A common problem I face is that the default row name from KNIME is in the format "Row0, Row1, Row2 .etc" wheras if you output a table from an R snippet the RowID is commonly "1,2,3". Obviously I could write some R code in the R snippet to add a new rowID to match the KNIME row ID format but it would be useful if the Joiner node also had the option to just join columns in numerical order rather than by matching. i.e.
Anyone know of a workaround?
(Answering my own post again)
If you use a RowID node, set 'Replace RowID with selected column values or create a new one' but then set New RowID column to ?<none> then it re-numbers the rows Row0, Row1 etc.
Then you can join the tables
I just wanted to let you know that I've created a very simple plugin (Column Concatenate) that does just that (simple row join). You can download the plugin update site from github:
The WorkFlowToolkit contains a few simple nodes that could be helpful. Let me know if you had the chance to use it.
I think the Column Append node does this in KNIME 2.8.
Shin: Could you add documentation what is in that update site? Also, I have not seen the sources. Where are they? Thanks, keep it up. :) gabor
Yes indeed it does, thanks for the tip! :)
I made these nodes for my personal convenience while using 2.7 (maybe the function already existed and I couldn't find it?), so I'm glad this feature is officially supported in 2.8. The upgrade has been wonderful so far, especially the error handling nodes.
There are other nodes in the toolkit, but I'll make sure to see if it exists in 2.8. I'll put up sources in the near future - they are ridiculously short.
Column append in knime2.8.1 does not work for non-marching rows. These are one row tables. The error is shown in below:
ERROR Column Appender Execute failed: Tables contain non-matching rows or are sorted differently, keys in row 0 do not match: "Row0" vs. "Row219"