Occasionally I'm asked to plot points that relate to sensitive data items. I would quite like to be able to add a varying level of random jitter to points when plotting so that point density is still evident to viewers but without giving away precise locations.

I could do this quite easily by adding/subtracting random values from the co-ordinates in Excel or SQL prior to loading data but this would make it a bit difficult for when I want to do precise analysis based on actual co-ordinates as I would have to swap between real and jittered values (which is just asking for a mistake to happen further down the road).

Is there a way of doing this in QGIS? I've had a look through the documentation and haven't spotted anything so my last hope is that someone knows of a plug-in somewhere.

I know of other techniques to protect sensitive data on maps (binning, assigning to centroid, heatmaps, contours etc.) but none of these are quite right for the data I'm trying to display (individuals who have used a service multiple times, with point size related to number of times they've used that service).


Aggregation to a boundary such as postal code then symbolize the polygons for bivariate data representation (number of users and service usage).

There are many methods for mapping bivariate data and I am sure others will chime in. One such way is to use chloropleth symbology on your polygons to represent average service usage and dot density within polygon to represent number of users.

This way you dont have to mess with your original data points and the method results are always replicable.

  • Great thanks, I hadn't thought about using a bivariate type symbol to display this. One thing I was looking at is the 'offset point' function in QGIS which will shift all points by a set distance but your suggestion is a lot cleaner. – Tumbledown Aug 3 '12 at 9:57

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.