This post may contain affiliate links. As an Amazon Associate, we earn from qualifying purchases.
Since 1994, the City Council and state legislature have gradually dismantled legal protections for New York City tenants, giving landlords an incentive to hike rents and evict lower-income tenants. We’ve mapped more than 450,000 New York City eviction cases filed between January 2013 and June 2015. Look up your building to see its recent eviction cases and whether it may be rent stabilized. Source: Evictions data from the New York City Public Advocate’s Office. Rent stabilization data derived from the New York City Department of Finance and taxbills.nyc. Building footprints from NYC OpenData. Additional property data provided by Rentlogic. Additional analysis by Stephen Werner.
Methodology: We geocoded eviction petitions to include property identifiers used by New York City’s finance and building departments. We then used those identifiers to link the data to additional information about each property where an eviction was filed, such as its age, size, building classification, tax benefits, owners and officers. To limit our results to eviction cases filed in private residential buildings, we removed public property owners, such as the New York City Housing Authority. We also removed commercial and industrial properties as identified by New York City Dept. of City Planning Land Use Categories.
As far as building it, first we loaded up all that data and joined it in a postgis database. Then we output a geojson of all the footprints joined to the data, which we processed with tippecanoe [https://github.com/mapbox/tippecanoe] to make vector tiles. Then we used mbutil [https://github.com/mapbox/mbutil] to output those as flatfiles and uploaded ’em to S3. We used the excellent Mapbox GL library to make the map using the vector tiles as the source of the footprints, and sandwiched those between Carto’s lovely label-free base map and label map raster tiles.
As far as geocoding to give you the info for your building, we use Google Places autocomplete to get a point for the address, then grab data from the matching vector tile and use turf [http://turfjs.org] to find the building that corresponds to the point you typed in.
