Hi all,
Could someone give me some advice on how to do spatial analysis in R? Here’s the scenario:
I want to map disease prevalence in a village, specifically the seroprevalance in 4-17 year old.
I have the seroprevelance data, as well as the GPS coordinates for the houses.
We already know that prevalence increases with age, but I want to know if there are particular hotspots of infection within the village. How would I do this in R? I want to remove the effect of increasing age from my analysis, as the distribution of age across the village is not uniform.
I’m very new to R – I’m able to import my data, but other than that it’s a new learning experience for me! Any advice offered would be most appreciated!
Thanks,
Tom
There are many text books on this topic, as well as many R packages. see https://cran.r-project.org/view=Spatial
For a start, see https://stackoverflow.com/questions/28672399/spatial-clustering-in-r-simple-example
Hi Thomas,
I think you can use Kernel density. Also, to do more kinds of spatial analysis I would sugest the use of GIS tecnology (QuantumGIS or ArcGIS).
have a nice day,
Best regards,
Patricia
Hello,
In order to detect hotspots, as Renaud Lancelot suggested it, density-based clustering algorithms such as DBSCAN would do the job.
This being said, these algorithms are quite sensitive, and one should quickly explore GPS data, in order to fix the two dbscan parameters.
If you want to add a third variable, there are declinations of DBSCAN, and here is a good review of these algorithms (applied on trajectories datas).
Good luck
Best regards
Colin
http://www.lume.ufrgs.br/bitstream/handle/10183/17024/000708303.pdf?sequence=1
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Software
- R Statistical Package