Hello!
I started using the Joiner (Labs) node today as a drop-in replacement for Joiner.
It’s faster and the ability to get out 3 different data streams is great - spares me from using a splitter right after the joiner.
However, in one case, Joiner (Labs) led to a data loss error.
Here were the conditions:
1 table has 49 rows x 33 columns. The second 110190 rows x 26 columns.
Join is over 2 fields. One is a guid the second an InChIKey.
Issue: data in 2 columns (at least) is set to missing for some records.
One data table is private data so I cannot just send the entire workflow. Maybe I can provide the relevant parts, if useful.
and with “old” Joiner there are no missing values for those columns/records? Do you know a way to reproduce it cause otherwise I’m not sure where to start? See here topic for bug reporting: Bug Reporting Best Practices
Maybe you can use dummy data for table where data is private?
Thanks, Ivan!
I think I’ve figured out why there’s data missing in the case of the new Joiner and not the older one:
The older one offered the option of match any of the selected column pairs, as far as I can see, the new one only matches all.
Is that correct? Is there a way to match any of the listed pairs of columns?