About
This site is the map of the OpenMobilityData repositories on GitHub: public records and open data on mobility in Montréal, kept with their provenance, and the code that reads them. Each entry opens a document or the folder that holds it; the folder's README says what the document is, where it came from and how every table and figure made from it can be reproduced.
The author
The repositories were created, and the derived documents written, by Rick Hoge, a researcher in the Faculty of Medicine at McGill University. The views expressed here are the author's own and do not reflect the position of any institution or organization. Unless otherwise indicated, analyses presented here should be considered preliminary pre-peer-review drafts - feedback from other researchers and members of the public is welcome in the GitHub 'Issues' board.
Nothing in this site should be considered an engineering opinion or legal advice.
How the record repositories are kept
The record repositories follow the same rules, stated in each README:
- One folder per document, request or dataset, with a
README.mdthat gives the issuer, the date, where and when the document was obtained and what it says;raw/holds the documents exactly as obtained and is never edited;derived/holds anything made from them, with the script beside each table or figure. - The READMEs describe and quote. They do not evaluate, recommend or impute motives.
- Every page reference in a public README links to the page where a web copy exists.
- Third-party data are named with a link and never redistributed unless their licence allows it; City open data (CC BY 4.0) is redistributed with credit.
- The full edit history is public by design.
What the labels mean
- Record, as issued: a document published or released by a public body or another issuer, kept exactly as obtained.
- Transcript, checked by ear: every passage has been verified against the recording.
- Machine draft, unchecked: an automatic transcription, published so the record is available at once. It may contain errors and is not to be quoted; the checked transcript replaces it when ready.
- Analysis: tables, figures or a note made by the author from the records, with the code.
- Preliminary, not peer reviewed: an analysis that has stabilized and has not been reviewed by peers.
- Draft: an analysis still under the author's review; its figures may change.
Use of AI
Much of the code and many of the drafts in these repositories were produced with an AI assistant, Anthropic's Claude, in interactive sessions directed by the author. The author chose the questions, supplied the sources, reviewed every output, and checked every quotation, page reference and figure against the documents; the author alone is responsible for the content. Commits to which the assistant contributed name it as co-author, so the division of work is visible in the history, and errors found in its work are corrected in the open like any other. Nothing in a raw/ folder is edited by anyone, human or machine. The practice follows the principles of the European Commission's « Living guidelines on the responsible use of generative AI in research » (document page): the researcher remains responsible for the output, the use of the tool is disclosed, sources and facts are verified, privacy and third-party rights are respected, and the tool is never an author.
Corrections
Everything is open to review: all source code, documents and data in these repositories are publicly visible, as is the complete change log of every file. Anyone with a concern is welcome to open an issue on GitHub in the repository concerned; every issue submitted is publicly visible, as is its answer.
This site
The pages are generated from one catalogue file by a short script, both in the site's repository. The site sets no cookies and loads nothing from other servers; it is hosted on GitHub Pages.