spatio 0.2.0

Blazingly fast spatio-temporal database library

Spatio for Python

Fast spatial database for Python, powered by Rust.

What is it?

Spatio is an embedded database for location data. Store points on a map, query "what's nearby", track movement over time - all with minimal setup.

Install

pip install spatio

Supports: Linux (x86_64, aarch64), macOS (Intel, Apple Silicon), Python 3.9-3.13.

Windows not supported - use WSL2, Docker, or a Linux VM.

Quick Start

import spatio

db = spatio.Spatio.memory()

nyc = spatio.Point(-74.0060, 40.7128)
db.insert_point("cities", nyc, b"New York")

# Find nearby (within 100km)
nearby = db.query_within_radius("cities", nyc, 100000, 10)
for point, data, distance in nearby:
    print(f"{data.decode()}: {distance/1000:.1f}km away")

Examples

Basic Operations

import spatio

# Create database
db = spatio.Spatio.memory()                # In-memory
db = spatio.Spatio.open("data.db")         # Persistent

# Store key-value data
db.insert(b"user:123", b"John Doe")
data = db.get(b"user:123")
db.delete(b"user:123")

# Store points
point = spatio.Point(-74.0060, 40.7128)
db.insert_point("pois", point, b"Restaurant")

# Query spatial
nearby = db.query_within_radius("pois", point, 1000.0, 10)
count = db.count_within_radius("pois", point, 1000.0)
exists = db.contains_point("pois", point, 1000.0)

Distance Calculations

nyc = spatio.Point(-74.0060, 40.7128)
london = spatio.Point(-0.1278, 51.5074)

# Haversine (fast, spherical Earth)
metric = spatio.DistanceMetric("haversine")
distance = db.distance_between(nyc, london, metric)
print(f"{distance/1000:.0f}km")  # ~5570km

# Geodesic (accurate, ellipsoidal Earth - Karney 2013)
metric = spatio.DistanceMetric("geodesic")
distance = db.distance_between(nyc, london, metric)

# Also available: "rhumb", "euclidean"

K-Nearest Neighbors

center = spatio.Point(-74.0, 40.7)

# Find 5 nearest within 50km
metric = spatio.DistanceMetric("haversine")
nearest = db.knn("cities", center, 5, 50000.0, metric)

for point, data, distance in nearest:
    print(f"{data.decode()}: {distance/1000:.1f}km")

Bounding Box Queries

# Find all points in rectangle
results = db.find_within_bounds(
    "cities",
    min_lat=40.0,
    min_lon=-75.0,
    max_lat=41.0,
    max_lon=-73.0,
    limit=100
)

# Check if any exist in box
has_points = db.intersects_bounds("cities", 40.0, -75.0, 41.0, -73.0)

Polygon Queries

# Points within polygon boundary
polygon = [
    (-74.0, 40.7),
    (-73.9, 40.7),
    (-73.9, 40.8),
    (-74.0, 40.8),
]
inside = db.query_within_polygon("pois", polygon, 100)

Trajectories

import time

# Track movement
trajectory = [
    (spatio.Point(-74.00, 40.71), int(time.time())),
    (spatio.Point(-74.01, 40.72), int(time.time()) + 60),
    (spatio.Point(-74.02, 40.73), int(time.time()) + 120),
]
db.insert_trajectory("vehicle:truck001", trajectory)

# Query movement history
start = int(time.time()) - 3600
end = int(time.time())
path = db.query_trajectory("vehicle:truck001", start, end)

for point, timestamp in path:
    print(f"At ({point.lon}, {point.lat}) at {timestamp}")

TTL (Auto-Expiration)

import spatio

db = spatio.Spatio.memory()

# Expire after 5 minutes
opts = spatio.SetOptions.with_ttl(300.0)
db.insert(b"session:abc", b"user data", opts)

# Returns None if expired
data = db.get(b"session:abc")

# Clean up expired items manually
removed = db.cleanup_expired()
print(f"Removed {removed} expired items")

Important: TTL is lazy. Expired items stay in memory until you call cleanup_expired() or they're overwritten. For long-running apps, clean up periodically to avoid memory leaks.

Configuration

config = spatio.Config()
db = spatio.Spatio.memory_with_config(config)

# That's it - most defaults are sensible
# See Rust docs for advanced options

API Reference

Spatio Class

# Create
Spatio.memory()
Spatio.open(path)
Spatio.memory_with_config(config)
Spatio.open_with_config(path, config)

# Key-value
insert(key, value, options=None)
get(key) -> bytes | None
delete(key) -> bytes | None

# Spatial
insert_point(prefix, point, value, options=None)
query_within_radius(prefix, center, radius_meters, limit)
count_within_radius(prefix, center, radius_meters)
contains_point(prefix, center, radius_meters)
find_within_bounds(prefix, min_lat, min_lon, max_lat, max_lon, limit)
intersects_bounds(prefix, min_lat, min_lon, max_lat, max_lon)
query_within_polygon(prefix, polygon_coords, limit)

# Distance
distance_between(point1, point2, metric)
knn(prefix, center, k, max_radius, metric)

# Trajectories
insert_trajectory(object_id, trajectory, options=None)
query_trajectory(object_id, start_time, end_time)

# Maintenance
cleanup_expired() -> int
stats() -> dict
close()

Point Class

Point(longitude, latitude, altitude=None)

# Properties
point.lon  # -180 to 180
point.lat  # -90 to 90
point.alt  # Always None (2D only for now)

# Methods
point.distance_to(other_point) -> float  # meters, Haversine

SetOptions Class

SetOptions.with_ttl(seconds)
SetOptions.with_expiration(unix_timestamp)

DistanceMetric Class

DistanceMetric("haversine")   # Fast spherical
DistanceMetric("geodesic")    # Accurate ellipsoidal
DistanceMetric("rhumb")       # Constant bearing
DistanceMetric("euclidean")   # Planar only

Performance Tips

1. Batch inserts when possible
2. Use namespaces to separate data types
3. Set reasonable limits on queries
4. Call cleanup_expired() periodically if using TTL
5. Use persistent storage for data you can't lose

Building from Source

git clone https://github.com/pkvartsianyi/spatio
cd spatio/py-spatio

# Install maturin
pip install maturin

# Build and install
maturin develop

# Or build wheel
maturin build --release

Platform Support

Supported:

• Linux: x86_64, aarch64 (manylinux)
• macOS: x86_64 (Intel), arm64 (Apple Silicon)
• Python: 3.9, 3.10, 3.11, 3.12, 3.13

Examples

Check the examples/ directory:

• basic_usage.py - Getting started
• performance_demo.py - Benchmarks
• trajectory_tracking.py - Movement tracking

Links

• GitHub: https://github.com/pkvartsianyi/spatio
• PyPI: https://pypi.org/project/spatio
• Rust docs: https://docs.rs/spatio
• Issues: https://github.com/pkvartsianyi/spatio/issues

License

MIT - see LICENSE