Hi,

When the data is not superimposed on each other and don't have unique number for identification then which points are you trying to compare?

can you post some more detail with screenshot to understand better

I have two datasets from different sources which are supposed to represent the same thing (but in reality, there are some little XY-differences). I try to quantify this differences.

I thought to compare the two shapefiles "cell by cell" (every 1m² for example). Then, I would be able to average the difference using the shape_polygon1 's outlines.

What I tried :

capture.png

Sorry I did not understand your requirement.

you have source shp file as geometry (polygon) feature and you are using Image transformer and compare cell-by-cell???

That's it. It looks easier for me to obtain a raster with z-attribute from the shapefile so as to compare z on the basis of xy-position (since I can't pair polygons). But I'm certainly on the wrong track.

Here is an example. The red polygons have a z-value, and black ones another z-value.

capture.png

Thanks

Hi @joju, in your screenshot, I can see that there are black polygons and red polygons, and they are overlapping with 1 to 1 relations. In the situation, I think the SpatialFilter can be used to merge attributes of a red polygon to a spatially related black polygon. vice versa.

Thanks a lot, it works for the case you quote. Moreveover, SpatialFilter allows me to separate 1 to 1 relation from other cases (one of my goals !)

However, I've also the following thorny case : one red building contains many black ones. I'd like to compare red mean z-value from each black ones.

There aren't M to N relation.Thank you, it solves the issue !