test script:
from simpleVDB import VectorDB
import numpy as np
# Create a new VectorDB with 128-dimensional vectors
db = VectorDB(dim=128)
# Batch insert vectors
vectors = [(i, np.random.randn(128)) for i in range(1000)]
db.batch_insert(vectors)
# Search for similar vectors
results = db.search(np.random.randn(128), k=5)
print("K-NN Search Results:", results)
# Range search
range_results = db.range_search(np.random.randn(128), radius=1.0)
print("Range Search Results:", range_results)
# Update a vector
db.update(500, np.random.randn(128))
# Delete a vector
db.delete(100)
# Optimize memory usage across devices
db.optimize_memory()
# Rebalance the tree
db.rebalance()
# Get database statistics
stats = db.get_stats()
print("Database Stats:", stats)
# Create and populate the database
db = VectorDB(dim=128)
vectors = [(i, np.random.randn(128)) for i in range(1000)]
db.batch_insert(vectors)
# Save the database
db.save("my_vector_db.pkl")
# Later, load the database
loaded_db = VectorDB.load("my_vector_db.pkl")
# Use the loaded database
results = loaded_db.search(np.random.randn(128), k=5)
print("Search Results:", results)simpleVDB/
├── __init__.py
├── core/
│ ├── __init__.py
│ ├── vector_db.py
│ ├── node.py
│ ├── cluster.py
│ ├── vector.py
│ └── device_manager.py
├── utils/
│ ├── __init__.py
│ ├── config.py
│ └── exceptions.py
├── operations/
│ ├── __init__.py
│ ├── insert.py
│ ├── delete.py
│ ├── search.py
│ └── update.py
├── backends/
│ ├── __init__.py
│ ├── cpu_backend.py
│ └── gpu_backend.py
└── tests/
├── __init__.py
└── test_vector_db.py