Section 2.8 Summary and further reading
In this chapter we explored the differences between attribute and spatial operations, and we made great use of the latter in practicing spatial manipulation of data. These spatial operations allow us to manipulate our data in a way that lets us study what is going on with crime at micro-places. At the start of the chapter, we introduced the idea of different crime places. To read further about crime attractors vs. crime generators, turn to the recommended readings by [112] and [122]. There have since been more developments, for example about crime radiators and absorbers as well (see [123] to learn more).
We covered how to subset points within an area, build buffers around geometries, count the number of points in a point layer that fall within each polygon in a polygon layer, find the nearest feature to a set of features, and turn non-spatial information such as an address into something we can map using geocoding. These spatial operations are just some of many, but we chose to cover as they have the most frequent application in crime analysis.
For those interested to learn more, spatial operations are typically discussed in standard GIS textbooks, such as those we have recommended in previous chapters. You could see, for example, chapter 9 of [124]. But probably the best follow-up to what we discuss here is chapter 4 and 5 of [115], for it provides a systematic introduction to how to perform these spatial operations with R. There is an online book in development by two giants of the R spatial community, Edzer Pebesma and Roger Bivand, which at the time of writing is best cited as [125]. The first chapters of the book provide a strong backbone to understand
sf objects in greater detail, coordinate systems, and key concepts for spatial data science.
