Week 12

Class Outline

GeoJSON and SQL through CartoDB
- REMINDER: Always:clap:use:clap:the:clap:SQL:clap:console:clap:
- You can use a REST client like Postman or Advanced REST Client
Carto Input/Output
- Result formats
- Carto backed maps
SQL Performance
- Ask for only the columns you need
- Geometry simplification
- Indices
- Save work in new tables
- Force CARTO to recognize new tables
- Use the cartodb_id in queries (this column is indexed by default). Example:

SELECT
  *
FROM
  <TABLE>
WHERE
  cartodb_id = <some_id>

Rasters
- Tiling for TMS
- PostGIS analysis (this is painful - I've always avoided it and you should too)
- Alternatives
Work on final project preparations + library/carto exploration for the remainder of class
Extras:
- Fast spatial joins on the frontend with rtrees (demo by Jeff Frankl) Note that these joins work with bounding boxes rather than complex shapes
- Working with topojson (demo by Jeff Frankl)
- Project work by the creator of Leaflet Quoting Jeff: "rbush can handle points and rectangles and can be updated, flatbush is faster than rbush but is static (can’t be updated with new geoms), kdbush is the fastest but is only for points"