chroma 向量库使用；查询、更新与插入

参考:
https://docs.trychroma.com/usage-guide
https://blog.csdn.net/weixin_46515328/article/details/131855650

chroma 支持保存文档、向量，和向量的ids值

##安装
pip install chromadb

1、使用

import chromadb

##这句可以数据库默认时刻保存，启动自动加载
chroma_client = chromadb.PersistentClient(path=".")
# chroma_client = chromadb.Client()

#创建Chroma数据集
collection = chroma_client.create_collection(name="my_collection2")

# 插入向量：
collection.add(
    embeddings=[[1.2, 2.3, 4.5], [6.7, 8.2, 9.2]],
    # documents=["This is a document", "This is another document"],
    metadatas=[{"source": "my_source"}, {"source": "my_source"}],
    ids=["id3", "id5"]
)

2、查询、更新与插入

# 查询Chroma中的数据
results = collection.query(
    # query_texts=["This is a query document"],
    query_embeddings = [[2.3, 4.5,6.8]],
    n_results=2,
    include = ["embeddings",'distances',"metadatas"]  ##include使用embeddings结果展示出来，不加结果为None
)
print(results)

##具体id查询
collection.get(
    ids=["id3"],
    where={"source": "my_source"},
    include = ["embeddings"]
)

##更新具体id
collection.update(
    ids=["id3"],
    embeddings=[[1.1, 2.3, 3.2]],
    metadatas=[{"source": "my_source"}],
  
)

collection.get(
    ids=["id3"],
    where={"source": "my_source"},
    include = ["embeddings"]
)

##插入新数据
collection.upsert(
    ids=["id18"],
    embeddings=[[1.5, 2.3, 3.2]],
    metadatas=[{"source": "my_source"}],
  
)

collection.get(
    ids=["id18"],
    where={"source": "my_source"},
    include = ["embeddings"]
)