Feature request #9124
Heatmap in percentage
|Pull Request or Patch supplied:||No||Resolution:||invalid|
|Easy fix?:||No||Copied to github as #:||17761|
Heatmaps would be far more convenient (easier to interpret in many contexts) if they could (optionally) be expressed in % rather than in absolute values (in inverse values, so one could create isolines including, say, 95% probability). The change seems trivial:
- calculate the maximum
- dividing the raster values by that
- adding an option to choose between absolute values and %
Also a note on the help on how to interpret cell values would make life easier for newbies.
#1 Updated by Nyall Dawson over 6 years ago
It's actually not quite that simple - the maximum value is not known until the entire heatmap is generated, it can't be calculated in advance.
The actual process would be:
1. Generate the heatmap
2. Scan for the max and min values within the heatmap (The minimum value would also need to be considered, since points could potentially have a negative weighting set resulting in raster values < 0)
3. Update the cell values with a value relative to the max/min values
I'm wondering if it would be better/more flexible for this to be done via a step in a processing script. Problem is that heatmaps can't be generated via scripts at the moment - this would require moving the heatmap code from a c++ plugin to core (something which should be done at some time in any case). This is on my (distant) radar, alongside adding additional hotspot types such as Getis Ord Gi*. I'm hoping to get funding or some time of sponsorship to tackle this.
#9 Updated by Nyall Dawson over 2 years ago
- Status changed from Open to Feedback
I think this should be closed, and replaced with a feature request for a processing algorithm which scales raster values to percent of max.
Heatmaps are part of processing now, so only the scaling algorithm is missing from this request.