Solved

Best Transformer for Conditional Join/Merge?

  • 8 September 2021
  • 2 replies
  • 11 views

Badge

Hi. I'm trying to get a single output from these two sources:

  • County Assessor Parcel Boundaries
  • Surveyed Parcel Boundaries

 

The goal is that, if the parcel is in the Surveyed Parcel Boundaries layer, we use its shape. If not, we keep the County Assessor layer shape. Only some of our records are in both sources. Some are in one and not the other.

 

Which transformer would be best for this scenario, and how would I go about setting it up? Thanks!

icon

Best answer by zzordo 9 September 2021, 19:25

View original

2 replies

Badge +22

Is there a common id between the datasets? If so I would try the ChangeDetector.

 

If not, then I would try something along the lines of using the County Assessor as the filters in a spatial filter (intersects) and keep only the Failed parcels from the surveyed along with the complete assessor dataset.

Badge

I managed to get something to work using the FeatureMerger Transformer instead. It works as follows:

  • First Delete Marshall and Wetzel County Parcels from Target Dataset
  • AttributeCreator and AttributeKeeper Transformers to Adjust and Narrow Down Columns (for example, reformat a Wetzel County Parcel APN to be something like WV-103-1-1-1, but Marshall County to be like WV-51-1-1-1)
  • Use EsriReprojector - Sources were USA Contiguous Alberts Equal Area Conic or NAD 1983 UTM Zone 17N US FT, Destination is GCS North American 1983
  • FeatureMerger Transformer with a Join On matching unique parcel identifiers (GIS_ID in Requestor, GIS_LINK in Supplier)
    • UnmergedRequestors were untouched by FeatureMerger, and skipped EsriReprojector at the next step since they had already been reprojected
    • Merged, UnusedSupplier, and <Rejected> all took attribution and/or new shapes from the surveyed tracts, and went through EsriReprojector (since they hadn't been reprojected yet)

ETL

Reply