Area Gap and overlap cleaner doesn't keep the inside polygon boundary

@nikkia do you have a small sample dataset and workspace that you can share with the group so we can try and reproduce the issue?

FMESampFME_ShareWorkspace.fmwtle.zip

Here are the files. I am trying to dissolve boundaries based on different fields.

Thanks

Nikki

Hi @nikkia The polygons get "lost" in Dissolver_POLICE, not the following AreaGapAndOverlapCleaner. Enable Feature Caching to see the results of each transformer.

Hello @danatsafe Thanks for you reply!

I want to dissolve polygons by different fields but I don't want to loose the topology.

if you can look at EMS layer object ID 46 and EMS name Front EMS has some part inside west EMS but it doesn't show the relationship correctly and selects the whole polygon ignoring the inner polygon. And it happens after area gap transformer before that they stay separate polygons.

i hope this clarifies what I am trying to achieve. check this snap shot below i don't want inner polygon to be party of the big one.

1565970120787.png

Hi @nikkia Maybe you should dissolve (group by) both POLICE and EMS?

I don't know how is that going to help. I am doing it separately. I need them as separate output. and both have different service areas.

Thanks

@nikkia The boundaries disappear because the values for either POLICE or EMS are the same on either side. Dissolvers change topology. Maybe try switching the Dissolvers? Have you looked at the feature caches from each transformer?

Dissolving by both EMS and Police doesn't help. It gives me multiple records of each EMS even after aggregation and give me not continuous polygons like attached.FME_multiRecord.PNGFme_Crevur.PNGFme_Crevur2.PNG

These 3 polygons should be one continuous polygon because they are next to each other and should have only one record after dissolve and aggregation since it is served by same EMS.

Yes i have been looking at feature caches after every transformer and found that after area gap i loose the inner polygon relationship as you see in the attached images in comparison to topology output.

FME_InnerPoly.PNGFMEBiggerPoly.PNGFMEInnerPolyAGap.PNGFMEBiggerPolyAGap.PNG

Thanks for your help!

Hi @nikkia Can you please file a case about this? I'll follow up from there. Thanks!

done.

Thanks for helping!

@nikkia Thanks for adding your sample data and workspace.

Are you able to use FME 2019.1?

When I recreate your workflow in FME 2019, I think I get the result you need. For FME 2019 we did a lot of work on the Dissolver: to improve both performance and the underlying algorithms. This also allows you to simplify the workspace and get a better result. Below are images from the attached workspace (FME 2019.1): dissolver.fmwt

Original Data

Dissolved by POLICE

Dissolved by EMC