PInecone return an error during embedding

[haihaowu]

Hi, I got an error when trying to use Pinecone and GPT to create embeddings. The error message pointed to @langchain/pinecone/dist/vectorstores.cjs:149:

const namespace = this.pineconeIndex.namespace(""); ^ TypeError: Cannot read properties of undefined (reading 'namespace')

in vectorstores.cjs., both vectors and documents are valid objects right before the statement
const namespace = this.pineconeIndex.namespace(optionsNamespace ?? "");

Here is the express.js code:
`
const { Pinecone } = require('@pinecone-database/pinecone');
const { PineconeStore } = require('@langchain/pinecone');
const { OpenAIEmbeddings } = require('@langchain/openai');

const pc = new Pinecone({
apiKey: process.env.PINECONE_API_KEY,
});

const pcIndex = pc.Index(process.env.PINECONE_INDEX_NAME);

const embeddings = new OpenAIEmbeddings({
openAIApiKey: process.env.OPENAI_API_KEY,
modelName: 'text-embedding-ada-002',
});

const namespace = pcIndex.namespace('');
console.log('Namespace: ', namespace);

await PineconeStore.fromDocuments(doc, embeddings, {
pcIndex,
});`

Note that the const namespace = pcIndex.namespace(''); in this was executed properly. Appreciate any help to resolve this issue.

langchain: 0.1.18
pinecone: 2.0.1
@langchain/pinecone: 0.0.3

[dosubot[bot]]

🤖

Hey there, @haihaowu! Great to see you back. Hope you're ready to tackle another puzzle together. 🛠😊

Thank you for providing the detailed information. The error message you're seeing is indicating that this.pineconeIndex is undefined at the time it's being accessed. This is likely because the pcIndex object you're passing to PineconeStore.fromDocuments is not being correctly recognized as an instance of PineconeIndex.

In your code, you're passing pcIndex directly as an argument to PineconeStore.fromDocuments:

await PineconeStore.fromDocuments(doc, embeddings, {
  pcIndex,
});

However, according to the LangChainJS codebase, the PineconeStore.fromDocuments method expects an object of type PineconeLibArgs as its third argument. This object should have a pineconeIndex property that holds the PineconeIndex instance.

Here's how you should structure your call to PineconeStore.fromDocuments:

await PineconeStore.fromDocuments(doc, embeddings, {
  pineconeIndex: pcIndex,
});

By making this change, you're correctly passing the PineconeIndex instance to PineconeStore.fromDocuments, and it should be able to access the namespace method without any issues.

Please give this a try and let me know if it resolves your issue. If you're still encountering problems, don't hesitate to provide more details so I can further assist you.

Sources

• libs/langchain-community/src/vectorstores/pinecone.ts

About Dosu

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.

[haihaowu]

It worked! Thanks