Dataset notes and licence¶

This directory supports Partle's study of catalogue extraction across 205 OpenStreetMap-derived storefront websites in Illinois and Pennsylvania.

Files¶

sampling-frame.csv contains the 205 store keys, names, states, and homepages used in both runs.
platform-survey-2026-08-31.json is the complete platform-and-catalogue prefilter output.
import-outcomes-2026-08-31.json is the complete controlled dry-run output. Catalogue retrieval was capped at five products per store and no product was written to a database.
store-outcomes-2026-08-31.csv joins the two outputs into one row per sampled store.
summary.json contains the historical aggregates and the new replication totals.
source-manifest.json records source commits, commands, timestamps, file hashes, and settings.
spot-check.md records the independent verification pass.
verify.py checks the frame, key coverage, aggregate calculations, and file hashes with the Python standard library.

Run python verify.py from this directory to check the packet.

The packet deliberately excludes product names, descriptions, images, email addresses, coordinates, and a catalogue dump. Store names and homepages are included so another reader can understand and check the classification.

Attribution and licence¶

OpenStreetMap data is available under the Open Data Commons Open Database License 1.0. The sampling frame and the store-level files derived from it are distributed under that same licence. The source websites and their contents remain subject to their own rights and terms; this packet does not relicense them.

The accompanying analysis is part of the Partle documentation. The repository's LICENSE governs Partle source code; it does not replace the Open Database License for the derived dataset files in this directory.