Release 2.2.0

What's new

• Addresses near a point: near= and within_km= filters on real_random_address(), real_random_addresses() and count(). near takes a (lat, lng) pair or an address from an earlier call.
• Groups of nearby addresses: real_random_cluster(count, within_km=...) returns addresses that lie close together, with no centre to pick.
• Nine new states: IL, ID, LA, IN, OH, PA, SC, NM and MN, bringing the dataset to 3,850 addresses across 29 states. Every source was checked against the publisher's own license terms, so every address is still public domain.
• CLI: new random-address count and random-address cluster commands, plus --near LAT,LNG and --within-km KM options.
• Friendlier inputs: a postal code can be passed as an int (6040 finds 06040), and extra spaces in a city name are ignored.
• Coverage page: COVERAGE.md lists what the dataset holds, state by state. It is generated from the data and checked in CI.
• Python 3.15 is supported.

Compatibility

• Everything from 2.1.1 keeps working unchanged.
• Python 3.11 is now the minimum. Python 3.10 reached end of life on 2026-10-01. Nothing breaks for anyone still on it: pip keeps installing 2.1.1 on Python 3.10, and 1.3.0 on 3.9 or older.

Installation

pip install --upgrade random-address

See the CHANGELOG for the full list of changes.

Release 2.1.1

What's new

• Added count() with optional state, city, and postal-code filters.
• Expanded address coverage with new geographic data.
• Added stronger dataset integrity and consistency validation.
• Improved ingestion validation and duplicate detection.
• Maintained full backward compatibility with the 2.x API.

Installation

pip install --upgrade random-address

v2.1.0

Merge pull request #11 from neosergio/release/v2.1.0

Release 2.1.0. The public API is unchanged. This is a dataset release…

v1.3.0

Add counting functions