+MongoDB Vector Support

[ranfysvalle02]

Description

Add MongoDB as a VectorDB.

Fixes # (issue)

#1166

Type of change

Please delete options that are not relevant.

• Bug fix (non-breaking change which fixes an issue)
• [X] New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Refactor (does not change functionality, e.g. code style improvements, linting)
• Documentation update

How Has This Been Tested?

There is working code that highlights how to use, and the output.

Please delete options that are not relevant.

• Unit Test
• Test Script

config = {
    "vector_store": {
        "provider": "mongodb",
        "config": {
            "mdb_uri": "mongodb://0.0.0.0/?directConnection=true",
            "dbname": "DEMO",
            "collection_name": "TEST",
            "get_embedding": get_embedding,
            "embedding_model_dims": len(get_embedding("0")),
        }
    }
}
m = Memory.from_config(config)

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• Any dependent changes have been merged and published in downstream modules
• I have checked my code and corrected any misspellings

Maintainer Checklist

• closes #xxxx (Replace xxxx with the GitHub issue number)
• Made sure Checks passed

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name	Status	Preview	Comments	Updated (UTC)
multimodal-demo	✅ Ready (Inspect)	Visit Preview	💬 Add feedback	Mar 12, 2025 3:19am

[ranfysvalle02]

@Dev-Khant could you please take a look at this one?

[Dev-Khant]

Please add the relevant docs and tests for it. Thanks.

[ranfysvalle02]

@Dev-Khant will add this soon! Sorry for delay

[CLAassistant]

All committers have signed the CLA.

[dgupta18]

Hi @Dev-Khant , this PR is ready for re-review when you get a second!

[coder-pikachu]

When would this be closed?

[Dev-Khant]

Hey @ranfysvalle02 Can you please resolve the merge conflicts so we can run the CI tests.

[ranfysvalle02]

@Dev-Khant -- merge conflicts resolved!!!

[ranfysvalle02]

@Dev-Khant - Just following up here, let me know if any issues

[ranfysvalle02]

@Dev-Khant --- We are about to have more merge conflicts -- Can you please advise if there is anything wrong with the MongoDB aspect of it or help get this merged please?

[Dev-Khant]

Hey, sorry the delay here. Left few comments once it's resolved we can merge this.

[Dev-Khant]

Also @ranfysvalle02 Can you please why the monogodb tests are failing?

[ranfysvalle02]

@Dev-Khant give me a bit to tackle all this -- will report back soon!

[ranfysvalle02]

Also @ranfysvalle02 Can you please why the monogodb tests are failing?

@Dev-Khant can you elaborate a bit more here? I'm trying to figure out what is going on here.

pytest test_mongodb.py
=============================================================================================== test session starts ================================================================================================
platform darwin -- Python 3.10.10, pytest-8.3.5, pluggy-1.5.0
rootdir: /Users/xxxx/dev/xxxxxx/mem0
configfile: pyproject.toml
plugins: anyio-4.9.0, langsmith-0.3.42
collected 10 items                                                                                                                                                                                                 

test_mongodb.py ..........                                                                                                                                                                                   [100%]

================================================================================================ 10 passed in 0.57s ================================================================================================

The provided test script errors out? Am I missing something here?

[ranfysvalle02]

do a docker docker run -p 27017:27017 mongodb/mongodb-atlas-local

then try to run the below script -- it should work... unless I'm doing something wrong here?

import logging  
import numpy as np  
from mongodb import MongoVector  # Ensure this matches the filename where MongoVector is saved  
  
# Configure logging  
logging.basicConfig(level=logging.INFO)  
logger = logging.getLogger(__name__)  
  
def test_mongo_vector():  
    # Configuration  
    db_name = 'test_db'  
    collection_name = 'test_vectors'  
    embedding_dims = 128  # Example embedding dimension  
  
    # Initialize the MongoVector instance  
    mongo_vector = MongoVector(  
        mongo_uri='mongodb://localhost:27017/?directConnection=true&serverSelectionTimeoutMS=2000',  # Adjust as needed
        db_name=db_name,  
        collection_name=collection_name,  
        embedding_model_dims=embedding_dims  
    )  
    logger.info('Initialized MongoVector instance.')  
  
    # Generate sample vectors and payloads  
    num_vectors = 10  
    vectors = [np.random.rand(embedding_dims).tolist() for _ in range(num_vectors)]  
    payloads = [{'payload_info': f'document_{i}'} for i in range(num_vectors)]  
    ids = [f'vector_{i}' for i in range(num_vectors)]  
  
    # Insert vectors into MongoDB  
    mongo_vector.insert(vectors=vectors, payloads=payloads, ids=ids)  
    logger.info('Inserted vectors into MongoDB.')  
  
    # Perform a vector search  
    query_vector = np.random.rand(embedding_dims).tolist()  
    search_results = mongo_vector.search(query='', query_vector=query_vector, limit=5)  
    logger.info(f'Search results:')  
    for result in search_results:  
        logger.info(f'ID: {result.id}, Score: {result.score}, Payload: {result.payload}')  
  
    # Update a vector  
    update_vector_id = ids[0]  
    new_vector = np.random.rand(embedding_dims).tolist()  
    new_payload = {'payload_info': 'updated_document'}  
    mongo_vector.update(vector_id=update_vector_id, vector=new_vector, payload=new_payload)  
    logger.info(f'Updated vector {update_vector_id}.')  
  
    # Retrieve an updated vector  
    retrieved_vector = mongo_vector.get(vector_id=update_vector_id)  
    logger.info(f'Retrieved vector: {retrieved_vector}')  
  
    # Delete a vector  
    delete_vector_id = ids[1]  
    mongo_vector.delete(vector_id=delete_vector_id)  
    logger.info(f'Deleted vector {delete_vector_id}.')  
  
    # List vectors in the collection  
    vector_list = mongo_vector.list(limit=5)  
    logger.info(f'List of vectors:')  
    for v in vector_list:  
        logger.info(f'ID: {v.id}, Payload: {v.payload}')  
  
    # Get collection information  
    collection_info = mongo_vector.col_info()  
    logger.info(f'Collection info: {collection_info}')  
  
    # Clean up by deleting the collection  
    mongo_vector.delete_col()  
    logger.info('Deleted the collection.')  
  
    # Close the database connection  
    del mongo_vector  
    logger.info('MongoDB connection closed.')  
  
if __name__ == '__main__':  
    test_mongo_vector()  

[Dev-Khant]

Hey @ranfysvalle02 We have a separate issue for CI tests, will fix it. But your changes looks good to me. Thanks for the contribution!