instantiate redis vector store with custom index schema

[shiamkumar7]

I am trying to instantiate the redis vector store with a custom field name instead of the default ones. I want to override the default "content" with "page". I am using 'from_documents' to create a redis vectorstore. I am looking for the correct way to instantiate the vector store so that all the "content" field name is replaced with a custom name

[dosubot[bot]]

Hey @shiamkumar7! I'm here to help you with any bugs, questions, or contributions while you wait for a human maintainer. Let's tackle this together!

To override the default field names in the Redis vector store when using the from_documents method, you can provide a custom index schema. This allows you to specify custom field names, such as replacing "content" with "page". Here's how you can do it:

1. Define a custom index schema in a YAML file or as a dictionary. For example, to replace "content" with "page":

text:
  - name: page

2. Use this schema when creating the Redis vector store:

from langchain_community.vectorstores import Redis
from langchain_community.embeddings.fake import FakeEmbeddings

# Define your documents and embeddings
docs = [...]  # your list of documents
embeddings = FakeEmbeddings(size=128)

# Define the custom index schema
index_schema = {
    "text": [
        {"name": "page"}
    ]
}

# Create the Redis vector store with the custom index schema
redis_client = redis.from_url("redis://127.0.0.1:6379")
rvs = Redis.from_documents(
    docs,
    embedding=embeddings,
    client=redis_client,
    index_name="my_vector_index",
    index_schema=index_schema
)

This setup will replace the default "content" field with "page" in your Redis vector store ^[1].

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Other}

[shiamkumar7]

@dosu
I tried a similar thing already. FYI, I have few other metadata as well. I want the data that is present in "page_content" to be present in the field "text" instead of the default value "content". Please refer below to the schema that I created
index_schema = {
"text": [{"name": "text"},
{"name":"source_file"},
{"name":"webUrl"}
],
"numeric":[{'name':'page'}]
I'm getting a warning as shown below,

index_schema does not match generated metadata schema.
If you meant to manually override the schema, please ignore this message.
index_schema: {'text': [{'name': 'text'}, {'name': 'source_file'}, {'name': 'webUrl'}], 'numeric': [{'name': 'page'}]}
generated_schema: {'text': [{'name': 'source_file'}, {'name': 'webUrl'}], 'numeric': [{'name': 'page'}], 'tag': []}