Geospatial data aggregation: cases & examples

Myeong Lee

---

0. Getting Ready

• This tutorial is based on R.
• Sample data is available in the data folder (a sample of Car2Go's location data)
• Sample code is availalbe at the .Rmd files.
• Currently, there are two sample files available (geo_agg_tutorial.Rmd for geospatial visualization and aggregation; and voronoi_overlay_sample.Rmd for generating Voronoi diagram based on real-world points).

1. Geospatial Points

• Loading the data as a dataframe (if lon/lat).
• Loading the data as SpatialPointsDataFrame (sp package).

2. Geospatial Polygons

• Sources might be GeoJSON, Shapefile, or KML.
• Loading the data as SpatialPolygonDataFrame (sp package).
• Many boundaries datasets available from Census or city data portals.

3. Basic Geospatial Operations

• raster package for intersection, union, and etc.
• It's possible to iterate through each polygon in a SpatialPolygonDataFrame.
• Example: Aggregation I: Points to Polygons

4. Density-based Viz

• Using ggplot and ggmap to visualize frequency using the color gradiation (polygon-based).

5. Generating Polygons

• Hexagons (in this example)
• Grids
• Voronoi diagram

6. Aggregation II: Polygons to Polygons

• Intersection and union are still available (raster)
• When aggregating different polygon layers, you need to prorate the value of each polygon based on the proportion of the area.