Good day all,
is it possible to create a table out of the 3 below nodes?
What I want is to obtain a table created with one column from each of the 3nodes:
Thank you!
Good day all,
is it possible to create a table out of the 3 below nodes?
What I want is to obtain a table created with one column from each of the 3nodes:
Thank you!
Hi @Mihnea , do all three tables have the same number of rows, or have a key in common? If there is no key but you just want to align the rows sequentially, then you can attach all three to a Column Appender:
Unless they have identical number of rows, and the same RowID, you’ll probably need to configure it to “Generate new ROWIDs”.
If you only want a subset of the columns (e.g. one column from each) appended, I would suggest adding Column Filters BEFORE the Column Appender, to restrict it to just the required columns.
On the other hand, if there are keys that need to be matched, then you could use two joiners. The first joiner joins the two Column Expressions nodes, and the second joiner joins the joiner to the Column Aggregator. Within the joiners, you would specify the required columns to be retained.
The choice of method really depends on your data and how it aligns in the above three nodes.
Thanks a lot @takbb
I just want to align the rows, no keys are needed, Column Appender sounds good.
Actually my task would require this type of appending for a lot more columns from many different nodes into this final table. Is that possible?
(I want to run many (300+) DQ checks on different columns of the same initial table and want to consolidate only the check results in one final table that would consist of the DQ check name as column and 0/1 as pass/fail in the rows)
Thank you
Now I see better what you mean.
I noticed I can add input port to the Column Appender (not sure what the max number of input ports is?)
And indeed, filtering before the Column Appender makes a lot of sense.
Thank you again!
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.