After running features through a transformer that does uses a Group By, is there a simple way to determine which destination features came from which source features? The best I could come up with is some combination of SpatialRelator and FeatureMerger, to do a spatial join on the result.

Edit: For example, I have a source dataset that feeds into a BUFFERER, that I set a group-by attribute on. For each BUFFERER output feature, I wish to associate it somehow with the subset of original features that donated themselves to it. Unfortunately I don't believe the BUFFERER has a List attribute in FME Desktop 2011. How would I do this?

Edit edit: bufferer, not dissolver.

  • Isn't there a attribute that you can expose that is called fme_basename or fme_filename or something like that? Couldn't you maybe expose that, toss it into a variable and then call it later? Don't have access to FME right now or I'd test it out. Apr 28, 2011 at 16:54
  • @Chad: I'm looking for the specific source polygons within a dataset, not just the source dataset as a whole. May 20, 2011 at 12:37

4 Answers 4


Using the List feature you can keep source features

FMEpedia is a excellent resource for this information.

specifically for you http://www.fmepedia.com/index.php/Group-Based_Transformers#Output_and_Output_Attributes

"Using a list. Many transformers have a list option so you can keep track of source attributes by keeping all the values. "

just create a list of source datasets before you set it in the groupby.

  • BUFFERER has no list attribute. May 20, 2011 at 12:37

The fme_feature_type attribute contains the value of the feature's original feature type (feature class, table name etc.).

  • I'm looking for specific polygons, not just the name of the source dataset. May 20, 2011 at 12:37

I think what you are looking for is the filterTypeFilter transformer. Give it a try.featureTypeFilter

  • I'm looking for specific source polygons, not the name of the source dataset. May 20, 2011 at 12:41

Rather than SpatialRelator and FeatureMerger, I would use the Aggregator and FeatureMerger. The Aggregator would let you use the same group-by as the Bufferer, but with the ability to create a list. Then the FeatureMerger would let you merge that list onto the buffered features. Of course there would need to be an ID of some sort to enable the merging.

  • This is the best answer so far. Unfortunately I also want the effect of having run the bufferer, which means I have to do the provenance assocation after the bufferer (bufferer and amalgamator will result in differently membered-groups) which means the amalgamator won't have any effect. May 20, 2011 at 12:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.