Hey Rainer,
No worries, glad to help. I can't speak to the 2 Quartz options as they are still in design. Once they are fleshed out, I believe the raw raster analysis is going to be the cleanest and most straight forward.
For now, try to create the feature table, either of grids polygons or maybe even point centroids (do this work in ArcMap). Once you have that, I would bring it into your Runtime app, and do the following:
1) Create signal handler for mouse clicked on the map
2) call Buffer on the mapPoint that is returned
3) Set up a query against the feature table. The query will need to have several things set:
- Where clause to "1=1" to return any record. Change this to some other SQL clause if you want to filter results.
- Set the geometry to your buffer polygon
- Set spatial relationship to either intersects or within, depending on what your requirements are
- Set outstatisitcs to gather the average, standard deviation, sum, etc. Here are the different out statistics you can set - ArcGIS Runtime SDK for Qt QML API: Enums Class Reference
4) Execute the query and create a signal handler for when it completes. The results will be aggregated values to whatever your outstatistics were set to, as opposed to individual values from the raw data
I do something similar in the following example. It creates the buffer, runs a query with the buffer, sets out statistics, and gets the aggregated data points back. The only difference is this runs a QueryTask, but this workflow should work against a feature table with queryFeatures. RuntimeQtSamples/main.qml at master · ldanzinger/RuntimeQtSamples · GitHub
Hope this gets you on the right track.
-Luke