Was just wondering if anyone could help me with a problem. I'm trying to map wellbeing scores (out of 10), using locations from a survey of 15,000 households, using the HEatmap command in QGIS. It's an unclustered sample, but obviously the density of points is dependent on the population density of the area. I'm using wellbeing score as a weight in the heatmap, but unfortunately every map I produce is effectively a map of population density. I have tried multiple ways of weighting the wellbeing score to factor in the differing densities of points (so scores in areas with only a few data points are weighted, up and scores in dense areas are weighted down), but the heatmap looks effectively the same every time. Is there any sensible way of doing this? Thanks.
Please read How to build effective heat-maps?
It seems like you are looking for Distributions of attribute values rather than Concentration of points. Therefore, the QGIS heatmap plugin is the wrong tool for the job since it only does concentration of points.
Try Raster | Analysis | Grid (Interpolation) instead.
Another solution could be to first generate a vector grid - could be a hex grid if you like it fancy - and then calculate e.g. the average wellbeing score for each cell and map that.
I realize this post is a year old, but I came across it when looking up information on kernel density estimations with QGIS. I've not tried this yet, but it appears from the QGIS documentation that the Heatmap Plugin can make use of the attributes of the original points via the "Use weight from field" option.
(would love to hear back from anyone with experience with this)