ClickHouse ERROR: Distance function argument of Annoy index must be of type String.

[sheiy]

System Info

langchain: 0.0.200 platform: macOS python: 3.10.11 clickhouse: version 23.5.2.7 (official build)

Who can help?

@hwchase17

Information

• [ ] The official example notebooks/scripts
• [X] My own modified scripts

Related Components

• [ ] LLMs/Chat Models
• [ ] Embedding Models
• [ ] Prompts / Prompt Templates / Prompt Selectors
• [ ] Output Parsers
• [ ] Document Loaders
• [X] Vector Stores / Retrievers
• [ ] Memory
• [ ] Agents / Agent Executors
• [ ] Tools / Toolkits
• [ ] Chains
• [ ] Callbacks/Tracing
• [ ] Async

Reproduction

Expected behavior

success write data to clickhouse

[devstein]

Generated by a 🤖

Anwser

To resolve the issue, update the index_param in ClickHouseSettings to use the metric parameter as a string:

index_param: Optional[Union[List, Dict]] = [100, f"'{metric}'"]

Also, ensure the metric parameter in the Annoy.from_texts() function is set to a string value, such as "angular" or "euclidean":

vector_store_v2 = Annoy.from_texts(
    texts, embeddings_func, metric="angular", n_trees=100, n_jobs=1
)

Sources

• langchain/vectorstores/clickhouse.py
• docs/modules/indexes/vectorstores/examples/annoy.ipynb

---

This response is meant to be helpful and save users time. It is not meant to be a precise solution.

Please leave a 👍 if this is helpful and 👎 if it is irrelevant

[sheiy]

@haoch do you have any idea?

[haoch]

@sheiy Will test with the clickhouse version you used and get back to you shortly.

[haoch]

@sheiy It will be fixed with #6385