X-Aggregator Execute failed: Input table's structure differs from reference (first iteration) table: different column counts

KNIME Analytics Platform
1

I'm doing cross validation and I'm doing some filtering on training data for each fold (correlation filter). So the number of columns vary per cycle, because, depending on the training sample, correlation between columns can change and so the number of columns filtered...

 

any ideas?

 

Thanks

2

yup, ensure that all variables are created at the end of each loop, you can use a reference validator node.