FME Desktop 2022: The “clip” tool fails to produce the expected result,

I erased one of the polygons that has issues at the level of AutoCAD. However, the issue persists to exist despite the fact that the “geometry validator” tool detects nothing

Which geometry issues in need to resolve so that the result gets correct?

Why the “geometry validator” tool fails to detect any issues while the result of the “clip” tool is incorrect?

The data and workbench are attached

Really weird. How come the result of clipping 13 polygons from 1 polygon can be as shown in the screen considering that fact that there are no geometric errors?

Hi @jamal​ ,

I think there is no bug involved on any transformer. The problem is coming from the INPUT data (which polygon coordinates has less precision than expected).

The fact that you then used a FileGDB to perform an ArcGIS Pro operation is not that a proof that something is happening in FME Desktop. What I mean is that the FileGDB interacts with your geometry and "implement" a snapping grid which modify the INPUT coordinates, luckily, it solves directly the precision issue that comes from the INPUT.

So, FME uses all the precision from your input, which ends up failing or giving you a bad output when performing complex spatial transformations. And once again, it does not mean that FME is now working well, is because is trying to do some spatial operations where topology must be fine in case you want to have a successful result.

There is a Transformer called Snapper, which acts likely FileGDB format is doing. For that reason, from your input, when polygon shared coordinates are too near, but not the same X,Y, we're forcing with this transformer to be the same, to be totally aligned. With that, you can continue with the rest of Spatial transformations without any issue.

Here is the Workbench that I've re-used from the one you uploaded:

On this case, I've used AreaOnAreaOverlayer instead of Clipper (which I guess could produce same output).

Here is the Output:

1_ Inputs:

2_Final:

Let me know if you have any doubts,

Juanma

Sure, here is the workbench attached. "Simple_solved.fmw".

Regarding the Snapper parameter, you'd need to adjust to your Reference system values, and to know "how much" your polygons are misplaced one to the other.

If you still find any issues, feel free to comment ;)

Best,

Juanma,

I’m still struggling in how to IDENTIFY polygons that have this issue and what this issue is all about that causes not to get the expected result from the clipper