Author notes. Motivation and target: country detection can be certainly solved with standard tools eg PostGIS, with 3 prerequisites: 1) you know the stack (right formats, resolutions, CRSs, have operational database or certain suitable library properly setup), 2) you have good vector data 3) enough compute for potentially non-trivial task. With this domain-specific solution none of the three is needed - you just import a library (Python and JS provided) and call one method. It is compact (sub-MB datasets for minimal use in eg web browser, for much better accuracy optional ~10MB download) and order(s) of magnitude faster than generic point-in-polygon SQL queries (benchmarks are provided): ~1M lookups per second. No online, no DB servers needed etc.
Comments
Author notes. Motivation and target: country detection can be certainly solved with standard tools eg PostGIS, with 3 prerequisites: 1) you know the stack (right formats, resolutions, CRSs, have operational database or certain suitable library properly setup), 2) you have good vector data 3) enough compute for potentially non-trivial task. With this domain-specific solution none of the three is needed - you just import a library (Python and JS provided) and call one method. It is compact (sub-MB datasets for minimal use in eg web browser, for much better accuracy optional ~10MB download) and order(s) of magnitude faster than generic point-in-polygon SQL queries (benchmarks are provided): ~1M lookups per second. No online, no DB servers needed etc.