Re: [postgis-users] Incorrect Result for ST_CountAgg with nodata pixels?

2016-07-13 Thread Michael Treglia
Hi All, Just to follow-up - looks like I've figured out the issue: In the count of total number of pixels per building, it seems I was calculating the total number of pixels per tile that intersected with the polygons, not the total number of pixels that actually intersected the polygon. For

[postgis-users] Incorrect Result for ST_CountAgg with nodata pixels?

2016-07-12 Thread Michael Treglia
Hi All, I'm working to do some overlay analyses between raster and vector layers, and having an issue getting an accurate count of pixels per polygon. My query is below - basically my raster layer has values ranging 1-5, with some nodata pixels (defined as such, with values of 0). I want to