Analysis & research on the camera dataset — notebooks, one-off studies, and writeups (coverage growth over time, operator/brand breakdowns, geographic density, etc.).
Status: empty for now. Conventions when work lands here:
- One subfolder per study, named
YYYY-MM-topic/, containing its code and a README with the question, method, and findings - Keep raw data out of git — pull from the public tile/data URLs (see the root README) or the R2 buckets, and write intermediate artifacts to the study folder (gitignored patterns cover
.geojson/.pmtiles) - Anything reusable across studies graduates to a shared
analysis/lib/