Follow-up exercise

To do

For this follow-up exercise, create an R markdown file that, when knitted, includes the code and output for the following, ideally as a pdf document:

  • To read the following file into R: https://github.com/profrichharris/profrichharris.github.io/raw/main/MandM/boundary%20files/Local_Authority.geojson. The file is in .geojson format. The data are sourced from here and give Index of Multiple Deprivation (IMD) data for English local authorities.
  • To create a neighbourhood list for the local authorities and then spatial weights. How you define neighbours is up to you.
  • To use the spatial weights to map clusters of high or low values of the variable RAvgScor. This is the rank of the average 2019 multiple deprivation score for each authority. (It isn’t therefore, really a continuous variable1 but we will treat it as though it is).

Conclude with a brief reflection on what, if anything, the map reveals about a possible geography of deprivation in England.

To help

All that you need is in the session ‘Measuring spatial autocorrelation’. Keep in mind that a rank with a low numeric value is high for deprivation (because 1 = most) so, rather confusingly perhaps, a cluster of high values is a cluster of low deprivation, and vice versa. Keep in mind, also, that some local authorities may not have any neighbours, depending upon how you define them, so you may need to change the zero.policy in some functions.

Footnotes

  1. it is ordinal↩︎