Milvus Range Search

"""
Milvus Range Search
===================

Demonstrates Milvus range-search parameters (`radius`, `range_filter`) in sync and async calls.
"""

import asyncio

from agno.agent import Agent
from agno.knowledge.knowledge import Knowledge
from agno.vectordb.milvus import Milvus

# ---------------------------------------------------------------------------
# Setup
# ---------------------------------------------------------------------------
vector_db = Milvus(
    collection="recipes_range_search",
    uri="/tmp/milvus_range.db",
)


# ---------------------------------------------------------------------------
# Create Knowledge Base
# ---------------------------------------------------------------------------
knowledge = Knowledge(
    name="My Milvus Range Search Knowledge Base",
    description="This demonstrates range-based search with radius and range_filter parameters",
    vector_db=vector_db,
)


# ---------------------------------------------------------------------------
# Create Agent
# ---------------------------------------------------------------------------
agent = Agent(knowledge=knowledge)


# ---------------------------------------------------------------------------
# Run Agent
# ---------------------------------------------------------------------------
async def main() -> None:
    knowledge.insert(
        name="Recipes",
        url="https://agno-public.s3.amazonaws.com/recipes/ThaiRecipes.pdf",
        metadata={"doc_type": "recipe_book"},
    )

    print("=" * 80)
    print("Example 1: Regular search (no radius/range_filter)")
    print("=" * 80)
    agent.print_response("How to make Tom Kha Gai", markdown=True)

    print("\n" + "=" * 80)
    print("Example 2: Search with radius parameter (minimum similarity threshold)")
    print("=" * 80)
    query = "How to make Thai curry"
    results = knowledge.vector_db.search(
        query=query,
        limit=5,
        search_params={"radius": 0.3},
    )
    print(f"\nFound {len(results)} documents with similarity >= 0.3:")
    for i, doc in enumerate(results, 1):
        print(f"\n{i}. {doc.name}")
        print(f"   Content preview: {doc.content[:100]}...")

    print("\n" + "=" * 80)
    print("Example 3: Search with radius and range_filter (similarity range)")
    print("=" * 80)
    results_with_range = knowledge.vector_db.search(
        query=query,
        limit=5,
        search_params={"radius": 0.3, "range_filter": 0.8},
    )
    print(
        f"\nFound {len(results_with_range)} documents with similarity in range [0.3, 0.8]:"
    )
    for i, doc in enumerate(results_with_range, 1):
        print(f"\n{i}. {doc.name}")
        print(f"   Content preview: {doc.content[:100]}...")

    print("\n" + "=" * 80)
    print("Example 4: Async search with range parameters")
    print("=" * 80)
    async_results = await knowledge.vector_db.async_search(
        query="Thai desserts",
        limit=3,
        search_params={"radius": 0.3, "range_filter": 0.9},
    )
    print(f"\nAsync search found {len(async_results)} documents:")
    for i, doc in enumerate(async_results, 1):
        print(f"\n{i}. {doc.name}")
        print(f"   Content preview: {doc.content[:100]}...")

    print("\n" + "=" * 80)
    print("Cleaning up...")
    print("=" * 80)
    vector_db.delete_by_metadata({"doc_type": "recipe_book"})
    print("Cleaned up successfully!")


if __name__ == "__main__":
    asyncio.run(main())

# Clone and setup repo
git clone https://github.com/agno-agi/agno.git
cd agno/cookbook/07_knowledge/vector_db/milvus_db

# Create and activate virtual environment
./scripts/demo_setup.sh
source .venvs/demo/bin/activate

python milvus_db_range_search.py