How to sum up columns of each same data set after groupby?

Hello everyone,
I am new here. I need your advice so much.
I need to sum up each model(Nitro,Miza,Paz).

After I use nodes as my understanding

I got the result like this:

But I need the sum of each model like this:

Please advise how to do it.
Thank you in advance for your help.

Hi @toonchainarong

Welcome to the KNIME forum!

I believe this can be solved using the “moving aggregation” or the “groupby” nodes and I would be happy to try it this way.
Could you please upload the table (as csv, excel or Knime table) that you are displaying in your first image ?



That’s a very kind of you.
I appreciate your help so much.

I’ve attached excel file already.

Thank you so much.need help.xlsx (13.3 KB)

Hi @toonchainarong

My pleasure. Please find a possible solution here attached. It was a bit more involved than I thought but the result is there:

20210426 Pikairos How to sum up columns of each same data set.knwf (127.9 KB)

Hope this helps.

All the best,


PS: A second alternative version based on a Java Snippet to generate Missing Values instead of 5 nodes after Excel reader in first posted version:

20210426 Pikairos How to sum up columns of each same data set (Java Snippet).knwf (109.5 KB)


@aworker - great workflow!

It has a whole bunch of simple but clever ideas wrapped up in this one single flow, and nicely explained by the comments too!


Thanks a lot @takbb for you kind comments !



Nice one @aworker Especially like description under every node and usage of row index throughout workflow :+1:
Another alternative to generate missing values is Rule Engine node. Simply don’t match rows where you want to have missing value. That’s just in case someone doesn’t know (simple) Java :wink:

NOT $Model$ = "" => $Model$



Many thanks @ipazin for your kind comments and for the hint to generate missing values :clap: :slight_smile:
I didn’t know that this was the way to generating them within the Rule Engine node but it makes definitely sense if no option is left as default (TRUE => “something else”) as usual. Great clue :slight_smile: !




Hello aworker,

OMG. You are so talented.
First of all, I need to apologize for late response.
And I’m very appreciate that you sacrifice the time doing this task for me.

That’s all help me so much.

Thank you so much aworker.

God bless you!!


Hello @toonchainarong

Thank you very much for your so kind message. It was really a pleasure to help and having your positive feedback is really rewarding. Do not hesitate to ask again.

If you think this question is fully answered, I would be grateful if you could validate the answer with the workflow attached as the solution.

Best wishes


1 Like

This topic was automatically closed 182 days after the last reply. New replies are no longer allowed.