尚未偵測到更新。GitHub 發布版本後會顯示在這裡。
依最近已完成的 8 小時觀測回合間的 GitHub 星號增量排序
個人 GitHub 帳號下公開儲存庫的擁有者可以登入並驗證所有權，自行管理介紹，無需管理員核准。
登入並產生驗證檔案，在 30 分鐘內將其提交至此儲存庫的預設分支。其他分支或分叉儲存庫中的檔案無效。然後在下方檢查檔案，即可無需管理員核准直接管理專案。請保留檔案以維持管理權限。
如需登入，將前往登入頁面。登入後請回到此處，再次點擊所選的驗證按鈕。
在專案資料夾中執行 /nomorevibe verify，即可驗證所有權並自行管理介紹。
一分鐘完成驗證正在載入完整 README。 以下為已儲存的摘錄。
Interactive 3D maps of parcel-level property data — land value, improvement
value, value per acre, improvement-to-land ratio, surface parking, and more —
for cities across the US and beyond.
Maintained by the Center for Land Economics.檢視完整 README ↗Interactive 3D maps of parcel-level property data — land value, improvement value, value per acre, improvement-to-land ratio, surface parking, and more — for cities across the US and beyond.
Live site: civicmapper.org
Maintained by the Center for Land Economics.
Each city's parcel data is cleaned and exported to GeoParquet by a Python ETL notebook or script. Larger cities also get PMTiles vector tiles (with an H3 hexagon overlay) so they render at any scale. The frontend loads these files over HTTP and renders them with MapLibre GL; a small Express proxy serves the files with CORS and rate limiting in deployed environments.
| Path | What it is |
|---|---|
viz/ | The map frontend (Vite + MapLibre). See viz/README.md |
api/ | Express data proxy. See api/README.md |
data/ | Per-city ETL notebooks, shared Python utilities, post-processing scripts. See data/README.md |
docs/ | Design docs and playbooks (add a city, parquet format, …) |
scripts/ | Repo-level utilities (e.g. headless smoke test of every deployed city) |
site/ | Static landing page for civicmapper.org |
.github/ | CI/CD workflows (deploy, ETL runs, smoke tests) |
Prerequisites: Node 20+ (frontend/API), Python 3.11+ (only needed for the ETL side).
The one-shot way — starts the API on port 8080 and the Vite dev server on port 5173, installing dependencies on first run:
./start-local.sh # macOS / Linux
start-local.bat # Windows (opens two terminal windows)
Or manually:
# Terminal 1 — API (optional for local dev; the frontend has its own dev proxy)
cd api
npm install
npm start # http://localhost:8080
# Terminal 2 — frontend
cd viz
npm install
npm run dev # http://localhost:5173
Open http://localhost:5173 — it takes you to the city picker
(cities.html); each city opens as app.html?city=<key>.
No environment variables are required for a basic local run: in dev mode the Vite server proxies dataset requests to a hosted data endpoint, so cities load out of the box. To run against your own data, either:
<city>-<state>-parcels.parquet file into viz/public/ — in dev the
app prefers a local copy of the current city's file when one exists, orVITE_PARQUET_BASE_URL (and friends — see
viz/README.md) at your own parquet/PMTiles hosting.For the ETL side (producing those parquet/PMTiles files yourself), follow data/README.md.
The end-to-end procedure — sourcing parcel data, ETL, PMTiles, frontend registration, deployment — is documented in docs/add-city-playbook.md. The required output format is specified in docs/parcel-parquet-format.md.
Want a city but don't want to build it? Open a city request issue.
Civic Mapper works like a library: the engine is generic, and a deployment is the engine plus configuration pointing at your data hosting. To stand up your own instance:
VITE_PARQUET_BASE_URL (and friends;
the full list with docs is in
viz/env/.env.example). ETL-side variables are in
data/.env.example. Secrets never live in this repo —
deployments supply them via environment / CI secrets.viz/src/cities/<key>.json + viz/src/dictionaries/<key>.json; no shared
files to edit..github/workflows/ are the maintainers'
deploy pipelines and are guarded by repository owner, so they skip
automatically on forks (no failing runs). Copy one as a starting point for
your own target and supply your own repository secrets.Parcel and assessment data come from each jurisdiction's open-data programs. Per-source licenses and credits are listed in ATTRIBUTION.md.
See CONTRIBUTING.md. Security reports: see SECURITY.md.
MIT.