Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.
Originally posted by: WizardousHi
I have a DB query in my graph which gets data from a teradata db. The source data has decimal columns with specific formats (like '99999999999999' or (10,2)). The data i wanted to use long type (id column) comes with type double(like 8.6907191001e+12) to the graph and when i convert it with a filter node using id.long() syntax, some duplicate results occur.
To prevent duplicates i get data as string manually with sql words and then convert to long(as seen on
Convert_with_no_duplicates.brg), the result was fine. But i am still not sure if it is a bug or it just has to be like this or even i have faults on converting or acquiring data. Are there better or more accurate ways of converting?
Thank you
Kind regards.
Ozgun