Skip to main content

I have two set of polygons A and B. After passing by dissolver transformer, it doesn't dissolve all the polygons. Instead it only dssolves in a random fashion. This results in missing of the polygons in dissolved polygon shape.

Polygon A:

Polygon B:

After dissolving A, B. I get this output. Clearly it is missing few of the polygons in dissolved shape.

Also shown below - Highlighting the dissolved in dotted(with coordinates) and missing the other polys with no dots.

Used parameters:

Any thoughts please?

Tagging in @takashi, @erik_jan, @TiaAtSafe, @Mark2AtSafe, @RylanAtSafe

A quick thought - change the Input Ordered to No. That might be causing the assumption that data arrives in an order that it doesn't. If that helps, let us know because I would want to check with development about it.


A quick thought - change the Input Ordered to No. That might be causing the assumption that data arrives in an order that it doesn't. If that helps, let us know because I would want to check with development about it.

Also check the log window. When I try this transformer I see:

 

 

2018-06-01 14:39:22| 0.7| 0.0|STATS |Dissolver(PolygonDissolveFactory_base_factory): Dissolving 436 input polygon(s) across 1 group(s)

 

 

Does your log have this? Does it show the correct numbers?

 

 

Also look for something like:

 

 

2018-06-01 14:39:23| 1.0| 0.0|STATS |Dissolver(PolygonDissolveFactory_base_factory): Completed polygon dissolve on all groups

 

2018-06-01 14:39:23| 1.0| 0.0|INFORM|Dissolver_AREA Feature Recorder -1 2147614722(RecorderFactory): Recorded 1 feature(s) to file `xxxxx'

 

 

Do the numbers there look correct too? Any other warnings?

 


Hi @Mark2AtSafe,

I did change the Input Ordered to No and it's still same behavior. Infact this time, more polygons are disappeared now.

Regardless of this change, I do see the above observations in log window:

Here the 20811 input polygon number is right. However don't know the exact number of dissolved polygons. I found this difference when I visualized it.

  • Dissolver(PolygonDissolveFactory_base_factory): Dissolving 20811 input polygon(s) across 1 group(s)
  • 2018-06-01 15:54:02| 183.9| 0.7|STATS |Dissolver(PolygonDissolveFactory_base_factory): Completed polygon dissolve on all groups

     


    2018-06-01 15:54:02| 183.9| 0.0|STATS |Dissolver AREA Splitter(TeeFactory): Cloned 371 input feature(s) into 742 output feature(s)

     


    2018-06-01 15:54:02| 183.9| 0.0|STATS |Dissolver_AREA_0_ETD2D9TM6RM= Feature Counter -1 61(TeeFactory): Cloned 371 input feature(s) into 371 output feature(s)

     


    2018-06-01 15:54:02| 183.9| 0.0|STATS |Dissolver_AREA_1_L04hQmr69h0= Feature Counter -1 176(TeeFactory): Cloned 371 input feature(s) into 371 output feature(s)

     


    2018-06-01 15:54:02| 183.9| 0.0|STATS |Dissolver_<REJECTED> Feature Counter -1 2147745849(TeeFactory): Cloned 0 input feature(s) into 0 output feature(s)

     


     


Hi @fmeuser_gc, I remember that the Dissolver could not work as expected in some special cases, including the case where the target areas were ellipses or elliptical polygons overlapping each other.

What version of FME are you using? Hopefully it has been improved in the latest version.

I'd strongly advise trying FME 2018.x (release or 2018.1 beta) with a Dissolver set to have Tolerance at Automatic.


Hi @fmeuser_gc, I remember that the Dissolver could not work as expected in some special cases, including the case where the target areas were ellipses or elliptical polygons overlapping each other.

What version of FME are you using? Hopefully it has been improved in the latest version.
Hi @takashi, we are using 2017.1 version. So do you think this is expected?

 


I'd strongly advise trying FME 2018.x (release or 2018.1 beta) with a Dissolver set to have Tolerance at Automatic.

Hi @daleatsafe, Are you sure it's fixed in 2018.1 beta version?

 

Our FME server is in 2015.1, We recently had an ArcGIS upgrade to 10.5 which means it needs higher than 2015.1 version.

 

So I was trying in 2017.1 desktop and found this issue. I guess it also exists in 2015.1.

 

 

Not sure if I push the workspace from 2017/2018.1 version desktop workspace to 2015.1 server works or not? Do you think it works?

 

 


Reply