Why PyWorldAtlas?

Geographic software often begins with a network request, an API key, or a large scientific stack. PyWorldAtlas takes a different approach: a useful world reference should still work in a classroom with unreliable Wi-Fi, on a small computer, inside a reproducible build, or during a flight.

The product promise

PyWorldAtlas is designed around four constraints:

Offline by default

Package use never makes a network request. Source downloads happen only in the separate development pipeline.

Small enough to understand

The runtime uses the Python standard library and one normalized SQLite database. There is no ORM, dataframe layer, or GIS framework hidden beneath a simple call.

Missing values remain missing

Published values must be traceable to a documented source. An unavailable scalar remains None and an unavailable collection remains empty.

Traceable to sources

Country identity and regional classification come from UN M49. Profile, capital, and populated-place records come from GeoNames. Complete local display names come from Unicode CLDR, while reviewed national official forms come from UNGEGN. English formal names use a public-domain Factbook base with eight exact reviewed UN or Wikidata exceptions. Source references are exposed on profiles and local-name records.

Useful for learning

Readable country summaries, flag emoji, multilingual names, friendly coordinate labels, city discovery, reproducible samples, flashcards, and multiple-choice questions turn the same attributed profiles into reusable lesson material without adding a game framework or hidden data source.

Who is it for?

  • Students learning Python through real geographic data.

  • Teachers preparing offline lessons and exercises.

  • Developers prototyping quizzes, dashboards, or data tools.

  • People working with limited connectivity or older hardware.

  • Curious programmers who want an approachable atlas they can inspect.

What it is not

PyWorldAtlas is not a live-news service, professional GIS engine, routing system, or replacement for authoritative statistical agencies. It is a scoped, offline reference with explicit source and coverage boundaries.

It is also not a live commentary or opinion service. Source conventions are attributed and their limitations are documented. The formal editorial commitments and community standard are in Educational purpose and editorial policy.

What is bundled

Dataset 2026.10.27.1

Records

Coverage

Countries and areas

248

Primary capitals

241 / 248

Populated places

6,265

Selected local-language identities

248 / 248

Sourced English formal names

240 / 248

Anthem titles

234 / 248

Reviewed source-listed mottos

32 / 248

Travel basics (plugs, mains supply, driving side)

245 / 248

English demonym profiles

227 / 248

Country timezone profiles

246 / 248, 417 records

Country-language metadata

722 records across 245 profiles

Reviewed land-border relationships

319

Total-area profiles

248 / 248

Land area and coastline

238 / 248 each

Highest and lowest points

240 / 248 each

Source-listed river records

188 across 80 profiles

Source-listed lake records

187 across 69 profiles

Plain-language climate summaries

240 / 248

Köppen-Geiger climate profiles

241 / 248

Overview and Standard 3D maps

248 / 248 each