3

Is there a way to (preferably in QGIS that I actually can use) analyze two sets of point layers and describe the correlation in the espective cummulations?

For instance, I have points that experts have chosen as dangerous and then I have points that regular people have chosen as dangerous. I can do a visual comparison by putting heatmap and clusters above eachother. I would like to find out how much these correlate and also designate clusters that both groups deem dangerous.

The points visualised to better understand:

enter image description here

Improve this question
1
  • 2
    This is a bivariate point pattern statistics question, and I don't think QGIS has the capability except via an interface to R, and the spatstat package. To answer your question also you'd have to phrase your question in terms of statistical distributions in order to calculate if there's significant correlation or not under the assumed model. This looks further complicated (but not impossible) because your points seem constrained to a line (road?) network. Consult a statistician...
    – Spacedman
    Oct 18, 2021 at 15:42

1 Answer 1

Reset to default
1

One way is to rasterize one of the point layers. Then use SAGA -> Geostatistics -> Multiple Regression Analysis (Points/Raster). It will give you the correlation between the two dispersions.

Improve this answer
1
  • 1
    Nope, if you mean saga-gis.org/saga_tool_doc/2.2.0/statistics_regression_1.html it correlates a grid with attributes of a set of points - for example maybe you have point rainfall measurements and an elevation grid. Here the correlation is between two point pattern locations. If you rasterize one, you've just got some point locations left.
    – Spacedman
    Oct 18, 2021 at 15:55

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.