Difference between add_texts and from_texts

[Abe410]

Hi all

I was working with the Pinecone vector store, and in the following file

https://github.com/langchain-ai/langchain/blob/master/libs/langchain/langchain/vectorstores/pinecone.py#L394

I saw two functions here:

1- add_texts
2- Class method from_texts

What is the difference between both?

Secondly, previously there used to be a method Pinecone.from_documents as of version 0.0.177. Has that been deprecated?

Thanks

[dosubot[bot]]

🤖

Hi,

The add_texts function and the class method from_texts in the Pinecone vector store of the LangChain repository both serve to add texts to the vector store, but they do so in different ways.

The add_texts function is an instance method that takes an iterable of strings (texts), optional metadatas, ids, and a namespace, and adds these texts to the vector store. It first embeds the texts, then creates documents with the embeddings, metadata, and ids, and finally upserts these documents to the Pinecone index. The function returns a list of ids from adding the texts into the vectorstore.

On the other hand, the from_texts class method is a constructor that creates a Pinecone wrapper from raw documents. It takes a list of strings (texts), an Embeddings object, optional metadatas, ids, a batch size, a text key, an index name, a namespace, and upsert kwargs. It first checks if the provided index exists, then for each batch of texts, it creates ids if not provided, creates embeddings, prepares metadata, and upserts the batch to the Pinecone index. The method returns a Pinecone instance.

As for the Pinecone.from_documents method, it is not mentioned in the provided context, so it's not possible to determine whether it has been deprecated since version 0.0.177.

You can find more details about these methods in the Pinecone vector store file of the LangChain repository.

I hope this helps! Let me know if you have any other questions.

Sources

• Pinecone in docs is outdated
• libs/langchain/langchain/vectorstores/pinecone.py

---

This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.

Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.

If you want to continue the conversation, start your reply with @dosu-bot.