Weighted mean in attribute union with QGIS

Question

I have joined Argentina's census data wich has small area divisions, with electoral disctrict data wich has a bigger area than those of the census. In QGIS, I've calculated centroids for the census data and joined through the "join attribute by location (summary) tool".

Census data i aggregated to get the population in electoral disctrict, but i've found no way to caculate a census index which i would want to take to the electoral level by doin a weighted average:

(sum (index * total_pop)) / sum(total_pop)

The question sums as: Is there a way to add a statistic function (a user edited one) to the spatial join of the census centroids to the electoral disctrict polygons?

Answer 1

I suppose you have a polygon layer called district and a point layer census with attributes named index, pop (for population) and id for the polygon feature from the layer district that the point is within. To calculate for each district separately the sum of pop muliplied by index, use this expression:

sum( index*pop, group_by:=id)

---

In case you don't yet have an attribute id that links to the district the the census-centroid is in, create it using Join attribute by location or by creating a new field with field calculator and this expression (instead of name use any attribute name of the district layer - or $id if you have no attributes there):

array_first (overlay_within( 'districts', name))