 和命令快速构建应用程序。使用 MemoryDB，您的所有数据都存储在内存中，这使您能够实现微秒级读取和个位数毫秒写入延迟以及高吞吐量。MemoryDB 还使用多可用区事务日志跨多个可用区（AZs）持久化存储数据，以实现快速故障转移、数据库恢复和节点重启。

## MemoryDB 的向量搜索

MemoryDB 的向量搜索扩展了 MemoryDB 的功能。向量搜索可与现有的 MemoryDB 功能结合使用。不使用向量搜索的应用程序不受其存在的影响。向量搜索在 MemoryDB 可用的所有区域中均可用。您可以使用现有的 MemoryDB 数据或 Redis OSS API 构建机器学习和生成式 AI 用例，例如检索增强生成、异常检测、文档检索和实时推荐。

* Redis 哈希和 `JSON` 中的多字段索引
* 向量相似度搜索（使用 `HNSW` (ANN) 或 `FLAT` (KNN)）
* 向量范围搜索（例如，查找查询向量半径内的所有向量）
* 无性能损失的增量索引

## 设置

### 安装 Redis Python 客户端

`Redis-py` 是一个可用于连接到 MemoryDB 的 Python 客户端

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
pip install -qU  redis langchain-aws
```

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
from langchain_aws.embeddings import BedrockEmbeddings

embeddings = BedrockEmbeddings()
```

### MemoryDB 连接

有效的 Redis URL 模式如下：

1. `redis://`  - 连接到 Redis 集群，未加密
2. `rediss://` - 连接到 Redis 集群，带有 TLS 加密

有关其他连接参数的更多信息，请参见 [redis-py 文档](https://redis-py.readthedocs.io/en/stable/connections.html)。

### 示例数据

首先，我们将描述一些示例数据，以便展示 Redis 向量存储的各种属性。

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
metadata = [
    {
        "user": "john",
        "age": 18,
        "job": "engineer",
        "credit_score": "high",
    },
    {
        "user": "derrick",
        "age": 45,
        "job": "doctor",
        "credit_score": "low",
    },
    {
        "user": "nancy",
        "age": 94,
        "job": "doctor",
        "credit_score": "high",
    },
    {
        "user": "tyler",
        "age": 100,
        "job": "engineer",
        "credit_score": "high",
    },
    {
        "user": "joe",
        "age": 35,
        "job": "dentist",
        "credit_score": "medium",
    },
]
texts = ["foo", "foo", "foo", "bar", "bar"]
index_name = "users"
```

### 创建 MemoryDB 向量存储

InMemoryVectorStore 实例可以使用以下方法初始化

* `InMemoryVectorStore.__init__` - 直接初始化
* `InMemoryVectorStore.from_documents` - 从 `LangChain.docstore.Document` 对象列表初始化
* `InMemoryVectorStore.from_texts` - 从文本列表初始化（可选带元数据）
* `InMemoryVectorStore.from_existing_index` - 从现有 MemoryDB 索引初始化

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
from langchain_aws.vectorstores.inmemorydb import InMemoryVectorStore

vds = InMemoryVectorStore.from_texts(
    embeddings,
    redis_url="rediss://cluster_endpoint:6379/ssl=True ssl_cert_reqs=none",
)
```

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
vds.index_name
```

```text theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
'users'
```

## 查询

根据您的用例，有多种方式可以查询 `InMemoryVectorStore` 实现：

* `similarity_search`: 查找与给定向量最相似的向量。
* `similarity_search_with_score`: 查找与给定向量最相似的向量并返回向量距离
* `similarity_search_limit_score`: 查找与给定向量最相似的向量并将结果数量限制为 `score_threshold`
* `similarity_search_with_relevance_scores`: 查找与给定向量最相似的向量并返回向量相似度
* `max_marginal_relevance_search`: 查找与给定向量最相似的向量，同时优化多样性

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
results = vds.similarity_search("foo")
print(results[0].page_content)
```

```text theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
foo
```

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
# with scores (distances)
results = vds.similarity_search_with_score("foo", k=5)
for result in results:
    print(f"Content: {result[0].page_content} --- Score: {result[1]}")
```

```text theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
Content: foo --- Score: 0.0
Content: foo --- Score: 0.0
Content: foo --- Score: 0.0
Content: bar --- Score: 0.1566
Content: bar --- Score: 0.1566
```

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
# limit the vector distance that can be returned
results = vds.similarity_search_with_score("foo", k=5, distance_threshold=0.1)
for result in results:
    print(f"Content: {result[0].page_content} --- Score: {result[1]}")
```

```text theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
Content: foo --- Score: 0.0
Content: foo --- Score: 0.0
Content: foo --- Score: 0.0
```

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
# with scores
results = vds.similarity_search_with_relevance_scores("foo", k=5)
for result in results:
    print(f"Content: {result[0].page_content} --- Similarity: {result[1]}")
```

```text theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
Content: foo --- Similarity: 1.0
Content: foo --- Similarity: 1.0
Content: foo --- Similarity: 1.0
Content: bar --- Similarity: 0.8434
Content: bar --- Similarity: 0.8434
```

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
# you can also add new documents as follows
new_document = ["baz"]
new_metadata = [{"user": "sam", "age": 50, "job": "janitor", "credit_score": "high"}]
# both the document and metadata must be lists
vds.add_texts(new_document, new_metadata)
```

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
['doc:users:b9c71d62a0a34241a37950b448dafd38']
```

## 将 MemoryDB 用作检索器

这里我们概述了将向量存储用作检索器的不同选项。

我们可以使用三种不同的搜索方法来执行检索。默认情况下，它将使用语义相似度。

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
query = "foo"
results = vds.similarity_search_with_score(query, k=3, return_metadata=True)

for result in results:
    print("Content:", result[0].page_content, " --- Score: ", result[1])
```

```text theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
Content: foo  --- Score:  0.0
Content: foo  --- Score:  0.0
Content: foo  --- Score:  0.0
```

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
retriever = vds.as_retriever(search_type="similarity", search_kwargs={"k": 4})
```

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
docs = retriever.invoke(query)
docs
```

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
[Document(page_content='foo', metadata={'id': 'doc:users_modified:988ecca7574048e396756efc0e79aeca', 'user': 'john', 'job': 'engineer', 'credit_score': 'high', 'age': '18'}),
 Document(page_content='foo', metadata={'id': 'doc:users_modified:009b1afeb4084cc6bdef858c7a99b48e', 'user': 'derrick', 'job': 'doctor', 'credit_score': 'low', 'age': '45'}),
 Document(page_content='foo', metadata={'id': 'doc:users_modified:7087cee9be5b4eca93c30fbdd09a2731', 'user': 'nancy', 'job': 'doctor', 'credit_score': 'high', 'age': '94'}),
 Document(page_content='bar', metadata={'id': 'doc:users_modified:01ef6caac12b42c28ad870aefe574253', 'user': 'tyler', 'job': 'engineer', 'credit_score': 'high', 'age': '100'})]
```

还有 `similarity_distance_threshold` 检索器，允许用户指定向量距离

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
retriever = vds.as_retriever(
    search_type="similarity_distance_threshold",
    search_kwargs={"k": 4, "distance_threshold": 0.1},
)
```

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
docs = retriever.invoke(query)
docs
```

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
[Document(page_content='foo', metadata={'id': 'doc:users_modified:988ecca7574048e396756efc0e79aeca', 'user': 'john', 'job': 'engineer', 'credit_score': 'high', 'age': '18'}),
 Document(page_content='foo', metadata={'id': 'doc:users_modified:009b1afeb4084cc6bdef858c7a99b48e', 'user': 'derrick', 'job': 'doctor', 'credit_score': 'low', 'age': '45'}),
 Document(page_content='foo', metadata={'id': 'doc:users_modified:7087cee9be5b4eca93c30fbdd09a2731', 'user': 'nancy', 'job': 'doctor', 'credit_score': 'high', 'age': '94'})]
```

最后，`similarity_score_threshold` 允许用户定义相似文档的最小分数

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
retriever = vds.as_retriever(
    search_type="similarity_score_threshold",
    search_kwargs={"score_threshold": 0.9, "k": 10},
)
```

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
retriever.invoke("foo")
```

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
[Document(page_content='foo', metadata={'id': 'doc:users_modified:988ecca7574048e396756efc0e79aeca', 'user': 'john', 'job': 'engineer', 'credit_score': 'high', 'age': '18'}),
 Document(page_content='foo', metadata={'id': 'doc:users_modified:009b1afeb4084cc6bdef858c7a99b48e', 'user': 'derrick', 'job': 'doctor', 'credit_score': 'low', 'age': '45'}),
 Document(page_content='foo', metadata={'id': 'doc:users_modified:7087cee9be5b4eca93c30fbdd09a2731', 'user': 'nancy', 'job': 'doctor', 'credit_score': 'high', 'age': '94'})]
```

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
retriever.invoke("foo")
```

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
[Document(page_content='foo', metadata={'id': 'doc:users:8f6b673b390647809d510112cde01a27', 'user': 'john', 'job': 'engineer', 'credit_score': 'high', 'age': '18'}),
 Document(page_content='bar', metadata={'id': 'doc:users:93521560735d42328b48c9c6f6418d6a', 'user': 'tyler', 'job': 'engineer', 'credit_score': 'high', 'age': '100'}),
 Document(page_content='foo', metadata={'id': 'doc:users:125ecd39d07845eabf1a699d44134a5b', 'user': 'nancy', 'job': 'doctor', 'credit_score': 'high', 'age': '94'}),
 Document(page_content='foo', metadata={'id': 'doc:users:d6200ab3764c466082fde3eaab972a2a', 'user': 'derrick', 'job': 'doctor', 'credit_score': 'low', 'age': '45'})]
```

## 删除索引

要删除您的条目，必须通过它们的键来寻址它们。

```python theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
# delete the indices too
InMemoryVectorStore.drop_index(
    index_name="users", delete_documents=True, redis_url="redis://localhost:6379"
)
InMemoryVectorStore.drop_index(
    index_name="users_modified",
    delete_documents=True,
    redis_url="redis://localhost:6379",
)
```

```text theme={"theme":{"light":"catppuccin-latte","dark":"catppuccin-mocha"}}
True
```

***

<div className="source-links">
  <Callout icon="edit">
    [Edit this page on GitHub](https://github.com/langchain-ai/docs/edit/main/src/i18n\zh-CN\oss\python\integrations\vectorstores\memorydb.mdx) or [file an issue](https://github.com/langchain-ai/docs/issues/new/choose).
  </Callout>

  <Callout icon="terminal-2">
    [Connect these docs](/use-these-docs) to Claude, VSCode, and more via MCP for real-time answers.
  </Callout>
</div>