If a relationship class works as expected on two data sets (call them A and B), where it shows the related rows between the two, why would it not work establishing a relationship on tables A and D, where D is the merging of the original table B and a new table C?
All tables B, C, and D have a column that has a key that matches to a primary key column in table A, but sometimes they do not have a corresponding match. I have verified that if I do that with the original table B (where the joined column is null) the establishment of the relationship is successful and rows that have the keyed relationship are returned.
To create table D, we create the table via the geoprocessing tools, iterate and add the rows in table B to D, then do the same with table C. Is there another step that has to happen? Table D appears to have all of the relevant data, and if I try to then create a relationship class, it does not error, but nothing is returned.
Solved! Go to Solution.
If you post the schema of tables A and D, along with the geoprocessing command that you're using to create the relationship class, we can try to help you out.
You might have better luck however, if you post in the Geoprocessing space rather than here in the SDK space.
--Rich
If you post the schema of tables A and D, along with the geoprocessing command that you're using to create the relationship class, we can try to help you out.
You might have better luck however, if you post in the Geoprocessing space rather than here in the SDK space.
--Rich
It was our process of merging tables, there was a single thread of logic that generated schema where the columns were different (text vs guid). It didn't cause any errors in the relationship class construction, which confused me until I started pulling data together for a response. Thanks.