Hello, new to FME here and am looking for some advice on which transformers might work for a workbench that I am trying to create. I have 2 road line feature classes stored in an SDE geodatabase. The lines store similar information but are created and maintained by two different departments (they have different road name standards and the geometry does not match). I would like to compare the lines and identify which ones from Roads_A are missing from Roads_B.
My main concern at this point is identifying missing subdivisions like the one in the screen capture below.
Page 1 / 1
Hello @srijneke , how are you?
I suggest you the use the transformer ChangeDetector.
Thanks in Advance,
Danilo
This thread may be useful to you, there are a few approaches mentioned utilizing bufferer, neighbor finder, or fuzzy parallel line finder:
Yes, thank you I saw that post as well. I was hoping fuzzy parallel line finder would work but I can’t seem to get the inputs right. As far as I can tell, all it did was append the two datasets together. I didn’t see an attribute that indicated which lines were the same or different.
I used the ChangeDetector transformer with a tolerance of 10 to compare the geometry between the 2 datasets. The output is about half of the features. Would it be possible to now use that output and compare the street names? i.e. identify segments with street names that do not match.
I see some strange results from the ChangeDetector. The original datasets are the pink and orange lines, the results of the ChangeDetector are the green dashed lines. In my example below, I wouldn’t expect any green lines since all of the lines are less than 10m apart. Why are these features included?
How have you configured your change detector?
See this link to see if the suggestions here help:
Thanks for passing along that thread. These are my current settings:
