1

I have two sets of point data, structured like below.

Table A 
long | lat | class 
 ... | ... | A
 ... | ... | A
 ... | ... | A

Table B
long | lat | class 
 ... | ... | B
 ... | ... | B
 ... | ... | B

Both tables aim to capture locations of the same certain things, but Table B is much more granular than Table A. The result is that Table B is quite large (higher n) and Table A is more manageable.

I want to see if Table A is representative of Table B (in other words, is Table A "good enough" to represent table B?). How might I approach this? My thought was to see if they two classes are spatially correlated for a certain geography, but I don't know how I would approach this. I'm using R.

1

This looks like a classic "point pattern analysis" problem. Your hypothesis is something like both sets of points come from the same underlying intensity process, or that table A is a random subset of tables A + B.

Various tests exist for testing hypotheses like this, mostly implemented in the spatstat package. Textbooks by Baddeley et al or Diggle or Cressie are good references.

Another option is the spatialkernel package which can show where on a map the intensity of pattern A is significantly higher than pattern B - for example to show areas of significantly high crime taking into account diversity in population density.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.