milvus_lite_collection_setup_insert_and_similarity_search.py

1from pymilvus import MilvusClient
2
3# 1. Set up a local Milvus Lite instance
4client = MilvusClient("milvus_demo.db")
5
6# 2. Create a collection in quick setup mode
7# This automatically handles schema and index creation
8if client.has_collection(collection_name="demo_collection"):
9    client.drop_collection(collection_name="demo_collection")
10
11client.create_collection(
12    collection_name="demo_collection",
13    dimension=5  # The vectors we will use have 5 dimensions
14)
15
16# 3. Prepare data (ID, vector, and optional metadata)
17data = [
18    {"id": 0, "vector": [0.35, 0.08, 0.11, 0.44, 0.06], "color": "pink"},
19    {"id": 1, "vector": [0.23, 0.47, 0.20, 0.05, 0.04], "color": "red"},
20    {"id": 2, "vector": [0.24, 0.06, 0.38, 0.32, 0.00], "color": "blue"},
21    {"id": 3, "vector": [0.44, 0.31, 0.18, 0.06, 0.18], "color": "grey"},
22    {"id": 4, "vector": [0.22, 0.30, 0.39, 0.41, 0.03], "color": "orange"},
23]
24
25# 4. Insert data
26res = client.insert(
27    collection_name="demo_collection",
28    data=data
29)
30
31print(f"Insert result: {res}")
32
33# 5. Perform a vector similarity search
34query_vectors = [[0.35, 0.08, 0.11, 0.44, 0.06]]
35
36search_res = client.search(
37    collection_name="demo_collection",
38    data=query_vectors,
39    limit=3,  # Return top 3 results
40    output_fields=["color"]  # Include the 'color' field in results
41)
42
43print("Search results:")
44for hits in search_res:
45    for hit in hits:
46        print(f"Hit: {hit}")