How to use it to search for similar array of numbers?

[adivoj]

Can anyone tell me please how I can use it given 10 numbers to answer with similar array? My database consists of 40 (0 to 40) keys each having 100 arrays of 10 numbers from 0 to 40. Actually I need the key where the similar array was found.

[dosubot[bot]]

🤖

Hello,

To find a similar array in your database using LangChain, you can use the similarity_search method of the Tair class in the langchain.vectorstores.tair module. This method takes an array as an argument and returns a list of documents that are most similar to the given array.

Here is an example of how you can use this method:

# Assuming `tair` is an instance of `Tair`
# and `array` is your array of 10 numbers
documents = tair.similarity_search(query=array, k=1)

In this example, k=1 means that you want to find the 1 most similar document. The documents variable will be a list of Document objects. Each Document object has a metadata attribute which you can use to get the key of the document in your database.

Here is how you can get the key:

key = documents[0].metadata['key']

Please note that this assumes that the metadata of each document in your database has a 'key' field. If the key is stored in a different field, you should replace 'key' with the name of that field.

Also, please note that the similarity_search method uses the cosine similarity to find the most similar documents. If your arrays of numbers are not normalized, you might want to normalize them before using this method.

For more information, you can refer to the following sources:

• libs/langchain/langchain/vectorstores/tencentvectordb.py
• libs/langchain/tests/integration_tests/vectorstores/test_myscale.py
• libs/langchain/langchain/vectorstores/tair.py

I hope this helps! If you have any other questions, feel free to ask.

Sources

• libs/langchain/langchain/vectorstores/tencentvectordb.py
• libs/langchain/tests/integration_tests/vectorstores/test_myscale.py
• libs/langchain/langchain/vectorstores/tair.py
• libs/langchain/langchain/evaluation/embedding_distance/base.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.