Skip to main content

I have a table called "alternate street names" that contains values such as: NORTH, NORTH MILL, NORTH, DRIVE, MILL, MILL DRIVE, DRIVE. (These are the permutations of 'NORTH MILL DRIVE'.)

I also have a table of "exclusions", such as: AVENUE, STREET, DRIVE, NORTH.

The objective is to only write the rows from "alternate street names" where the alternate name is not in "exclusions". Also note that the "exclusions" must be read at run time.

I did a lot of searching and trying different transformers, but could not get anything to work. This seems like a very simple task and I'm just overlooking something.

Use the FeatureMerger to merge both tables, based on the common join field.

Use alternative names as requester and the exclusions as supplier.

All not merged features are alternative street names not in the exclusions. All not referenced are exclusions not in the alternative names.


If the values you described are attributevalues then you only need touse a featuremerger on the attributevalues from each table.

As featuremerging is almost basic functionality and you said you did a lot of searching, it makes me believe that your table is constructed differently? For onstance you might be naming strings rather then attirbute values?


Hi t1,

As already mentioned by erik_jan and gio you can use the FeatureMerger to tackle this problem. I noticed that you have duplicates in the list containing the alternative street names. Try to remove those (if that is the functionality you want) before using the FeatureMerger using a sampler with a group by (see screenshot below).

If the list of alternative street names and the exclusions is rather large, you can improve the performance by using the suppliers first-option of the FeatureMerger. But make sure all suppliers arrive before the first requestor! This can be done by reading the data from the exclusions first. Ensure this by dragging this reader on top of the reader reading the street names.


If the values you described are attributevalues then you only need touse a featuremerger on the attributevalues from each table.

As featuremerging is almost basic functionality and you said you did a lot of searching, it makes me believe that your table is constructed differently? For onstance you might be naming strings rather then attirbute values?

FeatureMerger was one of the first things I tried, but it did not work as expected. It works like a champ now however, and I have no idea what was wrong before. Thanks for your help.


Reply