Memgraph属性图索引¶

Memgraph是一个开源的图数据库，专为实时流处理和快速分析存储数据而构建。

在运行Memgraph之前，请确保Docker已在后台运行。首次试用Memgraph Platform（Memgraph数据库 + MAGE库 + Memgraph Lab）的最快方法是运行以下命令

适用于Linux/macOS

curl https://install.memgraph.com | sh

适用于Windows

iwr https://windows.memgraph.com | iex

从这里，您可以在http://localhost:3000/上查看Memgraph的可视化工具Memgraph Lab，或者使用该应用的桌面版本。

In [ ]

%pip install llama-index llama-index-graph-stores-memgraph

%pip install llama-index llama-index-graph-stores-memgraph

环境设置¶

In [ ]

import os

os.environ[
    "OPENAI_API_KEY"
] = "sk-proj-..."  # Replace with your OpenAI API key

import os os.environ[ "OPENAI_API_KEY" ] = "sk-proj-..." # 替换为您的OpenAI API密钥

创建数据目录并下载我们将在此示例中用作输入数据的Paul Graham文章。

In [ ]

import urllib.request

os.makedirs("data/paul_graham/", exist_ok=True)

url = "https://raw.githubusercontent.com/run-llama/llama_index/main/docs/docs/examples/data/paul_graham/paul_graham_essay.txt"
output_path = "data/paul_graham/paul_graham_essay.txt"
urllib.request.urlretrieve(url, output_path)

import urllib.request os.makedirs("data/paul_graham/", exist_ok=True) url = "https://raw.githubusercontent.com/run-llama/llama_index/main/docs/docs/examples/data/paul_graham/paul_graham_essay.txt" output_path = "data/paul_graham/paul_graham_essay.txt" urllib.request.urlretrieve(url, output_path)

In [ ]

import nest_asyncio

nest_asyncio.apply()

import nest_asyncio nest_asyncio.apply()

读取文件，替换单引号，保存修改后的内容，并使用SimpleDirectoryReader加载文档数据

In [ ]

from llama_index.core import SimpleDirectoryReader

with open(output_path, "r", encoding="utf-8") as file:
    content = file.read()

with open(output_path, "w", encoding="utf-8") as file:
    file.write(content)

documents = SimpleDirectoryReader("./data/paul_graham/").load_data()

from llama_index.core import SimpleDirectoryReader with open(output_path, "r", encoding="utf-8") as file: content = file.read() with open(output_path, "w", encoding="utf-8") as file: file.write(content) documents = SimpleDirectoryReader("./data/paul_graham/").load_data()

设置Memgraph连接¶

通过提供数据库凭据来设置您的图存储类。

In [ ]

from llama_index.graph_stores.memgraph import MemgraphPropertyGraphStore

username = ""  # Enter your Memgraph username (default "")
password = ""  # Enter your Memgraph password (default "")
url = ""  # Specify the connection URL, e.g., 'bolt://localhost:7687'

graph_store = MemgraphPropertyGraphStore(
    username=username,
    password=password,
    url=url,
)

from llama_index.graph_stores.memgraph import MemgraphPropertyGraphStore username = "" # 输入您的Memgraph用户名（默认为""） password = "" # 输入您的Memgraph密码（默认为""） url = "" # 指定连接URL，例如 'bolt://localhost:7687' graph_store = MemgraphPropertyGraphStore( username=username, password=password, url=url, )

索引构建¶

In [ ]

from llama_index.core import PropertyGraphIndex
from llama_index.embeddings.openai import OpenAIEmbedding
from llama_index.llms.openai import OpenAI
from llama_index.core.indices.property_graph import SchemaLLMPathExtractor

index = PropertyGraphIndex.from_documents(
    documents,
    embed_model=OpenAIEmbedding(model_name="text-embedding-ada-002"),
    kg_extractors=[
        SchemaLLMPathExtractor(
            llm=OpenAI(model="gpt-3.5-turbo", temperature=0.0)
        )
    ],
    property_graph_store=graph_store,
    show_progress=True,
)

from llama_index.core import PropertyGraphIndex from llama_index.embeddings.openai import OpenAIEmbedding from llama_index.llms.openai import OpenAI from llama_index.core.indices.property_graph import SchemaLLMPathExtractor index = PropertyGraphIndex.from_documents( documents, embed_model=OpenAIEmbedding(model_name="text-embedding-ada-002"), kg_extractors=[ SchemaLLMPathExtractor( llm=OpenAI(model="gpt-3.5-turbo", temperature=0.0) ) ], property_graph_store=graph_store, show_progress=True, )

图创建完成后，我们可以访问http://localhost:3000/在UI中探索它。

可视化整个图的最简单方法是运行类似于以下的Cypher命令

MATCH p=()-[]-() RETURN p;

此命令匹配图中的所有可能路径并返回整个图。

要可视化图的架构，请访问“图架构”标签页，并根据新创建的图生成新的架构。

要删除整个图，请使用

MATCH (n) DETACH DELETE n;

查询与检索¶

In [ ]

retriever = index.as_retriever(include_text=False)

# Example query: "What happened at Interleaf and Viaweb?"
nodes = retriever.retrieve("What happened at Interleaf and Viaweb?")

# Output results
print("Query Results:")
for node in nodes:
    print(node.text)

# Alternatively, using a query engine
query_engine = index.as_query_engine(include_text=True)

# Perform a query and print the detailed response
response = query_engine.query("What happened at Interleaf and Viaweb?")
print("\nDetailed Query Response:")
print(str(response))

retriever = index.as_retriever(include_text=False) # 示例查询："What happened at Interleaf and Viaweb?" nodes = retriever.retrieve("What happened at Interleaf and Viaweb?") # 输出结果 print("Query Results:") for node in nodes: print(node.text) # 或者，使用查询引擎 query_engine = index.as_query_engine(include_text=True) # 执行查询并打印详细响应 response = query_engine.query("What happened at Interleaf and Viaweb?") print("\nDetailed Query Response:") print(str(response))

从现有图加载¶

如果您有一个现有图（无论是使用LlamaIndex创建的还是其他方式创建的），我们可以连接并使用它！

注意： 如果您的图是在LlamaIndex之外创建的，最实用的检索器将是文本到Cypher或Cypher模板。其他检索器依赖于LlamaIndex插入的属性。

In [ ]

llm = OpenAI(model="gpt-4", temperature=0.0)
kg_extractors = [SchemaLLMPathExtractor(llm=llm)]

index = PropertyGraphIndex.from_existing(
    property_graph_store=graph_store,
    kg_extractors=kg_extractors,
    embed_model=OpenAIEmbedding(model_name="text-embedding-ada-002"),
    show_progress=True,
)

llm = OpenAI(model="gpt-4", temperature=0.0) kg_extractors = [SchemaLLMPathExtractor(llm=llm)] index = PropertyGraphIndex.from_existing( property_graph_store=graph_store, kg_extractors=kg_extractors, embed_model=OpenAIEmbedding(model_name="text-embedding-ada-002"), show_progress=True, )