I have built the first classification based on the entry_filter using row filters.
how can i let the tool build the grouping automatically using the entry_filter column instead of buildin 4 different row filters?
Your data isn’t clear. Where’s the “entry_filter column” ? If its not proprietary, it would be helpful if you could share your data and give a more detailed description of the desired output.
I want to classify the rows into separe nodes based on entry pattern. i created it manually using row filter and hard coding the pattern string.
Today, there are 4 possibile category but this list can expand
Without knowing what your downstream processing looks like, its difficult to make any solid suggestions. The only way I can think of to avoid multiple row filters is to use a loop with a GroupBy node grouping by the entry_pattern and creating a list of the exit-names then ungroup inside the loop. The problem is what to do with the individual entry_pattern groups. You could either process them inside the loop or write them to files to be retrieved later. In either case it might be easier to use multiple row filters even if it doesn’t seem very elegant. Again without knowing your bigger picture I’m at something of a loss. I’m working on a loop approach which writes the individual entry_pattern groups to files.
@Suresh_Murthy an instrument could indeed be a loop. But as @rfeigel has said it is difficult to know without a concept of what it is you want to do and what the role of the input and exit pattern is.