ChangeDetector Still Flagging Geometry Changes Even with Lenient Matching and Vector Tolerance

Just a few ideas that come to my mind:

-Have you checked the change report lists? What are those saying?

-Have you tried using a Matcher instead of ChangeDetector for a simpler approach, check if that grants the same results?

-without knowing your geometries it is kinda hard, but maybe you find a way to refine/validate/reset them to get the same ordering/orientations

Hope any of that helps, cheers!

Can you extract a couple of the ‘non-matching’ geometries and post here?

I am also running into this issue. Albeit it’s a small set of values.

I think it may be stemming from having the “Simplify Geometry” option enabled in the writer, so when I rerun the dataset with fresh data, the “revised” pipeline has un-simplified geometry, maybe if its simplified before going to the ChangeDetector it may match up better?

Does anyone know of a transformer that performs the same operation as the “Simplify Geometry”? Wondering if that will solve the issue.

Also Max, thank you for mentioning the lists you can output with the ChangeDetector, I didn’t even know that was an option, I am going to write those results to a excel sheet and have it act as a change log!

Hi ​@aawad.93 and ​@schulte.a 

It might be the measures or elevation.

If you change the ChangeDetector's Check Geometry-> Match Geometry setting from FULL to 3D you will turn off measure matching. This will confirm if the issue is with mismatched measures.  
If you set it to 2D you can rule out elevations as well.