How much of NYC's open data is well formed, who publishes it, and how stale it is.
Coverage is not completeness — Quality scores are computed a batch at a time by a nightly worker, so a dataset with no score has not been assessed yet — it does not mean it scored badly.
Freshness is the publisher's, not ours — “Updated” reflects when the city last changed the dataset on the portal. A stale dataset is a fact about NYC's publishing, not about this app's sync.
Distribution of overall grade across scored datasets.
Where the catalog is weakest, out of 100.
Breakdowns
Dataset count by attribution. Hover a bar for its average quality.
Socrata's own domain category.
Time since the source last updated the dataset.
How each dataset carries location. Only point and lat/lng render on the map today.
Overall quality score, out of 100.
Then pick any dataset to aggregate inside it, live. Counts are aggregated by NYC Open Data itself, so grouping a 38-million-row dataset costs the same as a small one and nothing large is downloaded.